diff options
| -rw-r--r-- | main/dahdi-linux-vserver/APKBUILD | 4 | ||||
| -rw-r--r-- | main/linux-vserver/APKBUILD | 10 | ||||
| -rw-r--r-- | main/linux-vserver/patch-3.0.10-vs2.3.2.1-unofficial.diff | 26517 | ||||
| -rw-r--r-- | main/open-vm-tools-vserver/APKBUILD | 4 |
4 files changed, 26526 insertions, 9 deletions
diff --git a/main/dahdi-linux-vserver/APKBUILD b/main/dahdi-linux-vserver/APKBUILD index 76db9981f70..a4691a081c2 100644 --- a/main/dahdi-linux-vserver/APKBUILD +++ b/main/dahdi-linux-vserver/APKBUILD @@ -4,8 +4,8 @@ _flavor=${FLAVOR:-vserver} _realname=dahdi-linux _name=$_realname-$_flavor -_kver=3.0.9 -_kpkgrel=1 +_kver=3.0.10 +_kpkgrel=0 _myvsver=vs2.3.2.1 # verify the kernel version before entering chroot diff --git a/main/linux-vserver/APKBUILD b/main/linux-vserver/APKBUILD index b9ad3a39649..c2b0e0bbf4d 100644 --- a/main/linux-vserver/APKBUILD +++ b/main/linux-vserver/APKBUILD @@ -2,8 +2,8 @@ _flavor=vserver pkgname=linux-${_flavor} -pkgver=3.0.9 -pkgrel=1 +pkgver=3.0.10 +pkgrel=0 _vsver=vs2.3.2.1 if [ "${pkgver##*.*.*}" = "$pkgver" ]; then @@ -21,7 +21,7 @@ _config=${config:-kernelconfig.${CARCH}} install= source="http://www.kernel.org/pub/linux/kernel/v3.0/linux-$_kernver.tar.bz2 http://www.kernel.org/pub/linux/kernel/v3.0/patch-$pkgver.bz2 - http://vserver.13thfloor.at/Experimental/patch-$pkgver-$_vsver.diff + patch-3.0.10-vs2.3.2.1-unofficial.diff kernelconfig.x86 kernelconfig.x86_64 " @@ -132,7 +132,7 @@ dev() { } md5sums="398e95866794def22b12dfbc15ce89c0 linux-3.0.tar.bz2 -0154d21e63d3f14fc1084cdb130fab2d patch-3.0.9.bz2 -da9ad65521398d86afb61223bf4d19a1 patch-3.0.9-vs2.3.2.1.diff +8e0391dd935d838ee3023ab9daa6b447 patch-3.0.10.bz2 +6d2fd60b055272429dde5e4043280fcb patch-3.0.10-vs2.3.2.1-unofficial.diff fdf2737ad7b77bcfe4290ef2b0112138 kernelconfig.x86 f82d261dc117ced42e2a3dcd4b6850a5 kernelconfig.x86_64" diff --git a/main/linux-vserver/patch-3.0.10-vs2.3.2.1-unofficial.diff b/main/linux-vserver/patch-3.0.10-vs2.3.2.1-unofficial.diff new file mode 100644 index 00000000000..2c09d43edfe --- /dev/null +++ b/main/linux-vserver/patch-3.0.10-vs2.3.2.1-unofficial.diff @@ -0,0 +1,26517 @@ +diff -NurpP --minimal linux-3.0.9/Documentation/vserver/debug.txt linux-3.0.9-vs2.3.2.1/Documentation/vserver/debug.txt +--- linux-3.0.9/Documentation/vserver/debug.txt 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/Documentation/vserver/debug.txt 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,154 @@ ++ ++debug_cvirt: ++ ++ 2 4 "vx_map_tgid: %p/%llx: %d -> %d" ++ "vx_rmap_tgid: %p/%llx: %d -> %d" ++ ++debug_dlim: ++ ++ 0 1 "ALLOC (%p,#%d)%c inode (%d)" ++ "FREE (%p,#%d)%c inode" ++ 1 2 "ALLOC (%p,#%d)%c %lld bytes (%d)" ++ "FREE (%p,#%d)%c %lld bytes" ++ 2 4 "ADJUST: %lld,%lld on %ld,%ld [mult=%d]" ++ 3 8 "ext3_has_free_blocks(%p): %lu<%lu+1, %c, %u!=%u r=%d" ++ "ext3_has_free_blocks(%p): free=%lu, root=%lu" ++ "rcu_free_dl_info(%p)" ++ 4 10 "alloc_dl_info(%p,%d) = %p" ++ "dealloc_dl_info(%p)" ++ "get_dl_info(%p[#%d.%d])" ++ "put_dl_info(%p[#%d.%d])" ++ 5 20 "alloc_dl_info(%p,%d)*" ++ 6 40 "__hash_dl_info: %p[#%d]" ++ "__unhash_dl_info: %p[#%d]" ++ 7 80 "locate_dl_info(%p,#%d) = %p" ++ ++debug_misc: ++ ++ 0 1 "destroy_dqhash: %p [#0x%08x] c=%d" ++ "new_dqhash: %p [#0x%08x]" ++ "vroot[%d]_clr_dev: dev=%p[%lu,%d:%d]" ++ "vroot[%d]_get_real_bdev: dev=%p[%lu,%d:%d]" ++ "vroot[%d]_set_dev: dev=%p[%lu,%d:%d]" ++ "vroot_get_real_bdev not set" ++ 1 2 "cow_break_link(»%s«)" ++ "temp copy »%s«" ++ 2 4 "dentry_open(new): %p" ++ "dentry_open(old): %p" ++ "lookup_create(new): %p" ++ "old path »%s«" ++ "path_lookup(old): %d" ++ "vfs_create(new): %d" ++ "vfs_rename: %d" ++ "vfs_sendfile: %d" ++ 3 8 "fput(new_file=%p[#%d])" ++ "fput(old_file=%p[#%d])" ++ 4 10 "vx_info_kill(%p[#%d],%d,%d) = %d" ++ "vx_info_kill(%p[#%d],%d,%d)*" ++ 5 20 "vs_reboot(%p[#%d],%d)" ++ 6 40 "dropping task %p[#%u,%u] for %p[#%u,%u]" ++ ++debug_net: ++ ++ 2 4 "nx_addr_conflict(%p,%p) %d.%d,%d.%d" ++ 3 8 "inet_bind(%p) %d.%d.%d.%d, %d.%d.%d.%d, %d.%d.%d.%d" ++ "inet_bind(%p)* %p,%p;%lx %d.%d.%d.%d" ++ 4 10 "ip_route_connect(%p) %p,%p;%lx" ++ 5 20 "__addr_in_socket(%p,%d.%d.%d.%d) %p:%d.%d.%d.%d %p;%lx" ++ 6 40 "sk,egf: %p [#%d] (from %d)" ++ "sk,egn: %p [#%d] (from %d)" ++ "sk,req: %p [#%d] (from %d)" ++ "sk: %p [#%d] (from %d)" ++ "tw: %p [#%d] (from %d)" ++ 7 80 "__sock_recvmsg: %p[%p,%p,%p;%d]:%d/%d" ++ "__sock_sendmsg: %p[%p,%p,%p;%d]:%d/%d" ++ ++debug_nid: ++ ++ 0 1 "__lookup_nx_info(#%u): %p[#%u]" ++ "alloc_nx_info(%d) = %p" ++ "create_nx_info(%d) (dynamic rejected)" ++ "create_nx_info(%d) = %p (already there)" ++ "create_nx_info(%d) = %p (new)" ++ "dealloc_nx_info(%p)" ++ 1 2 "alloc_nx_info(%d)*" ++ "create_nx_info(%d)*" ++ 2 4 "get_nx_info(%p[#%d.%d])" ++ "put_nx_info(%p[#%d.%d])" ++ 3 8 "claim_nx_info(%p[#%d.%d.%d]) %p" ++ "clr_nx_info(%p[#%d.%d])" ++ "init_nx_info(%p[#%d.%d])" ++ "release_nx_info(%p[#%d.%d.%d]) %p" ++ "set_nx_info(%p[#%d.%d])" ++ 4 10 "__hash_nx_info: %p[#%d]" ++ "__nx_dynamic_id: [#%d]" ++ "__unhash_nx_info: %p[#%d.%d.%d]" ++ 5 20 "moved task %p into nxi:%p[#%d]" ++ "nx_migrate_task(%p,%p[#%d.%d.%d])" ++ "task_get_nx_info(%p)" ++ 6 40 "nx_clear_persistent(%p[#%d])" ++ ++debug_quota: ++ ++ 0 1 "quota_sync_dqh(%p,%d) discard inode %p" ++ 1 2 "quota_sync_dqh(%p,%d)" ++ "sync_dquots(%p,%d)" ++ "sync_dquots_dqh(%p,%d)" ++ 3 8 "do_quotactl(%p,%d,cmd=%d,id=%d,%p)" ++ ++debug_switch: ++ ++ 0 1 "vc: VCMD_%02d_%d[%d], %d,%p [%d,%d,%x,%x]" ++ 1 2 "vc: VCMD_%02d_%d[%d] = %08lx(%ld) [%d,%d]" ++ 4 10 "%s: (%s %s) returned %s with %d" ++ ++debug_tag: ++ ++ 7 80 "dx_parse_tag(»%s«): %d:#%d" ++ "dx_propagate_tag(%p[#%lu.%d]): %d,%d" ++ ++debug_xid: ++ ++ 0 1 "__lookup_vx_info(#%u): %p[#%u]" ++ "alloc_vx_info(%d) = %p" ++ "alloc_vx_info(%d)*" ++ "create_vx_info(%d) (dynamic rejected)" ++ "create_vx_info(%d) = %p (already there)" ++ "create_vx_info(%d) = %p (new)" ++ "dealloc_vx_info(%p)" ++ "loc_vx_info(%d) = %p (found)" ++ "loc_vx_info(%d) = %p (new)" ++ "loc_vx_info(%d) = %p (not available)" ++ 1 2 "create_vx_info(%d)*" ++ "loc_vx_info(%d)*" ++ 2 4 "get_vx_info(%p[#%d.%d])" ++ "put_vx_info(%p[#%d.%d])" ++ 3 8 "claim_vx_info(%p[#%d.%d.%d]) %p" ++ "clr_vx_info(%p[#%d.%d])" ++ "init_vx_info(%p[#%d.%d])" ++ "release_vx_info(%p[#%d.%d.%d]) %p" ++ "set_vx_info(%p[#%d.%d])" ++ 4 10 "__hash_vx_info: %p[#%d]" ++ "__unhash_vx_info: %p[#%d.%d.%d]" ++ "__vx_dynamic_id: [#%d]" ++ 5 20 "enter_vx_info(%p[#%d],%p) %p[#%d,%p]" ++ "leave_vx_info(%p[#%d,%p]) %p[#%d,%p]" ++ "moved task %p into vxi:%p[#%d]" ++ "task_get_vx_info(%p)" ++ "vx_migrate_task(%p,%p[#%d.%d])" ++ 6 40 "vx_clear_persistent(%p[#%d])" ++ "vx_exit_init(%p[#%d],%p[#%d,%d,%d])" ++ "vx_set_init(%p[#%d],%p[#%d,%d,%d])" ++ "vx_set_persistent(%p[#%d])" ++ "vx_set_reaper(%p[#%d],%p[#%d,%d])" ++ 7 80 "vx_child_reaper(%p[#%u,%u]) = %p[#%u,%u]" ++ ++ ++debug_limit: ++ ++ n 2^n "vx_acc_cres[%5d,%s,%2d]: %5d%s" ++ "vx_cres_avail[%5d,%s,%2d]: %5ld > %5d + %5d" ++ ++ m 2^m "vx_acc_page[%5d,%s,%2d]: %5d%s" ++ "vx_acc_pages[%5d,%s,%2d]: %5d += %5d" ++ "vx_pages_avail[%5d,%s,%2d]: %5ld > %5d + %5d" +diff -NurpP --minimal linux-3.0.9/Makefile linux-3.0.9-vs2.3.2.1/Makefile +--- linux-3.0.9/Makefile 2011-11-15 16:40:41.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/Makefile 2011-11-17 23:18:51.000000000 +0100 +@@ -1,7 +1,7 @@ + VERSION = 3 + PATCHLEVEL = 0 + SUBLEVEL = 10 +-EXTRAVERSION = ++EXTRAVERSION = -vs2.3.2.1 + NAME = Sneaky Weasel + + # *DOCUMENTATION* +diff -NurpP --minimal linux-3.0.9/arch/alpha/Kconfig linux-3.0.9-vs2.3.2.1/arch/alpha/Kconfig +--- linux-3.0.9/arch/alpha/Kconfig 2011-07-22 11:17:32.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/alpha/Kconfig 2011-06-10 22:11:24.000000000 +0200 +@@ -668,6 +668,8 @@ config DUMMY_CONSOLE + depends on VGA_HOSE + default y + ++source "kernel/vserver/Kconfig" ++ + source "security/Kconfig" + + source "crypto/Kconfig" +diff -NurpP --minimal linux-3.0.9/arch/alpha/kernel/entry.S linux-3.0.9-vs2.3.2.1/arch/alpha/kernel/entry.S +--- linux-3.0.9/arch/alpha/kernel/entry.S 2010-10-21 13:06:45.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/alpha/kernel/entry.S 2011-06-10 22:11:24.000000000 +0200 +@@ -860,24 +860,15 @@ sys_getxgid: + .globl sys_getxpid + .ent sys_getxpid + sys_getxpid: ++ lda $sp, -16($sp) ++ stq $26, 0($sp) + .prologue 0 +- ldq $2, TI_TASK($8) + +- /* See linux/kernel/timer.c sys_getppid for discussion +- about this loop. */ +- ldq $3, TASK_GROUP_LEADER($2) +- ldq $4, TASK_REAL_PARENT($3) +- ldl $0, TASK_TGID($2) +-1: ldl $1, TASK_TGID($4) +-#ifdef CONFIG_SMP +- mov $4, $5 +- mb +- ldq $3, TASK_GROUP_LEADER($2) +- ldq $4, TASK_REAL_PARENT($3) +- cmpeq $4, $5, $5 +- beq $5, 1b +-#endif +- stq $1, 80($sp) ++ lda $16, 96($sp) ++ jsr $26, do_getxpid ++ ldq $26, 0($sp) ++ ++ lda $sp, 16($sp) + ret + .end sys_getxpid + +diff -NurpP --minimal linux-3.0.9/arch/alpha/kernel/ptrace.c linux-3.0.9-vs2.3.2.1/arch/alpha/kernel/ptrace.c +--- linux-3.0.9/arch/alpha/kernel/ptrace.c 2011-01-05 21:48:40.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/arch/alpha/kernel/ptrace.c 2011-06-10 22:11:24.000000000 +0200 +@@ -13,6 +13,7 @@ + #include <linux/user.h> + #include <linux/security.h> + #include <linux/signal.h> ++#include <linux/vs_base.h> + + #include <asm/uaccess.h> + #include <asm/pgtable.h> +diff -NurpP --minimal linux-3.0.9/arch/alpha/kernel/systbls.S linux-3.0.9-vs2.3.2.1/arch/alpha/kernel/systbls.S +--- linux-3.0.9/arch/alpha/kernel/systbls.S 2011-07-22 11:17:32.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/alpha/kernel/systbls.S 2011-06-10 22:11:24.000000000 +0200 +@@ -446,7 +446,7 @@ sys_call_table: + .quad sys_stat64 /* 425 */ + .quad sys_lstat64 + .quad sys_fstat64 +- .quad sys_ni_syscall /* sys_vserver */ ++ .quad sys_vserver /* sys_vserver */ + .quad sys_ni_syscall /* sys_mbind */ + .quad sys_ni_syscall /* sys_get_mempolicy */ + .quad sys_ni_syscall /* sys_set_mempolicy */ +diff -NurpP --minimal linux-3.0.9/arch/alpha/kernel/traps.c linux-3.0.9-vs2.3.2.1/arch/alpha/kernel/traps.c +--- linux-3.0.9/arch/alpha/kernel/traps.c 2010-10-21 13:06:46.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/alpha/kernel/traps.c 2011-06-10 22:11:24.000000000 +0200 +@@ -183,7 +183,8 @@ die_if_kernel(char * str, struct pt_regs + #ifdef CONFIG_SMP + printk("CPU %d ", hard_smp_processor_id()); + #endif +- printk("%s(%d): %s %ld\n", current->comm, task_pid_nr(current), str, err); ++ printk("%s(%d[#%u]): %s %ld\n", current->comm, ++ task_pid_nr(current), current->xid, str, err); + dik_show_regs(regs, r9_15); + add_taint(TAINT_DIE); + dik_show_trace((unsigned long *)(regs+1)); +diff -NurpP --minimal linux-3.0.9/arch/arm/Kconfig linux-3.0.9-vs2.3.2.1/arch/arm/Kconfig +--- linux-3.0.9/arch/arm/Kconfig 2011-11-15 16:40:41.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/arch/arm/Kconfig 2011-10-18 13:51:13.000000000 +0200 +@@ -2049,6 +2049,8 @@ source "fs/Kconfig" + + source "arch/arm/Kconfig.debug" + ++source "kernel/vserver/Kconfig" ++ + source "security/Kconfig" + + source "crypto/Kconfig" +diff -NurpP --minimal linux-3.0.9/arch/arm/kernel/calls.S linux-3.0.9-vs2.3.2.1/arch/arm/kernel/calls.S +--- linux-3.0.9/arch/arm/kernel/calls.S 2011-07-22 11:17:32.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/arm/kernel/calls.S 2011-06-10 22:11:24.000000000 +0200 +@@ -322,7 +322,7 @@ + /* 310 */ CALL(sys_request_key) + CALL(sys_keyctl) + CALL(ABI(sys_semtimedop, sys_oabi_semtimedop)) +-/* vserver */ CALL(sys_ni_syscall) ++ CALL(sys_vserver) + CALL(sys_ioprio_set) + /* 315 */ CALL(sys_ioprio_get) + CALL(sys_inotify_init) +diff -NurpP --minimal linux-3.0.9/arch/arm/kernel/process.c linux-3.0.9-vs2.3.2.1/arch/arm/kernel/process.c +--- linux-3.0.9/arch/arm/kernel/process.c 2011-05-22 16:16:47.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/arm/kernel/process.c 2011-06-10 22:11:24.000000000 +0200 +@@ -315,7 +315,8 @@ void __show_regs(struct pt_regs *regs) + void show_regs(struct pt_regs * regs) + { + printk("\n"); +- printk("Pid: %d, comm: %20s\n", task_pid_nr(current), current->comm); ++ printk("Pid: %d[#%u], comm: %20s\n", ++ task_pid_nr(current), current->xid, current->comm); + __show_regs(regs); + __backtrace(); + } +diff -NurpP --minimal linux-3.0.9/arch/arm/kernel/traps.c linux-3.0.9-vs2.3.2.1/arch/arm/kernel/traps.c +--- linux-3.0.9/arch/arm/kernel/traps.c 2011-07-22 11:17:32.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/arm/kernel/traps.c 2011-06-22 12:39:12.000000000 +0200 +@@ -242,8 +242,8 @@ static int __die(const char *str, int er + + print_modules(); + __show_regs(regs); +- printk(KERN_EMERG "Process %.*s (pid: %d, stack limit = 0x%p)\n", +- TASK_COMM_LEN, tsk->comm, task_pid_nr(tsk), thread + 1); ++ printk(KERN_EMERG "Process %.*s (pid: %d:#%u, stack limit = 0x%p)\n", ++ TASK_COMM_LEN, tsk->comm, task_pid_nr(tsk), tsk->xid, thread + 1); + + if (!user_mode(regs) || in_interrupt()) { + dump_mem(KERN_EMERG, "Stack: ", regs->ARM_sp, +diff -NurpP --minimal linux-3.0.9/arch/cris/Kconfig linux-3.0.9-vs2.3.2.1/arch/cris/Kconfig +--- linux-3.0.9/arch/cris/Kconfig 2011-07-22 11:17:35.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/cris/Kconfig 2011-06-10 22:11:24.000000000 +0200 +@@ -678,6 +678,8 @@ source "drivers/staging/Kconfig" + + source "arch/cris/Kconfig.debug" + ++source "kernel/vserver/Kconfig" ++ + source "security/Kconfig" + + source "crypto/Kconfig" +diff -NurpP --minimal linux-3.0.9/arch/frv/kernel/kernel_thread.S linux-3.0.9-vs2.3.2.1/arch/frv/kernel/kernel_thread.S +--- linux-3.0.9/arch/frv/kernel/kernel_thread.S 2008-12-25 00:26:37.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/arch/frv/kernel/kernel_thread.S 2011-06-10 22:11:24.000000000 +0200 +@@ -37,7 +37,7 @@ kernel_thread: + + # start by forking the current process, but with shared VM + setlos.p #__NR_clone,gr7 ; syscall number +- ori gr10,#CLONE_VM,gr8 ; first syscall arg [clone_flags] ++ ori gr10,#CLONE_KT,gr8 ; first syscall arg [clone_flags] + sethi.p #0xe4e4,gr9 ; second syscall arg [newsp] + setlo #0xe4e4,gr9 + setlos.p #0,gr10 ; third syscall arg [parent_tidptr] +diff -NurpP --minimal linux-3.0.9/arch/h8300/Kconfig linux-3.0.9-vs2.3.2.1/arch/h8300/Kconfig +--- linux-3.0.9/arch/h8300/Kconfig 2011-07-22 11:17:35.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/h8300/Kconfig 2011-06-10 22:11:24.000000000 +0200 +@@ -213,6 +213,8 @@ source "fs/Kconfig" + + source "arch/h8300/Kconfig.debug" + ++source "kernel/vserver/Kconfig" ++ + source "security/Kconfig" + + source "crypto/Kconfig" +diff -NurpP --minimal linux-3.0.9/arch/ia64/Kconfig linux-3.0.9-vs2.3.2.1/arch/ia64/Kconfig +--- linux-3.0.9/arch/ia64/Kconfig 2011-07-22 11:17:35.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/ia64/Kconfig 2011-06-10 22:11:24.000000000 +0200 +@@ -671,6 +671,8 @@ source "fs/Kconfig" + + source "arch/ia64/Kconfig.debug" + ++source "kernel/vserver/Kconfig" ++ + source "security/Kconfig" + + source "crypto/Kconfig" +diff -NurpP --minimal linux-3.0.9/arch/ia64/include/asm/tlb.h linux-3.0.9-vs2.3.2.1/arch/ia64/include/asm/tlb.h +--- linux-3.0.9/arch/ia64/include/asm/tlb.h 2011-07-22 11:17:35.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/ia64/include/asm/tlb.h 2011-06-10 22:11:24.000000000 +0200 +@@ -40,6 +40,7 @@ + #include <linux/mm.h> + #include <linux/pagemap.h> + #include <linux/swap.h> ++#include <linux/vs_memory.h> + + #include <asm/pgalloc.h> + #include <asm/processor.h> +diff -NurpP --minimal linux-3.0.9/arch/ia64/kernel/entry.S linux-3.0.9-vs2.3.2.1/arch/ia64/kernel/entry.S +--- linux-3.0.9/arch/ia64/kernel/entry.S 2011-07-22 11:17:35.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/ia64/kernel/entry.S 2011-06-10 22:11:24.000000000 +0200 +@@ -1714,7 +1714,7 @@ sys_call_table: + data8 sys_mq_notify + data8 sys_mq_getsetattr + data8 sys_kexec_load +- data8 sys_ni_syscall // reserved for vserver ++ data8 sys_vserver + data8 sys_waitid // 1270 + data8 sys_add_key + data8 sys_request_key +diff -NurpP --minimal linux-3.0.9/arch/ia64/kernel/perfmon.c linux-3.0.9-vs2.3.2.1/arch/ia64/kernel/perfmon.c +--- linux-3.0.9/arch/ia64/kernel/perfmon.c 2011-03-15 18:06:39.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/arch/ia64/kernel/perfmon.c 2011-06-10 22:11:24.000000000 +0200 +@@ -42,6 +42,7 @@ + #include <linux/completion.h> + #include <linux/tracehook.h> + #include <linux/slab.h> ++#include <linux/vs_memory.h> + + #include <asm/errno.h> + #include <asm/intrinsics.h> +diff -NurpP --minimal linux-3.0.9/arch/ia64/kernel/process.c linux-3.0.9-vs2.3.2.1/arch/ia64/kernel/process.c +--- linux-3.0.9/arch/ia64/kernel/process.c 2011-03-15 18:06:39.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/arch/ia64/kernel/process.c 2011-06-10 22:11:24.000000000 +0200 +@@ -109,8 +109,8 @@ show_regs (struct pt_regs *regs) + unsigned long ip = regs->cr_iip + ia64_psr(regs)->ri; + + print_modules(); +- printk("\nPid: %d, CPU %d, comm: %20s\n", task_pid_nr(current), +- smp_processor_id(), current->comm); ++ printk("\nPid: %d[#%u], CPU %d, comm: %20s\n", task_pid_nr(current), ++ current->xid, smp_processor_id(), current->comm); + printk("psr : %016lx ifs : %016lx ip : [<%016lx>] %s (%s)\n", + regs->cr_ipsr, regs->cr_ifs, ip, print_tainted(), + init_utsname()->release); +diff -NurpP --minimal linux-3.0.9/arch/ia64/kernel/ptrace.c linux-3.0.9-vs2.3.2.1/arch/ia64/kernel/ptrace.c +--- linux-3.0.9/arch/ia64/kernel/ptrace.c 2011-01-05 21:48:59.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/arch/ia64/kernel/ptrace.c 2011-06-10 22:11:24.000000000 +0200 +@@ -21,6 +21,7 @@ + #include <linux/regset.h> + #include <linux/elf.h> + #include <linux/tracehook.h> ++#include <linux/vs_base.h> + + #include <asm/pgtable.h> + #include <asm/processor.h> +diff -NurpP --minimal linux-3.0.9/arch/ia64/kernel/traps.c linux-3.0.9-vs2.3.2.1/arch/ia64/kernel/traps.c +--- linux-3.0.9/arch/ia64/kernel/traps.c 2010-07-07 18:31:01.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/ia64/kernel/traps.c 2011-06-10 22:11:24.000000000 +0200 +@@ -59,8 +59,9 @@ die (const char *str, struct pt_regs *re + put_cpu(); + + if (++die.lock_owner_depth < 3) { +- printk("%s[%d]: %s %ld [%d]\n", +- current->comm, task_pid_nr(current), str, err, ++die_counter); ++ printk("%s[%d[#%u]]: %s %ld [%d]\n", ++ current->comm, task_pid_nr(current), current->xid, ++ str, err, ++die_counter); + if (notify_die(DIE_OOPS, str, regs, err, 255, SIGSEGV) + != NOTIFY_STOP) + show_regs(regs); +@@ -323,8 +324,9 @@ handle_fpu_swa (int fp_fault, struct pt_ + if ((last.count & 15) < 5 && (ia64_fetchadd(1, &last.count, acq) & 15) < 5) { + last.time = current_jiffies + 5 * HZ; + printk(KERN_WARNING +- "%s(%d): floating-point assist fault at ip %016lx, isr %016lx\n", +- current->comm, task_pid_nr(current), regs->cr_iip + ia64_psr(regs)->ri, isr); ++ "%s(%d[#%u]): floating-point assist fault at ip %016lx, isr %016lx\n", ++ current->comm, task_pid_nr(current), current->xid, ++ regs->cr_iip + ia64_psr(regs)->ri, isr); + } + } + } +diff -NurpP --minimal linux-3.0.9/arch/ia64/mm/fault.c linux-3.0.9-vs2.3.2.1/arch/ia64/mm/fault.c +--- linux-3.0.9/arch/ia64/mm/fault.c 2011-07-22 11:17:35.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/ia64/mm/fault.c 2011-06-10 22:28:23.000000000 +0200 +@@ -11,6 +11,7 @@ + #include <linux/kprobes.h> + #include <linux/kdebug.h> + #include <linux/prefetch.h> ++#include <linux/vs_memory.h> + + #include <asm/pgtable.h> + #include <asm/processor.h> +diff -NurpP --minimal linux-3.0.9/arch/m32r/kernel/traps.c linux-3.0.9-vs2.3.2.1/arch/m32r/kernel/traps.c +--- linux-3.0.9/arch/m32r/kernel/traps.c 2009-12-03 20:01:57.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/arch/m32r/kernel/traps.c 2011-06-10 22:11:24.000000000 +0200 +@@ -196,8 +196,9 @@ static void show_registers(struct pt_reg + } else { + printk("SPI: %08lx\n", sp); + } +- printk("Process %s (pid: %d, process nr: %d, stackpage=%08lx)", +- current->comm, task_pid_nr(current), 0xffff & i, 4096+(unsigned long)current); ++ printk("Process %s (pid: %d[#%u], process nr: %d, stackpage=%08lx)", ++ current->comm, task_pid_nr(current), current->xid, ++ 0xffff & i, 4096+(unsigned long)current); + + /* + * When in-kernel, we also print out the stack and code at the +diff -NurpP --minimal linux-3.0.9/arch/m68k/Kconfig linux-3.0.9-vs2.3.2.1/arch/m68k/Kconfig +--- linux-3.0.9/arch/m68k/Kconfig 2011-07-22 11:17:35.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/m68k/Kconfig 2011-06-10 22:11:24.000000000 +0200 +@@ -241,6 +241,8 @@ source "fs/Kconfig" + + source "arch/m68k/Kconfig.debug" + ++source "kernel/vserver/Kconfig" ++ + source "security/Kconfig" + + source "crypto/Kconfig" +diff -NurpP --minimal linux-3.0.9/arch/mips/Kconfig linux-3.0.9-vs2.3.2.1/arch/mips/Kconfig +--- linux-3.0.9/arch/mips/Kconfig 2011-07-22 11:17:35.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/mips/Kconfig 2011-06-10 22:11:24.000000000 +0200 +@@ -2485,6 +2485,8 @@ source "fs/Kconfig" + + source "arch/mips/Kconfig.debug" + ++source "kernel/vserver/Kconfig" ++ + source "security/Kconfig" + + source "crypto/Kconfig" +diff -NurpP --minimal linux-3.0.9/arch/mips/kernel/ptrace.c linux-3.0.9-vs2.3.2.1/arch/mips/kernel/ptrace.c +--- linux-3.0.9/arch/mips/kernel/ptrace.c 2011-07-22 11:17:36.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/mips/kernel/ptrace.c 2011-06-10 22:11:24.000000000 +0200 +@@ -25,6 +25,7 @@ + #include <linux/security.h> + #include <linux/audit.h> + #include <linux/seccomp.h> ++#include <linux/vs_base.h> + + #include <asm/byteorder.h> + #include <asm/cpu.h> +@@ -263,6 +264,9 @@ long arch_ptrace(struct task_struct *chi + void __user *datavp = (void __user *) data; + unsigned long __user *datalp = (void __user *) data; + ++ if (!vx_check(vx_task_xid(child), VS_WATCH_P | VS_IDENT)) ++ goto out; ++ + switch (request) { + /* when I and D space are separate, these will need to be fixed. */ + case PTRACE_PEEKTEXT: /* read word at location addr. */ +diff -NurpP --minimal linux-3.0.9/arch/mips/kernel/scall32-o32.S linux-3.0.9-vs2.3.2.1/arch/mips/kernel/scall32-o32.S +--- linux-3.0.9/arch/mips/kernel/scall32-o32.S 2011-07-22 11:17:36.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/mips/kernel/scall32-o32.S 2011-06-10 22:11:24.000000000 +0200 +@@ -523,7 +523,7 @@ einval: li v0, -ENOSYS + sys sys_mq_timedreceive 5 + sys sys_mq_notify 2 /* 4275 */ + sys sys_mq_getsetattr 3 +- sys sys_ni_syscall 0 /* sys_vserver */ ++ sys sys_vserver 3 + sys sys_waitid 5 + sys sys_ni_syscall 0 /* available, was setaltroot */ + sys sys_add_key 5 /* 4280 */ +diff -NurpP --minimal linux-3.0.9/arch/mips/kernel/scall64-64.S linux-3.0.9-vs2.3.2.1/arch/mips/kernel/scall64-64.S +--- linux-3.0.9/arch/mips/kernel/scall64-64.S 2011-07-22 11:17:36.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/mips/kernel/scall64-64.S 2011-06-10 22:11:24.000000000 +0200 +@@ -362,7 +362,7 @@ sys_call_table: + PTR sys_mq_timedreceive + PTR sys_mq_notify + PTR sys_mq_getsetattr /* 5235 */ +- PTR sys_ni_syscall /* sys_vserver */ ++ PTR sys_vserver + PTR sys_waitid + PTR sys_ni_syscall /* available, was setaltroot */ + PTR sys_add_key +diff -NurpP --minimal linux-3.0.9/arch/mips/kernel/scall64-n32.S linux-3.0.9-vs2.3.2.1/arch/mips/kernel/scall64-n32.S +--- linux-3.0.9/arch/mips/kernel/scall64-n32.S 2011-07-22 11:17:36.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/mips/kernel/scall64-n32.S 2011-06-10 22:11:24.000000000 +0200 +@@ -361,7 +361,7 @@ EXPORT(sysn32_call_table) + PTR compat_sys_mq_timedreceive + PTR compat_sys_mq_notify + PTR compat_sys_mq_getsetattr +- PTR sys_ni_syscall /* 6240, sys_vserver */ ++ PTR sys32_vserver /* 6240 */ + PTR compat_sys_waitid + PTR sys_ni_syscall /* available, was setaltroot */ + PTR sys_add_key +diff -NurpP --minimal linux-3.0.9/arch/mips/kernel/scall64-o32.S linux-3.0.9-vs2.3.2.1/arch/mips/kernel/scall64-o32.S +--- linux-3.0.9/arch/mips/kernel/scall64-o32.S 2011-07-22 11:17:36.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/mips/kernel/scall64-o32.S 2011-06-10 22:11:24.000000000 +0200 +@@ -480,7 +480,7 @@ sys_call_table: + PTR compat_sys_mq_timedreceive + PTR compat_sys_mq_notify /* 4275 */ + PTR compat_sys_mq_getsetattr +- PTR sys_ni_syscall /* sys_vserver */ ++ PTR sys32_vserver + PTR sys_32_waitid + PTR sys_ni_syscall /* available, was setaltroot */ + PTR sys_add_key /* 4280 */ +diff -NurpP --minimal linux-3.0.9/arch/mips/kernel/traps.c linux-3.0.9-vs2.3.2.1/arch/mips/kernel/traps.c +--- linux-3.0.9/arch/mips/kernel/traps.c 2011-05-22 16:17:00.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/mips/kernel/traps.c 2011-06-10 22:11:24.000000000 +0200 +@@ -343,9 +343,10 @@ void show_registers(struct pt_regs *regs + + __show_regs(regs); + print_modules(); +- printk("Process %s (pid: %d, threadinfo=%p, task=%p, tls=%0*lx)\n", +- current->comm, current->pid, current_thread_info(), current, +- field, current_thread_info()->tp_value); ++ printk("Process %s (pid: %d:#%u, threadinfo=%p, task=%p, tls=%0*lx)\n", ++ current->comm, task_pid_nr(current), current->xid, ++ current_thread_info(), current, ++ field, current_thread_info()->tp_value); + if (cpu_has_userlocal) { + unsigned long tls; + +diff -NurpP --minimal linux-3.0.9/arch/parisc/Kconfig linux-3.0.9-vs2.3.2.1/arch/parisc/Kconfig +--- linux-3.0.9/arch/parisc/Kconfig 2011-07-22 11:17:36.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/parisc/Kconfig 2011-06-10 22:11:24.000000000 +0200 +@@ -279,6 +279,8 @@ source "fs/Kconfig" + + source "arch/parisc/Kconfig.debug" + ++source "kernel/vserver/Kconfig" ++ + source "security/Kconfig" + + source "crypto/Kconfig" +diff -NurpP --minimal linux-3.0.9/arch/parisc/kernel/syscall_table.S linux-3.0.9-vs2.3.2.1/arch/parisc/kernel/syscall_table.S +--- linux-3.0.9/arch/parisc/kernel/syscall_table.S 2011-11-15 16:40:42.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/arch/parisc/kernel/syscall_table.S 2011-08-29 03:45:07.000000000 +0200 +@@ -361,7 +361,7 @@ + ENTRY_COMP(mbind) /* 260 */ + ENTRY_COMP(get_mempolicy) + ENTRY_COMP(set_mempolicy) +- ENTRY_SAME(ni_syscall) /* 263: reserved for vserver */ ++ ENTRY_DIFF(vserver) + ENTRY_SAME(add_key) + ENTRY_SAME(request_key) /* 265 */ + ENTRY_SAME(keyctl) +diff -NurpP --minimal linux-3.0.9/arch/parisc/kernel/traps.c linux-3.0.9-vs2.3.2.1/arch/parisc/kernel/traps.c +--- linux-3.0.9/arch/parisc/kernel/traps.c 2009-09-10 15:25:40.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/parisc/kernel/traps.c 2011-06-10 22:11:24.000000000 +0200 +@@ -236,8 +236,9 @@ void die_if_kernel(char *str, struct pt_ + if (err == 0) + return; /* STFU */ + +- printk(KERN_CRIT "%s (pid %d): %s (code %ld) at " RFMT "\n", +- current->comm, task_pid_nr(current), str, err, regs->iaoq[0]); ++ printk(KERN_CRIT "%s (pid %d:#%u): %s (code %ld) at " RFMT "\n", ++ current->comm, task_pid_nr(current), current->xid, ++ str, err, regs->iaoq[0]); + #ifdef PRINT_USER_FAULTS + /* XXX for debugging only */ + show_regs(regs); +@@ -270,8 +271,8 @@ void die_if_kernel(char *str, struct pt_ + pdc_console_restart(); + + if (err) +- printk(KERN_CRIT "%s (pid %d): %s (code %ld)\n", +- current->comm, task_pid_nr(current), str, err); ++ printk(KERN_CRIT "%s (pid %d:#%u): %s (code %ld)\n", ++ current->comm, task_pid_nr(current), current->xid, str, err); + + /* Wot's wrong wif bein' racy? */ + if (current->thread.flags & PARISC_KERNEL_DEATH) { +diff -NurpP --minimal linux-3.0.9/arch/parisc/mm/fault.c linux-3.0.9-vs2.3.2.1/arch/parisc/mm/fault.c +--- linux-3.0.9/arch/parisc/mm/fault.c 2010-08-02 16:52:06.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/parisc/mm/fault.c 2011-06-10 22:11:24.000000000 +0200 +@@ -237,8 +237,9 @@ bad_area: + + #ifdef PRINT_USER_FAULTS + printk(KERN_DEBUG "\n"); +- printk(KERN_DEBUG "do_page_fault() pid=%d command='%s' type=%lu address=0x%08lx\n", +- task_pid_nr(tsk), tsk->comm, code, address); ++ printk(KERN_DEBUG "do_page_fault() pid=%d:#%u " ++ "command='%s' type=%lu address=0x%08lx\n", ++ task_pid_nr(tsk), tsk->xid, tsk->comm, code, address); + if (vma) { + printk(KERN_DEBUG "vm_start = 0x%08lx, vm_end = 0x%08lx\n", + vma->vm_start, vma->vm_end); +diff -NurpP --minimal linux-3.0.9/arch/powerpc/Kconfig linux-3.0.9-vs2.3.2.1/arch/powerpc/Kconfig +--- linux-3.0.9/arch/powerpc/Kconfig 2011-07-22 11:17:36.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/powerpc/Kconfig 2011-06-10 22:11:24.000000000 +0200 +@@ -978,6 +978,8 @@ source "lib/Kconfig" + + source "arch/powerpc/Kconfig.debug" + ++source "kernel/vserver/Kconfig" ++ + source "security/Kconfig" + + config KEYS_COMPAT +diff -NurpP --minimal linux-3.0.9/arch/powerpc/include/asm/unistd.h linux-3.0.9-vs2.3.2.1/arch/powerpc/include/asm/unistd.h +--- linux-3.0.9/arch/powerpc/include/asm/unistd.h 2011-07-22 11:17:40.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/powerpc/include/asm/unistd.h 2011-06-10 22:11:24.000000000 +0200 +@@ -275,7 +275,7 @@ + #endif + #define __NR_rtas 255 + #define __NR_sys_debug_setcontext 256 +-/* Number 257 is reserved for vserver */ ++#define __NR_vserver 257 + #define __NR_migrate_pages 258 + #define __NR_mbind 259 + #define __NR_get_mempolicy 260 +diff -NurpP --minimal linux-3.0.9/arch/powerpc/kernel/process.c linux-3.0.9-vs2.3.2.1/arch/powerpc/kernel/process.c +--- linux-3.0.9/arch/powerpc/kernel/process.c 2011-07-22 11:17:40.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/powerpc/kernel/process.c 2011-06-10 22:11:24.000000000 +0200 +@@ -656,8 +656,9 @@ void show_regs(struct pt_regs * regs) + #else + printk("DAR: "REG", DSISR: %08lx\n", regs->dar, regs->dsisr); + #endif +- printk("TASK = %p[%d] '%s' THREAD: %p", +- current, task_pid_nr(current), current->comm, task_thread_info(current)); ++ printk("TASK = %p[%d,#%u] '%s' THREAD: %p", ++ current, task_pid_nr(current), current->xid, ++ current->comm, task_thread_info(current)); + + #ifdef CONFIG_SMP + printk(" CPU: %d", raw_smp_processor_id()); +diff -NurpP --minimal linux-3.0.9/arch/powerpc/kernel/traps.c linux-3.0.9-vs2.3.2.1/arch/powerpc/kernel/traps.c +--- linux-3.0.9/arch/powerpc/kernel/traps.c 2011-07-22 11:17:40.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/powerpc/kernel/traps.c 2011-07-19 00:44:39.000000000 +0200 +@@ -1075,8 +1075,9 @@ void nonrecoverable_exception(struct pt_ + + void trace_syscall(struct pt_regs *regs) + { +- printk("Task: %p(%d), PC: %08lX/%08lX, Syscall: %3ld, Result: %s%ld %s\n", +- current, task_pid_nr(current), regs->nip, regs->link, regs->gpr[0], ++ printk("Task: %p(%d[#%u]), PC: %08lX/%08lX, Syscall: %3ld, Result: %s%ld %s\n", ++ current, task_pid_nr(current), current->xid, ++ regs->nip, regs->link, regs->gpr[0], + regs->ccr&0x10000000?"Error=":"", regs->gpr[3], print_tainted()); + } + +diff -NurpP --minimal linux-3.0.9/arch/powerpc/kernel/vdso.c linux-3.0.9-vs2.3.2.1/arch/powerpc/kernel/vdso.c +--- linux-3.0.9/arch/powerpc/kernel/vdso.c 2011-05-22 16:17:02.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/powerpc/kernel/vdso.c 2011-06-10 22:11:24.000000000 +0200 +@@ -23,6 +23,7 @@ + #include <linux/security.h> + #include <linux/bootmem.h> + #include <linux/memblock.h> ++#include <linux/vs_memory.h> + + #include <asm/pgtable.h> + #include <asm/system.h> +diff -NurpP --minimal linux-3.0.9/arch/s390/Kconfig linux-3.0.9-vs2.3.2.1/arch/s390/Kconfig +--- linux-3.0.9/arch/s390/Kconfig 2011-07-22 11:17:41.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/s390/Kconfig 2011-07-01 11:35:34.000000000 +0200 +@@ -628,6 +628,8 @@ source "fs/Kconfig" + + source "arch/s390/Kconfig.debug" + ++source "kernel/vserver/Kconfig" ++ + source "security/Kconfig" + + source "crypto/Kconfig" +diff -NurpP --minimal linux-3.0.9/arch/s390/include/asm/tlb.h linux-3.0.9-vs2.3.2.1/arch/s390/include/asm/tlb.h +--- linux-3.0.9/arch/s390/include/asm/tlb.h 2011-07-22 11:17:41.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/s390/include/asm/tlb.h 2011-06-15 02:40:14.000000000 +0200 +@@ -24,6 +24,8 @@ + #include <linux/mm.h> + #include <linux/pagemap.h> + #include <linux/swap.h> ++#include <linux/vs_memory.h> ++ + #include <asm/processor.h> + #include <asm/pgalloc.h> + #include <asm/tlbflush.h> +diff -NurpP --minimal linux-3.0.9/arch/s390/include/asm/unistd.h linux-3.0.9-vs2.3.2.1/arch/s390/include/asm/unistd.h +--- linux-3.0.9/arch/s390/include/asm/unistd.h 2011-07-22 11:17:41.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/s390/include/asm/unistd.h 2011-06-10 22:11:24.000000000 +0200 +@@ -202,7 +202,7 @@ + #define __NR_clock_gettime (__NR_timer_create+6) + #define __NR_clock_getres (__NR_timer_create+7) + #define __NR_clock_nanosleep (__NR_timer_create+8) +-/* Number 263 is reserved for vserver */ ++#define __NR_vserver 263 + #define __NR_statfs64 265 + #define __NR_fstatfs64 266 + #define __NR_remap_file_pages 267 +diff -NurpP --minimal linux-3.0.9/arch/s390/kernel/ptrace.c linux-3.0.9-vs2.3.2.1/arch/s390/kernel/ptrace.c +--- linux-3.0.9/arch/s390/kernel/ptrace.c 2011-11-15 16:40:42.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/arch/s390/kernel/ptrace.c 2011-11-15 17:37:04.000000000 +0100 +@@ -20,6 +20,7 @@ + #include <linux/regset.h> + #include <linux/tracehook.h> + #include <linux/seccomp.h> ++#include <linux/vs_base.h> + #include <trace/syscall.h> + #include <asm/compat.h> + #include <asm/segment.h> +diff -NurpP --minimal linux-3.0.9/arch/s390/kernel/syscalls.S linux-3.0.9-vs2.3.2.1/arch/s390/kernel/syscalls.S +--- linux-3.0.9/arch/s390/kernel/syscalls.S 2011-07-22 11:17:41.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/s390/kernel/syscalls.S 2011-06-10 22:11:24.000000000 +0200 +@@ -271,7 +271,7 @@ SYSCALL(sys_clock_settime,sys_clock_sett + SYSCALL(sys_clock_gettime,sys_clock_gettime,sys32_clock_gettime_wrapper) /* 260 */ + SYSCALL(sys_clock_getres,sys_clock_getres,sys32_clock_getres_wrapper) + SYSCALL(sys_clock_nanosleep,sys_clock_nanosleep,sys32_clock_nanosleep_wrapper) +-NI_SYSCALL /* reserved for vserver */ ++SYSCALL(sys_vserver,sys_vserver,sys32_vserver) + SYSCALL(sys_s390_fadvise64_64,sys_ni_syscall,sys32_fadvise64_64_wrapper) + SYSCALL(sys_statfs64,sys_statfs64,compat_sys_statfs64_wrapper) + SYSCALL(sys_fstatfs64,sys_fstatfs64,compat_sys_fstatfs64_wrapper) +diff -NurpP --minimal linux-3.0.9/arch/sh/Kconfig linux-3.0.9-vs2.3.2.1/arch/sh/Kconfig +--- linux-3.0.9/arch/sh/Kconfig 2011-07-22 11:17:41.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/sh/Kconfig 2011-07-19 00:44:39.000000000 +0200 +@@ -893,6 +893,8 @@ source "fs/Kconfig" + + source "arch/sh/Kconfig.debug" + ++source "kernel/vserver/Kconfig" ++ + source "security/Kconfig" + + source "crypto/Kconfig" +diff -NurpP --minimal linux-3.0.9/arch/sh/kernel/irq.c linux-3.0.9-vs2.3.2.1/arch/sh/kernel/irq.c +--- linux-3.0.9/arch/sh/kernel/irq.c 2011-07-22 11:17:41.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/sh/kernel/irq.c 2011-07-19 00:45:06.000000000 +0200 +@@ -14,6 +14,7 @@ + #include <linux/ftrace.h> + #include <linux/delay.h> + #include <linux/ratelimit.h> ++// #include <linux/vs_context.h> + #include <asm/processor.h> + #include <asm/machvec.h> + #include <asm/uaccess.h> +diff -NurpP --minimal linux-3.0.9/arch/sh/kernel/vsyscall/vsyscall.c linux-3.0.9-vs2.3.2.1/arch/sh/kernel/vsyscall/vsyscall.c +--- linux-3.0.9/arch/sh/kernel/vsyscall/vsyscall.c 2011-05-22 16:17:07.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/sh/kernel/vsyscall/vsyscall.c 2011-06-10 22:11:24.000000000 +0200 +@@ -18,6 +18,7 @@ + #include <linux/elf.h> + #include <linux/sched.h> + #include <linux/err.h> ++#include <linux/vs_memory.h> + + /* + * Should the kernel map a VDSO page into processes and pass its +diff -NurpP --minimal linux-3.0.9/arch/sparc/Kconfig linux-3.0.9-vs2.3.2.1/arch/sparc/Kconfig +--- linux-3.0.9/arch/sparc/Kconfig 2011-11-15 16:40:42.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/arch/sparc/Kconfig 2011-10-18 13:51:13.000000000 +0200 +@@ -602,6 +602,8 @@ source "fs/Kconfig" + + source "arch/sparc/Kconfig.debug" + ++source "kernel/vserver/Kconfig" ++ + source "security/Kconfig" + + source "crypto/Kconfig" +diff -NurpP --minimal linux-3.0.9/arch/sparc/include/asm/tlb_64.h linux-3.0.9-vs2.3.2.1/arch/sparc/include/asm/tlb_64.h +--- linux-3.0.9/arch/sparc/include/asm/tlb_64.h 2011-07-22 11:17:42.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/sparc/include/asm/tlb_64.h 2011-06-10 22:11:24.000000000 +0200 +@@ -3,6 +3,7 @@ + + #include <linux/swap.h> + #include <linux/pagemap.h> ++#include <linux/vs_memory.h> + #include <asm/pgalloc.h> + #include <asm/tlbflush.h> + #include <asm/mmu_context.h> +diff -NurpP --minimal linux-3.0.9/arch/sparc/include/asm/unistd.h linux-3.0.9-vs2.3.2.1/arch/sparc/include/asm/unistd.h +--- linux-3.0.9/arch/sparc/include/asm/unistd.h 2011-07-22 11:17:42.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/sparc/include/asm/unistd.h 2011-06-10 22:11:24.000000000 +0200 +@@ -335,7 +335,7 @@ + #define __NR_timer_getoverrun 264 + #define __NR_timer_delete 265 + #define __NR_timer_create 266 +-/* #define __NR_vserver 267 Reserved for VSERVER */ ++#define __NR_vserver 267 + #define __NR_io_setup 268 + #define __NR_io_destroy 269 + #define __NR_io_submit 270 +diff -NurpP --minimal linux-3.0.9/arch/sparc/kernel/systbls_32.S linux-3.0.9-vs2.3.2.1/arch/sparc/kernel/systbls_32.S +--- linux-3.0.9/arch/sparc/kernel/systbls_32.S 2011-07-22 11:17:42.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/sparc/kernel/systbls_32.S 2011-06-10 22:11:24.000000000 +0200 +@@ -70,7 +70,7 @@ sys_call_table: + /*250*/ .long sys_mremap, sys_sysctl, sys_getsid, sys_fdatasync, sys_nfsservctl + /*255*/ .long sys_sync_file_range, sys_clock_settime, sys_clock_gettime, sys_clock_getres, sys_clock_nanosleep + /*260*/ .long sys_sched_getaffinity, sys_sched_setaffinity, sys_timer_settime, sys_timer_gettime, sys_timer_getoverrun +-/*265*/ .long sys_timer_delete, sys_timer_create, sys_nis_syscall, sys_io_setup, sys_io_destroy ++/*265*/ .long sys_timer_delete, sys_timer_create, sys_vserver, sys_io_setup, sys_io_destroy + /*270*/ .long sys_io_submit, sys_io_cancel, sys_io_getevents, sys_mq_open, sys_mq_unlink + /*275*/ .long sys_mq_timedsend, sys_mq_timedreceive, sys_mq_notify, sys_mq_getsetattr, sys_waitid + /*280*/ .long sys_tee, sys_add_key, sys_request_key, sys_keyctl, sys_openat +diff -NurpP --minimal linux-3.0.9/arch/sparc/kernel/systbls_64.S linux-3.0.9-vs2.3.2.1/arch/sparc/kernel/systbls_64.S +--- linux-3.0.9/arch/sparc/kernel/systbls_64.S 2011-07-22 11:17:42.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/sparc/kernel/systbls_64.S 2011-06-10 22:11:24.000000000 +0200 +@@ -71,7 +71,7 @@ sys_call_table32: + /*250*/ .word sys_mremap, compat_sys_sysctl, sys32_getsid, sys_fdatasync, sys32_nfsservctl + .word sys32_sync_file_range, compat_sys_clock_settime, compat_sys_clock_gettime, compat_sys_clock_getres, sys32_clock_nanosleep + /*260*/ .word compat_sys_sched_getaffinity, compat_sys_sched_setaffinity, sys32_timer_settime, compat_sys_timer_gettime, sys_timer_getoverrun +- .word sys_timer_delete, compat_sys_timer_create, sys_ni_syscall, compat_sys_io_setup, sys_io_destroy ++ .word sys_timer_delete, compat_sys_timer_create, sys32_vserver, compat_sys_io_setup, sys_io_destroy + /*270*/ .word sys32_io_submit, sys_io_cancel, compat_sys_io_getevents, sys32_mq_open, sys_mq_unlink + .word compat_sys_mq_timedsend, compat_sys_mq_timedreceive, compat_sys_mq_notify, compat_sys_mq_getsetattr, compat_sys_waitid + /*280*/ .word sys32_tee, sys_add_key, sys_request_key, sys_keyctl, compat_sys_openat +@@ -148,7 +148,7 @@ sys_call_table: + /*250*/ .word sys_64_mremap, sys_sysctl, sys_getsid, sys_fdatasync, sys_nfsservctl + .word sys_sync_file_range, sys_clock_settime, sys_clock_gettime, sys_clock_getres, sys_clock_nanosleep + /*260*/ .word sys_sched_getaffinity, sys_sched_setaffinity, sys_timer_settime, sys_timer_gettime, sys_timer_getoverrun +- .word sys_timer_delete, sys_timer_create, sys_ni_syscall, sys_io_setup, sys_io_destroy ++ .word sys_timer_delete, sys_timer_create, sys_vserver, sys_io_setup, sys_io_destroy + /*270*/ .word sys_io_submit, sys_io_cancel, sys_io_getevents, sys_mq_open, sys_mq_unlink + .word sys_mq_timedsend, sys_mq_timedreceive, sys_mq_notify, sys_mq_getsetattr, sys_waitid + /*280*/ .word sys_tee, sys_add_key, sys_request_key, sys_keyctl, sys_openat +diff -NurpP --minimal linux-3.0.9/arch/um/Kconfig.rest linux-3.0.9-vs2.3.2.1/arch/um/Kconfig.rest +--- linux-3.0.9/arch/um/Kconfig.rest 2009-06-11 17:12:19.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/um/Kconfig.rest 2011-06-10 22:11:24.000000000 +0200 +@@ -18,6 +18,8 @@ source "drivers/connector/Kconfig" + + source "fs/Kconfig" + ++source "kernel/vserver/Kconfig" ++ + source "security/Kconfig" + + source "crypto/Kconfig" +diff -NurpP --minimal linux-3.0.9/arch/um/include/asm/tlb.h linux-3.0.9-vs2.3.2.1/arch/um/include/asm/tlb.h +--- linux-3.0.9/arch/um/include/asm/tlb.h 2011-07-22 11:17:42.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/um/include/asm/tlb.h 2011-06-10 22:11:24.000000000 +0200 +@@ -3,6 +3,7 @@ + + #include <linux/pagemap.h> + #include <linux/swap.h> ++#include <linux/vs_memory.h> + #include <asm/percpu.h> + #include <asm/pgalloc.h> + #include <asm/tlbflush.h> +diff -NurpP --minimal linux-3.0.9/arch/um/include/shared/kern_constants.h linux-3.0.9-vs2.3.2.1/arch/um/include/shared/kern_constants.h +--- linux-3.0.9/arch/um/include/shared/kern_constants.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/arch/um/include/shared/kern_constants.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1 @@ ++#include "../../../../include/generated/asm-offsets.h" +diff -NurpP --minimal linux-3.0.9/arch/um/include/shared/user_constants.h linux-3.0.9-vs2.3.2.1/arch/um/include/shared/user_constants.h +--- linux-3.0.9/arch/um/include/shared/user_constants.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/arch/um/include/shared/user_constants.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,40 @@ ++/* ++ * DO NOT MODIFY. ++ * ++ * This file was generated by arch/um/Makefile ++ * ++ */ ++ ++#define HOST_SC_CR2 176 /* offsetof(struct sigcontext, cr2) # */ ++#define HOST_SC_ERR 152 /* offsetof(struct sigcontext, err) # */ ++#define HOST_SC_TRAPNO 160 /* offsetof(struct sigcontext, trapno) # */ ++#define HOST_FP_SIZE 64 /* sizeof(struct _fpstate) / sizeof(unsigned long) # */ ++#define HOST_RBX 5 /* RBX # */ ++#define HOST_RCX 11 /* RCX # */ ++#define HOST_RDI 14 /* RDI # */ ++#define HOST_RSI 13 /* RSI # */ ++#define HOST_RDX 12 /* RDX # */ ++#define HOST_RBP 4 /* RBP # */ ++#define HOST_RAX 10 /* RAX # */ ++#define HOST_R8 9 /* R8 # */ ++#define HOST_R9 8 /* R9 # */ ++#define HOST_R10 7 /* R10 # */ ++#define HOST_R11 6 /* R11 # */ ++#define HOST_R12 3 /* R12 # */ ++#define HOST_R13 2 /* R13 # */ ++#define HOST_R14 1 /* R14 # */ ++#define HOST_R15 0 /* R15 # */ ++#define HOST_ORIG_RAX 15 /* ORIG_RAX # */ ++#define HOST_CS 17 /* CS # */ ++#define HOST_SS 20 /* SS # */ ++#define HOST_EFLAGS 18 /* EFLAGS # */ ++#define HOST_IP 16 /* RIP # */ ++#define HOST_SP 19 /* RSP # */ ++#define UM_FRAME_SIZE 216 /* sizeof(struct user_regs_struct) # */ ++#define UM_POLLIN 1 /* POLLIN # */ ++#define UM_POLLPRI 2 /* POLLPRI # */ ++#define UM_POLLOUT 4 /* POLLOUT # */ ++#define UM_PROT_READ 1 /* PROT_READ # */ ++#define UM_PROT_WRITE 2 /* PROT_WRITE # */ ++#define UM_PROT_EXEC 4 /* PROT_EXEC # */ ++ +diff -NurpP --minimal linux-3.0.9/arch/x86/Kconfig linux-3.0.9-vs2.3.2.1/arch/x86/Kconfig +--- linux-3.0.9/arch/x86/Kconfig 2011-07-22 11:17:42.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/x86/Kconfig 2011-07-22 11:20:39.000000000 +0200 +@@ -2159,6 +2159,8 @@ source "fs/Kconfig" + + source "arch/x86/Kconfig.debug" + ++source "kernel/vserver/Kconfig" ++ + source "security/Kconfig" + + source "crypto/Kconfig" +diff -NurpP --minimal linux-3.0.9/arch/x86/ia32/ia32entry.S linux-3.0.9-vs2.3.2.1/arch/x86/ia32/ia32entry.S +--- linux-3.0.9/arch/x86/ia32/ia32entry.S 2011-07-22 11:17:42.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/x86/ia32/ia32entry.S 2011-06-10 22:11:24.000000000 +0200 +@@ -776,7 +776,7 @@ ia32_sys_call_table: + .quad sys_tgkill /* 270 */ + .quad compat_sys_utimes + .quad sys32_fadvise64_64 +- .quad quiet_ni_syscall /* sys_vserver */ ++ .quad sys32_vserver + .quad sys_mbind + .quad compat_sys_get_mempolicy /* 275 */ + .quad sys_set_mempolicy +diff -NurpP --minimal linux-3.0.9/arch/x86/include/asm/unistd_64.h linux-3.0.9-vs2.3.2.1/arch/x86/include/asm/unistd_64.h +--- linux-3.0.9/arch/x86/include/asm/unistd_64.h 2011-07-22 11:17:43.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/x86/include/asm/unistd_64.h 2011-06-10 22:11:24.000000000 +0200 +@@ -535,7 +535,7 @@ __SYSCALL(__NR_tgkill, sys_tgkill) + #define __NR_utimes 235 + __SYSCALL(__NR_utimes, sys_utimes) + #define __NR_vserver 236 +-__SYSCALL(__NR_vserver, sys_ni_syscall) ++__SYSCALL(__NR_vserver, sys_vserver) + #define __NR_mbind 237 + __SYSCALL(__NR_mbind, sys_mbind) + #define __NR_set_mempolicy 238 +diff -NurpP --minimal linux-3.0.9/arch/x86/kernel/syscall_table_32.S linux-3.0.9-vs2.3.2.1/arch/x86/kernel/syscall_table_32.S +--- linux-3.0.9/arch/x86/kernel/syscall_table_32.S 2011-07-22 11:17:43.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/arch/x86/kernel/syscall_table_32.S 2011-06-10 22:11:24.000000000 +0200 +@@ -272,7 +272,7 @@ ENTRY(sys_call_table) + .long sys_tgkill /* 270 */ + .long sys_utimes + .long sys_fadvise64_64 +- .long sys_ni_syscall /* sys_vserver */ ++ .long sys_vserver + .long sys_mbind + .long sys_get_mempolicy + .long sys_set_mempolicy +diff -NurpP --minimal linux-3.0.9/block/genhd.c linux-3.0.9-vs2.3.2.1/block/genhd.c +--- linux-3.0.9/block/genhd.c 2011-11-15 16:40:42.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/block/genhd.c 2011-11-15 17:37:04.000000000 +0100 +@@ -1162,17 +1162,17 @@ static int diskstats_show(struct seq_fil + cpu = part_stat_lock(); + part_round_stats(cpu, hd); + part_stat_unlock(); +- seq_printf(seqf, "%4d %7d %s %lu %lu %llu " +- "%u %lu %lu %llu %u %u %u %u\n", ++ seq_printf(seqf, "%4d %7d %s %lu %lu %lu " ++ "%u %lu %lu %lu %u %u %u %u\n", + MAJOR(part_devt(hd)), MINOR(part_devt(hd)), + disk_name(gp, hd->partno, buf), + part_stat_read(hd, ios[READ]), + part_stat_read(hd, merges[READ]), +- (unsigned long long)part_stat_read(hd, sectors[READ]), ++ part_stat_read(hd, sectors[READ]), + jiffies_to_msecs(part_stat_read(hd, ticks[READ])), + part_stat_read(hd, ios[WRITE]), + part_stat_read(hd, merges[WRITE]), +- (unsigned long long)part_stat_read(hd, sectors[WRITE]), ++ part_stat_read(hd, sectors[WRITE]), + jiffies_to_msecs(part_stat_read(hd, ticks[WRITE])), + part_in_flight(hd), + jiffies_to_msecs(part_stat_read(hd, io_ticks)), +diff -NurpP --minimal linux-3.0.9/drivers/block/Kconfig linux-3.0.9-vs2.3.2.1/drivers/block/Kconfig +--- linux-3.0.9/drivers/block/Kconfig 2011-07-22 11:17:44.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/drivers/block/Kconfig 2011-06-10 22:11:24.000000000 +0200 +@@ -273,6 +273,13 @@ config BLK_DEV_CRYPTOLOOP + + source "drivers/block/drbd/Kconfig" + ++config BLK_DEV_VROOT ++ tristate "Virtual Root device support" ++ depends on QUOTACTL ++ ---help--- ++ Saying Y here will allow you to use quota/fs ioctls on a shared ++ partition within a virtual server without compromising security. ++ + config BLK_DEV_NBD + tristate "Network block device support" + depends on NET +diff -NurpP --minimal linux-3.0.9/drivers/block/Makefile linux-3.0.9-vs2.3.2.1/drivers/block/Makefile +--- linux-3.0.9/drivers/block/Makefile 2011-07-22 11:17:44.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/drivers/block/Makefile 2011-06-10 22:11:24.000000000 +0200 +@@ -34,6 +34,7 @@ obj-$(CONFIG_VIODASD) += viodasd.o + obj-$(CONFIG_BLK_DEV_SX8) += sx8.o + obj-$(CONFIG_BLK_DEV_UB) += ub.o + obj-$(CONFIG_BLK_DEV_HD) += hd.o ++obj-$(CONFIG_BLK_DEV_VROOT) += vroot.o + + obj-$(CONFIG_XEN_BLKDEV_FRONTEND) += xen-blkfront.o + obj-$(CONFIG_XEN_BLKDEV_BACKEND) += xen-blkback/ +diff -NurpP --minimal linux-3.0.9/drivers/block/loop.c linux-3.0.9-vs2.3.2.1/drivers/block/loop.c +--- linux-3.0.9/drivers/block/loop.c 2011-11-15 16:40:42.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/drivers/block/loop.c 2011-08-31 19:37:44.000000000 +0200 +@@ -75,6 +75,7 @@ + #include <linux/kthread.h> + #include <linux/splice.h> + #include <linux/sysfs.h> ++#include <linux/vs_context.h> + + #include <asm/uaccess.h> + +@@ -891,6 +892,7 @@ static int loop_set_fd(struct loop_devic + lo->lo_blocksize = lo_blocksize; + lo->lo_device = bdev; + lo->lo_flags = lo_flags; ++ lo->lo_xid = vx_current_xid(); + lo->lo_backing_file = file; + lo->transfer = transfer_none; + lo->ioctl = NULL; +@@ -1021,6 +1023,7 @@ static int loop_clr_fd(struct loop_devic + lo->lo_encrypt_key_size = 0; + lo->lo_flags = 0; + lo->lo_thread = NULL; ++ lo->lo_xid = 0; + memset(lo->lo_encrypt_key, 0, LO_KEY_SIZE); + memset(lo->lo_crypt_name, 0, LO_NAME_SIZE); + memset(lo->lo_file_name, 0, LO_NAME_SIZE); +@@ -1059,7 +1062,7 @@ loop_set_status(struct loop_device *lo, + + if (lo->lo_encrypt_key_size && + lo->lo_key_owner != uid && +- !capable(CAP_SYS_ADMIN)) ++ !vx_capable(CAP_SYS_ADMIN, VXC_ADMIN_CLOOP)) + return -EPERM; + if (lo->lo_state != Lo_bound) + return -ENXIO; +@@ -1143,7 +1146,8 @@ loop_get_status(struct loop_device *lo, + memcpy(info->lo_crypt_name, lo->lo_crypt_name, LO_NAME_SIZE); + info->lo_encrypt_type = + lo->lo_encryption ? lo->lo_encryption->number : 0; +- if (lo->lo_encrypt_key_size && capable(CAP_SYS_ADMIN)) { ++ if (lo->lo_encrypt_key_size && ++ vx_capable(CAP_SYS_ADMIN, VXC_ADMIN_CLOOP)) { + info->lo_encrypt_key_size = lo->lo_encrypt_key_size; + memcpy(info->lo_encrypt_key, lo->lo_encrypt_key, + lo->lo_encrypt_key_size); +@@ -1489,6 +1493,9 @@ static int lo_open(struct block_device * + { + struct loop_device *lo = bdev->bd_disk->private_data; + ++ if (!vx_check(lo->lo_xid, VS_IDENT|VS_HOSTID|VS_ADMIN_P)) ++ return -EACCES; ++ + mutex_lock(&lo->lo_ctl_mutex); + lo->lo_refcnt++; + mutex_unlock(&lo->lo_ctl_mutex); +diff -NurpP --minimal linux-3.0.9/drivers/block/vroot.c linux-3.0.9-vs2.3.2.1/drivers/block/vroot.c +--- linux-3.0.9/drivers/block/vroot.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/drivers/block/vroot.c 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,292 @@ ++/* ++ * linux/drivers/block/vroot.c ++ * ++ * written by Herbert Pötzl, 9/11/2002 ++ * ported to 2.6.10 by Herbert Pötzl, 30/12/2004 ++ * ++ * based on the loop.c code by Theodore Ts'o. ++ * ++ * Copyright (C) 2002-2007 by Herbert Pötzl. ++ * Redistribution of this file is permitted under the ++ * GNU General Public License. ++ * ++ */ ++ ++#include <linux/module.h> ++#include <linux/moduleparam.h> ++#include <linux/file.h> ++#include <linux/major.h> ++#include <linux/blkdev.h> ++#include <linux/slab.h> ++ ++#include <linux/vroot.h> ++#include <linux/vs_context.h> ++ ++ ++static int max_vroot = 8; ++ ++static struct vroot_device *vroot_dev; ++static struct gendisk **disks; ++ ++ ++static int vroot_set_dev( ++ struct vroot_device *vr, ++ struct block_device *bdev, ++ unsigned int arg) ++{ ++ struct block_device *real_bdev; ++ struct file *file; ++ struct inode *inode; ++ int error; ++ ++ error = -EBUSY; ++ if (vr->vr_state != Vr_unbound) ++ goto out; ++ ++ error = -EBADF; ++ file = fget(arg); ++ if (!file) ++ goto out; ++ ++ error = -EINVAL; ++ inode = file->f_dentry->d_inode; ++ ++ ++ if (S_ISBLK(inode->i_mode)) { ++ real_bdev = inode->i_bdev; ++ vr->vr_device = real_bdev; ++ __iget(real_bdev->bd_inode); ++ } else ++ goto out_fput; ++ ++ vxdprintk(VXD_CBIT(misc, 0), ++ "vroot[%d]_set_dev: dev=" VXF_DEV, ++ vr->vr_number, VXD_DEV(real_bdev)); ++ ++ vr->vr_state = Vr_bound; ++ error = 0; ++ ++ out_fput: ++ fput(file); ++ out: ++ return error; ++} ++ ++static int vroot_clr_dev( ++ struct vroot_device *vr, ++ struct block_device *bdev) ++{ ++ struct block_device *real_bdev; ++ ++ if (vr->vr_state != Vr_bound) ++ return -ENXIO; ++ if (vr->vr_refcnt > 1) /* we needed one fd for the ioctl */ ++ return -EBUSY; ++ ++ real_bdev = vr->vr_device; ++ ++ vxdprintk(VXD_CBIT(misc, 0), ++ "vroot[%d]_clr_dev: dev=" VXF_DEV, ++ vr->vr_number, VXD_DEV(real_bdev)); ++ ++ bdput(real_bdev); ++ vr->vr_state = Vr_unbound; ++ vr->vr_device = NULL; ++ return 0; ++} ++ ++ ++static int vr_ioctl(struct block_device *bdev, fmode_t mode, ++ unsigned int cmd, unsigned long arg) ++{ ++ struct vroot_device *vr = bdev->bd_disk->private_data; ++ int err; ++ ++ down(&vr->vr_ctl_mutex); ++ switch (cmd) { ++ case VROOT_SET_DEV: ++ err = vroot_set_dev(vr, bdev, arg); ++ break; ++ case VROOT_CLR_DEV: ++ err = vroot_clr_dev(vr, bdev); ++ break; ++ default: ++ err = -EINVAL; ++ break; ++ } ++ up(&vr->vr_ctl_mutex); ++ return err; ++} ++ ++static int vr_open(struct block_device *bdev, fmode_t mode) ++{ ++ struct vroot_device *vr = bdev->bd_disk->private_data; ++ ++ down(&vr->vr_ctl_mutex); ++ vr->vr_refcnt++; ++ up(&vr->vr_ctl_mutex); ++ return 0; ++} ++ ++static int vr_release(struct gendisk *disk, fmode_t mode) ++{ ++ struct vroot_device *vr = disk->private_data; ++ ++ down(&vr->vr_ctl_mutex); ++ --vr->vr_refcnt; ++ up(&vr->vr_ctl_mutex); ++ return 0; ++} ++ ++static struct block_device_operations vr_fops = { ++ .owner = THIS_MODULE, ++ .open = vr_open, ++ .release = vr_release, ++ .ioctl = vr_ioctl, ++}; ++ ++static int vroot_make_request(struct request_queue *q, struct bio *bio) ++{ ++ printk("vroot_make_request %p, %p\n", q, bio); ++ bio_io_error(bio); ++ return 0; ++} ++ ++struct block_device *__vroot_get_real_bdev(struct block_device *bdev) ++{ ++ struct inode *inode = bdev->bd_inode; ++ struct vroot_device *vr; ++ struct block_device *real_bdev; ++ int minor = iminor(inode); ++ ++ vr = &vroot_dev[minor]; ++ real_bdev = vr->vr_device; ++ ++ vxdprintk(VXD_CBIT(misc, 0), ++ "vroot[%d]_get_real_bdev: dev=" VXF_DEV, ++ vr->vr_number, VXD_DEV(real_bdev)); ++ ++ if (vr->vr_state != Vr_bound) ++ return ERR_PTR(-ENXIO); ++ ++ __iget(real_bdev->bd_inode); ++ return real_bdev; ++} ++ ++ ++ ++/* ++ * And now the modules code and kernel interface. ++ */ ++ ++module_param(max_vroot, int, 0); ++ ++MODULE_PARM_DESC(max_vroot, "Maximum number of vroot devices (1-256)"); ++MODULE_LICENSE("GPL"); ++MODULE_ALIAS_BLOCKDEV_MAJOR(VROOT_MAJOR); ++ ++MODULE_AUTHOR ("Herbert Pötzl"); ++MODULE_DESCRIPTION ("Virtual Root Device Mapper"); ++ ++ ++int __init vroot_init(void) ++{ ++ int err, i; ++ ++ if (max_vroot < 1 || max_vroot > 256) { ++ max_vroot = MAX_VROOT_DEFAULT; ++ printk(KERN_WARNING "vroot: invalid max_vroot " ++ "(must be between 1 and 256), " ++ "using default (%d)\n", max_vroot); ++ } ++ ++ if (register_blkdev(VROOT_MAJOR, "vroot")) ++ return -EIO; ++ ++ err = -ENOMEM; ++ vroot_dev = kmalloc(max_vroot * sizeof(struct vroot_device), GFP_KERNEL); ++ if (!vroot_dev) ++ goto out_mem1; ++ memset(vroot_dev, 0, max_vroot * sizeof(struct vroot_device)); ++ ++ disks = kmalloc(max_vroot * sizeof(struct gendisk *), GFP_KERNEL); ++ if (!disks) ++ goto out_mem2; ++ ++ for (i = 0; i < max_vroot; i++) { ++ disks[i] = alloc_disk(1); ++ if (!disks[i]) ++ goto out_mem3; ++ disks[i]->queue = blk_alloc_queue(GFP_KERNEL); ++ if (!disks[i]->queue) ++ goto out_mem3; ++ blk_queue_make_request(disks[i]->queue, vroot_make_request); ++ } ++ ++ for (i = 0; i < max_vroot; i++) { ++ struct vroot_device *vr = &vroot_dev[i]; ++ struct gendisk *disk = disks[i]; ++ ++ memset(vr, 0, sizeof(*vr)); ++ sema_init(&vr->vr_ctl_mutex, 1); ++ vr->vr_number = i; ++ disk->major = VROOT_MAJOR; ++ disk->first_minor = i; ++ disk->fops = &vr_fops; ++ sprintf(disk->disk_name, "vroot%d", i); ++ disk->private_data = vr; ++ } ++ ++ err = register_vroot_grb(&__vroot_get_real_bdev); ++ if (err) ++ goto out_mem3; ++ ++ for (i = 0; i < max_vroot; i++) ++ add_disk(disks[i]); ++ printk(KERN_INFO "vroot: loaded (max %d devices)\n", max_vroot); ++ return 0; ++ ++out_mem3: ++ while (i--) ++ put_disk(disks[i]); ++ kfree(disks); ++out_mem2: ++ kfree(vroot_dev); ++out_mem1: ++ unregister_blkdev(VROOT_MAJOR, "vroot"); ++ printk(KERN_ERR "vroot: ran out of memory\n"); ++ return err; ++} ++ ++void vroot_exit(void) ++{ ++ int i; ++ ++ if (unregister_vroot_grb(&__vroot_get_real_bdev)) ++ printk(KERN_WARNING "vroot: cannot unregister grb\n"); ++ ++ for (i = 0; i < max_vroot; i++) { ++ del_gendisk(disks[i]); ++ put_disk(disks[i]); ++ } ++ unregister_blkdev(VROOT_MAJOR, "vroot"); ++ ++ kfree(disks); ++ kfree(vroot_dev); ++} ++ ++module_init(vroot_init); ++module_exit(vroot_exit); ++ ++#ifndef MODULE ++ ++static int __init max_vroot_setup(char *str) ++{ ++ max_vroot = simple_strtol(str, NULL, 0); ++ return 1; ++} ++ ++__setup("max_vroot=", max_vroot_setup); ++ ++#endif ++ +diff -NurpP --minimal linux-3.0.9/drivers/infiniband/core/addr.c linux-3.0.9-vs2.3.2.1/drivers/infiniband/core/addr.c +--- linux-3.0.9/drivers/infiniband/core/addr.c 2011-07-22 11:17:45.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/drivers/infiniband/core/addr.c 2011-06-16 14:16:51.000000000 +0200 +@@ -252,7 +252,7 @@ static int addr6_resolve(struct sockaddr + + if (ipv6_addr_any(&fl6.saddr)) { + ret = ipv6_dev_get_saddr(&init_net, ip6_dst_idev(dst)->dev, +- &fl6.daddr, 0, &fl6.saddr); ++ &fl6.daddr, 0, &fl6.saddr, NULL); + if (ret) + goto put; + +diff -NurpP --minimal linux-3.0.9/drivers/infiniband/hw/ipath/ipath_user_pages.c linux-3.0.9-vs2.3.2.1/drivers/infiniband/hw/ipath/ipath_user_pages.c +--- linux-3.0.9/drivers/infiniband/hw/ipath/ipath_user_pages.c 2011-05-22 16:17:16.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/drivers/infiniband/hw/ipath/ipath_user_pages.c 2011-06-10 22:11:24.000000000 +0200 +@@ -35,6 +35,7 @@ + #include <linux/device.h> + #include <linux/slab.h> + #include <linux/sched.h> ++#include <linux/vs_memory.h> + + #include "ipath_kernel.h" + +diff -NurpP --minimal linux-3.0.9/drivers/md/dm-ioctl.c linux-3.0.9-vs2.3.2.1/drivers/md/dm-ioctl.c +--- linux-3.0.9/drivers/md/dm-ioctl.c 2011-05-22 16:17:18.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/drivers/md/dm-ioctl.c 2011-06-10 22:11:24.000000000 +0200 +@@ -16,6 +16,7 @@ + #include <linux/dm-ioctl.h> + #include <linux/hdreg.h> + #include <linux/compat.h> ++#include <linux/vs_context.h> + + #include <asm/uaccess.h> + +@@ -106,7 +107,8 @@ static struct hash_cell *__get_name_cell + unsigned int h = hash_str(str); + + list_for_each_entry (hc, _name_buckets + h, name_list) +- if (!strcmp(hc->name, str)) { ++ if (vx_check(dm_get_xid(hc->md), VS_WATCH_P | VS_IDENT) && ++ !strcmp(hc->name, str)) { + dm_get(hc->md); + return hc; + } +@@ -120,7 +122,8 @@ static struct hash_cell *__get_uuid_cell + unsigned int h = hash_str(str); + + list_for_each_entry (hc, _uuid_buckets + h, uuid_list) +- if (!strcmp(hc->uuid, str)) { ++ if (vx_check(dm_get_xid(hc->md), VS_WATCH_P | VS_IDENT) && ++ !strcmp(hc->uuid, str)) { + dm_get(hc->md); + return hc; + } +@@ -427,6 +430,9 @@ typedef int (*ioctl_fn)(struct dm_ioctl + + static int remove_all(struct dm_ioctl *param, size_t param_size) + { ++ if (!vx_check(0, VS_ADMIN)) ++ return -EPERM; ++ + dm_hash_remove_all(1); + param->data_size = 0; + return 0; +@@ -474,6 +480,8 @@ static int list_devices(struct dm_ioctl + */ + for (i = 0; i < NUM_BUCKETS; i++) { + list_for_each_entry (hc, _name_buckets + i, name_list) { ++ if (!vx_check(dm_get_xid(hc->md), VS_WATCH_P | VS_IDENT)) ++ continue; + needed += sizeof(struct dm_name_list); + needed += strlen(hc->name) + 1; + needed += ALIGN_MASK; +@@ -497,6 +505,8 @@ static int list_devices(struct dm_ioctl + */ + for (i = 0; i < NUM_BUCKETS; i++) { + list_for_each_entry (hc, _name_buckets + i, name_list) { ++ if (!vx_check(dm_get_xid(hc->md), VS_WATCH_P | VS_IDENT)) ++ continue; + if (old_nl) + old_nl->next = (uint32_t) ((void *) nl - + (void *) old_nl); +@@ -731,10 +741,11 @@ static struct hash_cell *__find_device_h + if (!md) + goto out; + +- mdptr = dm_get_mdptr(md); ++ if (vx_check(dm_get_xid(md), VS_WATCH_P | VS_IDENT)) ++ mdptr = dm_get_mdptr(md); ++ + if (!mdptr) + dm_put(md); +- + out: + return mdptr; + } +@@ -1577,8 +1588,8 @@ static int ctl_ioctl(uint command, struc + ioctl_fn fn = NULL; + size_t input_param_size; + +- /* only root can play with this */ +- if (!capable(CAP_SYS_ADMIN)) ++ /* only root and certain contexts can play with this */ ++ if (!vx_capable(CAP_SYS_ADMIN, VXC_ADMIN_MAPPER)) + return -EACCES; + + if (_IOC_TYPE(command) != DM_IOCTL) +diff -NurpP --minimal linux-3.0.9/drivers/md/dm.c linux-3.0.9-vs2.3.2.1/drivers/md/dm.c +--- linux-3.0.9/drivers/md/dm.c 2011-11-15 16:40:44.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/drivers/md/dm.c 2011-08-08 23:04:47.000000000 +0200 +@@ -20,6 +20,7 @@ + #include <linux/idr.h> + #include <linux/hdreg.h> + #include <linux/delay.h> ++#include <linux/vs_base.h> + + #include <trace/events/block.h> + +@@ -121,6 +122,7 @@ struct mapped_device { + rwlock_t map_lock; + atomic_t holders; + atomic_t open_count; ++ xid_t xid; + + unsigned long flags; + +@@ -334,6 +336,7 @@ int dm_deleting_md(struct mapped_device + static int dm_blk_open(struct block_device *bdev, fmode_t mode) + { + struct mapped_device *md; ++ int ret = -ENXIO; + + spin_lock(&_minor_lock); + +@@ -342,18 +345,19 @@ static int dm_blk_open(struct block_devi + goto out; + + if (test_bit(DMF_FREEING, &md->flags) || +- dm_deleting_md(md)) { +- md = NULL; ++ dm_deleting_md(md)) ++ goto out; ++ ++ ret = -EACCES; ++ if (!vx_check(md->xid, VS_IDENT|VS_HOSTID)) + goto out; +- } + + dm_get(md); + atomic_inc(&md->open_count); +- ++ ret = 0; + out: + spin_unlock(&_minor_lock); +- +- return md ? 0 : -ENXIO; ++ return ret; + } + + static int dm_blk_close(struct gendisk *disk, fmode_t mode) +@@ -574,6 +578,14 @@ int dm_set_geometry(struct mapped_device + return 0; + } + ++/* ++ * Get the xid associated with a dm device ++ */ ++xid_t dm_get_xid(struct mapped_device *md) ++{ ++ return md->xid; ++} ++ + /*----------------------------------------------------------------- + * CRUD START: + * A more elegant soln is in the works that uses the queue +@@ -1847,6 +1859,7 @@ static struct mapped_device *alloc_dev(i + INIT_LIST_HEAD(&md->uevent_list); + spin_lock_init(&md->uevent_lock); + ++ md->xid = vx_current_xid(); + md->queue = blk_alloc_queue(GFP_KERNEL); + if (!md->queue) + goto bad_queue; +diff -NurpP --minimal linux-3.0.9/drivers/md/dm.h linux-3.0.9-vs2.3.2.1/drivers/md/dm.h +--- linux-3.0.9/drivers/md/dm.h 2011-05-22 16:17:18.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/drivers/md/dm.h 2011-06-10 22:11:24.000000000 +0200 +@@ -41,6 +41,8 @@ struct dm_dev_internal { + struct dm_table; + struct dm_md_mempools; + ++xid_t dm_get_xid(struct mapped_device *md); ++ + /*----------------------------------------------------------------- + * Internal table functions. + *---------------------------------------------------------------*/ +diff -NurpP --minimal linux-3.0.9/drivers/net/tun.c linux-3.0.9-vs2.3.2.1/drivers/net/tun.c +--- linux-3.0.9/drivers/net/tun.c 2011-11-15 16:40:45.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/drivers/net/tun.c 2011-08-29 03:45:08.000000000 +0200 +@@ -64,6 +64,7 @@ + #include <linux/nsproxy.h> + #include <linux/virtio_net.h> + #include <linux/rcupdate.h> ++#include <linux/vs_network.h> + #include <net/net_namespace.h> + #include <net/netns/generic.h> + #include <net/rtnetlink.h> +@@ -121,6 +122,7 @@ struct tun_struct { + unsigned int flags; + uid_t owner; + gid_t group; ++ nid_t nid; + + struct net_device *dev; + u32 set_features; +@@ -905,6 +907,7 @@ static void tun_setup(struct net_device + + tun->owner = -1; + tun->group = -1; ++ tun->nid = current->nid; + + dev->ethtool_ops = &tun_ethtool_ops; + dev->destructor = tun_free_netdev; +@@ -1055,7 +1058,7 @@ static int tun_set_iff(struct net *net, + + if (((tun->owner != -1 && cred->euid != tun->owner) || + (tun->group != -1 && !in_egroup_p(tun->group))) && +- !capable(CAP_NET_ADMIN)) ++ !cap_raised(current_cap(), CAP_NET_ADMIN)) + return -EPERM; + err = security_tun_dev_attach(tun->socket.sk); + if (err < 0) +@@ -1069,7 +1072,7 @@ static int tun_set_iff(struct net *net, + char *name; + unsigned long flags = 0; + +- if (!capable(CAP_NET_ADMIN)) ++ if (!nx_capable(CAP_NET_ADMIN, NXC_TUN_CREATE)) + return -EPERM; + err = security_tun_dev_create(); + if (err < 0) +@@ -1137,6 +1140,9 @@ static int tun_set_iff(struct net *net, + + sk->sk_destruct = tun_sock_destruct; + ++ if (!nx_check(tun->nid, VS_IDENT | VS_HOSTID | VS_ADMIN_P)) ++ return -EPERM; ++ + err = tun_attach(tun, file); + if (err < 0) + goto failed; +@@ -1318,6 +1324,16 @@ static long __tun_chr_ioctl(struct file + tun_debug(KERN_INFO, tun, "group set to %d\n", tun->group); + break; + ++ case TUNSETNID: ++ if (!capable(CAP_CONTEXT)) ++ return -EPERM; ++ ++ /* Set nid owner of the device */ ++ tun->nid = (nid_t) arg; ++ ++ tun_debug(KERN_INFO, tun, "nid owner set to %u\n", tun->nid); ++ break; ++ + case TUNSETLINK: + /* Only allow setting the type when the interface is down */ + if (tun->dev->flags & IFF_UP) { +diff -NurpP --minimal linux-3.0.9/drivers/tty/sysrq.c linux-3.0.9-vs2.3.2.1/drivers/tty/sysrq.c +--- linux-3.0.9/drivers/tty/sysrq.c 2011-05-22 16:17:44.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/drivers/tty/sysrq.c 2011-06-10 22:11:24.000000000 +0200 +@@ -41,6 +41,7 @@ + #include <linux/oom.h> + #include <linux/slab.h> + #include <linux/input.h> ++#include <linux/vserver/debug.h> + + #include <asm/ptrace.h> + #include <asm/irq_regs.h> +@@ -395,6 +396,21 @@ static struct sysrq_key_op sysrq_unrt_op + .enable_mask = SYSRQ_ENABLE_RTNICE, + }; + ++ ++#ifdef CONFIG_VSERVER_DEBUG ++static void sysrq_handle_vxinfo(int key) ++{ ++ dump_vx_info_inactive((key == 'x') ? 0 : 1); ++} ++ ++static struct sysrq_key_op sysrq_showvxinfo_op = { ++ .handler = sysrq_handle_vxinfo, ++ .help_msg = "conteXt", ++ .action_msg = "Show Context Info", ++ .enable_mask = SYSRQ_ENABLE_DUMP, ++}; ++#endif ++ + /* Key Operations table and lock */ + static DEFINE_SPINLOCK(sysrq_key_table_lock); + +@@ -449,7 +465,11 @@ static struct sysrq_key_op *sysrq_key_ta + NULL, /* v */ + &sysrq_showstate_blocked_op, /* w */ + /* x: May be registered on ppc/powerpc for xmon */ ++#ifdef CONFIG_VSERVER_DEBUG ++ &sysrq_showvxinfo_op, /* x */ ++#else + NULL, /* x */ ++#endif + /* y: May be registered on sparc64 for global register dump */ + NULL, /* y */ + &sysrq_ftrace_dump_op, /* z */ +@@ -464,6 +484,8 @@ static int sysrq_key_table_key2index(int + retval = key - '0'; + else if ((key >= 'a') && (key <= 'z')) + retval = key + 10 - 'a'; ++ else if ((key >= 'A') && (key <= 'Z')) ++ retval = key + 10 - 'A'; + else + retval = -1; + return retval; +diff -NurpP --minimal linux-3.0.9/drivers/tty/tty_io.c linux-3.0.9-vs2.3.2.1/drivers/tty/tty_io.c +--- linux-3.0.9/drivers/tty/tty_io.c 2011-11-15 16:40:46.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/drivers/tty/tty_io.c 2011-11-15 17:37:05.000000000 +0100 +@@ -104,6 +104,7 @@ + + #include <linux/kmod.h> + #include <linux/nsproxy.h> ++#include <linux/vs_pid.h> + + #undef TTY_DEBUG_HANGUP + +@@ -2080,7 +2081,8 @@ static int tiocsti(struct tty_struct *tt + char ch, mbz = 0; + struct tty_ldisc *ld; + +- if ((current->signal->tty != tty) && !capable(CAP_SYS_ADMIN)) ++ if (((current->signal->tty != tty) && ++ !vx_capable(CAP_SYS_ADMIN, VXC_TIOCSTI))) + return -EPERM; + if (get_user(ch, p)) + return -EFAULT; +@@ -2368,6 +2370,7 @@ static int tiocspgrp(struct tty_struct * + return -ENOTTY; + if (get_user(pgrp_nr, p)) + return -EFAULT; ++ pgrp_nr = vx_rmap_pid(pgrp_nr); + if (pgrp_nr < 0) + return -EINVAL; + rcu_read_lock(); +diff -NurpP --minimal linux-3.0.9/fs/attr.c linux-3.0.9-vs2.3.2.1/fs/attr.c +--- linux-3.0.9/fs/attr.c 2011-07-22 11:18:05.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/attr.c 2011-06-10 22:11:24.000000000 +0200 +@@ -13,6 +13,9 @@ + #include <linux/fsnotify.h> + #include <linux/fcntl.h> + #include <linux/security.h> ++#include <linux/proc_fs.h> ++#include <linux/devpts_fs.h> ++#include <linux/vs_tag.h> + + /** + * inode_change_ok - check if attribute changes to an inode are allowed +@@ -73,6 +76,10 @@ int inode_change_ok(const struct inode * + return -EPERM; + } + ++ /* check for inode tag permission */ ++ if (dx_permission(inode, MAY_WRITE)) ++ return -EACCES; ++ + return 0; + } + EXPORT_SYMBOL(inode_change_ok); +@@ -143,6 +150,8 @@ void setattr_copy(struct inode *inode, c + inode->i_uid = attr->ia_uid; + if (ia_valid & ATTR_GID) + inode->i_gid = attr->ia_gid; ++ if ((ia_valid & ATTR_TAG) && IS_TAGGED(inode)) ++ inode->i_tag = attr->ia_tag; + if (ia_valid & ATTR_ATIME) + inode->i_atime = timespec_trunc(attr->ia_atime, + inode->i_sb->s_time_gran); +@@ -170,7 +179,8 @@ int notify_change(struct dentry * dentry + struct timespec now; + unsigned int ia_valid = attr->ia_valid; + +- if (ia_valid & (ATTR_MODE | ATTR_UID | ATTR_GID | ATTR_TIMES_SET)) { ++ if (ia_valid & (ATTR_MODE | ATTR_UID | ATTR_GID | ++ ATTR_TAG | ATTR_TIMES_SET)) { + if (IS_IMMUTABLE(inode) || IS_APPEND(inode)) + return -EPERM; + } +diff -NurpP --minimal linux-3.0.9/fs/binfmt_aout.c linux-3.0.9-vs2.3.2.1/fs/binfmt_aout.c +--- linux-3.0.9/fs/binfmt_aout.c 2010-10-21 13:07:47.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/binfmt_aout.c 2011-06-10 22:11:24.000000000 +0200 +@@ -25,6 +25,7 @@ + #include <linux/init.h> + #include <linux/coredump.h> + #include <linux/slab.h> ++#include <linux/vs_memory.h> + + #include <asm/system.h> + #include <asm/uaccess.h> +diff -NurpP --minimal linux-3.0.9/fs/binfmt_elf.c linux-3.0.9-vs2.3.2.1/fs/binfmt_elf.c +--- linux-3.0.9/fs/binfmt_elf.c 2011-11-15 16:40:46.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/binfmt_elf.c 2011-11-15 17:37:06.000000000 +0100 +@@ -32,6 +32,7 @@ + #include <linux/elf.h> + #include <linux/utsname.h> + #include <linux/coredump.h> ++#include <linux/vs_memory.h> + #include <asm/uaccess.h> + #include <asm/param.h> + #include <asm/page.h> +diff -NurpP --minimal linux-3.0.9/fs/binfmt_flat.c linux-3.0.9-vs2.3.2.1/fs/binfmt_flat.c +--- linux-3.0.9/fs/binfmt_flat.c 2011-07-22 11:18:05.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/binfmt_flat.c 2011-06-10 22:11:24.000000000 +0200 +@@ -35,6 +35,7 @@ + #include <linux/init.h> + #include <linux/flat.h> + #include <linux/syscalls.h> ++#include <linux/vs_memory.h> + + #include <asm/byteorder.h> + #include <asm/system.h> +diff -NurpP --minimal linux-3.0.9/fs/binfmt_som.c linux-3.0.9-vs2.3.2.1/fs/binfmt_som.c +--- linux-3.0.9/fs/binfmt_som.c 2010-02-25 11:52:04.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/binfmt_som.c 2011-06-10 22:11:24.000000000 +0200 +@@ -28,6 +28,7 @@ + #include <linux/shm.h> + #include <linux/personality.h> + #include <linux/init.h> ++#include <linux/vs_memory.h> + + #include <asm/uaccess.h> + #include <asm/pgtable.h> +diff -NurpP --minimal linux-3.0.9/fs/block_dev.c linux-3.0.9-vs2.3.2.1/fs/block_dev.c +--- linux-3.0.9/fs/block_dev.c 2011-11-15 16:40:46.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/block_dev.c 2011-11-15 17:37:06.000000000 +0100 +@@ -25,6 +25,7 @@ + #include <linux/namei.h> + #include <linux/log2.h> + #include <linux/kmemleak.h> ++#include <linux/vs_device.h> + #include <asm/uaccess.h> + #include "internal.h" + +@@ -553,6 +554,7 @@ struct block_device *bdget(dev_t dev) + bdev->bd_invalidated = 0; + inode->i_mode = S_IFBLK; + inode->i_rdev = dev; ++ inode->i_mdev = dev; + inode->i_bdev = bdev; + inode->i_data.a_ops = &def_blk_aops; + mapping_set_gfp_mask(&inode->i_data, GFP_USER); +@@ -599,6 +601,11 @@ EXPORT_SYMBOL(bdput); + static struct block_device *bd_acquire(struct inode *inode) + { + struct block_device *bdev; ++ dev_t mdev; ++ ++ if (!vs_map_blkdev(inode->i_rdev, &mdev, DATTR_OPEN)) ++ return NULL; ++ inode->i_mdev = mdev; + + spin_lock(&bdev_lock); + bdev = inode->i_bdev; +@@ -609,7 +616,7 @@ static struct block_device *bd_acquire(s + } + spin_unlock(&bdev_lock); + +- bdev = bdget(inode->i_rdev); ++ bdev = bdget(mdev); + if (bdev) { + spin_lock(&bdev_lock); + if (!inode->i_bdev) { +diff -NurpP --minimal linux-3.0.9/fs/btrfs/ctree.h linux-3.0.9-vs2.3.2.1/fs/btrfs/ctree.h +--- linux-3.0.9/fs/btrfs/ctree.h 2011-07-22 11:18:05.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/btrfs/ctree.h 2011-07-19 00:44:39.000000000 +0200 +@@ -600,11 +600,14 @@ struct btrfs_inode_item { + /* modification sequence number for NFS */ + __le64 sequence; + ++ __le16 tag; + /* + * a little future expansion, for more than this we can + * just grow the inode item and version it + */ +- __le64 reserved[4]; ++ __le16 reserved16; ++ __le32 reserved32; ++ __le64 reserved[3]; + struct btrfs_timespec atime; + struct btrfs_timespec ctime; + struct btrfs_timespec mtime; +@@ -1359,6 +1362,8 @@ struct btrfs_ioctl_defrag_range_args { + #define BTRFS_MOUNT_AUTO_DEFRAG (1 << 16) + #define BTRFS_MOUNT_INODE_MAP_CACHE (1 << 17) + ++#define BTRFS_MOUNT_TAGGED (1 << 24) ++ + #define btrfs_clear_opt(o, opt) ((o) &= ~BTRFS_MOUNT_##opt) + #define btrfs_set_opt(o, opt) ((o) |= BTRFS_MOUNT_##opt) + #define btrfs_test_opt(root, opt) ((root)->fs_info->mount_opt & \ +@@ -1568,6 +1573,7 @@ BTRFS_SETGET_FUNCS(inode_block_group, st + BTRFS_SETGET_FUNCS(inode_nlink, struct btrfs_inode_item, nlink, 32); + BTRFS_SETGET_FUNCS(inode_uid, struct btrfs_inode_item, uid, 32); + BTRFS_SETGET_FUNCS(inode_gid, struct btrfs_inode_item, gid, 32); ++BTRFS_SETGET_FUNCS(inode_tag, struct btrfs_inode_item, tag, 16); + BTRFS_SETGET_FUNCS(inode_mode, struct btrfs_inode_item, mode, 32); + BTRFS_SETGET_FUNCS(inode_rdev, struct btrfs_inode_item, rdev, 64); + BTRFS_SETGET_FUNCS(inode_flags, struct btrfs_inode_item, flags, 64); +@@ -1621,6 +1627,10 @@ BTRFS_SETGET_FUNCS(extent_flags, struct + + BTRFS_SETGET_FUNCS(extent_refs_v0, struct btrfs_extent_item_v0, refs, 32); + ++#define BTRFS_INODE_IXUNLINK (1 << 24) ++#define BTRFS_INODE_BARRIER (1 << 25) ++#define BTRFS_INODE_COW (1 << 26) ++ + + BTRFS_SETGET_FUNCS(tree_block_level, struct btrfs_tree_block_info, level, 8); + +@@ -2595,6 +2605,7 @@ extern const struct dentry_operations bt + long btrfs_ioctl(struct file *file, unsigned int cmd, unsigned long arg); + void btrfs_update_iflags(struct inode *inode); + void btrfs_inherit_iflags(struct inode *inode, struct inode *dir); ++int btrfs_sync_flags(struct inode *inode, int, int); + int btrfs_defrag_file(struct inode *inode, struct file *file, + struct btrfs_ioctl_defrag_range_args *range, + u64 newer_than, unsigned long max_pages); +diff -NurpP --minimal linux-3.0.9/fs/btrfs/disk-io.c linux-3.0.9-vs2.3.2.1/fs/btrfs/disk-io.c +--- linux-3.0.9/fs/btrfs/disk-io.c 2011-07-22 11:18:05.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/btrfs/disk-io.c 2011-06-22 12:39:15.000000000 +0200 +@@ -1765,6 +1765,9 @@ struct btrfs_root *open_ctree(struct sup + goto fail_alloc; + } + ++ if (btrfs_test_opt(tree_root, TAGGED)) ++ sb->s_flags |= MS_TAGGED; ++ + features = btrfs_super_incompat_flags(disk_super) & + ~BTRFS_FEATURE_INCOMPAT_SUPP; + if (features) { +diff -NurpP --minimal linux-3.0.9/fs/btrfs/inode.c linux-3.0.9-vs2.3.2.1/fs/btrfs/inode.c +--- linux-3.0.9/fs/btrfs/inode.c 2011-11-15 16:40:46.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/btrfs/inode.c 2011-10-18 13:51:13.000000000 +0200 +@@ -38,6 +38,7 @@ + #include <linux/falloc.h> + #include <linux/slab.h> + #include <linux/ratelimit.h> ++#include <linux/vs_tag.h> + #include "compat.h" + #include "ctree.h" + #include "disk-io.h" +@@ -2508,6 +2509,8 @@ static void btrfs_read_locked_inode(stru + struct btrfs_key location; + int maybe_acls; + u32 rdev; ++ uid_t uid; ++ gid_t gid; + int ret; + bool filled = false; + +@@ -2540,8 +2543,13 @@ static void btrfs_read_locked_inode(stru + + inode->i_mode = btrfs_inode_mode(leaf, inode_item); + inode->i_nlink = btrfs_inode_nlink(leaf, inode_item); +- inode->i_uid = btrfs_inode_uid(leaf, inode_item); +- inode->i_gid = btrfs_inode_gid(leaf, inode_item); ++ ++ uid = btrfs_inode_uid(leaf, inode_item); ++ gid = btrfs_inode_gid(leaf, inode_item); ++ inode->i_uid = INOTAG_UID(DX_TAG(inode), uid, gid); ++ inode->i_gid = INOTAG_GID(DX_TAG(inode), uid, gid); ++ inode->i_tag = INOTAG_TAG(DX_TAG(inode), uid, gid, ++ btrfs_inode_tag(leaf, inode_item)); + btrfs_i_size_write(inode, btrfs_inode_size(leaf, inode_item)); + + tspec = btrfs_inode_atime(inode_item); +@@ -2624,6 +2632,9 @@ static void fill_inode_item(struct btrfs + struct btrfs_inode_item *item, + struct inode *inode) + { ++ uid_t uid = TAGINO_UID(DX_TAG(inode), inode->i_uid, inode->i_tag); ++ gid_t gid = TAGINO_GID(DX_TAG(inode), inode->i_gid, inode->i_tag); ++ + if (!leaf->map_token) + map_private_extent_buffer(leaf, (unsigned long)item, + sizeof(struct btrfs_inode_item), +@@ -2631,8 +2642,11 @@ static void fill_inode_item(struct btrfs + &leaf->map_start, &leaf->map_len, + KM_USER1); + +- btrfs_set_inode_uid(leaf, item, inode->i_uid); +- btrfs_set_inode_gid(leaf, item, inode->i_gid); ++ btrfs_set_inode_uid(leaf, item, uid); ++ btrfs_set_inode_gid(leaf, item, gid); ++#ifdef CONFIG_TAGGING_INTERN ++ btrfs_set_inode_tag(leaf, item, inode->i_tag); ++#endif + btrfs_set_inode_size(leaf, item, BTRFS_I(inode)->disk_i_size); + btrfs_set_inode_mode(leaf, item, inode->i_mode); + btrfs_set_inode_nlink(leaf, item, inode->i_nlink); +@@ -7360,6 +7374,7 @@ static const struct inode_operations btr + .listxattr = btrfs_listxattr, + .removexattr = btrfs_removexattr, + .permission = btrfs_permission, ++ .sync_flags = btrfs_sync_flags, + }; + static const struct inode_operations btrfs_dir_ro_inode_operations = { + .lookup = btrfs_lookup, +@@ -7432,6 +7447,7 @@ static const struct inode_operations btr + .removexattr = btrfs_removexattr, + .permission = btrfs_permission, + .fiemap = btrfs_fiemap, ++ .sync_flags = btrfs_sync_flags, + }; + static const struct inode_operations btrfs_special_inode_operations = { + .getattr = btrfs_getattr, +diff -NurpP --minimal linux-3.0.9/fs/btrfs/ioctl.c linux-3.0.9-vs2.3.2.1/fs/btrfs/ioctl.c +--- linux-3.0.9/fs/btrfs/ioctl.c 2011-07-22 11:18:05.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/btrfs/ioctl.c 2011-06-22 12:39:15.000000000 +0200 +@@ -70,10 +70,13 @@ static unsigned int btrfs_flags_to_ioctl + { + unsigned int iflags = 0; + +- if (flags & BTRFS_INODE_SYNC) +- iflags |= FS_SYNC_FL; + if (flags & BTRFS_INODE_IMMUTABLE) + iflags |= FS_IMMUTABLE_FL; ++ if (flags & BTRFS_INODE_IXUNLINK) ++ iflags |= FS_IXUNLINK_FL; ++ ++ if (flags & BTRFS_INODE_SYNC) ++ iflags |= FS_SYNC_FL; + if (flags & BTRFS_INODE_APPEND) + iflags |= FS_APPEND_FL; + if (flags & BTRFS_INODE_NODUMP) +@@ -90,28 +93,78 @@ static unsigned int btrfs_flags_to_ioctl + else if (flags & BTRFS_INODE_NOCOMPRESS) + iflags |= FS_NOCOMP_FL; + ++ if (flags & BTRFS_INODE_BARRIER) ++ iflags |= FS_BARRIER_FL; ++ if (flags & BTRFS_INODE_COW) ++ iflags |= FS_COW_FL; + return iflags; + } + + /* +- * Update inode->i_flags based on the btrfs internal flags. ++ * Update inode->i_(v)flags based on the btrfs internal flags. + */ + void btrfs_update_iflags(struct inode *inode) + { + struct btrfs_inode *ip = BTRFS_I(inode); + +- inode->i_flags &= ~(S_SYNC|S_APPEND|S_IMMUTABLE|S_NOATIME|S_DIRSYNC); ++ inode->i_flags &= ~(S_IMMUTABLE | S_IXUNLINK | ++ S_SYNC | S_APPEND | S_NOATIME | S_DIRSYNC); + +- if (ip->flags & BTRFS_INODE_SYNC) +- inode->i_flags |= S_SYNC; + if (ip->flags & BTRFS_INODE_IMMUTABLE) + inode->i_flags |= S_IMMUTABLE; ++ if (ip->flags & BTRFS_INODE_IXUNLINK) ++ inode->i_flags |= S_IXUNLINK; ++ ++ if (ip->flags & BTRFS_INODE_SYNC) ++ inode->i_flags |= S_SYNC; + if (ip->flags & BTRFS_INODE_APPEND) + inode->i_flags |= S_APPEND; + if (ip->flags & BTRFS_INODE_NOATIME) + inode->i_flags |= S_NOATIME; + if (ip->flags & BTRFS_INODE_DIRSYNC) + inode->i_flags |= S_DIRSYNC; ++ ++ inode->i_vflags &= ~(V_BARRIER | V_COW); ++ ++ if (ip->flags & BTRFS_INODE_BARRIER) ++ inode->i_vflags |= V_BARRIER; ++ if (ip->flags & BTRFS_INODE_COW) ++ inode->i_vflags |= V_COW; ++} ++ ++/* ++ * Update btrfs internal flags from inode->i_(v)flags. ++ */ ++void btrfs_update_flags(struct inode *inode) ++{ ++ struct btrfs_inode *ip = BTRFS_I(inode); ++ ++ unsigned int flags = inode->i_flags; ++ unsigned int vflags = inode->i_vflags; ++ ++ ip->flags &= ~(BTRFS_INODE_SYNC | BTRFS_INODE_APPEND | ++ BTRFS_INODE_IMMUTABLE | BTRFS_INODE_IXUNLINK | ++ BTRFS_INODE_NOATIME | BTRFS_INODE_DIRSYNC | ++ BTRFS_INODE_BARRIER | BTRFS_INODE_COW); ++ ++ if (flags & S_IMMUTABLE) ++ ip->flags |= BTRFS_INODE_IMMUTABLE; ++ if (flags & S_IXUNLINK) ++ ip->flags |= BTRFS_INODE_IXUNLINK; ++ ++ if (flags & S_SYNC) ++ ip->flags |= BTRFS_INODE_SYNC; ++ if (flags & S_APPEND) ++ ip->flags |= BTRFS_INODE_APPEND; ++ if (flags & S_NOATIME) ++ ip->flags |= BTRFS_INODE_NOATIME; ++ if (flags & S_DIRSYNC) ++ ip->flags |= BTRFS_INODE_DIRSYNC; ++ ++ if (vflags & V_BARRIER) ++ ip->flags |= BTRFS_INODE_BARRIER; ++ if (vflags & V_COW) ++ ip->flags |= BTRFS_INODE_COW; + } + + /* +@@ -129,7 +182,7 @@ void btrfs_inherit_iflags(struct inode * + flags = BTRFS_I(dir)->flags; + + if (S_ISREG(inode->i_mode)) +- flags &= ~BTRFS_INODE_DIRSYNC; ++ flags &= ~(BTRFS_INODE_DIRSYNC | BTRFS_INODE_BARRIER); + else if (!S_ISDIR(inode->i_mode)) + flags &= (BTRFS_INODE_NODUMP | BTRFS_INODE_NOATIME); + +@@ -137,6 +190,30 @@ void btrfs_inherit_iflags(struct inode * + btrfs_update_iflags(inode); + } + ++int btrfs_sync_flags(struct inode *inode, int flags, int vflags) ++{ ++ struct btrfs_inode *ip = BTRFS_I(inode); ++ struct btrfs_root *root = ip->root; ++ struct btrfs_trans_handle *trans; ++ int ret; ++ ++ trans = btrfs_join_transaction(root); ++ BUG_ON(!trans); ++ ++ inode->i_flags = flags; ++ inode->i_vflags = vflags; ++ btrfs_update_flags(inode); ++ ++ ret = btrfs_update_inode(trans, root, inode); ++ BUG_ON(ret); ++ ++ btrfs_update_iflags(inode); ++ inode->i_ctime = CURRENT_TIME; ++ btrfs_end_transaction(trans, root); ++ ++ return 0; ++} ++ + static int btrfs_ioctl_getflags(struct file *file, void __user *arg) + { + struct btrfs_inode *ip = BTRFS_I(file->f_path.dentry->d_inode); +@@ -188,7 +265,8 @@ static int btrfs_ioctl_setflags(struct f + + flags = btrfs_mask_flags(inode->i_mode, flags); + oldflags = btrfs_flags_to_ioctl(ip->flags); +- if ((flags ^ oldflags) & (FS_APPEND_FL | FS_IMMUTABLE_FL)) { ++ if ((flags ^ oldflags) & (FS_APPEND_FL | ++ FS_IMMUTABLE_FL | FS_IXUNLINK_FL)) { + if (!capable(CAP_LINUX_IMMUTABLE)) { + ret = -EPERM; + goto out_unlock; +@@ -199,14 +277,19 @@ static int btrfs_ioctl_setflags(struct f + if (ret) + goto out_unlock; + +- if (flags & FS_SYNC_FL) +- ip->flags |= BTRFS_INODE_SYNC; +- else +- ip->flags &= ~BTRFS_INODE_SYNC; + if (flags & FS_IMMUTABLE_FL) + ip->flags |= BTRFS_INODE_IMMUTABLE; + else + ip->flags &= ~BTRFS_INODE_IMMUTABLE; ++ if (flags & FS_IXUNLINK_FL) ++ ip->flags |= BTRFS_INODE_IXUNLINK; ++ else ++ ip->flags &= ~BTRFS_INODE_IXUNLINK; ++ ++ if (flags & FS_SYNC_FL) ++ ip->flags |= BTRFS_INODE_SYNC; ++ else ++ ip->flags &= ~BTRFS_INODE_SYNC; + if (flags & FS_APPEND_FL) + ip->flags |= BTRFS_INODE_APPEND; + else +diff -NurpP --minimal linux-3.0.9/fs/btrfs/super.c linux-3.0.9-vs2.3.2.1/fs/btrfs/super.c +--- linux-3.0.9/fs/btrfs/super.c 2011-07-22 11:18:05.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/btrfs/super.c 2011-07-19 00:44:39.000000000 +0200 +@@ -162,7 +162,7 @@ enum { + Opt_notreelog, Opt_ratio, Opt_flushoncommit, Opt_discard, + Opt_space_cache, Opt_clear_cache, Opt_user_subvol_rm_allowed, + Opt_enospc_debug, Opt_subvolrootid, Opt_defrag, +- Opt_inode_cache, Opt_err, ++ Opt_inode_cache, Opt_tag, Opt_notag, Opt_tagid, Opt_err, + }; + + static match_table_t tokens = { +@@ -195,6 +195,9 @@ static match_table_t tokens = { + {Opt_subvolrootid, "subvolrootid=%d"}, + {Opt_defrag, "autodefrag"}, + {Opt_inode_cache, "inode_cache"}, ++ {Opt_tag, "tag"}, ++ {Opt_notag, "notag"}, ++ {Opt_tagid, "tagid=%u"}, + {Opt_err, NULL}, + }; + +@@ -381,6 +384,22 @@ int btrfs_parse_options(struct btrfs_roo + printk(KERN_INFO "btrfs: enabling auto defrag"); + btrfs_set_opt(info->mount_opt, AUTO_DEFRAG); + break; ++#ifndef CONFIG_TAGGING_NONE ++ case Opt_tag: ++ printk(KERN_INFO "btrfs: use tagging\n"); ++ btrfs_set_opt(info->mount_opt, TAGGED); ++ break; ++ case Opt_notag: ++ printk(KERN_INFO "btrfs: disabled tagging\n"); ++ btrfs_clear_opt(info->mount_opt, TAGGED); ++ break; ++#endif ++#ifdef CONFIG_PROPAGATE ++ case Opt_tagid: ++ /* use args[0] */ ++ btrfs_set_opt(info->mount_opt, TAGGED); ++ break; ++#endif + case Opt_err: + printk(KERN_INFO "btrfs: unrecognized mount option " + "'%s'\n", p); +@@ -907,6 +926,12 @@ static int btrfs_remount(struct super_bl + if (ret) + return -EINVAL; + ++ if (btrfs_test_opt(root, TAGGED) && !(sb->s_flags & MS_TAGGED)) { ++ printk("btrfs: %s: tagging not permitted on remount.\n", ++ sb->s_id); ++ return -EINVAL; ++ } ++ + if ((*flags & MS_RDONLY) == (sb->s_flags & MS_RDONLY)) + return 0; + +diff -NurpP --minimal linux-3.0.9/fs/char_dev.c linux-3.0.9-vs2.3.2.1/fs/char_dev.c +--- linux-3.0.9/fs/char_dev.c 2011-03-15 18:07:31.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/char_dev.c 2011-06-10 22:11:24.000000000 +0200 +@@ -21,6 +21,8 @@ + #include <linux/mutex.h> + #include <linux/backing-dev.h> + #include <linux/tty.h> ++#include <linux/vs_context.h> ++#include <linux/vs_device.h> + + #include "internal.h" + +@@ -371,14 +373,21 @@ static int chrdev_open(struct inode *ino + struct cdev *p; + struct cdev *new = NULL; + int ret = 0; ++ dev_t mdev; ++ ++ if (!vs_map_chrdev(inode->i_rdev, &mdev, DATTR_OPEN)) ++ return -EPERM; ++ inode->i_mdev = mdev; + + spin_lock(&cdev_lock); + p = inode->i_cdev; + if (!p) { + struct kobject *kobj; + int idx; ++ + spin_unlock(&cdev_lock); +- kobj = kobj_lookup(cdev_map, inode->i_rdev, &idx); ++ ++ kobj = kobj_lookup(cdev_map, mdev, &idx); + if (!kobj) + return -ENXIO; + new = container_of(kobj, struct cdev, kobj); +diff -NurpP --minimal linux-3.0.9/fs/dcache.c linux-3.0.9-vs2.3.2.1/fs/dcache.c +--- linux-3.0.9/fs/dcache.c 2011-07-22 11:18:05.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/dcache.c 2011-07-22 11:20:39.000000000 +0200 +@@ -36,6 +36,7 @@ + #include <linux/bit_spinlock.h> + #include <linux/rculist_bl.h> + #include <linux/prefetch.h> ++#include <linux/vs_limit.h> + #include "internal.h" + + /* +@@ -479,6 +480,8 @@ int d_invalidate(struct dentry * dentry) + spin_lock(&dentry->d_lock); + } + ++ vx_dentry_dec(dentry); ++ + /* + * Somebody else still using it? + * +@@ -506,6 +509,7 @@ EXPORT_SYMBOL(d_invalidate); + static inline void __dget_dlock(struct dentry *dentry) + { + dentry->d_count++; ++ vx_dentry_inc(dentry); + } + + static inline void __dget(struct dentry *dentry) +@@ -1266,6 +1270,9 @@ struct dentry *d_alloc(struct dentry * p + struct dentry *dentry; + char *dname; + ++ if (!vx_dentry_avail(1)) ++ return NULL; ++ + dentry = kmem_cache_alloc(dentry_cache, GFP_KERNEL); + if (!dentry) + return NULL; +@@ -1288,6 +1295,7 @@ struct dentry *d_alloc(struct dentry * p + + dentry->d_count = 1; + dentry->d_flags = 0; ++ vx_dentry_inc(dentry); + spin_lock_init(&dentry->d_lock); + seqcount_init(&dentry->d_seq); + dentry->d_inode = NULL; +@@ -1945,6 +1953,7 @@ struct dentry *__d_lookup(struct dentry + } + + dentry->d_count++; ++ vx_dentry_inc(dentry); + found = dentry; + spin_unlock(&dentry->d_lock); + break; +diff -NurpP --minimal linux-3.0.9/fs/devpts/inode.c linux-3.0.9-vs2.3.2.1/fs/devpts/inode.c +--- linux-3.0.9/fs/devpts/inode.c 2011-05-22 16:17:50.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/devpts/inode.c 2011-06-10 22:11:24.000000000 +0200 +@@ -25,6 +25,7 @@ + #include <linux/parser.h> + #include <linux/fsnotify.h> + #include <linux/seq_file.h> ++#include <linux/vs_base.h> + + #define DEVPTS_DEFAULT_MODE 0600 + /* +@@ -36,6 +37,20 @@ + #define DEVPTS_DEFAULT_PTMX_MODE 0000 + #define PTMX_MINOR 2 + ++static int devpts_permission(struct inode *inode, int mask, unsigned int flags) ++{ ++ int ret = -EACCES; ++ ++ /* devpts is xid tagged */ ++ if (vx_check((xid_t)inode->i_tag, VS_WATCH_P | VS_IDENT)) ++ ret = generic_permission(inode, mask, flags, NULL); ++ return ret; ++} ++ ++static struct inode_operations devpts_file_inode_operations = { ++ .permission = devpts_permission, ++}; ++ + extern int pty_limit; /* Config limit on Unix98 ptys */ + static DEFINE_MUTEX(allocated_ptys_lock); + +@@ -263,6 +278,34 @@ static int devpts_show_options(struct se + return 0; + } + ++static int devpts_filter(struct dentry *de) ++{ ++ xid_t xid = 0; ++ ++ /* devpts is xid tagged */ ++ if (de && de->d_inode) ++ xid = (xid_t)de->d_inode->i_tag; ++#ifdef CONFIG_VSERVER_WARN_DEVPTS ++ else ++ vxwprintk_task(1, "devpts " VS_Q("%.*s") " without inode.", ++ de->d_name.len, de->d_name.name); ++#endif ++ return vx_check(xid, VS_WATCH_P | VS_IDENT); ++} ++ ++static int devpts_readdir(struct file * filp, void * dirent, filldir_t filldir) ++{ ++ return dcache_readdir_filter(filp, dirent, filldir, devpts_filter); ++} ++ ++static struct file_operations devpts_dir_operations = { ++ .open = dcache_dir_open, ++ .release = dcache_dir_close, ++ .llseek = dcache_dir_lseek, ++ .read = generic_read_dir, ++ .readdir = devpts_readdir, ++}; ++ + static const struct super_operations devpts_sops = { + .statfs = simple_statfs, + .remount_fs = devpts_remount, +@@ -302,12 +345,15 @@ devpts_fill_super(struct super_block *s, + inode = new_inode(s); + if (!inode) + goto free_fsi; ++ + inode->i_ino = 1; + inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME; + inode->i_mode = S_IFDIR | S_IRUGO | S_IXUGO | S_IWUSR; + inode->i_op = &simple_dir_inode_operations; +- inode->i_fop = &simple_dir_operations; ++ inode->i_fop = &devpts_dir_operations; + inode->i_nlink = 2; ++ /* devpts is xid tagged */ ++ inode->i_tag = (tag_t)vx_current_xid(); + + s->s_root = d_alloc_root(inode); + if (s->s_root) +@@ -494,6 +540,9 @@ int devpts_pty_new(struct inode *ptmx_in + inode->i_gid = opts->setgid ? opts->gid : current_fsgid(); + inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME; + init_special_inode(inode, S_IFCHR|opts->mode, device); ++ /* devpts is xid tagged */ ++ inode->i_tag = (tag_t)vx_current_xid(); ++ inode->i_op = &devpts_file_inode_operations; + inode->i_private = tty; + tty->driver_data = inode; + +diff -NurpP --minimal linux-3.0.9/fs/ext2/balloc.c linux-3.0.9-vs2.3.2.1/fs/ext2/balloc.c +--- linux-3.0.9/fs/ext2/balloc.c 2011-05-22 16:17:51.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/ext2/balloc.c 2011-06-10 22:11:24.000000000 +0200 +@@ -701,7 +701,6 @@ ext2_try_to_allocate(struct super_block + start = 0; + end = EXT2_BLOCKS_PER_GROUP(sb); + } +- + BUG_ON(start > EXT2_BLOCKS_PER_GROUP(sb)); + + repeat: +diff -NurpP --minimal linux-3.0.9/fs/ext2/ext2.h linux-3.0.9-vs2.3.2.1/fs/ext2/ext2.h +--- linux-3.0.9/fs/ext2/ext2.h 2011-05-22 16:17:51.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/ext2/ext2.h 2011-06-10 22:11:24.000000000 +0200 +@@ -126,6 +126,7 @@ extern void ext2_set_inode_flags(struct + extern void ext2_get_inode_flags(struct ext2_inode_info *); + extern int ext2_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, + u64 start, u64 len); ++extern int ext2_sync_flags(struct inode *, int, int); + + /* ioctl.c */ + extern long ext2_ioctl(struct file *, unsigned int, unsigned long); +diff -NurpP --minimal linux-3.0.9/fs/ext2/file.c linux-3.0.9-vs2.3.2.1/fs/ext2/file.c +--- linux-3.0.9/fs/ext2/file.c 2010-08-02 16:52:48.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/ext2/file.c 2011-06-10 22:11:24.000000000 +0200 +@@ -104,4 +104,5 @@ const struct inode_operations ext2_file_ + .setattr = ext2_setattr, + .check_acl = ext2_check_acl, + .fiemap = ext2_fiemap, ++ .sync_flags = ext2_sync_flags, + }; +diff -NurpP --minimal linux-3.0.9/fs/ext2/ialloc.c linux-3.0.9-vs2.3.2.1/fs/ext2/ialloc.c +--- linux-3.0.9/fs/ext2/ialloc.c 2011-05-22 16:17:51.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/ext2/ialloc.c 2011-06-10 22:11:24.000000000 +0200 +@@ -17,6 +17,7 @@ + #include <linux/backing-dev.h> + #include <linux/buffer_head.h> + #include <linux/random.h> ++#include <linux/vs_tag.h> + #include "ext2.h" + #include "xattr.h" + #include "acl.h" +@@ -549,6 +550,7 @@ got: + inode->i_mode = mode; + inode->i_uid = current_fsuid(); + inode->i_gid = dir->i_gid; ++ inode->i_tag = dx_current_fstag(sb); + } else + inode_init_owner(inode, dir, mode); + +diff -NurpP --minimal linux-3.0.9/fs/ext2/inode.c linux-3.0.9-vs2.3.2.1/fs/ext2/inode.c +--- linux-3.0.9/fs/ext2/inode.c 2011-05-22 16:17:51.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/ext2/inode.c 2011-06-10 22:11:24.000000000 +0200 +@@ -32,6 +32,7 @@ + #include <linux/mpage.h> + #include <linux/fiemap.h> + #include <linux/namei.h> ++#include <linux/vs_tag.h> + #include "ext2.h" + #include "acl.h" + #include "xip.h" +@@ -1167,7 +1168,7 @@ static void ext2_truncate_blocks(struct + return; + if (ext2_inode_is_fast_symlink(inode)) + return; +- if (IS_APPEND(inode) || IS_IMMUTABLE(inode)) ++ if (IS_APPEND(inode) || IS_IXORUNLINK(inode)) + return; + __ext2_truncate_blocks(inode, offset); + } +@@ -1256,36 +1257,61 @@ void ext2_set_inode_flags(struct inode * + { + unsigned int flags = EXT2_I(inode)->i_flags; + +- inode->i_flags &= ~(S_SYNC|S_APPEND|S_IMMUTABLE|S_NOATIME|S_DIRSYNC); ++ inode->i_flags &= ~(S_IMMUTABLE | S_IXUNLINK | ++ S_SYNC | S_APPEND | S_NOATIME | S_DIRSYNC); ++ ++ ++ if (flags & EXT2_IMMUTABLE_FL) ++ inode->i_flags |= S_IMMUTABLE; ++ if (flags & EXT2_IXUNLINK_FL) ++ inode->i_flags |= S_IXUNLINK; ++ + if (flags & EXT2_SYNC_FL) + inode->i_flags |= S_SYNC; + if (flags & EXT2_APPEND_FL) + inode->i_flags |= S_APPEND; +- if (flags & EXT2_IMMUTABLE_FL) +- inode->i_flags |= S_IMMUTABLE; + if (flags & EXT2_NOATIME_FL) + inode->i_flags |= S_NOATIME; + if (flags & EXT2_DIRSYNC_FL) + inode->i_flags |= S_DIRSYNC; ++ ++ inode->i_vflags &= ~(V_BARRIER | V_COW); ++ ++ if (flags & EXT2_BARRIER_FL) ++ inode->i_vflags |= V_BARRIER; ++ if (flags & EXT2_COW_FL) ++ inode->i_vflags |= V_COW; + } + + /* Propagate flags from i_flags to EXT2_I(inode)->i_flags */ + void ext2_get_inode_flags(struct ext2_inode_info *ei) + { + unsigned int flags = ei->vfs_inode.i_flags; ++ unsigned int vflags = ei->vfs_inode.i_vflags; ++ ++ ei->i_flags &= ~(EXT2_SYNC_FL | EXT2_APPEND_FL | ++ EXT2_IMMUTABLE_FL | EXT2_IXUNLINK_FL | ++ EXT2_NOATIME_FL | EXT2_DIRSYNC_FL | ++ EXT2_BARRIER_FL | EXT2_COW_FL); ++ ++ if (flags & S_IMMUTABLE) ++ ei->i_flags |= EXT2_IMMUTABLE_FL; ++ if (flags & S_IXUNLINK) ++ ei->i_flags |= EXT2_IXUNLINK_FL; + +- ei->i_flags &= ~(EXT2_SYNC_FL|EXT2_APPEND_FL| +- EXT2_IMMUTABLE_FL|EXT2_NOATIME_FL|EXT2_DIRSYNC_FL); + if (flags & S_SYNC) + ei->i_flags |= EXT2_SYNC_FL; + if (flags & S_APPEND) + ei->i_flags |= EXT2_APPEND_FL; +- if (flags & S_IMMUTABLE) +- ei->i_flags |= EXT2_IMMUTABLE_FL; + if (flags & S_NOATIME) + ei->i_flags |= EXT2_NOATIME_FL; + if (flags & S_DIRSYNC) + ei->i_flags |= EXT2_DIRSYNC_FL; ++ ++ if (vflags & V_BARRIER) ++ ei->i_flags |= EXT2_BARRIER_FL; ++ if (vflags & V_COW) ++ ei->i_flags |= EXT2_COW_FL; + } + + struct inode *ext2_iget (struct super_block *sb, unsigned long ino) +@@ -1295,6 +1321,8 @@ struct inode *ext2_iget (struct super_bl + struct ext2_inode *raw_inode; + struct inode *inode; + long ret = -EIO; ++ uid_t uid; ++ gid_t gid; + int n; + + inode = iget_locked(sb, ino); +@@ -1313,12 +1341,17 @@ struct inode *ext2_iget (struct super_bl + } + + inode->i_mode = le16_to_cpu(raw_inode->i_mode); +- inode->i_uid = (uid_t)le16_to_cpu(raw_inode->i_uid_low); +- inode->i_gid = (gid_t)le16_to_cpu(raw_inode->i_gid_low); ++ uid = (uid_t)le16_to_cpu(raw_inode->i_uid_low); ++ gid = (gid_t)le16_to_cpu(raw_inode->i_gid_low); + if (!(test_opt (inode->i_sb, NO_UID32))) { +- inode->i_uid |= le16_to_cpu(raw_inode->i_uid_high) << 16; +- inode->i_gid |= le16_to_cpu(raw_inode->i_gid_high) << 16; ++ uid |= le16_to_cpu(raw_inode->i_uid_high) << 16; ++ gid |= le16_to_cpu(raw_inode->i_gid_high) << 16; + } ++ inode->i_uid = INOTAG_UID(DX_TAG(inode), uid, gid); ++ inode->i_gid = INOTAG_GID(DX_TAG(inode), uid, gid); ++ inode->i_tag = INOTAG_TAG(DX_TAG(inode), uid, gid, ++ le16_to_cpu(raw_inode->i_raw_tag)); ++ + inode->i_nlink = le16_to_cpu(raw_inode->i_links_count); + inode->i_size = le32_to_cpu(raw_inode->i_size); + inode->i_atime.tv_sec = (signed)le32_to_cpu(raw_inode->i_atime); +@@ -1416,8 +1449,8 @@ static int __ext2_write_inode(struct ino + struct ext2_inode_info *ei = EXT2_I(inode); + struct super_block *sb = inode->i_sb; + ino_t ino = inode->i_ino; +- uid_t uid = inode->i_uid; +- gid_t gid = inode->i_gid; ++ uid_t uid = TAGINO_UID(DX_TAG(inode), inode->i_uid, inode->i_tag); ++ gid_t gid = TAGINO_GID(DX_TAG(inode), inode->i_gid, inode->i_tag); + struct buffer_head * bh; + struct ext2_inode * raw_inode = ext2_get_inode(sb, ino, &bh); + int n; +@@ -1453,6 +1486,9 @@ static int __ext2_write_inode(struct ino + raw_inode->i_uid_high = 0; + raw_inode->i_gid_high = 0; + } ++#ifdef CONFIG_TAGGING_INTERN ++ raw_inode->i_raw_tag = cpu_to_le16(inode->i_tag); ++#endif + raw_inode->i_links_count = cpu_to_le16(inode->i_nlink); + raw_inode->i_size = cpu_to_le32(inode->i_size); + raw_inode->i_atime = cpu_to_le32(inode->i_atime.tv_sec); +@@ -1533,7 +1569,8 @@ int ext2_setattr(struct dentry *dentry, + if (is_quota_modification(inode, iattr)) + dquot_initialize(inode); + if ((iattr->ia_valid & ATTR_UID && iattr->ia_uid != inode->i_uid) || +- (iattr->ia_valid & ATTR_GID && iattr->ia_gid != inode->i_gid)) { ++ (iattr->ia_valid & ATTR_GID && iattr->ia_gid != inode->i_gid) || ++ (iattr->ia_valid & ATTR_TAG && iattr->ia_tag != inode->i_tag)) { + error = dquot_transfer(inode, iattr); + if (error) + return error; +diff -NurpP --minimal linux-3.0.9/fs/ext2/ioctl.c linux-3.0.9-vs2.3.2.1/fs/ext2/ioctl.c +--- linux-3.0.9/fs/ext2/ioctl.c 2011-05-22 16:17:51.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/ext2/ioctl.c 2011-06-10 22:11:24.000000000 +0200 +@@ -17,6 +17,16 @@ + #include <asm/uaccess.h> + + ++int ext2_sync_flags(struct inode *inode, int flags, int vflags) ++{ ++ inode->i_flags = flags; ++ inode->i_vflags = vflags; ++ ext2_get_inode_flags(EXT2_I(inode)); ++ inode->i_ctime = CURRENT_TIME_SEC; ++ mark_inode_dirty(inode); ++ return 0; ++} ++ + long ext2_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) + { + struct inode *inode = filp->f_dentry->d_inode; +@@ -51,6 +61,11 @@ long ext2_ioctl(struct file *filp, unsig + + flags = ext2_mask_flags(inode->i_mode, flags); + ++ if (IS_BARRIER(inode)) { ++ vxwprintk_task(1, "messing with the barrier."); ++ return -EACCES; ++ } ++ + mutex_lock(&inode->i_mutex); + /* Is it quota file? Do not allow user to mess with it */ + if (IS_NOQUOTA(inode)) { +@@ -66,7 +81,9 @@ long ext2_ioctl(struct file *filp, unsig + * + * This test looks nicer. Thanks to Pauline Middelink + */ +- if ((flags ^ oldflags) & (EXT2_APPEND_FL | EXT2_IMMUTABLE_FL)) { ++ if ((oldflags & EXT2_IMMUTABLE_FL) || ++ ((flags ^ oldflags) & (EXT2_APPEND_FL | ++ EXT2_IMMUTABLE_FL | EXT2_IXUNLINK_FL))) { + if (!capable(CAP_LINUX_IMMUTABLE)) { + mutex_unlock(&inode->i_mutex); + ret = -EPERM; +@@ -74,7 +91,7 @@ long ext2_ioctl(struct file *filp, unsig + } + } + +- flags = flags & EXT2_FL_USER_MODIFIABLE; ++ flags &= EXT2_FL_USER_MODIFIABLE; + flags |= oldflags & ~EXT2_FL_USER_MODIFIABLE; + ei->i_flags = flags; + mutex_unlock(&inode->i_mutex); +diff -NurpP --minimal linux-3.0.9/fs/ext2/namei.c linux-3.0.9-vs2.3.2.1/fs/ext2/namei.c +--- linux-3.0.9/fs/ext2/namei.c 2011-05-22 16:17:51.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/ext2/namei.c 2011-06-10 22:11:24.000000000 +0200 +@@ -32,6 +32,7 @@ + + #include <linux/pagemap.h> + #include <linux/quotaops.h> ++#include <linux/vs_tag.h> + #include "ext2.h" + #include "xattr.h" + #include "acl.h" +@@ -75,6 +76,7 @@ static struct dentry *ext2_lookup(struct + return ERR_PTR(-EIO); + } else { + return ERR_CAST(inode); ++ dx_propagate_tag(nd, inode); + } + } + } +@@ -413,6 +415,7 @@ const struct inode_operations ext2_dir_i + #endif + .setattr = ext2_setattr, + .check_acl = ext2_check_acl, ++ .sync_flags = ext2_sync_flags, + }; + + const struct inode_operations ext2_special_inode_operations = { +diff -NurpP --minimal linux-3.0.9/fs/ext2/super.c linux-3.0.9-vs2.3.2.1/fs/ext2/super.c +--- linux-3.0.9/fs/ext2/super.c 2011-07-22 11:18:05.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/ext2/super.c 2011-06-10 22:11:24.000000000 +0200 +@@ -394,7 +394,8 @@ enum { + Opt_err_ro, Opt_nouid32, Opt_nocheck, Opt_debug, + Opt_oldalloc, Opt_orlov, Opt_nobh, Opt_user_xattr, Opt_nouser_xattr, + Opt_acl, Opt_noacl, Opt_xip, Opt_ignore, Opt_err, Opt_quota, +- Opt_usrquota, Opt_grpquota, Opt_reservation, Opt_noreservation ++ Opt_usrquota, Opt_grpquota, Opt_reservation, Opt_noreservation, ++ Opt_tag, Opt_notag, Opt_tagid + }; + + static const match_table_t tokens = { +@@ -422,6 +423,9 @@ static const match_table_t tokens = { + {Opt_acl, "acl"}, + {Opt_noacl, "noacl"}, + {Opt_xip, "xip"}, ++ {Opt_tag, "tag"}, ++ {Opt_notag, "notag"}, ++ {Opt_tagid, "tagid=%u"}, + {Opt_grpquota, "grpquota"}, + {Opt_ignore, "noquota"}, + {Opt_quota, "quota"}, +@@ -492,6 +496,20 @@ static int parse_options(char *options, + case Opt_nouid32: + set_opt (sbi->s_mount_opt, NO_UID32); + break; ++#ifndef CONFIG_TAGGING_NONE ++ case Opt_tag: ++ set_opt (sbi->s_mount_opt, TAGGED); ++ break; ++ case Opt_notag: ++ clear_opt (sbi->s_mount_opt, TAGGED); ++ break; ++#endif ++#ifdef CONFIG_PROPAGATE ++ case Opt_tagid: ++ /* use args[0] */ ++ set_opt (sbi->s_mount_opt, TAGGED); ++ break; ++#endif + case Opt_nocheck: + clear_opt (sbi->s_mount_opt, CHECK); + break; +@@ -850,6 +868,8 @@ static int ext2_fill_super(struct super_ + if (!parse_options((char *) data, sb)) + goto failed_mount; + ++ if (EXT2_SB(sb)->s_mount_opt & EXT2_MOUNT_TAGGED) ++ sb->s_flags |= MS_TAGGED; + sb->s_flags = (sb->s_flags & ~MS_POSIXACL) | + ((EXT2_SB(sb)->s_mount_opt & EXT2_MOUNT_POSIX_ACL) ? + MS_POSIXACL : 0); +@@ -1224,6 +1244,14 @@ static int ext2_remount (struct super_bl + goto restore_opts; + } + ++ if ((sbi->s_mount_opt & EXT2_MOUNT_TAGGED) && ++ !(sb->s_flags & MS_TAGGED)) { ++ printk("EXT2-fs: %s: tagging not permitted on remount.\n", ++ sb->s_id); ++ err = -EINVAL; ++ goto restore_opts; ++ } ++ + sb->s_flags = (sb->s_flags & ~MS_POSIXACL) | + ((sbi->s_mount_opt & EXT2_MOUNT_POSIX_ACL) ? MS_POSIXACL : 0); + +diff -NurpP --minimal linux-3.0.9/fs/ext3/file.c linux-3.0.9-vs2.3.2.1/fs/ext3/file.c +--- linux-3.0.9/fs/ext3/file.c 2010-07-07 18:31:51.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/ext3/file.c 2011-06-10 22:11:24.000000000 +0200 +@@ -81,5 +81,6 @@ const struct inode_operations ext3_file_ + #endif + .check_acl = ext3_check_acl, + .fiemap = ext3_fiemap, ++ .sync_flags = ext3_sync_flags, + }; + +diff -NurpP --minimal linux-3.0.9/fs/ext3/ialloc.c linux-3.0.9-vs2.3.2.1/fs/ext3/ialloc.c +--- linux-3.0.9/fs/ext3/ialloc.c 2011-05-22 16:17:52.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/ext3/ialloc.c 2011-06-10 22:11:24.000000000 +0200 +@@ -23,6 +23,7 @@ + #include <linux/buffer_head.h> + #include <linux/random.h> + #include <linux/bitops.h> ++#include <linux/vs_tag.h> + + #include <asm/byteorder.h> + +@@ -532,6 +533,7 @@ got: + inode->i_mode = mode; + inode->i_uid = current_fsuid(); + inode->i_gid = dir->i_gid; ++ inode->i_tag = dx_current_fstag(sb); + } else + inode_init_owner(inode, dir, mode); + +diff -NurpP --minimal linux-3.0.9/fs/ext3/inode.c linux-3.0.9-vs2.3.2.1/fs/ext3/inode.c +--- linux-3.0.9/fs/ext3/inode.c 2011-07-22 11:18:05.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/ext3/inode.c 2011-06-10 22:11:24.000000000 +0200 +@@ -38,6 +38,7 @@ + #include <linux/bio.h> + #include <linux/fiemap.h> + #include <linux/namei.h> ++#include <linux/vs_tag.h> + #include "xattr.h" + #include "acl.h" + +@@ -2391,7 +2392,7 @@ static void ext3_free_branches(handle_t + + int ext3_can_truncate(struct inode *inode) + { +- if (IS_APPEND(inode) || IS_IMMUTABLE(inode)) ++ if (IS_APPEND(inode) || IS_IXORUNLINK(inode)) + return 0; + if (S_ISREG(inode->i_mode)) + return 1; +@@ -2775,36 +2776,60 @@ void ext3_set_inode_flags(struct inode * + { + unsigned int flags = EXT3_I(inode)->i_flags; + +- inode->i_flags &= ~(S_SYNC|S_APPEND|S_IMMUTABLE|S_NOATIME|S_DIRSYNC); ++ inode->i_flags &= ~(S_IMMUTABLE | S_IXUNLINK | ++ S_SYNC | S_APPEND | S_NOATIME | S_DIRSYNC); ++ ++ if (flags & EXT3_IMMUTABLE_FL) ++ inode->i_flags |= S_IMMUTABLE; ++ if (flags & EXT3_IXUNLINK_FL) ++ inode->i_flags |= S_IXUNLINK; ++ + if (flags & EXT3_SYNC_FL) + inode->i_flags |= S_SYNC; + if (flags & EXT3_APPEND_FL) + inode->i_flags |= S_APPEND; +- if (flags & EXT3_IMMUTABLE_FL) +- inode->i_flags |= S_IMMUTABLE; + if (flags & EXT3_NOATIME_FL) + inode->i_flags |= S_NOATIME; + if (flags & EXT3_DIRSYNC_FL) + inode->i_flags |= S_DIRSYNC; ++ ++ inode->i_vflags &= ~(V_BARRIER | V_COW); ++ ++ if (flags & EXT3_BARRIER_FL) ++ inode->i_vflags |= V_BARRIER; ++ if (flags & EXT3_COW_FL) ++ inode->i_vflags |= V_COW; + } + + /* Propagate flags from i_flags to EXT3_I(inode)->i_flags */ + void ext3_get_inode_flags(struct ext3_inode_info *ei) + { + unsigned int flags = ei->vfs_inode.i_flags; ++ unsigned int vflags = ei->vfs_inode.i_vflags; ++ ++ ei->i_flags &= ~(EXT3_SYNC_FL | EXT3_APPEND_FL | ++ EXT3_IMMUTABLE_FL | EXT3_IXUNLINK_FL | ++ EXT3_NOATIME_FL | EXT3_DIRSYNC_FL | ++ EXT3_BARRIER_FL | EXT3_COW_FL); ++ ++ if (flags & S_IMMUTABLE) ++ ei->i_flags |= EXT3_IMMUTABLE_FL; ++ if (flags & S_IXUNLINK) ++ ei->i_flags |= EXT3_IXUNLINK_FL; + +- ei->i_flags &= ~(EXT3_SYNC_FL|EXT3_APPEND_FL| +- EXT3_IMMUTABLE_FL|EXT3_NOATIME_FL|EXT3_DIRSYNC_FL); + if (flags & S_SYNC) + ei->i_flags |= EXT3_SYNC_FL; + if (flags & S_APPEND) + ei->i_flags |= EXT3_APPEND_FL; +- if (flags & S_IMMUTABLE) +- ei->i_flags |= EXT3_IMMUTABLE_FL; + if (flags & S_NOATIME) + ei->i_flags |= EXT3_NOATIME_FL; + if (flags & S_DIRSYNC) + ei->i_flags |= EXT3_DIRSYNC_FL; ++ ++ if (vflags & V_BARRIER) ++ ei->i_flags |= EXT3_BARRIER_FL; ++ if (vflags & V_COW) ++ ei->i_flags |= EXT3_COW_FL; + } + + struct inode *ext3_iget(struct super_block *sb, unsigned long ino) +@@ -2818,6 +2843,8 @@ struct inode *ext3_iget(struct super_blo + transaction_t *transaction; + long ret; + int block; ++ uid_t uid; ++ gid_t gid; + + inode = iget_locked(sb, ino); + if (!inode) +@@ -2834,12 +2861,17 @@ struct inode *ext3_iget(struct super_blo + bh = iloc.bh; + raw_inode = ext3_raw_inode(&iloc); + inode->i_mode = le16_to_cpu(raw_inode->i_mode); +- inode->i_uid = (uid_t)le16_to_cpu(raw_inode->i_uid_low); +- inode->i_gid = (gid_t)le16_to_cpu(raw_inode->i_gid_low); ++ uid = (uid_t)le16_to_cpu(raw_inode->i_uid_low); ++ gid = (gid_t)le16_to_cpu(raw_inode->i_gid_low); + if(!(test_opt (inode->i_sb, NO_UID32))) { +- inode->i_uid |= le16_to_cpu(raw_inode->i_uid_high) << 16; +- inode->i_gid |= le16_to_cpu(raw_inode->i_gid_high) << 16; ++ uid |= le16_to_cpu(raw_inode->i_uid_high) << 16; ++ gid |= le16_to_cpu(raw_inode->i_gid_high) << 16; + } ++ inode->i_uid = INOTAG_UID(DX_TAG(inode), uid, gid); ++ inode->i_gid = INOTAG_GID(DX_TAG(inode), uid, gid); ++ inode->i_tag = INOTAG_TAG(DX_TAG(inode), uid, gid, ++ le16_to_cpu(raw_inode->i_raw_tag)); ++ + inode->i_nlink = le16_to_cpu(raw_inode->i_links_count); + inode->i_size = le32_to_cpu(raw_inode->i_size); + inode->i_atime.tv_sec = (signed)le32_to_cpu(raw_inode->i_atime); +@@ -2994,6 +3026,8 @@ static int ext3_do_update_inode(handle_t + struct ext3_inode *raw_inode = ext3_raw_inode(iloc); + struct ext3_inode_info *ei = EXT3_I(inode); + struct buffer_head *bh = iloc->bh; ++ uid_t uid = TAGINO_UID(DX_TAG(inode), inode->i_uid, inode->i_tag); ++ gid_t gid = TAGINO_GID(DX_TAG(inode), inode->i_gid, inode->i_tag); + int err = 0, rc, block; + + again: +@@ -3008,29 +3042,32 @@ again: + ext3_get_inode_flags(ei); + raw_inode->i_mode = cpu_to_le16(inode->i_mode); + if(!(test_opt(inode->i_sb, NO_UID32))) { +- raw_inode->i_uid_low = cpu_to_le16(low_16_bits(inode->i_uid)); +- raw_inode->i_gid_low = cpu_to_le16(low_16_bits(inode->i_gid)); ++ raw_inode->i_uid_low = cpu_to_le16(low_16_bits(uid)); ++ raw_inode->i_gid_low = cpu_to_le16(low_16_bits(gid)); + /* + * Fix up interoperability with old kernels. Otherwise, old inodes get + * re-used with the upper 16 bits of the uid/gid intact + */ + if(!ei->i_dtime) { + raw_inode->i_uid_high = +- cpu_to_le16(high_16_bits(inode->i_uid)); ++ cpu_to_le16(high_16_bits(uid)); + raw_inode->i_gid_high = +- cpu_to_le16(high_16_bits(inode->i_gid)); ++ cpu_to_le16(high_16_bits(gid)); + } else { + raw_inode->i_uid_high = 0; + raw_inode->i_gid_high = 0; + } + } else { + raw_inode->i_uid_low = +- cpu_to_le16(fs_high2lowuid(inode->i_uid)); ++ cpu_to_le16(fs_high2lowuid(uid)); + raw_inode->i_gid_low = +- cpu_to_le16(fs_high2lowgid(inode->i_gid)); ++ cpu_to_le16(fs_high2lowgid(gid)); + raw_inode->i_uid_high = 0; + raw_inode->i_gid_high = 0; + } ++#ifdef CONFIG_TAGGING_INTERN ++ raw_inode->i_raw_tag = cpu_to_le16(inode->i_tag); ++#endif + raw_inode->i_links_count = cpu_to_le16(inode->i_nlink); + raw_inode->i_size = cpu_to_le32(ei->i_disksize); + raw_inode->i_atime = cpu_to_le32(inode->i_atime.tv_sec); +@@ -3190,7 +3227,8 @@ int ext3_setattr(struct dentry *dentry, + if (is_quota_modification(inode, attr)) + dquot_initialize(inode); + if ((ia_valid & ATTR_UID && attr->ia_uid != inode->i_uid) || +- (ia_valid & ATTR_GID && attr->ia_gid != inode->i_gid)) { ++ (ia_valid & ATTR_GID && attr->ia_gid != inode->i_gid) || ++ (ia_valid & ATTR_TAG && attr->ia_tag != inode->i_tag)) { + handle_t *handle; + + /* (user+group)*(old+new) structure, inode write (sb, +@@ -3212,6 +3250,8 @@ int ext3_setattr(struct dentry *dentry, + inode->i_uid = attr->ia_uid; + if (attr->ia_valid & ATTR_GID) + inode->i_gid = attr->ia_gid; ++ if ((attr->ia_valid & ATTR_TAG) && IS_TAGGED(inode)) ++ inode->i_tag = attr->ia_tag; + error = ext3_mark_inode_dirty(handle, inode); + ext3_journal_stop(handle); + } +diff -NurpP --minimal linux-3.0.9/fs/ext3/ioctl.c linux-3.0.9-vs2.3.2.1/fs/ext3/ioctl.c +--- linux-3.0.9/fs/ext3/ioctl.c 2011-05-22 16:17:52.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/ext3/ioctl.c 2011-06-10 22:11:24.000000000 +0200 +@@ -8,6 +8,7 @@ + */ + + #include <linux/fs.h> ++#include <linux/mount.h> + #include <linux/jbd.h> + #include <linux/capability.h> + #include <linux/ext3_fs.h> +@@ -17,6 +18,34 @@ + #include <linux/compat.h> + #include <asm/uaccess.h> + ++ ++int ext3_sync_flags(struct inode *inode, int flags, int vflags) ++{ ++ handle_t *handle = NULL; ++ struct ext3_iloc iloc; ++ int err; ++ ++ handle = ext3_journal_start(inode, 1); ++ if (IS_ERR(handle)) ++ return PTR_ERR(handle); ++ ++ if (IS_SYNC(inode)) ++ handle->h_sync = 1; ++ err = ext3_reserve_inode_write(handle, inode, &iloc); ++ if (err) ++ goto flags_err; ++ ++ inode->i_flags = flags; ++ inode->i_vflags = vflags; ++ ext3_get_inode_flags(EXT3_I(inode)); ++ inode->i_ctime = CURRENT_TIME_SEC; ++ ++ err = ext3_mark_iloc_dirty(handle, inode, &iloc); ++flags_err: ++ ext3_journal_stop(handle); ++ return err; ++} ++ + long ext3_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) + { + struct inode *inode = filp->f_dentry->d_inode; +@@ -50,6 +79,11 @@ long ext3_ioctl(struct file *filp, unsig + + flags = ext3_mask_flags(inode->i_mode, flags); + ++ if (IS_BARRIER(inode)) { ++ vxwprintk_task(1, "messing with the barrier."); ++ return -EACCES; ++ } ++ + mutex_lock(&inode->i_mutex); + + /* Is it quota file? Do not allow user to mess with it */ +@@ -68,7 +102,9 @@ long ext3_ioctl(struct file *filp, unsig + * + * This test looks nicer. Thanks to Pauline Middelink + */ +- if ((flags ^ oldflags) & (EXT3_APPEND_FL | EXT3_IMMUTABLE_FL)) { ++ if ((oldflags & EXT3_IMMUTABLE_FL) || ++ ((flags ^ oldflags) & (EXT3_APPEND_FL | ++ EXT3_IMMUTABLE_FL | EXT3_IXUNLINK_FL))) { + if (!capable(CAP_LINUX_IMMUTABLE)) + goto flags_out; + } +@@ -93,7 +129,7 @@ long ext3_ioctl(struct file *filp, unsig + if (err) + goto flags_err; + +- flags = flags & EXT3_FL_USER_MODIFIABLE; ++ flags &= EXT3_FL_USER_MODIFIABLE; + flags |= oldflags & ~EXT3_FL_USER_MODIFIABLE; + ei->i_flags = flags; + +diff -NurpP --minimal linux-3.0.9/fs/ext3/namei.c linux-3.0.9-vs2.3.2.1/fs/ext3/namei.c +--- linux-3.0.9/fs/ext3/namei.c 2011-11-15 16:40:46.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/ext3/namei.c 2011-08-29 03:45:09.000000000 +0200 +@@ -36,6 +36,7 @@ + #include <linux/quotaops.h> + #include <linux/buffer_head.h> + #include <linux/bio.h> ++#include <linux/vs_tag.h> + + #include "namei.h" + #include "xattr.h" +@@ -923,6 +924,7 @@ restart: + if (bh) + ll_rw_block(READ_META, 1, &bh); + } ++ dx_propagate_tag(nd, inode); + } + if ((bh = bh_use[ra_ptr++]) == NULL) + goto next; +@@ -2536,6 +2538,7 @@ const struct inode_operations ext3_dir_i + .removexattr = generic_removexattr, + #endif + .check_acl = ext3_check_acl, ++ .sync_flags = ext3_sync_flags, + }; + + const struct inode_operations ext3_special_inode_operations = { +diff -NurpP --minimal linux-3.0.9/fs/ext3/super.c linux-3.0.9-vs2.3.2.1/fs/ext3/super.c +--- linux-3.0.9/fs/ext3/super.c 2011-07-22 11:18:05.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/ext3/super.c 2011-06-10 22:11:24.000000000 +0200 +@@ -821,7 +821,8 @@ enum { + Opt_usrjquota, Opt_grpjquota, Opt_offusrjquota, Opt_offgrpjquota, + Opt_jqfmt_vfsold, Opt_jqfmt_vfsv0, Opt_jqfmt_vfsv1, Opt_quota, + Opt_noquota, Opt_ignore, Opt_barrier, Opt_nobarrier, Opt_err, +- Opt_resize, Opt_usrquota, Opt_grpquota ++ Opt_resize, Opt_usrquota, Opt_grpquota, ++ Opt_tag, Opt_notag, Opt_tagid + }; + + static const match_table_t tokens = { +@@ -878,6 +879,9 @@ static const match_table_t tokens = { + {Opt_barrier, "barrier"}, + {Opt_nobarrier, "nobarrier"}, + {Opt_resize, "resize"}, ++ {Opt_tag, "tag"}, ++ {Opt_notag, "notag"}, ++ {Opt_tagid, "tagid=%u"}, + {Opt_err, NULL}, + }; + +@@ -1030,6 +1034,20 @@ static int parse_options (char *options, + case Opt_nouid32: + set_opt (sbi->s_mount_opt, NO_UID32); + break; ++#ifndef CONFIG_TAGGING_NONE ++ case Opt_tag: ++ set_opt (sbi->s_mount_opt, TAGGED); ++ break; ++ case Opt_notag: ++ clear_opt (sbi->s_mount_opt, TAGGED); ++ break; ++#endif ++#ifdef CONFIG_PROPAGATE ++ case Opt_tagid: ++ /* use args[0] */ ++ set_opt (sbi->s_mount_opt, TAGGED); ++ break; ++#endif + case Opt_nocheck: + clear_opt (sbi->s_mount_opt, CHECK); + break; +@@ -1724,6 +1742,9 @@ static int ext3_fill_super (struct super + NULL, 0)) + goto failed_mount; + ++ if (EXT3_SB(sb)->s_mount_opt & EXT3_MOUNT_TAGGED) ++ sb->s_flags |= MS_TAGGED; ++ + sb->s_flags = (sb->s_flags & ~MS_POSIXACL) | + (test_opt(sb, POSIX_ACL) ? MS_POSIXACL : 0); + +@@ -2604,6 +2625,14 @@ static int ext3_remount (struct super_bl + if (test_opt(sb, ABORT)) + ext3_abort(sb, __func__, "Abort forced by user"); + ++ if ((sbi->s_mount_opt & EXT3_MOUNT_TAGGED) && ++ !(sb->s_flags & MS_TAGGED)) { ++ printk("EXT3-fs: %s: tagging not permitted on remount.\n", ++ sb->s_id); ++ err = -EINVAL; ++ goto restore_opts; ++ } ++ + sb->s_flags = (sb->s_flags & ~MS_POSIXACL) | + (test_opt(sb, POSIX_ACL) ? MS_POSIXACL : 0); + +diff -NurpP --minimal linux-3.0.9/fs/ext4/ext4.h linux-3.0.9-vs2.3.2.1/fs/ext4/ext4.h +--- linux-3.0.9/fs/ext4/ext4.h 2011-11-15 16:40:46.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/ext4/ext4.h 2011-11-15 17:37:06.000000000 +0100 +@@ -351,8 +351,12 @@ struct flex_groups { + #define EXT4_EXTENTS_FL 0x00080000 /* Inode uses extents */ + #define EXT4_EA_INODE_FL 0x00200000 /* Inode used for large EA */ + #define EXT4_EOFBLOCKS_FL 0x00400000 /* Blocks allocated beyond EOF */ ++#define EXT4_IXUNLINK_FL 0x08000000 /* Immutable invert on unlink */ + #define EXT4_RESERVED_FL 0x80000000 /* reserved for ext4 lib */ + ++#define EXT4_BARRIER_FL 0x04000000 /* Barrier for chroot() */ ++#define EXT4_COW_FL 0x20000000 /* Copy on Write marker */ ++ + #define EXT4_FL_USER_VISIBLE 0x004BDFFF /* User visible flags */ + #define EXT4_FL_USER_MODIFIABLE 0x004B80FF /* User modifiable flags */ + +@@ -609,7 +613,8 @@ struct ext4_inode { + __le16 l_i_file_acl_high; + __le16 l_i_uid_high; /* these 2 fields */ + __le16 l_i_gid_high; /* were reserved2[0] */ +- __u32 l_i_reserved2; ++ __le16 l_i_tag; /* Context Tag */ ++ __u16 l_i_reserved2; + } linux2; + struct { + __le16 h_i_reserved1; /* Obsoleted fragment number/size which are removed in ext4 */ +@@ -727,6 +732,7 @@ do { \ + #define i_gid_low i_gid + #define i_uid_high osd2.linux2.l_i_uid_high + #define i_gid_high osd2.linux2.l_i_gid_high ++#define i_raw_tag osd2.linux2.l_i_tag + #define i_reserved2 osd2.linux2.l_i_reserved2 + + #elif defined(__GNU__) +@@ -903,6 +909,7 @@ struct ext4_inode_info { + #define EXT4_MOUNT_POSIX_ACL 0x08000 /* POSIX Access Control Lists */ + #define EXT4_MOUNT_NO_AUTO_DA_ALLOC 0x10000 /* No auto delalloc mapping */ + #define EXT4_MOUNT_BARRIER 0x20000 /* Use block barriers */ ++#define EXT4_MOUNT_TAGGED 0x40000 /* Enable Context Tags */ + #define EXT4_MOUNT_QUOTA 0x80000 /* Some quota option set */ + #define EXT4_MOUNT_USRQUOTA 0x100000 /* "old" user quota */ + #define EXT4_MOUNT_GRPQUOTA 0x200000 /* "old" group quota */ +@@ -2173,6 +2180,7 @@ extern int ext4_map_blocks(handle_t *han + struct ext4_map_blocks *map, int flags); + extern int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, + __u64 start, __u64 len); ++extern int ext4_sync_flags(struct inode *, int, int); + /* move_extent.c */ + extern int ext4_move_extents(struct file *o_filp, struct file *d_filp, + __u64 start_orig, __u64 start_donor, +diff -NurpP --minimal linux-3.0.9/fs/ext4/file.c linux-3.0.9-vs2.3.2.1/fs/ext4/file.c +--- linux-3.0.9/fs/ext4/file.c 2011-07-22 11:18:05.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/ext4/file.c 2011-06-10 22:11:24.000000000 +0200 +@@ -282,5 +282,6 @@ const struct inode_operations ext4_file_ + #endif + .check_acl = ext4_check_acl, + .fiemap = ext4_fiemap, ++ .sync_flags = ext4_sync_flags, + }; + +diff -NurpP --minimal linux-3.0.9/fs/ext4/ialloc.c linux-3.0.9-vs2.3.2.1/fs/ext4/ialloc.c +--- linux-3.0.9/fs/ext4/ialloc.c 2011-05-22 16:17:52.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/ext4/ialloc.c 2011-06-10 22:11:24.000000000 +0200 +@@ -22,6 +22,7 @@ + #include <linux/random.h> + #include <linux/bitops.h> + #include <linux/blkdev.h> ++#include <linux/vs_tag.h> + #include <asm/byteorder.h> + + #include "ext4.h" +@@ -992,6 +993,7 @@ got: + inode->i_mode = mode; + inode->i_uid = current_fsuid(); + inode->i_gid = dir->i_gid; ++ inode->i_tag = dx_current_fstag(sb); + } else + inode_init_owner(inode, dir, mode); + +diff -NurpP --minimal linux-3.0.9/fs/ext4/inode.c linux-3.0.9-vs2.3.2.1/fs/ext4/inode.c +--- linux-3.0.9/fs/ext4/inode.c 2011-11-15 16:40:46.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/ext4/inode.c 2011-11-15 17:37:06.000000000 +0100 +@@ -42,6 +42,7 @@ + #include <linux/printk.h> + #include <linux/slab.h> + #include <linux/ratelimit.h> ++#include <linux/vs_tag.h> + + #include "ext4_jbd2.h" + #include "xattr.h" +@@ -4816,41 +4817,64 @@ void ext4_set_inode_flags(struct inode * + { + unsigned int flags = EXT4_I(inode)->i_flags; + +- inode->i_flags &= ~(S_SYNC|S_APPEND|S_IMMUTABLE|S_NOATIME|S_DIRSYNC); ++ inode->i_flags &= ~(S_IMMUTABLE | S_IXUNLINK | ++ S_SYNC | S_APPEND | S_NOATIME | S_DIRSYNC); ++ ++ if (flags & EXT4_IMMUTABLE_FL) ++ inode->i_flags |= S_IMMUTABLE; ++ if (flags & EXT4_IXUNLINK_FL) ++ inode->i_flags |= S_IXUNLINK; ++ + if (flags & EXT4_SYNC_FL) + inode->i_flags |= S_SYNC; + if (flags & EXT4_APPEND_FL) + inode->i_flags |= S_APPEND; +- if (flags & EXT4_IMMUTABLE_FL) +- inode->i_flags |= S_IMMUTABLE; + if (flags & EXT4_NOATIME_FL) + inode->i_flags |= S_NOATIME; + if (flags & EXT4_DIRSYNC_FL) + inode->i_flags |= S_DIRSYNC; ++ ++ inode->i_vflags &= ~(V_BARRIER | V_COW); ++ ++ if (flags & EXT4_BARRIER_FL) ++ inode->i_vflags |= V_BARRIER; ++ if (flags & EXT4_COW_FL) ++ inode->i_vflags |= V_COW; + } + + /* Propagate flags from i_flags to EXT4_I(inode)->i_flags */ + void ext4_get_inode_flags(struct ext4_inode_info *ei) + { +- unsigned int vfs_fl; ++ unsigned int vfs_fl, vfs_vf; + unsigned long old_fl, new_fl; + + do { + vfs_fl = ei->vfs_inode.i_flags; ++ vfs_vf = ei->vfs_inode.i_vflags; + old_fl = ei->i_flags; + new_fl = old_fl & ~(EXT4_SYNC_FL|EXT4_APPEND_FL| + EXT4_IMMUTABLE_FL|EXT4_NOATIME_FL| +- EXT4_DIRSYNC_FL); ++ EXT4_DIRSYNC_FL|EXT4_BARRIER_FL| ++ EXT4_COW_FL); ++ ++ if (vfs_fl & S_IMMUTABLE) ++ new_fl |= EXT4_IMMUTABLE_FL; ++ if (vfs_fl & S_IXUNLINK) ++ new_fl |= EXT4_IXUNLINK_FL; ++ + if (vfs_fl & S_SYNC) + new_fl |= EXT4_SYNC_FL; + if (vfs_fl & S_APPEND) + new_fl |= EXT4_APPEND_FL; +- if (vfs_fl & S_IMMUTABLE) +- new_fl |= EXT4_IMMUTABLE_FL; + if (vfs_fl & S_NOATIME) + new_fl |= EXT4_NOATIME_FL; + if (vfs_fl & S_DIRSYNC) + new_fl |= EXT4_DIRSYNC_FL; ++ ++ if (vfs_vf & V_BARRIER) ++ new_fl |= EXT4_BARRIER_FL; ++ if (vfs_vf & V_COW) ++ new_fl |= EXT4_COW_FL; + } while (cmpxchg(&ei->i_flags, old_fl, new_fl) != old_fl); + } + +@@ -4886,6 +4910,8 @@ struct inode *ext4_iget(struct super_blo + journal_t *journal = EXT4_SB(sb)->s_journal; + long ret; + int block; ++ uid_t uid; ++ gid_t gid; + + inode = iget_locked(sb, ino); + if (!inode) +@@ -4901,12 +4927,16 @@ struct inode *ext4_iget(struct super_blo + goto bad_inode; + raw_inode = ext4_raw_inode(&iloc); + inode->i_mode = le16_to_cpu(raw_inode->i_mode); +- inode->i_uid = (uid_t)le16_to_cpu(raw_inode->i_uid_low); +- inode->i_gid = (gid_t)le16_to_cpu(raw_inode->i_gid_low); ++ uid = (uid_t)le16_to_cpu(raw_inode->i_uid_low); ++ gid = (gid_t)le16_to_cpu(raw_inode->i_gid_low); + if (!(test_opt(inode->i_sb, NO_UID32))) { +- inode->i_uid |= le16_to_cpu(raw_inode->i_uid_high) << 16; +- inode->i_gid |= le16_to_cpu(raw_inode->i_gid_high) << 16; ++ uid |= le16_to_cpu(raw_inode->i_uid_high) << 16; ++ gid |= le16_to_cpu(raw_inode->i_gid_high) << 16; + } ++ inode->i_uid = INOTAG_UID(DX_TAG(inode), uid, gid); ++ inode->i_gid = INOTAG_GID(DX_TAG(inode), uid, gid); ++ inode->i_tag = INOTAG_TAG(DX_TAG(inode), uid, gid, ++ le16_to_cpu(raw_inode->i_raw_tag)); + inode->i_nlink = le16_to_cpu(raw_inode->i_links_count); + + ext4_clear_state_flags(ei); /* Only relevant on 32-bit archs */ +@@ -5125,6 +5155,8 @@ static int ext4_do_update_inode(handle_t + struct ext4_inode *raw_inode = ext4_raw_inode(iloc); + struct ext4_inode_info *ei = EXT4_I(inode); + struct buffer_head *bh = iloc->bh; ++ uid_t uid = TAGINO_UID(DX_TAG(inode), inode->i_uid, inode->i_tag); ++ gid_t gid = TAGINO_GID(DX_TAG(inode), inode->i_gid, inode->i_tag); + int err = 0, rc, block; + + /* For fields not not tracking in the in-memory inode, +@@ -5135,29 +5167,32 @@ static int ext4_do_update_inode(handle_t + ext4_get_inode_flags(ei); + raw_inode->i_mode = cpu_to_le16(inode->i_mode); + if (!(test_opt(inode->i_sb, NO_UID32))) { +- raw_inode->i_uid_low = cpu_to_le16(low_16_bits(inode->i_uid)); +- raw_inode->i_gid_low = cpu_to_le16(low_16_bits(inode->i_gid)); ++ raw_inode->i_uid_low = cpu_to_le16(low_16_bits(uid)); ++ raw_inode->i_gid_low = cpu_to_le16(low_16_bits(gid)); + /* + * Fix up interoperability with old kernels. Otherwise, old inodes get + * re-used with the upper 16 bits of the uid/gid intact + */ + if (!ei->i_dtime) { + raw_inode->i_uid_high = +- cpu_to_le16(high_16_bits(inode->i_uid)); ++ cpu_to_le16(high_16_bits(uid)); + raw_inode->i_gid_high = +- cpu_to_le16(high_16_bits(inode->i_gid)); ++ cpu_to_le16(high_16_bits(gid)); + } else { + raw_inode->i_uid_high = 0; + raw_inode->i_gid_high = 0; + } + } else { + raw_inode->i_uid_low = +- cpu_to_le16(fs_high2lowuid(inode->i_uid)); ++ cpu_to_le16(fs_high2lowuid(uid)); + raw_inode->i_gid_low = +- cpu_to_le16(fs_high2lowgid(inode->i_gid)); ++ cpu_to_le16(fs_high2lowgid(gid)); + raw_inode->i_uid_high = 0; + raw_inode->i_gid_high = 0; + } ++#ifdef CONFIG_TAGGING_INTERN ++ raw_inode->i_raw_tag = cpu_to_le16(inode->i_tag); ++#endif + raw_inode->i_links_count = cpu_to_le16(inode->i_nlink); + + EXT4_INODE_SET_XTIME(i_ctime, inode, raw_inode); +@@ -5343,7 +5378,8 @@ int ext4_setattr(struct dentry *dentry, + if (is_quota_modification(inode, attr)) + dquot_initialize(inode); + if ((ia_valid & ATTR_UID && attr->ia_uid != inode->i_uid) || +- (ia_valid & ATTR_GID && attr->ia_gid != inode->i_gid)) { ++ (ia_valid & ATTR_GID && attr->ia_gid != inode->i_gid) || ++ (ia_valid & ATTR_TAG && attr->ia_tag != inode->i_tag)) { + handle_t *handle; + + /* (user+group)*(old+new) structure, inode write (sb, +@@ -5365,6 +5401,8 @@ int ext4_setattr(struct dentry *dentry, + inode->i_uid = attr->ia_uid; + if (attr->ia_valid & ATTR_GID) + inode->i_gid = attr->ia_gid; ++ if ((attr->ia_valid & ATTR_TAG) && IS_TAGGED(inode)) ++ inode->i_tag = attr->ia_tag; + error = ext4_mark_inode_dirty(handle, inode); + ext4_journal_stop(handle); + } +diff -NurpP --minimal linux-3.0.9/fs/ext4/ioctl.c linux-3.0.9-vs2.3.2.1/fs/ext4/ioctl.c +--- linux-3.0.9/fs/ext4/ioctl.c 2011-05-22 16:17:52.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/ext4/ioctl.c 2011-06-10 22:11:24.000000000 +0200 +@@ -14,10 +14,39 @@ + #include <linux/compat.h> + #include <linux/mount.h> + #include <linux/file.h> ++#include <linux/vs_tag.h> + #include <asm/uaccess.h> + #include "ext4_jbd2.h" + #include "ext4.h" + ++ ++int ext4_sync_flags(struct inode *inode, int flags, int vflags) ++{ ++ handle_t *handle = NULL; ++ struct ext4_iloc iloc; ++ int err; ++ ++ handle = ext4_journal_start(inode, 1); ++ if (IS_ERR(handle)) ++ return PTR_ERR(handle); ++ ++ if (IS_SYNC(inode)) ++ ext4_handle_sync(handle); ++ err = ext4_reserve_inode_write(handle, inode, &iloc); ++ if (err) ++ goto flags_err; ++ ++ inode->i_flags = flags; ++ inode->i_vflags = vflags; ++ ext4_get_inode_flags(EXT4_I(inode)); ++ inode->i_ctime = ext4_current_time(inode); ++ ++ err = ext4_mark_iloc_dirty(handle, inode, &iloc); ++flags_err: ++ ext4_journal_stop(handle); ++ return err; ++} ++ + long ext4_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) + { + struct inode *inode = filp->f_dentry->d_inode; +@@ -50,6 +79,11 @@ long ext4_ioctl(struct file *filp, unsig + + flags = ext4_mask_flags(inode->i_mode, flags); + ++ if (IS_BARRIER(inode)) { ++ vxwprintk_task(1, "messing with the barrier."); ++ return -EACCES; ++ } ++ + err = -EPERM; + mutex_lock(&inode->i_mutex); + /* Is it quota file? Do not allow user to mess with it */ +@@ -67,7 +101,9 @@ long ext4_ioctl(struct file *filp, unsig + * + * This test looks nicer. Thanks to Pauline Middelink + */ +- if ((flags ^ oldflags) & (EXT4_APPEND_FL | EXT4_IMMUTABLE_FL)) { ++ if ((oldflags & EXT4_IMMUTABLE_FL) || ++ ((flags ^ oldflags) & (EXT4_APPEND_FL | ++ EXT4_IMMUTABLE_FL | EXT4_IXUNLINK_FL))) { + if (!capable(CAP_LINUX_IMMUTABLE)) + goto flags_out; + } +diff -NurpP --minimal linux-3.0.9/fs/ext4/namei.c linux-3.0.9-vs2.3.2.1/fs/ext4/namei.c +--- linux-3.0.9/fs/ext4/namei.c 2011-11-15 16:40:46.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/ext4/namei.c 2011-11-15 17:37:06.000000000 +0100 +@@ -34,6 +34,7 @@ + #include <linux/quotaops.h> + #include <linux/buffer_head.h> + #include <linux/bio.h> ++#include <linux/vs_tag.h> + #include "ext4.h" + #include "ext4_jbd2.h" + +@@ -924,6 +925,7 @@ restart: + if (bh) + ll_rw_block(READ_META, 1, &bh); + } ++ dx_propagate_tag(nd, inode); + } + if ((bh = bh_use[ra_ptr++]) == NULL) + goto next; +@@ -2598,6 +2600,7 @@ const struct inode_operations ext4_dir_i + #endif + .check_acl = ext4_check_acl, + .fiemap = ext4_fiemap, ++ .sync_flags = ext4_sync_flags, + }; + + const struct inode_operations ext4_special_inode_operations = { +diff -NurpP --minimal linux-3.0.9/fs/ext4/super.c linux-3.0.9-vs2.3.2.1/fs/ext4/super.c +--- linux-3.0.9/fs/ext4/super.c 2011-11-15 16:40:46.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/ext4/super.c 2011-08-31 19:37:44.000000000 +0200 +@@ -1293,6 +1293,7 @@ enum { + Opt_dioread_nolock, Opt_dioread_lock, + Opt_discard, Opt_nodiscard, + Opt_init_inode_table, Opt_noinit_inode_table, ++ Opt_tag, Opt_notag, Opt_tagid + }; + + static const match_table_t tokens = { +@@ -1368,6 +1369,9 @@ static const match_table_t tokens = { + {Opt_init_inode_table, "init_itable=%u"}, + {Opt_init_inode_table, "init_itable"}, + {Opt_noinit_inode_table, "noinit_itable"}, ++ {Opt_tag, "tag"}, ++ {Opt_notag, "notag"}, ++ {Opt_tagid, "tagid=%u"}, + {Opt_err, NULL}, + }; + +@@ -1536,6 +1540,20 @@ static int parse_options(char *options, + case Opt_nouid32: + set_opt(sb, NO_UID32); + break; ++#ifndef CONFIG_TAGGING_NONE ++ case Opt_tag: ++ set_opt(sb, TAGGED); ++ break; ++ case Opt_notag: ++ clear_opt(sb, TAGGED); ++ break; ++#endif ++#ifdef CONFIG_PROPAGATE ++ case Opt_tagid: ++ /* use args[0] */ ++ set_opt(sb, TAGGED); ++ break; ++#endif + case Opt_debug: + set_opt(sb, DEBUG); + break; +@@ -3193,6 +3211,9 @@ static int ext4_fill_super(struct super_ + &journal_ioprio, NULL, 0)) + goto failed_mount; + ++ if (EXT4_SB(sb)->s_mount_opt & EXT4_MOUNT_TAGGED) ++ sb->s_flags |= MS_TAGGED; ++ + sb->s_flags = (sb->s_flags & ~MS_POSIXACL) | + (test_opt(sb, POSIX_ACL) ? MS_POSIXACL : 0); + +@@ -4323,6 +4344,14 @@ static int ext4_remount(struct super_blo + if (sbi->s_mount_flags & EXT4_MF_FS_ABORTED) + ext4_abort(sb, "Abort forced by user"); + ++ if ((sbi->s_mount_opt & EXT4_MOUNT_TAGGED) && ++ !(sb->s_flags & MS_TAGGED)) { ++ printk("EXT4-fs: %s: tagging not permitted on remount.\n", ++ sb->s_id); ++ err = -EINVAL; ++ goto restore_opts; ++ } ++ + sb->s_flags = (sb->s_flags & ~MS_POSIXACL) | + (test_opt(sb, POSIX_ACL) ? MS_POSIXACL : 0); + +diff -NurpP --minimal linux-3.0.9/fs/fcntl.c linux-3.0.9-vs2.3.2.1/fs/fcntl.c +--- linux-3.0.9/fs/fcntl.c 2011-05-22 16:17:52.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/fcntl.c 2011-06-10 22:11:24.000000000 +0200 +@@ -20,6 +20,7 @@ + #include <linux/signal.h> + #include <linux/rcupdate.h> + #include <linux/pid_namespace.h> ++#include <linux/vs_limit.h> + + #include <asm/poll.h> + #include <asm/siginfo.h> +@@ -103,6 +104,8 @@ SYSCALL_DEFINE3(dup3, unsigned int, oldf + + if (tofree) + filp_close(tofree, files); ++ else ++ vx_openfd_inc(newfd); /* fd was unused */ + + return newfd; + +@@ -447,6 +450,8 @@ SYSCALL_DEFINE3(fcntl, unsigned int, fd, + filp = fget_raw(fd); + if (!filp) + goto out; ++ if (!vx_files_avail(1)) ++ goto out; + + if (unlikely(filp->f_mode & FMODE_PATH)) { + if (!check_fcntl_cmd(cmd)) { +diff -NurpP --minimal linux-3.0.9/fs/file.c linux-3.0.9-vs2.3.2.1/fs/file.c +--- linux-3.0.9/fs/file.c 2011-05-22 16:17:52.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/file.c 2011-06-10 22:11:24.000000000 +0200 +@@ -21,6 +21,7 @@ + #include <linux/spinlock.h> + #include <linux/rcupdate.h> + #include <linux/workqueue.h> ++#include <linux/vs_limit.h> + + struct fdtable_defer { + spinlock_t lock; +@@ -359,6 +360,8 @@ struct files_struct *dup_fd(struct files + struct file *f = *old_fds++; + if (f) { + get_file(f); ++ /* TODO: sum it first for check and performance */ ++ vx_openfd_inc(open_files - i); + } else { + /* + * The fd may be claimed in the fd bitmap but not yet +@@ -466,6 +469,7 @@ repeat: + else + FD_CLR(fd, fdt->close_on_exec); + error = fd; ++ vx_openfd_inc(fd); + #if 1 + /* Sanity check */ + if (rcu_dereference_raw(fdt->fd[fd]) != NULL) { +diff -NurpP --minimal linux-3.0.9/fs/file_table.c linux-3.0.9-vs2.3.2.1/fs/file_table.c +--- linux-3.0.9/fs/file_table.c 2011-05-22 16:17:52.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/file_table.c 2011-06-10 22:11:24.000000000 +0200 +@@ -24,6 +24,8 @@ + #include <linux/percpu_counter.h> + #include <linux/percpu.h> + #include <linux/ima.h> ++#include <linux/vs_limit.h> ++#include <linux/vs_context.h> + + #include <asm/atomic.h> + +@@ -135,6 +137,8 @@ struct file *get_empty_filp(void) + spin_lock_init(&f->f_lock); + eventpoll_init_file(f); + /* f->f_version: 0 */ ++ f->f_xid = vx_current_xid(); ++ vx_files_inc(f); + return f; + + over: +@@ -253,6 +257,8 @@ static void __fput(struct file *file) + } + fops_put(file->f_op); + put_pid(file->f_owner.pid); ++ vx_files_dec(file); ++ file->f_xid = 0; + file_sb_list_del(file); + if ((file->f_mode & (FMODE_READ | FMODE_WRITE)) == FMODE_READ) + i_readcount_dec(inode); +@@ -383,6 +389,8 @@ void put_filp(struct file *file) + { + if (atomic_long_dec_and_test(&file->f_count)) { + security_file_free(file); ++ vx_files_dec(file); ++ file->f_xid = 0; + file_sb_list_del(file); + file_free(file); + } +diff -NurpP --minimal linux-3.0.9/fs/fs_struct.c linux-3.0.9-vs2.3.2.1/fs/fs_struct.c +--- linux-3.0.9/fs/fs_struct.c 2011-03-15 18:07:31.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/fs_struct.c 2011-06-10 22:11:24.000000000 +0200 +@@ -4,6 +4,7 @@ + #include <linux/path.h> + #include <linux/slab.h> + #include <linux/fs_struct.h> ++#include <linux/vserver/global.h> + #include "internal.h" + + static inline void path_get_longterm(struct path *path) +@@ -96,6 +97,7 @@ void free_fs_struct(struct fs_struct *fs + { + path_put_longterm(&fs->root); + path_put_longterm(&fs->pwd); ++ atomic_dec(&vs_global_fs); + kmem_cache_free(fs_cachep, fs); + } + +@@ -135,6 +137,7 @@ struct fs_struct *copy_fs_struct(struct + fs->pwd = old->pwd; + path_get_longterm(&fs->pwd); + spin_unlock(&old->lock); ++ atomic_inc(&vs_global_fs); + } + return fs; + } +diff -NurpP --minimal linux-3.0.9/fs/gfs2/file.c linux-3.0.9-vs2.3.2.1/fs/gfs2/file.c +--- linux-3.0.9/fs/gfs2/file.c 2011-07-22 11:18:05.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/gfs2/file.c 2011-06-10 22:11:24.000000000 +0200 +@@ -134,6 +134,9 @@ static const u32 fsflags_to_gfs2[32] = { + [7] = GFS2_DIF_NOATIME, + [12] = GFS2_DIF_EXHASH, + [14] = GFS2_DIF_INHERIT_JDATA, ++ [27] = GFS2_DIF_IXUNLINK, ++ [26] = GFS2_DIF_BARRIER, ++ [29] = GFS2_DIF_COW, + }; + + static const u32 gfs2_to_fsflags[32] = { +@@ -143,6 +146,9 @@ static const u32 gfs2_to_fsflags[32] = { + [gfs2fl_NoAtime] = FS_NOATIME_FL, + [gfs2fl_ExHash] = FS_INDEX_FL, + [gfs2fl_InheritJdata] = FS_JOURNAL_DATA_FL, ++ [gfs2fl_IXUnlink] = FS_IXUNLINK_FL, ++ [gfs2fl_Barrier] = FS_BARRIER_FL, ++ [gfs2fl_Cow] = FS_COW_FL, + }; + + static int gfs2_get_flags(struct file *filp, u32 __user *ptr) +@@ -173,10 +179,16 @@ void gfs2_set_inode_flags(struct inode * + { + struct gfs2_inode *ip = GFS2_I(inode); + unsigned int flags = inode->i_flags; ++ unsigned int vflags = inode->i_vflags; ++ ++ flags &= ~(S_IMMUTABLE | S_IXUNLINK | ++ S_SYNC | S_APPEND | S_NOATIME | S_DIRSYNC); + +- flags &= ~(S_SYNC|S_APPEND|S_IMMUTABLE|S_NOATIME|S_DIRSYNC); + if (ip->i_diskflags & GFS2_DIF_IMMUTABLE) + flags |= S_IMMUTABLE; ++ if (ip->i_diskflags & GFS2_DIF_IXUNLINK) ++ flags |= S_IXUNLINK; ++ + if (ip->i_diskflags & GFS2_DIF_APPENDONLY) + flags |= S_APPEND; + if (ip->i_diskflags & GFS2_DIF_NOATIME) +@@ -184,6 +196,43 @@ void gfs2_set_inode_flags(struct inode * + if (ip->i_diskflags & GFS2_DIF_SYNC) + flags |= S_SYNC; + inode->i_flags = flags; ++ ++ vflags &= ~(V_BARRIER | V_COW); ++ ++ if (ip->i_diskflags & GFS2_DIF_BARRIER) ++ vflags |= V_BARRIER; ++ if (ip->i_diskflags & GFS2_DIF_COW) ++ vflags |= V_COW; ++ inode->i_vflags = vflags; ++} ++ ++void gfs2_get_inode_flags(struct inode *inode) ++{ ++ struct gfs2_inode *ip = GFS2_I(inode); ++ unsigned int flags = inode->i_flags; ++ unsigned int vflags = inode->i_vflags; ++ ++ ip->i_diskflags &= ~(GFS2_DIF_APPENDONLY | ++ GFS2_DIF_NOATIME | GFS2_DIF_SYNC | ++ GFS2_DIF_IMMUTABLE | GFS2_DIF_IXUNLINK | ++ GFS2_DIF_BARRIER | GFS2_DIF_COW); ++ ++ if (flags & S_IMMUTABLE) ++ ip->i_diskflags |= GFS2_DIF_IMMUTABLE; ++ if (flags & S_IXUNLINK) ++ ip->i_diskflags |= GFS2_DIF_IXUNLINK; ++ ++ if (flags & S_APPEND) ++ ip->i_diskflags |= GFS2_DIF_APPENDONLY; ++ if (flags & S_NOATIME) ++ ip->i_diskflags |= GFS2_DIF_NOATIME; ++ if (flags & S_SYNC) ++ ip->i_diskflags |= GFS2_DIF_SYNC; ++ ++ if (vflags & V_BARRIER) ++ ip->i_diskflags |= GFS2_DIF_BARRIER; ++ if (vflags & V_COW) ++ ip->i_diskflags |= GFS2_DIF_COW; + } + + /* Flags that can be set by user space */ +@@ -295,6 +344,37 @@ static int gfs2_set_flags(struct file *f + return do_gfs2_set_flags(filp, gfsflags, ~GFS2_DIF_JDATA); + } + ++int gfs2_sync_flags(struct inode *inode, int flags, int vflags) ++{ ++ struct gfs2_inode *ip = GFS2_I(inode); ++ struct gfs2_sbd *sdp = GFS2_SB(inode); ++ struct buffer_head *bh; ++ struct gfs2_holder gh; ++ int error; ++ ++ error = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh); ++ if (error) ++ return error; ++ error = gfs2_trans_begin(sdp, RES_DINODE, 0); ++ if (error) ++ goto out; ++ error = gfs2_meta_inode_buffer(ip, &bh); ++ if (error) ++ goto out_trans_end; ++ gfs2_trans_add_bh(ip->i_gl, bh, 1); ++ inode->i_flags = flags; ++ inode->i_vflags = vflags; ++ gfs2_get_inode_flags(inode); ++ gfs2_dinode_out(ip, bh->b_data); ++ brelse(bh); ++ gfs2_set_aops(inode); ++out_trans_end: ++ gfs2_trans_end(sdp); ++out: ++ gfs2_glock_dq_uninit(&gh); ++ return error; ++} ++ + static long gfs2_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) + { + switch(cmd) { +diff -NurpP --minimal linux-3.0.9/fs/gfs2/inode.h linux-3.0.9-vs2.3.2.1/fs/gfs2/inode.h +--- linux-3.0.9/fs/gfs2/inode.h 2011-07-22 11:18:05.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/gfs2/inode.h 2011-06-10 22:11:24.000000000 +0200 +@@ -120,6 +120,7 @@ extern const struct file_operations gfs2 + extern const struct file_operations gfs2_dir_fops_nolock; + + extern void gfs2_set_inode_flags(struct inode *inode); ++extern int gfs2_sync_flags(struct inode *inode, int flags, int vflags); + + #ifdef CONFIG_GFS2_FS_LOCKING_DLM + extern const struct file_operations gfs2_file_fops; +diff -NurpP --minimal linux-3.0.9/fs/inode.c linux-3.0.9-vs2.3.2.1/fs/inode.c +--- linux-3.0.9/fs/inode.c 2011-07-22 11:18:05.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/inode.c 2011-07-01 11:35:34.000000000 +0200 +@@ -26,6 +26,7 @@ + #include <linux/ima.h> + #include <linux/cred.h> + #include <linux/buffer_head.h> /* for inode_has_buffers */ ++#include <linux/vs_tag.h> + #include "internal.h" + + /* +@@ -146,6 +147,9 @@ int inode_init_always(struct super_block + struct address_space *const mapping = &inode->i_data; + + inode->i_sb = sb; ++ ++ /* essential because of inode slab reuse */ ++ inode->i_tag = 0; + inode->i_blkbits = sb->s_blocksize_bits; + inode->i_flags = 0; + atomic_set(&inode->i_count, 1); +@@ -166,6 +170,7 @@ int inode_init_always(struct super_block + inode->i_bdev = NULL; + inode->i_cdev = NULL; + inode->i_rdev = 0; ++ inode->i_mdev = 0; + inode->dirtied_when = 0; + + if (security_inode_alloc(inode)) +@@ -404,6 +409,8 @@ void __insert_inode_hash(struct inode *i + } + EXPORT_SYMBOL(__insert_inode_hash); + ++EXPORT_SYMBOL_GPL(__iget); ++ + /** + * remove_inode_hash - remove an inode from the hash + * @inode: inode to unhash +@@ -1643,9 +1650,11 @@ void init_special_inode(struct inode *in + if (S_ISCHR(mode)) { + inode->i_fop = &def_chr_fops; + inode->i_rdev = rdev; ++ inode->i_mdev = rdev; + } else if (S_ISBLK(mode)) { + inode->i_fop = &def_blk_fops; + inode->i_rdev = rdev; ++ inode->i_mdev = rdev; + } else if (S_ISFIFO(mode)) + inode->i_fop = &def_fifo_fops; + else if (S_ISSOCK(mode)) +@@ -1674,6 +1683,7 @@ void inode_init_owner(struct inode *inod + } else + inode->i_gid = current_fsgid(); + inode->i_mode = mode; ++ inode->i_tag = dx_current_fstag(inode->i_sb); + } + EXPORT_SYMBOL(inode_init_owner); + +diff -NurpP --minimal linux-3.0.9/fs/ioctl.c linux-3.0.9-vs2.3.2.1/fs/ioctl.c +--- linux-3.0.9/fs/ioctl.c 2011-05-22 16:17:52.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/ioctl.c 2011-06-10 22:11:24.000000000 +0200 +@@ -15,6 +15,9 @@ + #include <linux/writeback.h> + #include <linux/buffer_head.h> + #include <linux/falloc.h> ++#include <linux/proc_fs.h> ++#include <linux/vserver/inode.h> ++#include <linux/vs_tag.h> + + #include <asm/ioctls.h> + +diff -NurpP --minimal linux-3.0.9/fs/ioprio.c linux-3.0.9-vs2.3.2.1/fs/ioprio.c +--- linux-3.0.9/fs/ioprio.c 2011-01-05 21:50:24.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/ioprio.c 2011-06-10 22:11:24.000000000 +0200 +@@ -27,6 +27,7 @@ + #include <linux/syscalls.h> + #include <linux/security.h> + #include <linux/pid_namespace.h> ++#include <linux/vs_base.h> + + int set_task_ioprio(struct task_struct *task, int ioprio) + { +@@ -119,6 +120,8 @@ SYSCALL_DEFINE3(ioprio_set, int, which, + else + pgrp = find_vpid(who); + do_each_pid_thread(pgrp, PIDTYPE_PGID, p) { ++ if (!vx_check(p->xid, VS_ADMIN_P | VS_IDENT)) ++ continue; + ret = set_task_ioprio(p, ioprio); + if (ret) + break; +@@ -208,6 +211,8 @@ SYSCALL_DEFINE2(ioprio_get, int, which, + else + pgrp = find_vpid(who); + do_each_pid_thread(pgrp, PIDTYPE_PGID, p) { ++ if (!vx_check(p->xid, VS_ADMIN_P | VS_IDENT)) ++ continue; + tmpio = get_task_ioprio(p); + if (tmpio < 0) + continue; +diff -NurpP --minimal linux-3.0.9/fs/jfs/file.c linux-3.0.9-vs2.3.2.1/fs/jfs/file.c +--- linux-3.0.9/fs/jfs/file.c 2011-07-22 11:18:05.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/jfs/file.c 2011-07-01 11:35:34.000000000 +0200 +@@ -102,7 +102,8 @@ int jfs_setattr(struct dentry *dentry, s + if (is_quota_modification(inode, iattr)) + dquot_initialize(inode); + if ((iattr->ia_valid & ATTR_UID && iattr->ia_uid != inode->i_uid) || +- (iattr->ia_valid & ATTR_GID && iattr->ia_gid != inode->i_gid)) { ++ (iattr->ia_valid & ATTR_GID && iattr->ia_gid != inode->i_gid) || ++ (iattr->ia_valid & ATTR_TAG && iattr->ia_tag != inode->i_tag)) { + rc = dquot_transfer(inode, iattr); + if (rc) + return rc; +@@ -133,6 +134,7 @@ const struct inode_operations jfs_file_i + #ifdef CONFIG_JFS_POSIX_ACL + .check_acl = jfs_check_acl, + #endif ++ .sync_flags = jfs_sync_flags, + }; + + const struct file_operations jfs_file_operations = { +diff -NurpP --minimal linux-3.0.9/fs/jfs/ioctl.c linux-3.0.9-vs2.3.2.1/fs/jfs/ioctl.c +--- linux-3.0.9/fs/jfs/ioctl.c 2011-05-22 16:17:52.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/jfs/ioctl.c 2011-06-10 22:11:24.000000000 +0200 +@@ -11,6 +11,7 @@ + #include <linux/mount.h> + #include <linux/time.h> + #include <linux/sched.h> ++#include <linux/mount.h> + #include <asm/current.h> + #include <asm/uaccess.h> + +@@ -52,6 +53,16 @@ static long jfs_map_ext2(unsigned long f + } + + ++int jfs_sync_flags(struct inode *inode, int flags, int vflags) ++{ ++ inode->i_flags = flags; ++ inode->i_vflags = vflags; ++ jfs_get_inode_flags(JFS_IP(inode)); ++ inode->i_ctime = CURRENT_TIME_SEC; ++ mark_inode_dirty(inode); ++ return 0; ++} ++ + long jfs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) + { + struct inode *inode = filp->f_dentry->d_inode; +@@ -85,6 +96,11 @@ long jfs_ioctl(struct file *filp, unsign + if (!S_ISDIR(inode->i_mode)) + flags &= ~JFS_DIRSYNC_FL; + ++ if (IS_BARRIER(inode)) { ++ vxwprintk_task(1, "messing with the barrier."); ++ return -EACCES; ++ } ++ + /* Is it quota file? Do not allow user to mess with it */ + if (IS_NOQUOTA(inode)) { + err = -EPERM; +@@ -102,8 +118,8 @@ long jfs_ioctl(struct file *filp, unsign + * the relevant capability. + */ + if ((oldflags & JFS_IMMUTABLE_FL) || +- ((flags ^ oldflags) & +- (JFS_APPEND_FL | JFS_IMMUTABLE_FL))) { ++ ((flags ^ oldflags) & (JFS_APPEND_FL | ++ JFS_IMMUTABLE_FL | JFS_IXUNLINK_FL))) { + if (!capable(CAP_LINUX_IMMUTABLE)) { + mutex_unlock(&inode->i_mutex); + err = -EPERM; +@@ -111,7 +127,7 @@ long jfs_ioctl(struct file *filp, unsign + } + } + +- flags = flags & JFS_FL_USER_MODIFIABLE; ++ flags &= JFS_FL_USER_MODIFIABLE; + flags |= oldflags & ~JFS_FL_USER_MODIFIABLE; + jfs_inode->mode2 = flags; + +diff -NurpP --minimal linux-3.0.9/fs/jfs/jfs_dinode.h linux-3.0.9-vs2.3.2.1/fs/jfs/jfs_dinode.h +--- linux-3.0.9/fs/jfs/jfs_dinode.h 2008-12-25 00:26:37.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/jfs/jfs_dinode.h 2011-06-10 22:11:24.000000000 +0200 +@@ -161,9 +161,13 @@ struct dinode { + + #define JFS_APPEND_FL 0x01000000 /* writes to file may only append */ + #define JFS_IMMUTABLE_FL 0x02000000 /* Immutable file */ ++#define JFS_IXUNLINK_FL 0x08000000 /* Immutable invert on unlink */ + +-#define JFS_FL_USER_VISIBLE 0x03F80000 +-#define JFS_FL_USER_MODIFIABLE 0x03F80000 ++#define JFS_BARRIER_FL 0x04000000 /* Barrier for chroot() */ ++#define JFS_COW_FL 0x20000000 /* Copy on Write marker */ ++ ++#define JFS_FL_USER_VISIBLE 0x07F80000 ++#define JFS_FL_USER_MODIFIABLE 0x07F80000 + #define JFS_FL_INHERIT 0x03C80000 + + /* These are identical to EXT[23]_IOC_GETFLAGS/SETFLAGS */ +diff -NurpP --minimal linux-3.0.9/fs/jfs/jfs_filsys.h linux-3.0.9-vs2.3.2.1/fs/jfs/jfs_filsys.h +--- linux-3.0.9/fs/jfs/jfs_filsys.h 2008-12-25 00:26:37.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/jfs/jfs_filsys.h 2011-06-10 22:11:24.000000000 +0200 +@@ -263,6 +263,7 @@ + #define JFS_NAME_MAX 255 + #define JFS_PATH_MAX BPSIZE + ++#define JFS_TAGGED 0x00800000 /* Context Tagging */ + + /* + * file system state (superblock state) +diff -NurpP --minimal linux-3.0.9/fs/jfs/jfs_imap.c linux-3.0.9-vs2.3.2.1/fs/jfs/jfs_imap.c +--- linux-3.0.9/fs/jfs/jfs_imap.c 2011-07-22 11:18:05.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/jfs/jfs_imap.c 2011-07-01 11:35:34.000000000 +0200 +@@ -46,6 +46,7 @@ + #include <linux/pagemap.h> + #include <linux/quotaops.h> + #include <linux/slab.h> ++#include <linux/vs_tag.h> + + #include "jfs_incore.h" + #include "jfs_inode.h" +@@ -3058,6 +3059,8 @@ static int copy_from_dinode(struct dinod + { + struct jfs_inode_info *jfs_ip = JFS_IP(ip); + struct jfs_sb_info *sbi = JFS_SBI(ip->i_sb); ++ uid_t uid; ++ gid_t gid; + + jfs_ip->fileset = le32_to_cpu(dip->di_fileset); + jfs_ip->mode2 = le32_to_cpu(dip->di_mode); +@@ -3078,14 +3081,18 @@ static int copy_from_dinode(struct dinod + } + ip->i_nlink = le32_to_cpu(dip->di_nlink); + +- jfs_ip->saved_uid = le32_to_cpu(dip->di_uid); ++ uid = le32_to_cpu(dip->di_uid); ++ gid = le32_to_cpu(dip->di_gid); ++ ip->i_tag = INOTAG_TAG(DX_TAG(ip), uid, gid, 0); ++ ++ jfs_ip->saved_uid = INOTAG_UID(DX_TAG(ip), uid, gid); + if (sbi->uid == -1) + ip->i_uid = jfs_ip->saved_uid; + else { + ip->i_uid = sbi->uid; + } + +- jfs_ip->saved_gid = le32_to_cpu(dip->di_gid); ++ jfs_ip->saved_gid = INOTAG_GID(DX_TAG(ip), uid, gid); + if (sbi->gid == -1) + ip->i_gid = jfs_ip->saved_gid; + else { +@@ -3150,14 +3157,12 @@ static void copy_to_dinode(struct dinode + dip->di_size = cpu_to_le64(ip->i_size); + dip->di_nblocks = cpu_to_le64(PBLK2LBLK(ip->i_sb, ip->i_blocks)); + dip->di_nlink = cpu_to_le32(ip->i_nlink); +- if (sbi->uid == -1) +- dip->di_uid = cpu_to_le32(ip->i_uid); +- else +- dip->di_uid = cpu_to_le32(jfs_ip->saved_uid); +- if (sbi->gid == -1) +- dip->di_gid = cpu_to_le32(ip->i_gid); +- else +- dip->di_gid = cpu_to_le32(jfs_ip->saved_gid); ++ ++ dip->di_uid = cpu_to_le32(TAGINO_UID(DX_TAG(ip), ++ (sbi->uid == -1) ? ip->i_uid : jfs_ip->saved_uid, ip->i_tag)); ++ dip->di_gid = cpu_to_le32(TAGINO_GID(DX_TAG(ip), ++ (sbi->gid == -1) ? ip->i_gid : jfs_ip->saved_gid, ip->i_tag)); ++ + jfs_get_inode_flags(jfs_ip); + /* + * mode2 is only needed for storing the higher order bits. +diff -NurpP --minimal linux-3.0.9/fs/jfs/jfs_inode.c linux-3.0.9-vs2.3.2.1/fs/jfs/jfs_inode.c +--- linux-3.0.9/fs/jfs/jfs_inode.c 2010-08-02 16:52:49.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/jfs/jfs_inode.c 2011-06-10 22:11:24.000000000 +0200 +@@ -18,6 +18,7 @@ + + #include <linux/fs.h> + #include <linux/quotaops.h> ++#include <linux/vs_tag.h> + #include "jfs_incore.h" + #include "jfs_inode.h" + #include "jfs_filsys.h" +@@ -30,29 +31,46 @@ void jfs_set_inode_flags(struct inode *i + { + unsigned int flags = JFS_IP(inode)->mode2; + +- inode->i_flags &= ~(S_IMMUTABLE | S_APPEND | +- S_NOATIME | S_DIRSYNC | S_SYNC); ++ inode->i_flags &= ~(S_IMMUTABLE | S_IXUNLINK | ++ S_SYNC | S_APPEND | S_NOATIME | S_DIRSYNC); + + if (flags & JFS_IMMUTABLE_FL) + inode->i_flags |= S_IMMUTABLE; ++ if (flags & JFS_IXUNLINK_FL) ++ inode->i_flags |= S_IXUNLINK; ++ ++ if (flags & JFS_SYNC_FL) ++ inode->i_flags |= S_SYNC; + if (flags & JFS_APPEND_FL) + inode->i_flags |= S_APPEND; + if (flags & JFS_NOATIME_FL) + inode->i_flags |= S_NOATIME; + if (flags & JFS_DIRSYNC_FL) + inode->i_flags |= S_DIRSYNC; +- if (flags & JFS_SYNC_FL) +- inode->i_flags |= S_SYNC; ++ ++ inode->i_vflags &= ~(V_BARRIER | V_COW); ++ ++ if (flags & JFS_BARRIER_FL) ++ inode->i_vflags |= V_BARRIER; ++ if (flags & JFS_COW_FL) ++ inode->i_vflags |= V_COW; + } + + void jfs_get_inode_flags(struct jfs_inode_info *jfs_ip) + { + unsigned int flags = jfs_ip->vfs_inode.i_flags; ++ unsigned int vflags = jfs_ip->vfs_inode.i_vflags; ++ ++ jfs_ip->mode2 &= ~(JFS_IMMUTABLE_FL | JFS_IXUNLINK_FL | ++ JFS_APPEND_FL | JFS_NOATIME_FL | ++ JFS_DIRSYNC_FL | JFS_SYNC_FL | ++ JFS_BARRIER_FL | JFS_COW_FL); + +- jfs_ip->mode2 &= ~(JFS_IMMUTABLE_FL | JFS_APPEND_FL | JFS_NOATIME_FL | +- JFS_DIRSYNC_FL | JFS_SYNC_FL); + if (flags & S_IMMUTABLE) + jfs_ip->mode2 |= JFS_IMMUTABLE_FL; ++ if (flags & S_IXUNLINK) ++ jfs_ip->mode2 |= JFS_IXUNLINK_FL; ++ + if (flags & S_APPEND) + jfs_ip->mode2 |= JFS_APPEND_FL; + if (flags & S_NOATIME) +@@ -61,6 +79,11 @@ void jfs_get_inode_flags(struct jfs_inod + jfs_ip->mode2 |= JFS_DIRSYNC_FL; + if (flags & S_SYNC) + jfs_ip->mode2 |= JFS_SYNC_FL; ++ ++ if (vflags & V_BARRIER) ++ jfs_ip->mode2 |= JFS_BARRIER_FL; ++ if (vflags & V_COW) ++ jfs_ip->mode2 |= JFS_COW_FL; + } + + /* +diff -NurpP --minimal linux-3.0.9/fs/jfs/jfs_inode.h linux-3.0.9-vs2.3.2.1/fs/jfs/jfs_inode.h +--- linux-3.0.9/fs/jfs/jfs_inode.h 2011-07-22 11:18:05.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/jfs/jfs_inode.h 2011-06-10 22:11:24.000000000 +0200 +@@ -39,6 +39,7 @@ extern struct dentry *jfs_fh_to_dentry(s + extern struct dentry *jfs_fh_to_parent(struct super_block *sb, struct fid *fid, + int fh_len, int fh_type); + extern void jfs_set_inode_flags(struct inode *); ++extern int jfs_sync_flags(struct inode *, int, int); + extern int jfs_get_block(struct inode *, sector_t, struct buffer_head *, int); + extern int jfs_setattr(struct dentry *, struct iattr *); + +diff -NurpP --minimal linux-3.0.9/fs/jfs/namei.c linux-3.0.9-vs2.3.2.1/fs/jfs/namei.c +--- linux-3.0.9/fs/jfs/namei.c 2011-05-22 16:17:53.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/jfs/namei.c 2011-06-10 22:11:24.000000000 +0200 +@@ -22,6 +22,7 @@ + #include <linux/ctype.h> + #include <linux/quotaops.h> + #include <linux/exportfs.h> ++#include <linux/vs_tag.h> + #include "jfs_incore.h" + #include "jfs_superblock.h" + #include "jfs_inode.h" +@@ -1486,6 +1487,7 @@ static struct dentry *jfs_lookup(struct + return ERR_CAST(ip); + } + ++ dx_propagate_tag(nd, ip); + return d_splice_alias(ip, dentry); + } + +@@ -1550,6 +1552,7 @@ const struct inode_operations jfs_dir_in + #ifdef CONFIG_JFS_POSIX_ACL + .check_acl = jfs_check_acl, + #endif ++ .sync_flags = jfs_sync_flags, + }; + + const struct file_operations jfs_dir_operations = { +diff -NurpP --minimal linux-3.0.9/fs/jfs/super.c linux-3.0.9-vs2.3.2.1/fs/jfs/super.c +--- linux-3.0.9/fs/jfs/super.c 2011-05-22 16:17:53.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/jfs/super.c 2011-06-10 22:11:24.000000000 +0200 +@@ -198,7 +198,8 @@ static void jfs_put_super(struct super_b + enum { + Opt_integrity, Opt_nointegrity, Opt_iocharset, Opt_resize, + Opt_resize_nosize, Opt_errors, Opt_ignore, Opt_err, Opt_quota, +- Opt_usrquota, Opt_grpquota, Opt_uid, Opt_gid, Opt_umask ++ Opt_usrquota, Opt_grpquota, Opt_uid, Opt_gid, Opt_umask, ++ Opt_tag, Opt_notag, Opt_tagid + }; + + static const match_table_t tokens = { +@@ -208,6 +209,10 @@ static const match_table_t tokens = { + {Opt_resize, "resize=%u"}, + {Opt_resize_nosize, "resize"}, + {Opt_errors, "errors=%s"}, ++ {Opt_tag, "tag"}, ++ {Opt_notag, "notag"}, ++ {Opt_tagid, "tagid=%u"}, ++ {Opt_tag, "tagxid"}, + {Opt_ignore, "noquota"}, + {Opt_ignore, "quota"}, + {Opt_usrquota, "usrquota"}, +@@ -342,6 +347,20 @@ static int parse_options(char *options, + } + break; + } ++#ifndef CONFIG_TAGGING_NONE ++ case Opt_tag: ++ *flag |= JFS_TAGGED; ++ break; ++ case Opt_notag: ++ *flag &= JFS_TAGGED; ++ break; ++#endif ++#ifdef CONFIG_PROPAGATE ++ case Opt_tagid: ++ /* use args[0] */ ++ *flag |= JFS_TAGGED; ++ break; ++#endif + default: + printk("jfs: Unrecognized mount option \"%s\" " + " or missing value\n", p); +@@ -373,6 +392,12 @@ static int jfs_remount(struct super_bloc + return -EINVAL; + } + ++ if ((flag & JFS_TAGGED) && !(sb->s_flags & MS_TAGGED)) { ++ printk(KERN_ERR "JFS: %s: tagging not permitted on remount.\n", ++ sb->s_id); ++ return -EINVAL; ++ } ++ + if (newLVSize) { + if (sb->s_flags & MS_RDONLY) { + printk(KERN_ERR +@@ -455,6 +480,9 @@ static int jfs_fill_super(struct super_b + #ifdef CONFIG_JFS_POSIX_ACL + sb->s_flags |= MS_POSIXACL; + #endif ++ /* map mount option tagxid */ ++ if (sbi->flag & JFS_TAGGED) ++ sb->s_flags |= MS_TAGGED; + + if (newLVSize) { + printk(KERN_ERR "resize option for remount only\n"); +diff -NurpP --minimal linux-3.0.9/fs/libfs.c linux-3.0.9-vs2.3.2.1/fs/libfs.c +--- linux-3.0.9/fs/libfs.c 2011-07-22 11:18:05.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/libfs.c 2011-07-22 11:20:39.000000000 +0200 +@@ -133,7 +133,8 @@ static inline unsigned char dt_type(stru + * both impossible due to the lock on directory. + */ + +-int dcache_readdir(struct file * filp, void * dirent, filldir_t filldir) ++static inline int do_dcache_readdir_filter(struct file *filp, ++ void *dirent, filldir_t filldir, int (*filter)(struct dentry *dentry)) + { + struct dentry *dentry = filp->f_path.dentry; + struct dentry *cursor = filp->private_data; +@@ -164,6 +165,8 @@ int dcache_readdir(struct file * filp, v + for (p=q->next; p != &dentry->d_subdirs; p=p->next) { + struct dentry *next; + next = list_entry(p, struct dentry, d_u.d_child); ++ if (filter && !filter(next)) ++ continue; + spin_lock_nested(&next->d_lock, DENTRY_D_LOCK_NESTED); + if (!simple_positive(next)) { + spin_unlock(&next->d_lock); +@@ -190,6 +193,17 @@ int dcache_readdir(struct file * filp, v + return 0; + } + ++int dcache_readdir(struct file *filp, void *dirent, filldir_t filldir) ++{ ++ return do_dcache_readdir_filter(filp, dirent, filldir, NULL); ++} ++ ++int dcache_readdir_filter(struct file *filp, void *dirent, filldir_t filldir, ++ int (*filter)(struct dentry *)) ++{ ++ return do_dcache_readdir_filter(filp, dirent, filldir, filter); ++} ++ + ssize_t generic_read_dir(struct file *filp, char __user *buf, size_t siz, loff_t *ppos) + { + return -EISDIR; +@@ -965,6 +979,7 @@ EXPORT_SYMBOL(dcache_dir_close); + EXPORT_SYMBOL(dcache_dir_lseek); + EXPORT_SYMBOL(dcache_dir_open); + EXPORT_SYMBOL(dcache_readdir); ++EXPORT_SYMBOL(dcache_readdir_filter); + EXPORT_SYMBOL(generic_read_dir); + EXPORT_SYMBOL(mount_pseudo); + EXPORT_SYMBOL(simple_write_begin); +diff -NurpP --minimal linux-3.0.9/fs/locks.c linux-3.0.9-vs2.3.2.1/fs/locks.c +--- linux-3.0.9/fs/locks.c 2011-07-22 11:18:05.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/locks.c 2011-07-19 00:51:58.000000000 +0200 +@@ -126,6 +126,8 @@ + #include <linux/time.h> + #include <linux/rcupdate.h> + #include <linux/pid_namespace.h> ++#include <linux/vs_base.h> ++#include <linux/vs_limit.h> + + #include <asm/uaccess.h> + +@@ -171,13 +173,19 @@ static void locks_init_lock_always(struc + fl->fl_flags = 0; + fl->fl_type = 0; + fl->fl_start = fl->fl_end = 0; ++ fl->fl_xid = -1; + } + ++ + /* Allocate an empty lock structure. */ + struct file_lock *locks_alloc_lock(void) + { +- struct file_lock *fl = kmem_cache_alloc(filelock_cache, GFP_KERNEL); ++ struct file_lock *fl; ++ ++ if (!vx_locks_avail(1)) ++ return NULL; + ++ fl = kmem_cache_alloc(filelock_cache, GFP_KERNEL); + if (fl) + locks_init_lock_always(fl); + +@@ -208,6 +216,7 @@ void locks_free_lock(struct file_lock *f + BUG_ON(!list_empty(&fl->fl_block)); + BUG_ON(!list_empty(&fl->fl_link)); + ++ vx_locks_dec(fl); + locks_release_private(fl); + kmem_cache_free(filelock_cache, fl); + } +@@ -272,6 +281,7 @@ void locks_copy_lock(struct file_lock *n + new->fl_file = fl->fl_file; + new->fl_ops = fl->fl_ops; + new->fl_lmops = fl->fl_lmops; ++ new->fl_xid = fl->fl_xid; + + locks_copy_private(new, fl); + } +@@ -310,6 +320,11 @@ static int flock_make_lock(struct file * + fl->fl_flags = FL_FLOCK; + fl->fl_type = type; + fl->fl_end = OFFSET_MAX; ++ ++ vxd_assert(filp->f_xid == vx_current_xid(), ++ "f_xid(%d) == current(%d)", filp->f_xid, vx_current_xid()); ++ fl->fl_xid = filp->f_xid; ++ vx_locks_inc(fl); + + *lock = fl; + return 0; +@@ -459,6 +474,7 @@ static int lease_init(struct file *filp, + + fl->fl_owner = current->files; + fl->fl_pid = current->tgid; ++ fl->fl_xid = vx_current_xid(); + + fl->fl_file = filp; + fl->fl_flags = FL_LEASE; +@@ -478,6 +494,11 @@ static struct file_lock *lease_alloc(str + if (fl == NULL) + return ERR_PTR(error); + ++ fl->fl_xid = vx_current_xid(); ++ if (filp) ++ vxd_assert(filp->f_xid == fl->fl_xid, ++ "f_xid(%d) == fl_xid(%d)", filp->f_xid, fl->fl_xid); ++ vx_locks_inc(fl); + error = lease_init(filp, type, fl); + if (error) { + locks_free_lock(fl); +@@ -779,6 +800,7 @@ static int flock_lock_file(struct file * + lock_flocks(); + } + ++ new_fl->fl_xid = -1; + find_conflict: + for_each_lock(inode, before) { + struct file_lock *fl = *before; +@@ -799,6 +821,7 @@ find_conflict: + goto out; + locks_copy_lock(new_fl, request); + locks_insert_lock(before, new_fl); ++ vx_locks_inc(new_fl); + new_fl = NULL; + error = 0; + +@@ -809,7 +832,8 @@ out: + return error; + } + +-static int __posix_lock_file(struct inode *inode, struct file_lock *request, struct file_lock *conflock) ++static int __posix_lock_file(struct inode *inode, struct file_lock *request, ++ struct file_lock *conflock, xid_t xid) + { + struct file_lock *fl; + struct file_lock *new_fl = NULL; +@@ -819,6 +843,8 @@ static int __posix_lock_file(struct inod + struct file_lock **before; + int error, added = 0; + ++ vxd_assert(xid == vx_current_xid(), ++ "xid(%d) == current(%d)", xid, vx_current_xid()); + /* + * We may need two file_lock structures for this operation, + * so we get them in advance to avoid races. +@@ -829,7 +855,11 @@ static int __posix_lock_file(struct inod + (request->fl_type != F_UNLCK || + request->fl_start != 0 || request->fl_end != OFFSET_MAX)) { + new_fl = locks_alloc_lock(); ++ new_fl->fl_xid = xid; ++ vx_locks_inc(new_fl); + new_fl2 = locks_alloc_lock(); ++ new_fl2->fl_xid = xid; ++ vx_locks_inc(new_fl2); + } + + lock_flocks(); +@@ -1028,7 +1058,8 @@ static int __posix_lock_file(struct inod + int posix_lock_file(struct file *filp, struct file_lock *fl, + struct file_lock *conflock) + { +- return __posix_lock_file(filp->f_path.dentry->d_inode, fl, conflock); ++ return __posix_lock_file(filp->f_path.dentry->d_inode, ++ fl, conflock, filp->f_xid); + } + EXPORT_SYMBOL(posix_lock_file); + +@@ -1118,7 +1149,7 @@ int locks_mandatory_area(int read_write, + fl.fl_end = offset + count - 1; + + for (;;) { +- error = __posix_lock_file(inode, &fl, NULL); ++ error = __posix_lock_file(inode, &fl, NULL, filp->f_xid); + if (error != FILE_LOCK_DEFERRED) + break; + error = wait_event_interruptible(fl.fl_wait, !fl.fl_next); +@@ -1431,6 +1462,7 @@ int generic_setlease(struct file *filp, + goto out; + + locks_insert_lock(before, lease); ++ vx_locks_inc(lease); + return 0; + + out: +@@ -1815,6 +1847,11 @@ int fcntl_setlk(unsigned int fd, struct + if (file_lock == NULL) + return -ENOLCK; + ++ vxd_assert(filp->f_xid == vx_current_xid(), ++ "f_xid(%d) == current(%d)", filp->f_xid, vx_current_xid()); ++ file_lock->fl_xid = filp->f_xid; ++ vx_locks_inc(file_lock); ++ + /* + * This might block, so we do it before checking the inode. + */ +@@ -1933,6 +1970,11 @@ int fcntl_setlk64(unsigned int fd, struc + if (file_lock == NULL) + return -ENOLCK; + ++ vxd_assert(filp->f_xid == vx_current_xid(), ++ "f_xid(%d) == current(%d)", filp->f_xid, vx_current_xid()); ++ file_lock->fl_xid = filp->f_xid; ++ vx_locks_inc(file_lock); ++ + /* + * This might block, so we do it before checking the inode. + */ +@@ -2198,8 +2240,11 @@ static int locks_show(struct seq_file *f + + lock_get_status(f, fl, *((loff_t *)f->private), ""); + +- list_for_each_entry(bfl, &fl->fl_block, fl_block) ++ list_for_each_entry(bfl, &fl->fl_block, fl_block) { ++ if (!vx_check(fl->fl_xid, VS_WATCH_P | VS_IDENT)) ++ continue; + lock_get_status(f, bfl, *((loff_t *)f->private), " ->"); ++ } + + return 0; + } +diff -NurpP --minimal linux-3.0.9/fs/namei.c linux-3.0.9-vs2.3.2.1/fs/namei.c +--- linux-3.0.9/fs/namei.c 2011-11-15 16:40:47.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/namei.c 2011-11-16 14:27:47.000000000 +0100 +@@ -32,6 +32,14 @@ + #include <linux/fcntl.h> + #include <linux/device_cgroup.h> + #include <linux/fs_struct.h> ++#include <linux/proc_fs.h> ++#include <linux/vserver/inode.h> ++#include <linux/vs_base.h> ++#include <linux/vs_tag.h> ++#include <linux/vs_cowbl.h> ++#include <linux/vs_device.h> ++#include <linux/vs_context.h> ++#include <linux/pid_namespace.h> + #include <asm/uaccess.h> + + #include "internal.h" +@@ -175,6 +183,89 @@ void putname(const char *name) + EXPORT_SYMBOL(putname); + #endif + ++static inline int dx_barrier(const struct inode *inode) ++{ ++ if (IS_BARRIER(inode) && !vx_check(0, VS_ADMIN | VS_WATCH)) { ++ vxwprintk_task(1, "did hit the barrier."); ++ return 1; ++ } ++ return 0; ++} ++ ++static int __dx_permission(const struct inode *inode, int mask) ++{ ++ if (dx_barrier(inode)) ++ return -EACCES; ++ ++ if (inode->i_sb->s_magic == DEVPTS_SUPER_MAGIC) { ++ /* devpts is xid tagged */ ++ if (S_ISDIR(inode->i_mode) || ++ vx_check((xid_t)inode->i_tag, VS_IDENT | VS_WATCH_P)) ++ return 0; ++ ++ /* just pretend we didn't find anything */ ++ return -ENOENT; ++ } ++ else if (inode->i_sb->s_magic == PROC_SUPER_MAGIC) { ++ struct proc_dir_entry *de = PDE(inode); ++ ++ if (de && !vx_hide_check(0, de->vx_flags)) ++ goto out; ++ ++ if ((mask & (MAY_WRITE | MAY_APPEND))) { ++ struct pid *pid; ++ struct task_struct *tsk; ++ ++ if (vx_check(0, VS_ADMIN | VS_WATCH_P) || ++ vx_flags(VXF_STATE_SETUP, 0)) ++ return 0; ++ ++ pid = PROC_I(inode)->pid; ++ if (!pid) ++ goto out; ++ ++ rcu_read_lock(); ++ tsk = pid_task(pid, PIDTYPE_PID); ++ vxdprintk(VXD_CBIT(tag, 0), "accessing %p[#%u]", ++ tsk, (tsk ? vx_task_xid(tsk) : 0)); ++ if (tsk && ++ vx_check(vx_task_xid(tsk), VS_IDENT | VS_WATCH_P)) { ++ rcu_read_unlock(); ++ return 0; ++ } ++ rcu_read_unlock(); ++ } ++ else { ++ /* FIXME: Should we block some entries here? */ ++ return 0; ++ } ++ } ++ else { ++ if (dx_notagcheck(inode->i_sb) || ++ dx_check(inode->i_tag, DX_HOSTID | DX_ADMIN | DX_WATCH | ++ DX_IDENT)) ++ return 0; ++ } ++ ++out: ++ return -EACCES; ++} ++ ++int dx_permission(const struct inode *inode, int mask) ++{ ++ int ret = __dx_permission(inode, mask); ++ if (unlikely(ret)) { ++#ifndef CONFIG_VSERVER_WARN_DEVPTS ++ if (inode->i_sb->s_magic != DEVPTS_SUPER_MAGIC) ++#endif ++ vxwprintk_task(1, ++ "denied [0x%x] access to inode %s:%p[#%d,%lu]", ++ mask, inode->i_sb->s_id, inode, inode->i_tag, ++ inode->i_ino); ++ } ++ return ret; ++} ++ + /* + * This does basic POSIX ACL permission checking + */ +@@ -285,10 +376,14 @@ int inode_permission(struct inode *inode + /* + * Nobody gets write access to an immutable file. + */ +- if (IS_IMMUTABLE(inode)) ++ if (IS_IMMUTABLE(inode) && !IS_COW(inode)) + return -EACCES; + } + ++ retval = dx_permission(inode, mask); ++ if (retval) ++ return retval; ++ + if (inode->i_op->permission) + retval = inode->i_op->permission(inode, mask, 0); + else +@@ -584,6 +679,9 @@ static inline int exec_permission(struct + int ret; + struct user_namespace *ns = inode_userns(inode); + ++ if (dx_barrier(inode)) ++ return -EACCES; ++ + if (inode->i_op->permission) { + ret = inode->i_op->permission(inode, MAY_EXEC, flags); + } else { +@@ -1095,7 +1193,8 @@ static void follow_dotdot(struct nameida + + if (nd->path.dentry == nd->root.dentry && + nd->path.mnt == nd->root.mnt) { +- break; ++ /* for sane '/' avoid follow_mount() */ ++ return; + } + if (nd->path.dentry != nd->path.mnt->mnt_root) { + /* rare case of legitimate dget_parent()... */ +@@ -1178,6 +1277,9 @@ static int do_lookup(struct nameidata *n + goto unlazy; + } + } ++ ++ /* FIXME: check dx permission */ ++ + path->mnt = mnt; + path->dentry = dentry; + if (unlikely(!__follow_mount_rcu(nd, path, inode))) +@@ -1226,6 +1328,8 @@ retry: + } + } + ++ /* FIXME: check dx permission */ ++ + path->mnt = mnt; + path->dentry = dentry; + err = follow_managed(path, nd->flags); +@@ -1853,7 +1957,7 @@ static int may_delete(struct inode *dir, + if (IS_APPEND(dir)) + return -EPERM; + if (check_sticky(dir, victim->d_inode)||IS_APPEND(victim->d_inode)|| +- IS_IMMUTABLE(victim->d_inode) || IS_SWAPFILE(victim->d_inode)) ++ IS_IXORUNLINK(victim->d_inode) || IS_SWAPFILE(victim->d_inode)) + return -EPERM; + if (isdir) { + if (!S_ISDIR(victim->d_inode->i_mode)) +@@ -1980,6 +2084,15 @@ static int may_open(struct path *path, i + break; + } + ++#ifdef CONFIG_VSERVER_COWBL ++ if (IS_COW(inode) && ++ ((flag & O_ACCMODE) != O_RDONLY)) { ++ if (IS_COW_LINK(inode)) ++ return -EMLINK; ++ inode->i_flags &= ~(S_IXUNLINK|S_IMMUTABLE); ++ mark_inode_dirty(inode); ++ } ++#endif + error = inode_permission(inode, acc_mode); + if (error) + return error; +@@ -2224,6 +2337,16 @@ ok: + } + common: + error = may_open(&nd->path, acc_mode, open_flag); ++#ifdef CONFIG_VSERVER_COWBL ++ if (error == -EMLINK) { ++ struct dentry *dentry; ++ dentry = cow_break_link(pathname); ++ if (IS_ERR(dentry)) ++ error = PTR_ERR(dentry); ++ else ++ dput(dentry); ++ } ++#endif + if (error) + goto exit; + filp = nameidata_to_filp(nd); +@@ -2266,6 +2389,7 @@ static struct file *path_openat(int dfd, + struct path path; + int error; + ++restart: + filp = get_empty_filp(); + if (!filp) + return ERR_PTR(-ENFILE); +@@ -2303,6 +2427,17 @@ static struct file *path_openat(int dfd, + filp = do_last(nd, &path, op, pathname); + put_link(nd, &link, cookie); + } ++ ++#ifdef CONFIG_VSERVER_COWBL ++ if (filp == ERR_PTR(-EMLINK)) { ++ if (nd->root.mnt && !(nd->flags & LOOKUP_ROOT)) ++ path_put(&nd->root); ++ if (base) ++ fput(base); ++ release_open_intent(nd); ++ goto restart; ++ } ++#endif + out: + if (nd->root.mnt && !(nd->flags & LOOKUP_ROOT)) + path_put(&nd->root); +@@ -2886,7 +3021,7 @@ int vfs_link(struct dentry *old_dentry, + /* + * A link to an append-only or immutable file cannot be created. + */ +- if (IS_APPEND(inode) || IS_IMMUTABLE(inode)) ++ if (IS_APPEND(inode) || IS_IXORUNLINK(inode)) + return -EPERM; + if (!dir->i_op->link) + return -EPERM; +@@ -3274,6 +3409,227 @@ int vfs_follow_link(struct nameidata *nd + return __vfs_follow_link(nd, link); + } + ++ ++#ifdef CONFIG_VSERVER_COWBL ++ ++static inline ++long do_cow_splice(struct file *in, struct file *out, size_t len) ++{ ++ loff_t ppos = 0; ++ ++ return do_splice_direct(in, &ppos, out, len, 0); ++} ++ ++struct dentry *cow_break_link(const char *pathname) ++{ ++ int ret, mode, pathlen, redo = 0; ++ struct nameidata old_nd, dir_nd; ++ struct path old_path; ++ struct dentry *dir, *old_dentry, *new_dentry = NULL; ++ struct file *old_file; ++ struct file *new_file; ++ char *to, *path, pad='\251'; ++ loff_t size; ++ ++ vxdprintk(VXD_CBIT(misc, 1), ++ "cow_break_link(" VS_Q("%s") ")", pathname); ++ path = kmalloc(PATH_MAX, GFP_KERNEL); ++ ret = -ENOMEM; ++ if (!path) ++ goto out; ++ ++ /* old_nd will have refs to dentry and mnt */ ++ ret = do_path_lookup(AT_FDCWD, pathname, LOOKUP_FOLLOW, &old_nd); ++ vxdprintk(VXD_CBIT(misc, 2), ++ "do_path_lookup(old): %d [r=%d]", ++ ret, mnt_get_count(old_nd.path.mnt)); ++ if (ret < 0) ++ goto out_free_path; ++ ++ old_path = old_nd.path; ++ old_dentry = old_path.dentry; ++ mode = old_dentry->d_inode->i_mode; ++ ++ to = d_path(&old_path, path, PATH_MAX-2); ++ pathlen = strlen(to); ++ vxdprintk(VXD_CBIT(misc, 2), ++ "old path " VS_Q("%s") " [%p:" VS_Q("%.*s") ":%d]", to, ++ old_dentry, ++ old_dentry->d_name.len, old_dentry->d_name.name, ++ old_dentry->d_name.len); ++ ++ to[pathlen + 1] = 0; ++retry: ++ new_dentry = NULL; ++ to[pathlen] = pad--; ++ ret = -ELOOP; ++ if (pad <= '\240') ++ goto out_rel_old; ++ ++ vxdprintk(VXD_CBIT(misc, 1), "temp copy " VS_Q("%s"), to); ++ /* dir_nd will have refs to dentry and mnt */ ++ ret = do_path_lookup(AT_FDCWD, to, ++ LOOKUP_PARENT | LOOKUP_OPEN | LOOKUP_CREATE, &dir_nd); ++ vxdprintk(VXD_CBIT(misc, 2), "do_path_lookup(new): %d", ret); ++ if (ret < 0) ++ goto retry; ++ ++ /* this puppy downs the inode mutex */ ++ new_dentry = lookup_create(&dir_nd, 0); ++ if (!new_dentry || IS_ERR(new_dentry)) { ++ path_put(&dir_nd.path); ++ vxdprintk(VXD_CBIT(misc, 2), ++ "lookup_create(new): failed with %ld", ++ PTR_ERR(new_dentry)); ++ mutex_unlock(&dir_nd.path.dentry->d_inode->i_mutex); ++ goto retry; ++ } ++ vxdprintk(VXD_CBIT(misc, 2), ++ "lookup_create(new): %p [" VS_Q("%.*s") ":%d]", ++ new_dentry, ++ new_dentry->d_name.len, new_dentry->d_name.name, ++ new_dentry->d_name.len); ++ ++ dir = dir_nd.path.dentry; ++ ++ ret = vfs_create(dir->d_inode, new_dentry, mode, &dir_nd); ++ vxdprintk(VXD_CBIT(misc, 2), ++ "vfs_create(new): %d", ret); ++ if (ret == -EEXIST) { ++ mutex_unlock(&dir->d_inode->i_mutex); ++ path_put(&dir_nd.path); ++ dput(new_dentry); ++ goto retry; ++ } ++ else if (ret < 0) ++ goto out_unlock_new; ++ ++ /* drop out early, ret passes ENOENT */ ++ ret = -ENOENT; ++ if ((redo = d_unhashed(old_dentry))) ++ goto out_unlock_new; ++ ++ path_get(&old_path); ++ /* this one cleans up the dentry/mnt in case of failure */ ++ old_file = dentry_open(old_dentry, old_path.mnt, ++ O_RDONLY, current_cred()); ++ vxdprintk(VXD_CBIT(misc, 2), ++ "dentry_open(old): %p", old_file); ++ if (IS_ERR(old_file)) { ++ ret = PTR_ERR(old_file); ++ goto out_unlock_new; ++ } ++ ++ dget(new_dentry); ++ mntget(old_path.mnt); ++ /* this one cleans up the dentry/mnt in case of failure */ ++ new_file = dentry_open(new_dentry, old_path.mnt, ++ O_WRONLY, current_cred()); ++ vxdprintk(VXD_CBIT(misc, 2), ++ "dentry_open(new): %p", new_file); ++ if (IS_ERR(new_file)) { ++ ret = PTR_ERR(new_file); ++ goto out_fput_old; ++ } ++ ++ size = i_size_read(old_file->f_dentry->d_inode); ++ ret = do_cow_splice(old_file, new_file, size); ++ vxdprintk(VXD_CBIT(misc, 2), "do_splice_direct: %d", ret); ++ if (ret < 0) { ++ goto out_fput_both; ++ } else if (ret < size) { ++ ret = -ENOSPC; ++ goto out_fput_both; ++ } else { ++ struct inode *old_inode = old_dentry->d_inode; ++ struct inode *new_inode = new_dentry->d_inode; ++ struct iattr attr = { ++ .ia_uid = old_inode->i_uid, ++ .ia_gid = old_inode->i_gid, ++ .ia_valid = ATTR_UID | ATTR_GID ++ }; ++ ++ setattr_copy(new_inode, &attr); ++ mark_inode_dirty(new_inode); ++ } ++ ++ mutex_lock(&old_dentry->d_inode->i_sb->s_vfs_rename_mutex); ++ ++ /* drop out late */ ++ ret = -ENOENT; ++ if ((redo = d_unhashed(old_dentry))) ++ goto out_unlock; ++ ++ vxdprintk(VXD_CBIT(misc, 2), ++ "vfs_rename: [" VS_Q("%*s") ":%d] -> [" VS_Q("%*s") ":%d]", ++ new_dentry->d_name.len, new_dentry->d_name.name, ++ new_dentry->d_name.len, ++ old_dentry->d_name.len, old_dentry->d_name.name, ++ old_dentry->d_name.len); ++ ret = vfs_rename(dir_nd.path.dentry->d_inode, new_dentry, ++ old_dentry->d_parent->d_inode, old_dentry); ++ vxdprintk(VXD_CBIT(misc, 2), "vfs_rename: %d", ret); ++ ++out_unlock: ++ mutex_unlock(&old_dentry->d_inode->i_sb->s_vfs_rename_mutex); ++ ++out_fput_both: ++ vxdprintk(VXD_CBIT(misc, 3), ++ "fput(new_file=%p[#%ld])", new_file, ++ atomic_long_read(&new_file->f_count)); ++ fput(new_file); ++ ++out_fput_old: ++ vxdprintk(VXD_CBIT(misc, 3), ++ "fput(old_file=%p[#%ld])", old_file, ++ atomic_long_read(&old_file->f_count)); ++ fput(old_file); ++ ++out_unlock_new: ++ mutex_unlock(&dir->d_inode->i_mutex); ++ if (!ret) ++ goto out_redo; ++ ++ /* error path cleanup */ ++ vfs_unlink(dir->d_inode, new_dentry); ++ ++out_redo: ++ if (!redo) ++ goto out_rel_both; ++ /* lookup dentry once again */ ++ /* old_nd.path is freed as old_path in out_rel_old */ ++ ret = do_path_lookup(AT_FDCWD, pathname, LOOKUP_FOLLOW, &old_nd); ++ if (ret) ++ goto out_rel_both; ++ ++ dput(new_dentry); ++ new_dentry = old_nd.path.dentry; ++ vxdprintk(VXD_CBIT(misc, 2), ++ "do_path_lookup(redo): %p [" VS_Q("%.*s") ":%d]", ++ new_dentry, ++ new_dentry->d_name.len, new_dentry->d_name.name, ++ new_dentry->d_name.len); ++ dget(new_dentry); ++ ++out_rel_both: ++ path_put(&dir_nd.path); ++out_rel_old: ++ path_put(&old_path); ++out_free_path: ++ kfree(path); ++out: ++ if (ret) { ++ dput(new_dentry); ++ new_dentry = ERR_PTR(ret); ++ } ++ vxdprintk(VXD_CBIT(misc, 3), ++ "cow_break_link returning with %p [r=%d]", ++ new_dentry, mnt_get_count(old_nd.path.mnt)); ++ return new_dentry; ++} ++ ++#endif ++ + /* get the link contents into pagecache */ + static char *page_getlink(struct dentry * dentry, struct page **ppage) + { +diff -NurpP --minimal linux-3.0.9/fs/namespace.c linux-3.0.9-vs2.3.2.1/fs/namespace.c +--- linux-3.0.9/fs/namespace.c 2011-11-15 16:40:47.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/namespace.c 2011-11-15 17:37:39.000000000 +0100 +@@ -31,6 +31,11 @@ + #include <linux/idr.h> + #include <linux/fs_struct.h> + #include <linux/fsnotify.h> ++#include <linux/vs_base.h> ++#include <linux/vs_context.h> ++#include <linux/vs_tag.h> ++#include <linux/vserver/space.h> ++#include <linux/vserver/global.h> + #include <asm/uaccess.h> + #include <asm/unistd.h> + #include "pnode.h" +@@ -679,6 +684,10 @@ vfs_kern_mount(struct file_system_type * + if (!type) + return ERR_PTR(-ENODEV); + ++ if ((type->fs_flags & FS_BINARY_MOUNTDATA) && ++ !vx_capable(CAP_SYS_ADMIN, VXC_BINARY_MOUNT)) ++ return ERR_PTR(-EPERM); ++ + mnt = alloc_vfsmnt(name); + if (!mnt) + return ERR_PTR(-ENOMEM); +@@ -724,6 +733,7 @@ static struct vfsmount *clone_mnt(struct + mnt->mnt_root = dget(root); + mnt->mnt_mountpoint = mnt->mnt_root; + mnt->mnt_parent = mnt; ++ mnt->mnt_tag = old->mnt_tag; + + if (flag & CL_SLAVE) { + list_add(&mnt->mnt_slave, &old->mnt_slave_list); +@@ -852,6 +862,31 @@ static inline void mangle(struct seq_fil + seq_escape(m, s, " \t\n\\"); + } + ++static int mnt_is_reachable(struct vfsmount *mnt) ++{ ++ struct path root; ++ struct dentry *point; ++ int ret; ++ ++ if (mnt == mnt->mnt_ns->root) ++ return 1; ++ ++ br_read_lock(vfsmount_lock); ++ root = current->fs->root; ++ point = root.dentry; ++ ++ while ((mnt != mnt->mnt_parent) && (mnt != root.mnt)) { ++ point = mnt->mnt_mountpoint; ++ mnt = mnt->mnt_parent; ++ } ++ ++ ret = (mnt == root.mnt) && is_subdir(point, root.dentry); ++ ++ br_read_unlock(vfsmount_lock); ++ ++ return ret; ++} ++ + /* + * Simple .show_options callback for filesystems which don't want to + * implement more complex mount option showing. +@@ -954,6 +989,8 @@ static int show_sb_opts(struct seq_file + { MS_SYNCHRONOUS, ",sync" }, + { MS_DIRSYNC, ",dirsync" }, + { MS_MANDLOCK, ",mand" }, ++ { MS_TAGGED, ",tag" }, ++ { MS_NOTAGCHECK, ",notagcheck" }, + { 0, NULL } + }; + const struct proc_fs_info *fs_infop; +@@ -1000,16 +1037,26 @@ static int show_vfsmnt(struct seq_file * + int err = 0; + struct path mnt_path = { .dentry = mnt->mnt_root, .mnt = mnt }; + +- if (mnt->mnt_sb->s_op->show_devname) { +- err = mnt->mnt_sb->s_op->show_devname(m, mnt); +- if (err) +- goto out; ++ if (vx_flags(VXF_HIDE_MOUNT, 0)) ++ return SEQ_SKIP; ++ if (!mnt_is_reachable(mnt) && !vx_check(0, VS_WATCH_P)) ++ return SEQ_SKIP; ++ ++ if (!vx_check(0, VS_ADMIN|VS_WATCH) && ++ mnt == current->fs->root.mnt) { ++ seq_puts(m, "/dev/root / "); + } else { +- mangle(m, mnt->mnt_devname ? mnt->mnt_devname : "none"); ++ if (mnt->mnt_sb->s_op->show_devname) { ++ err = mnt->mnt_sb->s_op->show_devname(m, mnt); ++ if (err) ++ goto out; ++ } else { ++ mangle(m, mnt->mnt_devname ? mnt->mnt_devname : "none"); ++ } ++ seq_putc(m, ' '); ++ seq_path(m, &mnt_path, " \t\n\\"); ++ seq_putc(m, ' '); + } +- seq_putc(m, ' '); +- seq_path(m, &mnt_path, " \t\n\\"); +- seq_putc(m, ' '); + show_type(m, mnt->mnt_sb); + seq_puts(m, __mnt_is_readonly(mnt) ? " ro" : " rw"); + err = show_sb_opts(m, mnt->mnt_sb); +@@ -1039,6 +1086,11 @@ static int show_mountinfo(struct seq_fil + struct path root = p->root; + int err = 0; + ++ if (vx_flags(VXF_HIDE_MOUNT, 0)) ++ return SEQ_SKIP; ++ if (!mnt_is_reachable(mnt) && !vx_check(0, VS_WATCH_P)) ++ return SEQ_SKIP; ++ + seq_printf(m, "%i %i %u:%u ", mnt->mnt_id, mnt->mnt_parent->mnt_id, + MAJOR(sb->s_dev), MINOR(sb->s_dev)); + if (sb->s_op->show_path) +@@ -1107,22 +1159,32 @@ static int show_vfsstat(struct seq_file + struct path mnt_path = { .dentry = mnt->mnt_root, .mnt = mnt }; + int err = 0; + +- /* device */ +- if (mnt->mnt_sb->s_op->show_devname) { +- seq_puts(m, "device "); +- err = mnt->mnt_sb->s_op->show_devname(m, mnt); ++ if (vx_flags(VXF_HIDE_MOUNT, 0)) ++ return SEQ_SKIP; ++ if (!mnt_is_reachable(mnt) && !vx_check(0, VS_WATCH_P)) ++ return SEQ_SKIP; ++ ++ if (!vx_check(0, VS_ADMIN|VS_WATCH) && ++ mnt == current->fs->root.mnt) { ++ seq_puts(m, "device /dev/root mounted on / "); + } else { +- if (mnt->mnt_devname) { ++ /* device */ ++ if (mnt->mnt_sb->s_op->show_devname) { + seq_puts(m, "device "); +- mangle(m, mnt->mnt_devname); +- } else +- seq_puts(m, "no device"); +- } ++ err = mnt->mnt_sb->s_op->show_devname(m, mnt); ++ } else { ++ if (mnt->mnt_devname) { ++ seq_puts(m, "device "); ++ mangle(m, mnt->mnt_devname); ++ } else ++ seq_puts(m, "no device"); ++ } + +- /* mount point */ +- seq_puts(m, " mounted on "); +- seq_path(m, &mnt_path, " \t\n\\"); +- seq_putc(m, ' '); ++ /* mount point */ ++ seq_puts(m, " mounted on "); ++ seq_path(m, &mnt_path, " \t\n\\"); ++ seq_putc(m, ' '); ++ } + + /* file system type */ + seq_puts(m, "with fstype "); +@@ -1381,7 +1443,7 @@ SYSCALL_DEFINE2(umount, char __user *, n + goto dput_and_out; + + retval = -EPERM; +- if (!capable(CAP_SYS_ADMIN)) ++ if (!vx_capable(CAP_SYS_ADMIN, VXC_SECURE_MOUNT)) + goto dput_and_out; + + retval = do_umount(path.mnt, flags); +@@ -1407,7 +1469,7 @@ SYSCALL_DEFINE1(oldumount, char __user * + + static int mount_is_safe(struct path *path) + { +- if (capable(CAP_SYS_ADMIN)) ++ if (vx_capable(CAP_SYS_ADMIN, VXC_SECURE_MOUNT)) + return 0; + return -EPERM; + #ifdef notyet +@@ -1717,7 +1779,7 @@ static int do_change_type(struct path *p + int type; + int err = 0; + +- if (!capable(CAP_SYS_ADMIN)) ++ if (!vx_capable(CAP_SYS_ADMIN, VXC_NAMESPACE)) + return -EPERM; + + if (path->dentry != path->mnt->mnt_root) +@@ -1733,6 +1795,7 @@ static int do_change_type(struct path *p + if (err) + goto out_unlock; + } ++ // mnt->mnt_flags = mnt_flags; + + br_write_lock(vfsmount_lock); + for (m = mnt; m; m = (recurse ? next_mnt(m, mnt) : NULL)) +@@ -1748,12 +1811,14 @@ static int do_change_type(struct path *p + * do loopback mount. + */ + static int do_loopback(struct path *path, char *old_name, +- int recurse) ++ tag_t tag, unsigned long flags, int mnt_flags) + { + LIST_HEAD(umount_list); + struct path old_path; + struct vfsmount *mnt = NULL; + int err = mount_is_safe(path); ++ int recurse = flags & MS_REC; ++ + if (err) + return err; + if (!old_name || !*old_name) +@@ -1819,12 +1884,12 @@ static int change_mount_flags(struct vfs + * on it - tough luck. + */ + static int do_remount(struct path *path, int flags, int mnt_flags, +- void *data) ++ void *data, xid_t xid) + { + int err; + struct super_block *sb = path->mnt->mnt_sb; + +- if (!capable(CAP_SYS_ADMIN)) ++ if (!vx_capable(CAP_SYS_ADMIN, VXC_SECURE_REMOUNT)) + return -EPERM; + + if (!check_mnt(path->mnt)) +@@ -1872,7 +1937,7 @@ static int do_move_mount(struct path *pa + struct path old_path, parent_path; + struct vfsmount *p; + int err = 0; +- if (!capable(CAP_SYS_ADMIN)) ++ if (!vx_capable(CAP_SYS_ADMIN, VXC_SECURE_MOUNT)) + return -EPERM; + if (!old_name || !*old_name) + return -EINVAL; +@@ -2023,7 +2088,7 @@ static int do_new_mount(struct path *pat + return -EINVAL; + + /* we need capabilities... */ +- if (!capable(CAP_SYS_ADMIN)) ++ if (!vx_capable(CAP_SYS_ADMIN, VXC_SECURE_MOUNT)) + return -EPERM; + + mnt = do_kern_mount(type, flags, name, data); +@@ -2292,6 +2357,7 @@ long do_mount(char *dev_name, char *dir_ + struct path path; + int retval = 0; + int mnt_flags = 0; ++ tag_t tag = 0; + + /* Discard magic */ + if ((flags & MS_MGC_MSK) == MS_MGC_VAL) +@@ -2319,6 +2385,12 @@ long do_mount(char *dev_name, char *dir_ + if (!(flags & MS_NOATIME)) + mnt_flags |= MNT_RELATIME; + ++ if (dx_parse_tag(data_page, &tag, 1, &mnt_flags, &flags)) { ++ /* FIXME: bind and re-mounts get the tag flag? */ ++ if (flags & (MS_BIND|MS_REMOUNT)) ++ flags |= MS_TAGID; ++ } ++ + /* Separate the per-mountpoint flags */ + if (flags & MS_NOSUID) + mnt_flags |= MNT_NOSUID; +@@ -2335,15 +2407,17 @@ long do_mount(char *dev_name, char *dir_ + if (flags & MS_RDONLY) + mnt_flags |= MNT_READONLY; + ++ if (!capable(CAP_SYS_ADMIN)) ++ mnt_flags |= MNT_NODEV; + flags &= ~(MS_NOSUID | MS_NOEXEC | MS_NODEV | MS_ACTIVE | MS_BORN | + MS_NOATIME | MS_NODIRATIME | MS_RELATIME| MS_KERNMOUNT | + MS_STRICTATIME); + + if (flags & MS_REMOUNT) + retval = do_remount(&path, flags & ~MS_REMOUNT, mnt_flags, +- data_page); ++ data_page, tag); + else if (flags & MS_BIND) +- retval = do_loopback(&path, dev_name, flags & MS_REC); ++ retval = do_loopback(&path, dev_name, tag, flags, mnt_flags); + else if (flags & (MS_SHARED | MS_PRIVATE | MS_SLAVE | MS_UNBINDABLE)) + retval = do_change_type(&path, flags); + else if (flags & MS_MOVE) +@@ -2443,6 +2517,7 @@ static struct mnt_namespace *dup_mnt_ns( + q = next_mnt(q, new_ns->root); + } + up_write(&namespace_sem); ++ atomic_inc(&vs_global_mnt_ns); + + if (rootmnt) + mntput(rootmnt); +@@ -2582,9 +2657,10 @@ SYSCALL_DEFINE2(pivot_root, const char _ + goto out3; + + error = -EINVAL; +- if (IS_MNT_SHARED(old.mnt) || +- IS_MNT_SHARED(new.mnt->mnt_parent) || +- IS_MNT_SHARED(root.mnt->mnt_parent)) ++ if ((IS_MNT_SHARED(old.mnt) || ++ IS_MNT_SHARED(new.mnt->mnt_parent) || ++ IS_MNT_SHARED(root.mnt->mnt_parent)) && ++ !vx_flags(VXF_STATE_SETUP, 0)) + goto out4; + if (!check_mnt(root.mnt) || !check_mnt(new.mnt)) + goto out4; +@@ -2716,6 +2792,7 @@ void put_mnt_ns(struct mnt_namespace *ns + br_write_unlock(vfsmount_lock); + up_write(&namespace_sem); + release_mounts(&umount_list); ++ atomic_dec(&vs_global_mnt_ns); + kfree(ns); + } + EXPORT_SYMBOL(put_mnt_ns); +diff -NurpP --minimal linux-3.0.9/fs/nfs/client.c linux-3.0.9-vs2.3.2.1/fs/nfs/client.c +--- linux-3.0.9/fs/nfs/client.c 2011-07-22 11:18:05.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/nfs/client.c 2011-06-10 22:11:24.000000000 +0200 +@@ -780,6 +780,9 @@ static int nfs_init_server_rpcclient(str + if (server->flags & NFS_MOUNT_SOFT) + server->client->cl_softrtry = 1; + ++ server->client->cl_tag = 0; ++ if (server->flags & NFS_MOUNT_TAGGED) ++ server->client->cl_tag = 1; + return 0; + } + +@@ -951,6 +954,10 @@ static void nfs_server_set_fsinfo(struct + server->acdirmin = server->acdirmax = 0; + } + ++ /* FIXME: needs fsinfo ++ if (server->flags & NFS_MOUNT_TAGGED) ++ sb->s_flags |= MS_TAGGED; */ ++ + server->maxfilesize = fsinfo->maxfilesize; + + server->time_delta = fsinfo->time_delta; +diff -NurpP --minimal linux-3.0.9/fs/nfs/dir.c linux-3.0.9-vs2.3.2.1/fs/nfs/dir.c +--- linux-3.0.9/fs/nfs/dir.c 2011-11-15 16:40:47.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/nfs/dir.c 2011-08-08 23:04:47.000000000 +0200 +@@ -35,6 +35,7 @@ + #include <linux/sched.h> + #include <linux/kmemleak.h> + #include <linux/xattr.h> ++#include <linux/vs_tag.h> + + #include "delegation.h" + #include "iostat.h" +@@ -1308,6 +1309,7 @@ static struct dentry *nfs_lookup(struct + if (IS_ERR(res)) + goto out_unblock_sillyrename; + ++ dx_propagate_tag(nd, inode); + no_entry: + res = d_materialise_unique(dentry, inode); + if (res != NULL) { +diff -NurpP --minimal linux-3.0.9/fs/nfs/inode.c linux-3.0.9-vs2.3.2.1/fs/nfs/inode.c +--- linux-3.0.9/fs/nfs/inode.c 2011-07-22 11:18:05.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/nfs/inode.c 2011-10-11 21:21:33.000000000 +0200 +@@ -38,6 +38,7 @@ + #include <linux/nfs_xdr.h> + #include <linux/slab.h> + #include <linux/compat.h> ++#include <linux/vs_tag.h> + + #include <asm/system.h> + #include <asm/uaccess.h> +@@ -273,6 +274,8 @@ nfs_fhget(struct super_block *sb, struct + if (inode->i_state & I_NEW) { + struct nfs_inode *nfsi = NFS_I(inode); + unsigned long now = jiffies; ++ uid_t uid; ++ gid_t gid; + + /* We set i_ino for the few things that still rely on it, + * such as stat(2) */ +@@ -321,8 +324,8 @@ nfs_fhget(struct super_block *sb, struct + nfsi->change_attr = 0; + inode->i_size = 0; + inode->i_nlink = 0; +- inode->i_uid = -2; +- inode->i_gid = -2; ++ uid = -2; ++ gid = -2; + inode->i_blocks = 0; + memset(nfsi->cookieverf, 0, sizeof(nfsi->cookieverf)); + +@@ -359,13 +362,13 @@ nfs_fhget(struct super_block *sb, struct + else if (nfs_server_capable(inode, NFS_CAP_NLINK)) + nfsi->cache_validity |= NFS_INO_INVALID_ATTR; + if (fattr->valid & NFS_ATTR_FATTR_OWNER) +- inode->i_uid = fattr->uid; ++ uid = fattr->uid; + else if (nfs_server_capable(inode, NFS_CAP_OWNER)) + nfsi->cache_validity |= NFS_INO_INVALID_ATTR + | NFS_INO_INVALID_ACCESS + | NFS_INO_INVALID_ACL; + if (fattr->valid & NFS_ATTR_FATTR_GROUP) +- inode->i_gid = fattr->gid; ++ gid = fattr->gid; + else if (nfs_server_capable(inode, NFS_CAP_OWNER_GROUP)) + nfsi->cache_validity |= NFS_INO_INVALID_ATTR + | NFS_INO_INVALID_ACCESS +@@ -378,6 +381,11 @@ nfs_fhget(struct super_block *sb, struct + */ + inode->i_blocks = nfs_calc_block_size(fattr->du.nfs3.used); + } ++ inode->i_uid = INOTAG_UID(DX_TAG(inode), uid, gid); ++ inode->i_gid = INOTAG_GID(DX_TAG(inode), uid, gid); ++ inode->i_tag = INOTAG_TAG(DX_TAG(inode), uid, gid, 0); ++ /* maybe fattr->xid someday */ ++ + nfsi->attrtimeo = NFS_MINATTRTIMEO(inode); + nfsi->attrtimeo_timestamp = now; + nfsi->access_cache = RB_ROOT; +@@ -494,6 +502,8 @@ void nfs_setattr_update_inode(struct ino + inode->i_uid = attr->ia_uid; + if ((attr->ia_valid & ATTR_GID) != 0) + inode->i_gid = attr->ia_gid; ++ if ((attr->ia_valid & ATTR_TAG) && IS_TAGGED(inode)) ++ inode->i_tag = attr->ia_tag; + NFS_I(inode)->cache_validity |= NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL; + spin_unlock(&inode->i_lock); + } +@@ -941,6 +951,9 @@ static int nfs_check_inode_attributes(st + struct nfs_inode *nfsi = NFS_I(inode); + loff_t cur_size, new_isize; + unsigned long invalid = 0; ++ uid_t uid; ++ gid_t gid; ++ tag_t tag; + + + /* Has the inode gone and changed behind our back? */ +@@ -964,13 +977,18 @@ static int nfs_check_inode_attributes(st + invalid |= NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE; + } + ++ uid = INOTAG_UID(DX_TAG(inode), fattr->uid, fattr->gid); ++ gid = INOTAG_GID(DX_TAG(inode), fattr->uid, fattr->gid); ++ tag = INOTAG_TAG(DX_TAG(inode), fattr->uid, fattr->gid, 0); ++ + /* Have any file permissions changed? */ + if ((fattr->valid & NFS_ATTR_FATTR_MODE) && (inode->i_mode & S_IALLUGO) != (fattr->mode & S_IALLUGO)) + invalid |= NFS_INO_INVALID_ATTR | NFS_INO_INVALID_ACCESS | NFS_INO_INVALID_ACL; +- if ((fattr->valid & NFS_ATTR_FATTR_OWNER) && inode->i_uid != fattr->uid) ++ if ((fattr->valid & NFS_ATTR_FATTR_OWNER) && uid != fattr->uid) + invalid |= NFS_INO_INVALID_ATTR | NFS_INO_INVALID_ACCESS | NFS_INO_INVALID_ACL; +- if ((fattr->valid & NFS_ATTR_FATTR_GROUP) && inode->i_gid != fattr->gid) ++ if ((fattr->valid & NFS_ATTR_FATTR_GROUP) && gid != fattr->gid) + invalid |= NFS_INO_INVALID_ATTR | NFS_INO_INVALID_ACCESS | NFS_INO_INVALID_ACL; ++ /* maybe check for tag too? */ + + /* Has the link count changed? */ + if ((fattr->valid & NFS_ATTR_FATTR_NLINK) && inode->i_nlink != fattr->nlink) +@@ -1205,6 +1223,9 @@ static int nfs_update_inode(struct inode + unsigned long invalid = 0; + unsigned long now = jiffies; + unsigned long save_cache_validity; ++ uid_t uid; ++ gid_t gid; ++ tag_t tag; + + dfprintk(VFS, "NFS: %s(%s/%ld ct=%d info=0x%x)\n", + __func__, inode->i_sb->s_id, inode->i_ino, +@@ -1312,6 +1333,9 @@ static int nfs_update_inode(struct inode + | NFS_INO_REVAL_PAGECACHE + | NFS_INO_REVAL_FORCED); + ++ uid = TAGINO_UID(DX_TAG(inode), inode->i_uid, inode->i_tag); ++ gid = TAGINO_GID(DX_TAG(inode), inode->i_gid, inode->i_tag); ++ tag = inode->i_tag; + + if (fattr->valid & NFS_ATTR_FATTR_ATIME) + memcpy(&inode->i_atime, &fattr->atime, sizeof(inode->i_atime)); +@@ -1333,9 +1357,9 @@ static int nfs_update_inode(struct inode + | NFS_INO_REVAL_FORCED); + + if (fattr->valid & NFS_ATTR_FATTR_OWNER) { +- if (inode->i_uid != fattr->uid) { ++ if (uid != fattr->uid) { + invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL; +- inode->i_uid = fattr->uid; ++ uid = fattr->uid; + } + } else if (server->caps & NFS_CAP_OWNER) + invalid |= save_cache_validity & (NFS_INO_INVALID_ATTR +@@ -1344,9 +1368,9 @@ static int nfs_update_inode(struct inode + | NFS_INO_REVAL_FORCED); + + if (fattr->valid & NFS_ATTR_FATTR_GROUP) { +- if (inode->i_gid != fattr->gid) { ++ if (gid != fattr->gid) { + invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL; +- inode->i_gid = fattr->gid; ++ gid = fattr->gid; + } + } else if (server->caps & NFS_CAP_OWNER_GROUP) + invalid |= save_cache_validity & (NFS_INO_INVALID_ATTR +@@ -1354,6 +1378,10 @@ static int nfs_update_inode(struct inode + | NFS_INO_INVALID_ACL + | NFS_INO_REVAL_FORCED); + ++ inode->i_uid = INOTAG_UID(DX_TAG(inode), uid, gid); ++ inode->i_gid = INOTAG_GID(DX_TAG(inode), uid, gid); ++ inode->i_tag = INOTAG_TAG(DX_TAG(inode), uid, gid, tag); ++ + if (fattr->valid & NFS_ATTR_FATTR_NLINK) { + if (inode->i_nlink != fattr->nlink) { + invalid |= NFS_INO_INVALID_ATTR; +diff -NurpP --minimal linux-3.0.9/fs/nfs/nfs3xdr.c linux-3.0.9-vs2.3.2.1/fs/nfs/nfs3xdr.c +--- linux-3.0.9/fs/nfs/nfs3xdr.c 2011-03-15 18:07:32.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/nfs/nfs3xdr.c 2011-06-10 22:11:24.000000000 +0200 +@@ -20,6 +20,7 @@ + #include <linux/nfs3.h> + #include <linux/nfs_fs.h> + #include <linux/nfsacl.h> ++#include <linux/vs_tag.h> + #include "internal.h" + + #define NFSDBG_FACILITY NFSDBG_XDR +@@ -562,7 +563,8 @@ static __be32 *xdr_decode_nfstime3(__be3 + * set_mtime mtime; + * }; + */ +-static void encode_sattr3(struct xdr_stream *xdr, const struct iattr *attr) ++static void encode_sattr3(struct xdr_stream *xdr, ++ const struct iattr *attr, int tag) + { + u32 nbytes; + __be32 *p; +@@ -594,15 +596,19 @@ static void encode_sattr3(struct xdr_str + } else + *p++ = xdr_zero; + +- if (attr->ia_valid & ATTR_UID) { ++ if (attr->ia_valid & ATTR_UID || ++ (tag && (attr->ia_valid & ATTR_TAG))) { + *p++ = xdr_one; +- *p++ = cpu_to_be32(attr->ia_uid); ++ *p++ = cpu_to_be32(TAGINO_UID(tag, ++ attr->ia_uid, attr->ia_tag)); + } else + *p++ = xdr_zero; + +- if (attr->ia_valid & ATTR_GID) { ++ if (attr->ia_valid & ATTR_GID || ++ (tag && (attr->ia_valid & ATTR_TAG))) { + *p++ = xdr_one; +- *p++ = cpu_to_be32(attr->ia_gid); ++ *p++ = cpu_to_be32(TAGINO_GID(tag, ++ attr->ia_gid, attr->ia_tag)); + } else + *p++ = xdr_zero; + +@@ -878,7 +884,7 @@ static void nfs3_xdr_enc_setattr3args(st + const struct nfs3_sattrargs *args) + { + encode_nfs_fh3(xdr, args->fh); +- encode_sattr3(xdr, args->sattr); ++ encode_sattr3(xdr, args->sattr, req->rq_task->tk_client->cl_tag); + encode_sattrguard3(xdr, args); + } + +@@ -1028,13 +1034,13 @@ static void nfs3_xdr_enc_write3args(stru + * }; + */ + static void encode_createhow3(struct xdr_stream *xdr, +- const struct nfs3_createargs *args) ++ const struct nfs3_createargs *args, int tag) + { + encode_uint32(xdr, args->createmode); + switch (args->createmode) { + case NFS3_CREATE_UNCHECKED: + case NFS3_CREATE_GUARDED: +- encode_sattr3(xdr, args->sattr); ++ encode_sattr3(xdr, args->sattr, tag); + break; + case NFS3_CREATE_EXCLUSIVE: + encode_createverf3(xdr, args->verifier); +@@ -1049,7 +1055,7 @@ static void nfs3_xdr_enc_create3args(str + const struct nfs3_createargs *args) + { + encode_diropargs3(xdr, args->fh, args->name, args->len); +- encode_createhow3(xdr, args); ++ encode_createhow3(xdr, args, req->rq_task->tk_client->cl_tag); + } + + /* +@@ -1065,7 +1071,7 @@ static void nfs3_xdr_enc_mkdir3args(stru + const struct nfs3_mkdirargs *args) + { + encode_diropargs3(xdr, args->fh, args->name, args->len); +- encode_sattr3(xdr, args->sattr); ++ encode_sattr3(xdr, args->sattr, req->rq_task->tk_client->cl_tag); + } + + /* +@@ -1082,9 +1088,9 @@ static void nfs3_xdr_enc_mkdir3args(stru + * }; + */ + static void encode_symlinkdata3(struct xdr_stream *xdr, +- const struct nfs3_symlinkargs *args) ++ const struct nfs3_symlinkargs *args, int tag) + { +- encode_sattr3(xdr, args->sattr); ++ encode_sattr3(xdr, args->sattr, tag); + encode_nfspath3(xdr, args->pages, args->pathlen); + } + +@@ -1093,7 +1099,7 @@ static void nfs3_xdr_enc_symlink3args(st + const struct nfs3_symlinkargs *args) + { + encode_diropargs3(xdr, args->fromfh, args->fromname, args->fromlen); +- encode_symlinkdata3(xdr, args); ++ encode_symlinkdata3(xdr, args, req->rq_task->tk_client->cl_tag); + } + + /* +@@ -1121,24 +1127,24 @@ static void nfs3_xdr_enc_symlink3args(st + * }; + */ + static void encode_devicedata3(struct xdr_stream *xdr, +- const struct nfs3_mknodargs *args) ++ const struct nfs3_mknodargs *args, int tag) + { +- encode_sattr3(xdr, args->sattr); ++ encode_sattr3(xdr, args->sattr, tag); + encode_specdata3(xdr, args->rdev); + } + + static void encode_mknoddata3(struct xdr_stream *xdr, +- const struct nfs3_mknodargs *args) ++ const struct nfs3_mknodargs *args, int tag) + { + encode_ftype3(xdr, args->type); + switch (args->type) { + case NF3CHR: + case NF3BLK: +- encode_devicedata3(xdr, args); ++ encode_devicedata3(xdr, args, tag); + break; + case NF3SOCK: + case NF3FIFO: +- encode_sattr3(xdr, args->sattr); ++ encode_sattr3(xdr, args->sattr, tag); + break; + case NF3REG: + case NF3DIR: +@@ -1153,7 +1159,7 @@ static void nfs3_xdr_enc_mknod3args(stru + const struct nfs3_mknodargs *args) + { + encode_diropargs3(xdr, args->fh, args->name, args->len); +- encode_mknoddata3(xdr, args); ++ encode_mknoddata3(xdr, args, req->rq_task->tk_client->cl_tag); + } + + /* +diff -NurpP --minimal linux-3.0.9/fs/nfs/super.c linux-3.0.9-vs2.3.2.1/fs/nfs/super.c +--- linux-3.0.9/fs/nfs/super.c 2011-11-15 16:40:47.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/nfs/super.c 2011-11-15 17:37:07.000000000 +0100 +@@ -53,6 +53,7 @@ + #include <linux/nfs_xdr.h> + #include <linux/magic.h> + #include <linux/parser.h> ++#include <linux/vs_tag.h> + + #include <asm/system.h> + #include <asm/uaccess.h> +@@ -87,6 +88,7 @@ enum { + Opt_sharecache, Opt_nosharecache, + Opt_resvport, Opt_noresvport, + Opt_fscache, Opt_nofscache, ++ Opt_tag, Opt_notag, + + /* Mount options that take integer arguments */ + Opt_port, +@@ -100,6 +102,7 @@ enum { + Opt_mountvers, + Opt_nfsvers, + Opt_minorversion, ++ Opt_tagid, + + /* Mount options that take string arguments */ + Opt_sec, Opt_proto, Opt_mountproto, Opt_mounthost, +@@ -180,6 +183,10 @@ static const match_table_t nfs_mount_opt + { Opt_fscache_uniq, "fsc=%s" }, + { Opt_local_lock, "local_lock=%s" }, + ++ { Opt_tag, "tag" }, ++ { Opt_notag, "notag" }, ++ { Opt_tagid, "tagid=%u" }, ++ + { Opt_err, NULL } + }; + +@@ -650,6 +657,7 @@ static void nfs_show_mount_options(struc + { NFS_MOUNT_NORDIRPLUS, ",nordirplus", "" }, + { NFS_MOUNT_UNSHARED, ",nosharecache", "" }, + { NFS_MOUNT_NORESVPORT, ",noresvport", "" }, ++ { NFS_MOUNT_TAGGED, ",tag", "" }, + { 0, NULL, NULL } + }; + const struct proc_nfs_info *nfs_infop; +@@ -1198,6 +1206,14 @@ static int nfs_parse_mount_options(char + kfree(mnt->fscache_uniq); + mnt->fscache_uniq = NULL; + break; ++#ifndef CONFIG_TAGGING_NONE ++ case Opt_tag: ++ mnt->flags |= NFS_MOUNT_TAGGED; ++ break; ++ case Opt_notag: ++ mnt->flags &= ~NFS_MOUNT_TAGGED; ++ break; ++#endif + + /* + * options that take numeric values +@@ -1304,6 +1320,12 @@ static int nfs_parse_mount_options(char + goto out_invalid_value; + mnt->minorversion = option; + break; ++#ifdef CONFIG_PROPAGATE ++ case Opt_tagid: ++ /* use args[0] */ ++ nfs_data.flags |= NFS_MOUNT_TAGGED; ++ break; ++#endif + + /* + * options that take text values +diff -NurpP --minimal linux-3.0.9/fs/nfsd/auth.c linux-3.0.9-vs2.3.2.1/fs/nfsd/auth.c +--- linux-3.0.9/fs/nfsd/auth.c 2010-02-25 11:52:05.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/nfsd/auth.c 2011-06-10 22:11:24.000000000 +0200 +@@ -1,6 +1,7 @@ + /* Copyright (C) 1995, 1996 Olaf Kirch <okir@monad.swb.de> */ + + #include <linux/sched.h> ++#include <linux/vs_tag.h> + #include "nfsd.h" + #include "auth.h" + +@@ -36,6 +37,9 @@ int nfsd_setuser(struct svc_rqst *rqstp, + + new->fsuid = rqstp->rq_cred.cr_uid; + new->fsgid = rqstp->rq_cred.cr_gid; ++ /* FIXME: this desperately needs a tag :) ++ new->xid = (xid_t)INOTAG_TAG(DX_TAG_NFSD, cred.cr_uid, cred.cr_gid, 0); ++ */ + + rqgi = rqstp->rq_cred.cr_group_info; + +diff -NurpP --minimal linux-3.0.9/fs/nfsd/nfs3xdr.c linux-3.0.9-vs2.3.2.1/fs/nfsd/nfs3xdr.c +--- linux-3.0.9/fs/nfsd/nfs3xdr.c 2011-07-22 11:18:05.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/nfsd/nfs3xdr.c 2011-06-10 22:11:24.000000000 +0200 +@@ -7,6 +7,7 @@ + */ + + #include <linux/namei.h> ++#include <linux/vs_tag.h> + #include "xdr3.h" + #include "auth.h" + +@@ -95,6 +96,8 @@ static __be32 * + decode_sattr3(__be32 *p, struct iattr *iap) + { + u32 tmp; ++ uid_t uid = 0; ++ gid_t gid = 0; + + iap->ia_valid = 0; + +@@ -104,12 +107,15 @@ decode_sattr3(__be32 *p, struct iattr *i + } + if (*p++) { + iap->ia_valid |= ATTR_UID; +- iap->ia_uid = ntohl(*p++); ++ uid = ntohl(*p++); + } + if (*p++) { + iap->ia_valid |= ATTR_GID; +- iap->ia_gid = ntohl(*p++); ++ gid = ntohl(*p++); + } ++ iap->ia_uid = INOTAG_UID(DX_TAG_NFSD, uid, gid); ++ iap->ia_gid = INOTAG_GID(DX_TAG_NFSD, uid, gid); ++ iap->ia_tag = INOTAG_TAG(DX_TAG_NFSD, uid, gid, 0); + if (*p++) { + u64 newsize; + +@@ -165,8 +171,12 @@ encode_fattr3(struct svc_rqst *rqstp, __ + *p++ = htonl(nfs3_ftypes[(stat->mode & S_IFMT) >> 12]); + *p++ = htonl((u32) stat->mode); + *p++ = htonl((u32) stat->nlink); +- *p++ = htonl((u32) nfsd_ruid(rqstp, stat->uid)); +- *p++ = htonl((u32) nfsd_rgid(rqstp, stat->gid)); ++ *p++ = htonl((u32) nfsd_ruid(rqstp, ++ TAGINO_UID(0 /* FIXME: DX_TAG(dentry->d_inode) */, ++ stat->uid, stat->tag))); ++ *p++ = htonl((u32) nfsd_rgid(rqstp, ++ TAGINO_GID(0 /* FIXME: DX_TAG(dentry->d_inode) */, ++ stat->gid, stat->tag))); + if (S_ISLNK(stat->mode) && stat->size > NFS3_MAXPATHLEN) { + p = xdr_encode_hyper(p, (u64) NFS3_MAXPATHLEN); + } else { +diff -NurpP --minimal linux-3.0.9/fs/nfsd/nfs4xdr.c linux-3.0.9-vs2.3.2.1/fs/nfsd/nfs4xdr.c +--- linux-3.0.9/fs/nfsd/nfs4xdr.c 2011-11-15 16:40:47.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/nfsd/nfs4xdr.c 2011-11-15 17:37:07.000000000 +0100 +@@ -45,6 +45,7 @@ + #include <linux/statfs.h> + #include <linux/utsname.h> + #include <linux/sunrpc/svcauth_gss.h> ++#include <linux/vs_tag.h> + + #include "idmap.h" + #include "acl.h" +@@ -2102,14 +2103,18 @@ out_acl: + WRITE32(stat.nlink); + } + if (bmval1 & FATTR4_WORD1_OWNER) { +- status = nfsd4_encode_user(rqstp, stat.uid, &p, &buflen); ++ status = nfsd4_encode_user(rqstp, ++ TAGINO_UID(DX_TAG(dentry->d_inode), ++ stat.uid, stat.tag), &p, &buflen); + if (status == nfserr_resource) + goto out_resource; + if (status) + goto out; + } + if (bmval1 & FATTR4_WORD1_OWNER_GROUP) { +- status = nfsd4_encode_group(rqstp, stat.gid, &p, &buflen); ++ status = nfsd4_encode_group(rqstp, ++ TAGINO_GID(DX_TAG(dentry->d_inode), ++ stat.gid, stat.tag), &p, &buflen); + if (status == nfserr_resource) + goto out_resource; + if (status) +diff -NurpP --minimal linux-3.0.9/fs/nfsd/nfsxdr.c linux-3.0.9-vs2.3.2.1/fs/nfsd/nfsxdr.c +--- linux-3.0.9/fs/nfsd/nfsxdr.c 2011-05-22 16:17:53.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/nfsd/nfsxdr.c 2011-06-10 22:11:24.000000000 +0200 +@@ -6,6 +6,7 @@ + + #include "xdr.h" + #include "auth.h" ++#include <linux/vs_tag.h> + + #define NFSDDBG_FACILITY NFSDDBG_XDR + +@@ -88,6 +89,8 @@ static __be32 * + decode_sattr(__be32 *p, struct iattr *iap) + { + u32 tmp, tmp1; ++ uid_t uid = 0; ++ gid_t gid = 0; + + iap->ia_valid = 0; + +@@ -101,12 +104,15 @@ decode_sattr(__be32 *p, struct iattr *ia + } + if ((tmp = ntohl(*p++)) != (u32)-1) { + iap->ia_valid |= ATTR_UID; +- iap->ia_uid = tmp; ++ uid = tmp; + } + if ((tmp = ntohl(*p++)) != (u32)-1) { + iap->ia_valid |= ATTR_GID; +- iap->ia_gid = tmp; ++ gid = tmp; + } ++ iap->ia_uid = INOTAG_UID(DX_TAG_NFSD, uid, gid); ++ iap->ia_gid = INOTAG_GID(DX_TAG_NFSD, uid, gid); ++ iap->ia_tag = INOTAG_TAG(DX_TAG_NFSD, uid, gid, 0); + if ((tmp = ntohl(*p++)) != (u32)-1) { + iap->ia_valid |= ATTR_SIZE; + iap->ia_size = tmp; +@@ -151,8 +157,10 @@ encode_fattr(struct svc_rqst *rqstp, __b + *p++ = htonl(nfs_ftypes[type >> 12]); + *p++ = htonl((u32) stat->mode); + *p++ = htonl((u32) stat->nlink); +- *p++ = htonl((u32) nfsd_ruid(rqstp, stat->uid)); +- *p++ = htonl((u32) nfsd_rgid(rqstp, stat->gid)); ++ *p++ = htonl((u32) nfsd_ruid(rqstp, ++ TAGINO_UID(DX_TAG(dentry->d_inode), stat->uid, stat->tag))); ++ *p++ = htonl((u32) nfsd_rgid(rqstp, ++ TAGINO_GID(DX_TAG(dentry->d_inode), stat->gid, stat->tag))); + + if (S_ISLNK(type) && stat->size > NFS_MAXPATHLEN) { + *p++ = htonl(NFS_MAXPATHLEN); +diff -NurpP --minimal linux-3.0.9/fs/ocfs2/dlmglue.c linux-3.0.9-vs2.3.2.1/fs/ocfs2/dlmglue.c +--- linux-3.0.9/fs/ocfs2/dlmglue.c 2011-05-22 16:17:53.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/ocfs2/dlmglue.c 2011-06-10 22:11:24.000000000 +0200 +@@ -2041,6 +2041,7 @@ static void __ocfs2_stuff_meta_lvb(struc + lvb->lvb_iclusters = cpu_to_be32(oi->ip_clusters); + lvb->lvb_iuid = cpu_to_be32(inode->i_uid); + lvb->lvb_igid = cpu_to_be32(inode->i_gid); ++ lvb->lvb_itag = cpu_to_be16(inode->i_tag); + lvb->lvb_imode = cpu_to_be16(inode->i_mode); + lvb->lvb_inlink = cpu_to_be16(inode->i_nlink); + lvb->lvb_iatime_packed = +@@ -2091,6 +2092,7 @@ static void ocfs2_refresh_inode_from_lvb + + inode->i_uid = be32_to_cpu(lvb->lvb_iuid); + inode->i_gid = be32_to_cpu(lvb->lvb_igid); ++ inode->i_tag = be16_to_cpu(lvb->lvb_itag); + inode->i_mode = be16_to_cpu(lvb->lvb_imode); + inode->i_nlink = be16_to_cpu(lvb->lvb_inlink); + ocfs2_unpack_timespec(&inode->i_atime, +diff -NurpP --minimal linux-3.0.9/fs/ocfs2/dlmglue.h linux-3.0.9-vs2.3.2.1/fs/ocfs2/dlmglue.h +--- linux-3.0.9/fs/ocfs2/dlmglue.h 2010-10-21 13:07:50.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/ocfs2/dlmglue.h 2011-06-10 22:11:24.000000000 +0200 +@@ -46,7 +46,8 @@ struct ocfs2_meta_lvb { + __be16 lvb_inlink; + __be32 lvb_iattr; + __be32 lvb_igeneration; +- __be32 lvb_reserved2; ++ __be16 lvb_itag; ++ __be16 lvb_reserved2; + }; + + #define OCFS2_QINFO_LVB_VERSION 1 +diff -NurpP --minimal linux-3.0.9/fs/ocfs2/file.c linux-3.0.9-vs2.3.2.1/fs/ocfs2/file.c +--- linux-3.0.9/fs/ocfs2/file.c 2011-07-22 11:18:06.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/ocfs2/file.c 2011-06-10 22:43:33.000000000 +0200 +@@ -1111,7 +1111,7 @@ int ocfs2_setattr(struct dentry *dentry, + attr->ia_valid &= ~ATTR_SIZE; + + #define OCFS2_VALID_ATTRS (ATTR_ATIME | ATTR_MTIME | ATTR_CTIME | ATTR_SIZE \ +- | ATTR_GID | ATTR_UID | ATTR_MODE) ++ | ATTR_GID | ATTR_UID | ATTR_TAG | ATTR_MODE) + if (!(attr->ia_valid & OCFS2_VALID_ATTRS)) + return 0; + +diff -NurpP --minimal linux-3.0.9/fs/ocfs2/inode.c linux-3.0.9-vs2.3.2.1/fs/ocfs2/inode.c +--- linux-3.0.9/fs/ocfs2/inode.c 2011-05-22 16:17:53.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/ocfs2/inode.c 2011-06-10 22:11:24.000000000 +0200 +@@ -28,6 +28,7 @@ + #include <linux/highmem.h> + #include <linux/pagemap.h> + #include <linux/quotaops.h> ++#include <linux/vs_tag.h> + + #include <asm/byteorder.h> + +@@ -78,11 +79,13 @@ void ocfs2_set_inode_flags(struct inode + { + unsigned int flags = OCFS2_I(inode)->ip_attr; + +- inode->i_flags &= ~(S_IMMUTABLE | ++ inode->i_flags &= ~(S_IMMUTABLE | S_IXUNLINK | + S_SYNC | S_APPEND | S_NOATIME | S_DIRSYNC); + + if (flags & OCFS2_IMMUTABLE_FL) + inode->i_flags |= S_IMMUTABLE; ++ if (flags & OCFS2_IXUNLINK_FL) ++ inode->i_flags |= S_IXUNLINK; + + if (flags & OCFS2_SYNC_FL) + inode->i_flags |= S_SYNC; +@@ -92,25 +95,44 @@ void ocfs2_set_inode_flags(struct inode + inode->i_flags |= S_NOATIME; + if (flags & OCFS2_DIRSYNC_FL) + inode->i_flags |= S_DIRSYNC; ++ ++ inode->i_vflags &= ~(V_BARRIER | V_COW); ++ ++ if (flags & OCFS2_BARRIER_FL) ++ inode->i_vflags |= V_BARRIER; ++ if (flags & OCFS2_COW_FL) ++ inode->i_vflags |= V_COW; + } + + /* Propagate flags from i_flags to OCFS2_I(inode)->ip_attr */ + void ocfs2_get_inode_flags(struct ocfs2_inode_info *oi) + { + unsigned int flags = oi->vfs_inode.i_flags; ++ unsigned int vflags = oi->vfs_inode.i_vflags; ++ ++ oi->ip_attr &= ~(OCFS2_SYNC_FL | OCFS2_APPEND_FL | ++ OCFS2_IMMUTABLE_FL | OCFS2_IXUNLINK_FL | ++ OCFS2_NOATIME_FL | OCFS2_DIRSYNC_FL | ++ OCFS2_BARRIER_FL | OCFS2_COW_FL); ++ ++ if (flags & S_IMMUTABLE) ++ oi->ip_attr |= OCFS2_IMMUTABLE_FL; ++ if (flags & S_IXUNLINK) ++ oi->ip_attr |= OCFS2_IXUNLINK_FL; + +- oi->ip_attr &= ~(OCFS2_SYNC_FL|OCFS2_APPEND_FL| +- OCFS2_IMMUTABLE_FL|OCFS2_NOATIME_FL|OCFS2_DIRSYNC_FL); + if (flags & S_SYNC) + oi->ip_attr |= OCFS2_SYNC_FL; + if (flags & S_APPEND) + oi->ip_attr |= OCFS2_APPEND_FL; +- if (flags & S_IMMUTABLE) +- oi->ip_attr |= OCFS2_IMMUTABLE_FL; + if (flags & S_NOATIME) + oi->ip_attr |= OCFS2_NOATIME_FL; + if (flags & S_DIRSYNC) + oi->ip_attr |= OCFS2_DIRSYNC_FL; ++ ++ if (vflags & V_BARRIER) ++ oi->ip_attr |= OCFS2_BARRIER_FL; ++ if (vflags & V_COW) ++ oi->ip_attr |= OCFS2_COW_FL; + } + + struct inode *ocfs2_ilookup(struct super_block *sb, u64 blkno) +@@ -241,6 +263,8 @@ void ocfs2_populate_inode(struct inode * + struct super_block *sb; + struct ocfs2_super *osb; + int use_plocks = 1; ++ uid_t uid; ++ gid_t gid; + + sb = inode->i_sb; + osb = OCFS2_SB(sb); +@@ -269,8 +293,12 @@ void ocfs2_populate_inode(struct inode * + inode->i_generation = le32_to_cpu(fe->i_generation); + inode->i_rdev = huge_decode_dev(le64_to_cpu(fe->id1.dev1.i_rdev)); + inode->i_mode = le16_to_cpu(fe->i_mode); +- inode->i_uid = le32_to_cpu(fe->i_uid); +- inode->i_gid = le32_to_cpu(fe->i_gid); ++ uid = le32_to_cpu(fe->i_uid); ++ gid = le32_to_cpu(fe->i_gid); ++ inode->i_uid = INOTAG_UID(DX_TAG(inode), uid, gid); ++ inode->i_gid = INOTAG_GID(DX_TAG(inode), uid, gid); ++ inode->i_tag = INOTAG_TAG(DX_TAG(inode), uid, gid, ++ /* le16_to_cpu(raw_inode->i_raw_tag)i */ 0); + + /* Fast symlinks will have i_size but no allocated clusters. */ + if (S_ISLNK(inode->i_mode) && !fe->i_clusters) +diff -NurpP --minimal linux-3.0.9/fs/ocfs2/inode.h linux-3.0.9-vs2.3.2.1/fs/ocfs2/inode.h +--- linux-3.0.9/fs/ocfs2/inode.h 2011-01-05 21:50:26.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/ocfs2/inode.h 2011-06-10 22:11:24.000000000 +0200 +@@ -151,6 +151,7 @@ struct buffer_head *ocfs2_bread(struct i + + void ocfs2_set_inode_flags(struct inode *inode); + void ocfs2_get_inode_flags(struct ocfs2_inode_info *oi); ++int ocfs2_sync_flags(struct inode *inode, int, int); + + static inline blkcnt_t ocfs2_inode_sector_count(struct inode *inode) + { +diff -NurpP --minimal linux-3.0.9/fs/ocfs2/ioctl.c linux-3.0.9-vs2.3.2.1/fs/ocfs2/ioctl.c +--- linux-3.0.9/fs/ocfs2/ioctl.c 2011-07-22 11:18:06.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/ocfs2/ioctl.c 2011-06-10 22:11:24.000000000 +0200 +@@ -78,7 +78,41 @@ static int ocfs2_get_inode_attr(struct i + return status; + } + +-static int ocfs2_set_inode_attr(struct inode *inode, unsigned flags, ++int ocfs2_sync_flags(struct inode *inode, int flags, int vflags) ++{ ++ struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); ++ struct buffer_head *bh = NULL; ++ handle_t *handle = NULL; ++ int status; ++ ++ status = ocfs2_inode_lock(inode, &bh, 1); ++ if (status < 0) { ++ mlog_errno(status); ++ return status; ++ } ++ handle = ocfs2_start_trans(osb, OCFS2_INODE_UPDATE_CREDITS); ++ if (IS_ERR(handle)) { ++ status = PTR_ERR(handle); ++ mlog_errno(status); ++ goto bail_unlock; ++ } ++ ++ inode->i_flags = flags; ++ inode->i_vflags = vflags; ++ ocfs2_get_inode_flags(OCFS2_I(inode)); ++ ++ status = ocfs2_mark_inode_dirty(handle, inode, bh); ++ if (status < 0) ++ mlog_errno(status); ++ ++ ocfs2_commit_trans(osb, handle); ++bail_unlock: ++ ocfs2_inode_unlock(inode, 1); ++ brelse(bh); ++ return status; ++} ++ ++int ocfs2_set_inode_attr(struct inode *inode, unsigned flags, + unsigned mask) + { + struct ocfs2_inode_info *ocfs2_inode = OCFS2_I(inode); +@@ -103,6 +137,11 @@ static int ocfs2_set_inode_attr(struct i + if (!S_ISDIR(inode->i_mode)) + flags &= ~OCFS2_DIRSYNC_FL; + ++ if (IS_BARRIER(inode)) { ++ vxwprintk_task(1, "messing with the barrier."); ++ goto bail_unlock; ++ } ++ + handle = ocfs2_start_trans(osb, OCFS2_INODE_UPDATE_CREDITS); + if (IS_ERR(handle)) { + status = PTR_ERR(handle); +@@ -880,6 +919,7 @@ bail: + return status; + } + ++ + long ocfs2_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) + { + struct inode *inode = filp->f_path.dentry->d_inode; +diff -NurpP --minimal linux-3.0.9/fs/ocfs2/namei.c linux-3.0.9-vs2.3.2.1/fs/ocfs2/namei.c +--- linux-3.0.9/fs/ocfs2/namei.c 2011-05-22 16:17:53.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/ocfs2/namei.c 2011-06-10 22:11:24.000000000 +0200 +@@ -41,6 +41,7 @@ + #include <linux/slab.h> + #include <linux/highmem.h> + #include <linux/quotaops.h> ++#include <linux/vs_tag.h> + + #include <cluster/masklog.h> + +@@ -477,6 +478,7 @@ static int __ocfs2_mknod_locked(struct i + struct ocfs2_dinode *fe = NULL; + struct ocfs2_extent_list *fel; + u16 feat; ++ tag_t tag; + + *new_fe_bh = NULL; + +@@ -514,8 +516,11 @@ static int __ocfs2_mknod_locked(struct i + fe->i_suballoc_loc = cpu_to_le64(suballoc_loc); + fe->i_suballoc_bit = cpu_to_le16(suballoc_bit); + fe->i_suballoc_slot = cpu_to_le16(inode_ac->ac_alloc_slot); +- fe->i_uid = cpu_to_le32(inode->i_uid); +- fe->i_gid = cpu_to_le32(inode->i_gid); ++ ++ tag = dx_current_fstag(osb->sb); ++ fe->i_uid = cpu_to_le32(TAGINO_UID(DX_TAG(inode), inode->i_uid, tag)); ++ fe->i_gid = cpu_to_le32(TAGINO_GID(DX_TAG(inode), inode->i_gid, tag)); ++ inode->i_tag = tag; + fe->i_mode = cpu_to_le16(inode->i_mode); + if (S_ISCHR(inode->i_mode) || S_ISBLK(inode->i_mode)) + fe->id1.dev1.i_rdev = cpu_to_le64(huge_encode_dev(dev)); +diff -NurpP --minimal linux-3.0.9/fs/ocfs2/ocfs2.h linux-3.0.9-vs2.3.2.1/fs/ocfs2/ocfs2.h +--- linux-3.0.9/fs/ocfs2/ocfs2.h 2011-05-22 16:17:53.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/ocfs2/ocfs2.h 2011-06-10 22:11:24.000000000 +0200 +@@ -272,6 +272,7 @@ enum ocfs2_mount_options + writes */ + OCFS2_MOUNT_HB_NONE = 1 << 13, /* No heartbeat */ + OCFS2_MOUNT_HB_GLOBAL = 1 << 14, /* Global heartbeat */ ++ OCFS2_MOUNT_TAGGED = 1 << 15, /* use tagging */ + }; + + #define OCFS2_OSB_SOFT_RO 0x0001 +diff -NurpP --minimal linux-3.0.9/fs/ocfs2/ocfs2_fs.h linux-3.0.9-vs2.3.2.1/fs/ocfs2/ocfs2_fs.h +--- linux-3.0.9/fs/ocfs2/ocfs2_fs.h 2011-05-22 16:17:53.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/ocfs2/ocfs2_fs.h 2011-06-10 22:11:24.000000000 +0200 +@@ -266,6 +266,11 @@ + #define OCFS2_TOPDIR_FL FS_TOPDIR_FL /* Top of directory hierarchies*/ + #define OCFS2_RESERVED_FL FS_RESERVED_FL /* reserved for ext2 lib */ + ++#define OCFS2_IXUNLINK_FL FS_IXUNLINK_FL /* Immutable invert on unlink */ ++ ++#define OCFS2_BARRIER_FL FS_BARRIER_FL /* Barrier for chroot() */ ++#define OCFS2_COW_FL FS_COW_FL /* Copy on Write marker */ ++ + #define OCFS2_FL_VISIBLE FS_FL_USER_VISIBLE /* User visible flags */ + #define OCFS2_FL_MODIFIABLE FS_FL_USER_MODIFIABLE /* User modifiable flags */ + +diff -NurpP --minimal linux-3.0.9/fs/ocfs2/super.c linux-3.0.9-vs2.3.2.1/fs/ocfs2/super.c +--- linux-3.0.9/fs/ocfs2/super.c 2011-07-22 11:18:06.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/ocfs2/super.c 2011-06-15 02:40:14.000000000 +0200 +@@ -184,6 +184,7 @@ enum { + Opt_coherency_full, + Opt_resv_level, + Opt_dir_resv_level, ++ Opt_tag, Opt_notag, Opt_tagid, + Opt_err, + }; + +@@ -215,6 +216,9 @@ static const match_table_t tokens = { + {Opt_coherency_full, "coherency=full"}, + {Opt_resv_level, "resv_level=%u"}, + {Opt_dir_resv_level, "dir_resv_level=%u"}, ++ {Opt_tag, "tag"}, ++ {Opt_notag, "notag"}, ++ {Opt_tagid, "tagid=%u"}, + {Opt_err, NULL} + }; + +@@ -662,6 +666,13 @@ static int ocfs2_remount(struct super_bl + goto out; + } + ++ if ((osb->s_mount_opt & OCFS2_MOUNT_TAGGED) != ++ (parsed_options.mount_opt & OCFS2_MOUNT_TAGGED)) { ++ ret = -EINVAL; ++ mlog(ML_ERROR, "Cannot change tagging on remount\n"); ++ goto out; ++ } ++ + /* We're going to/from readonly mode. */ + if ((*flags & MS_RDONLY) != (sb->s_flags & MS_RDONLY)) { + /* Disable quota accounting before remounting RO */ +@@ -1177,6 +1188,9 @@ static int ocfs2_fill_super(struct super + + ocfs2_complete_mount_recovery(osb); + ++ if (osb->s_mount_opt & OCFS2_MOUNT_TAGGED) ++ sb->s_flags |= MS_TAGGED; ++ + if (ocfs2_mount_local(osb)) + snprintf(nodestr, sizeof(nodestr), "local"); + else +@@ -1506,6 +1520,20 @@ static int ocfs2_parse_options(struct su + option < OCFS2_MAX_RESV_LEVEL) + mopt->dir_resv_level = option; + break; ++#ifndef CONFIG_TAGGING_NONE ++ case Opt_tag: ++ mopt->mount_opt |= OCFS2_MOUNT_TAGGED; ++ break; ++ case Opt_notag: ++ mopt->mount_opt &= ~OCFS2_MOUNT_TAGGED; ++ break; ++#endif ++#ifdef CONFIG_PROPAGATE ++ case Opt_tagid: ++ /* use args[0] */ ++ mopt->mount_opt |= OCFS2_MOUNT_TAGGED; ++ break; ++#endif + default: + mlog(ML_ERROR, + "Unrecognized mount option \"%s\" " +diff -NurpP --minimal linux-3.0.9/fs/open.c linux-3.0.9-vs2.3.2.1/fs/open.c +--- linux-3.0.9/fs/open.c 2011-05-22 16:17:53.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/open.c 2011-06-10 22:11:24.000000000 +0200 +@@ -30,6 +30,11 @@ + #include <linux/fs_struct.h> + #include <linux/ima.h> + #include <linux/dnotify.h> ++#include <linux/vs_base.h> ++#include <linux/vs_limit.h> ++#include <linux/vs_tag.h> ++#include <linux/vs_cowbl.h> ++#include <linux/vserver/dlimit.h> + + #include "internal.h" + +@@ -494,6 +499,12 @@ SYSCALL_DEFINE3(fchmodat, int, dfd, cons + error = user_path_at(dfd, filename, LOOKUP_FOLLOW, &path); + if (error) + goto out; ++ ++#ifdef CONFIG_VSERVER_COWBL ++ error = cow_check_and_break(&path); ++ if (error) ++ goto dput_and_out; ++#endif + inode = path.dentry->d_inode; + + error = mnt_want_write(path.mnt); +@@ -531,11 +542,11 @@ static int chown_common(struct path *pat + newattrs.ia_valid = ATTR_CTIME; + if (user != (uid_t) -1) { + newattrs.ia_valid |= ATTR_UID; +- newattrs.ia_uid = user; ++ newattrs.ia_uid = dx_map_uid(user); + } + if (group != (gid_t) -1) { + newattrs.ia_valid |= ATTR_GID; +- newattrs.ia_gid = group; ++ newattrs.ia_gid = dx_map_gid(group); + } + if (!S_ISDIR(inode->i_mode)) + newattrs.ia_valid |= +@@ -560,6 +571,10 @@ SYSCALL_DEFINE3(chown, const char __user + error = mnt_want_write(path.mnt); + if (error) + goto out_release; ++#ifdef CONFIG_VSERVER_COWBL ++ error = cow_check_and_break(&path); ++ if (!error) ++#endif + error = chown_common(&path, user, group); + mnt_drop_write(path.mnt); + out_release: +@@ -587,6 +602,10 @@ SYSCALL_DEFINE5(fchownat, int, dfd, cons + error = mnt_want_write(path.mnt); + if (error) + goto out_release; ++#ifdef CONFIG_VSERVER_COWBL ++ error = cow_check_and_break(&path); ++ if (!error) ++#endif + error = chown_common(&path, user, group); + mnt_drop_write(path.mnt); + out_release: +@@ -606,6 +625,10 @@ SYSCALL_DEFINE3(lchown, const char __use + error = mnt_want_write(path.mnt); + if (error) + goto out_release; ++#ifdef CONFIG_VSERVER_COWBL ++ error = cow_check_and_break(&path); ++ if (!error) ++#endif + error = chown_common(&path, user, group); + mnt_drop_write(path.mnt); + out_release: +@@ -857,6 +880,7 @@ static void __put_unused_fd(struct files + __FD_CLR(fd, fdt->open_fds); + if (fd < files->next_fd) + files->next_fd = fd; ++ vx_openfd_dec(fd); + } + + void put_unused_fd(unsigned int fd) +diff -NurpP --minimal linux-3.0.9/fs/proc/array.c linux-3.0.9-vs2.3.2.1/fs/proc/array.c +--- linux-3.0.9/fs/proc/array.c 2011-07-22 11:18:06.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/proc/array.c 2011-06-10 22:11:24.000000000 +0200 +@@ -81,6 +81,8 @@ + #include <linux/pid_namespace.h> + #include <linux/ptrace.h> + #include <linux/tracehook.h> ++#include <linux/vs_context.h> ++#include <linux/vs_network.h> + + #include <asm/pgtable.h> + #include <asm/processor.h> +@@ -170,6 +172,9 @@ static inline void task_state(struct seq + rcu_read_lock(); + ppid = pid_alive(p) ? + task_tgid_nr_ns(rcu_dereference(p->real_parent), ns) : 0; ++ if (unlikely(vx_current_initpid(p->pid))) ++ ppid = 0; ++ + tpid = 0; + if (pid_alive(p)) { + struct task_struct *tracer = tracehook_tracer_task(p); +@@ -287,7 +292,7 @@ static inline void task_sig(struct seq_f + } + + static void render_cap_t(struct seq_file *m, const char *header, +- kernel_cap_t *a) ++ struct vx_info *vxi, kernel_cap_t *a) + { + unsigned __capi; + +@@ -312,10 +317,11 @@ static inline void task_cap(struct seq_f + cap_bset = cred->cap_bset; + rcu_read_unlock(); + +- render_cap_t(m, "CapInh:\t", &cap_inheritable); +- render_cap_t(m, "CapPrm:\t", &cap_permitted); +- render_cap_t(m, "CapEff:\t", &cap_effective); +- render_cap_t(m, "CapBnd:\t", &cap_bset); ++ /* FIXME: maybe move the p->vx_info masking to __task_cred() ? */ ++ render_cap_t(m, "CapInh:\t", p->vx_info, &cap_inheritable); ++ render_cap_t(m, "CapPrm:\t", p->vx_info, &cap_permitted); ++ render_cap_t(m, "CapEff:\t", p->vx_info, &cap_effective); ++ render_cap_t(m, "CapBnd:\t", p->vx_info, &cap_bset); + } + + static inline void task_context_switch_counts(struct seq_file *m, +@@ -337,6 +343,42 @@ static void task_cpus_allowed(struct seq + seq_putc(m, '\n'); + } + ++int proc_pid_nsproxy(struct seq_file *m, struct pid_namespace *ns, ++ struct pid *pid, struct task_struct *task) ++{ ++ seq_printf(m, "Proxy:\t%p(%c)\n" ++ "Count:\t%u\n" ++ "uts:\t%p(%c)\n" ++ "ipc:\t%p(%c)\n" ++ "mnt:\t%p(%c)\n" ++ "pid:\t%p(%c)\n" ++ "net:\t%p(%c)\n", ++ task->nsproxy, ++ (task->nsproxy == init_task.nsproxy ? 'I' : '-'), ++ atomic_read(&task->nsproxy->count), ++ task->nsproxy->uts_ns, ++ (task->nsproxy->uts_ns == init_task.nsproxy->uts_ns ? 'I' : '-'), ++ task->nsproxy->ipc_ns, ++ (task->nsproxy->ipc_ns == init_task.nsproxy->ipc_ns ? 'I' : '-'), ++ task->nsproxy->mnt_ns, ++ (task->nsproxy->mnt_ns == init_task.nsproxy->mnt_ns ? 'I' : '-'), ++ task->nsproxy->pid_ns, ++ (task->nsproxy->pid_ns == init_task.nsproxy->pid_ns ? 'I' : '-'), ++ task->nsproxy->net_ns, ++ (task->nsproxy->net_ns == init_task.nsproxy->net_ns ? 'I' : '-')); ++ return 0; ++} ++ ++void task_vs_id(struct seq_file *m, struct task_struct *task) ++{ ++ if (task_vx_flags(task, VXF_HIDE_VINFO, 0)) ++ return; ++ ++ seq_printf(m, "VxID: %d\n", vx_task_xid(task)); ++ seq_printf(m, "NxID: %d\n", nx_task_nid(task)); ++} ++ ++ + int proc_pid_status(struct seq_file *m, struct pid_namespace *ns, + struct pid *pid, struct task_struct *task) + { +@@ -353,6 +395,7 @@ int proc_pid_status(struct seq_file *m, + task_cap(m, task); + task_cpus_allowed(m, task); + cpuset_task_status_allowed(m, task); ++ task_vs_id(m, task); + task_context_switch_counts(m, task); + return 0; + } +@@ -462,6 +505,17 @@ static int do_task_stat(struct seq_file + /* convert nsec -> ticks */ + start_time = nsec_to_clock_t(start_time); + ++ /* fixup start time for virt uptime */ ++ if (vx_flags(VXF_VIRT_UPTIME, 0)) { ++ unsigned long long bias = ++ current->vx_info->cvirt.bias_clock; ++ ++ if (start_time > bias) ++ start_time -= bias; ++ else ++ start_time = 0; ++ } ++ + seq_printf(m, "%d (%s) %c %d %d %d %d %d %u %lu \ + %lu %lu %lu %lu %lu %ld %ld %ld %ld %d 0 %llu %lu %ld %lu %lu %lu %lu %lu \ + %lu %lu %lu %lu %lu %lu %lu %lu %d %d %u %u %llu %lu %ld\n", +diff -NurpP --minimal linux-3.0.9/fs/proc/base.c linux-3.0.9-vs2.3.2.1/fs/proc/base.c +--- linux-3.0.9/fs/proc/base.c 2011-11-15 16:40:47.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/proc/base.c 2011-11-15 17:37:07.000000000 +0100 +@@ -83,6 +83,8 @@ + #include <linux/pid_namespace.h> + #include <linux/fs_struct.h> + #include <linux/slab.h> ++#include <linux/vs_context.h> ++#include <linux/vs_network.h> + #ifdef CONFIG_HARDWALL + #include <asm/hardwall.h> + #endif +@@ -1102,11 +1104,16 @@ static ssize_t oom_adjust_write(struct f + goto err_task_lock; + } + +- if (oom_adjust < task->signal->oom_adj && !capable(CAP_SYS_RESOURCE)) { ++ if (oom_adjust < task->signal->oom_adj && ++ !vx_capable(CAP_SYS_RESOURCE, VXC_OOM_ADJUST)) { + err = -EACCES; + goto err_sighand; + } + ++ /* prevent guest processes from circumventing the oom killer */ ++ if (vx_current_xid() && (oom_adjust == OOM_DISABLE)) ++ oom_adjust = OOM_ADJUST_MIN; ++ + if (oom_adjust != task->signal->oom_adj) { + if (oom_adjust == OOM_DISABLE) + atomic_inc(&task->mm->oom_disable_count); +@@ -1275,7 +1282,7 @@ static ssize_t proc_loginuid_write(struc + ssize_t length; + uid_t loginuid; + +- if (!capable(CAP_AUDIT_CONTROL)) ++ if (!vx_capable(CAP_AUDIT_CONTROL, VXC_AUDIT_CONTROL)) + return -EPERM; + + rcu_read_lock(); +@@ -1722,6 +1729,8 @@ struct inode *proc_pid_make_inode(struct + inode->i_gid = cred->egid; + rcu_read_unlock(); + } ++ /* procfs is xid tagged */ ++ inode->i_tag = (tag_t)vx_task_xid(task); + security_task_to_inode(task, inode); + + out: +@@ -1758,6 +1767,8 @@ int pid_getattr(struct vfsmount *mnt, st + + /* dentry stuff */ + ++static unsigned name_to_int(struct dentry *dentry); ++ + /* + * Exceptional case: normally we are not allowed to unhash a busy + * directory. In this case, however, we can do it - no aliasing problems +@@ -1786,6 +1797,12 @@ int pid_revalidate(struct dentry *dentry + task = get_proc_task(inode); + + if (task) { ++ unsigned pid = name_to_int(dentry); ++ ++ if (pid != ~0U && pid != vx_map_pid(task->pid)) { ++ put_task_struct(task); ++ goto drop; ++ } + if ((inode->i_mode == (S_IFDIR|S_IRUGO|S_IXUGO)) || + task_dumpable(task)) { + rcu_read_lock(); +@@ -1802,6 +1819,7 @@ int pid_revalidate(struct dentry *dentry + put_task_struct(task); + return 1; + } ++drop: + d_drop(dentry); + return 0; + } +@@ -2291,6 +2309,13 @@ static struct dentry *proc_pident_lookup + if (!task) + goto out_no_task; + ++ /* TODO: maybe we can come up with a generic approach? */ ++ if (task_vx_flags(task, VXF_HIDE_VINFO, 0) && ++ (dentry->d_name.len == 5) && ++ (!memcmp(dentry->d_name.name, "vinfo", 5) || ++ !memcmp(dentry->d_name.name, "ninfo", 5))) ++ goto out; ++ + /* + * Yes, it does not scale. And it should not. Don't add + * new entries into /proc/<tgid>/ without very good reasons. +@@ -2676,7 +2701,7 @@ out_iput: + static struct dentry *proc_base_lookup(struct inode *dir, struct dentry *dentry) + { + struct dentry *error; +- struct task_struct *task = get_proc_task(dir); ++ struct task_struct *task = get_proc_task_real(dir); + const struct pid_entry *p, *last; + + error = ERR_PTR(-ENOENT); +@@ -2783,6 +2808,9 @@ static int proc_pid_personality(struct s + static const struct file_operations proc_task_operations; + static const struct inode_operations proc_task_inode_operations; + ++extern int proc_pid_vx_info(struct task_struct *, char *); ++extern int proc_pid_nx_info(struct task_struct *, char *); ++ + static const struct pid_entry tgid_base_stuff[] = { + DIR("task", S_IRUGO|S_IXUGO, proc_task_inode_operations, proc_task_operations), + DIR("fd", S_IRUSR|S_IXUSR, proc_fd_inode_operations, proc_fd_operations), +@@ -2846,6 +2874,8 @@ static const struct pid_entry tgid_base_ + #ifdef CONFIG_CGROUPS + REG("cgroup", S_IRUGO, proc_cgroup_operations), + #endif ++ INF("vinfo", S_IRUGO, proc_pid_vx_info), ++ INF("ninfo", S_IRUGO, proc_pid_nx_info), + INF("oom_score", S_IRUGO, proc_oom_score), + REG("oom_adj", S_IRUGO|S_IWUSR, proc_oom_adjust_operations), + REG("oom_score_adj", S_IRUGO|S_IWUSR, proc_oom_score_adj_operations), +@@ -2865,6 +2895,7 @@ static const struct pid_entry tgid_base_ + #ifdef CONFIG_HARDWALL + INF("hardwall", S_IRUGO, proc_pid_hardwall), + #endif ++ ONE("nsproxy", S_IRUGO, proc_pid_nsproxy), + }; + + static int proc_tgid_base_readdir(struct file * filp, +@@ -3057,7 +3088,7 @@ retry: + iter.task = NULL; + pid = find_ge_pid(iter.tgid, ns); + if (pid) { +- iter.tgid = pid_nr_ns(pid, ns); ++ iter.tgid = pid_unmapped_nr_ns(pid, ns); + iter.task = pid_task(pid, PIDTYPE_PID); + /* What we to know is if the pid we have find is the + * pid of a thread_group_leader. Testing for task +@@ -3087,7 +3118,7 @@ static int proc_pid_fill_cache(struct fi + struct tgid_iter iter) + { + char name[PROC_NUMBUF]; +- int len = snprintf(name, sizeof(name), "%d", iter.tgid); ++ int len = snprintf(name, sizeof(name), "%d", vx_map_tgid(iter.tgid)); + return proc_fill_cache(filp, dirent, filldir, name, len, + proc_pid_instantiate, iter.task, NULL); + } +@@ -3104,7 +3135,7 @@ int proc_pid_readdir(struct file * filp, + goto out_no_task; + nr = filp->f_pos - FIRST_PROCESS_ENTRY; + +- reaper = get_proc_task(filp->f_path.dentry->d_inode); ++ reaper = get_proc_task_real(filp->f_path.dentry->d_inode); + if (!reaper) + goto out_no_task; + +@@ -3121,6 +3152,8 @@ int proc_pid_readdir(struct file * filp, + iter.task; + iter.tgid += 1, iter = next_tgid(ns, iter)) { + filp->f_pos = iter.tgid + TGID_OFFSET; ++ if (!vx_proc_task_visible(iter.task)) ++ continue; + if (proc_pid_fill_cache(filp, dirent, filldir, iter) < 0) { + put_task_struct(iter.task); + goto out; +@@ -3274,6 +3307,8 @@ static struct dentry *proc_task_lookup(s + tid = name_to_int(dentry); + if (tid == ~0U) + goto out; ++ if (vx_current_initpid(tid)) ++ goto out; + + ns = dentry->d_sb->s_fs_info; + rcu_read_lock(); +diff -NurpP --minimal linux-3.0.9/fs/proc/generic.c linux-3.0.9-vs2.3.2.1/fs/proc/generic.c +--- linux-3.0.9/fs/proc/generic.c 2011-07-22 11:18:06.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/proc/generic.c 2011-06-10 22:11:24.000000000 +0200 +@@ -22,6 +22,7 @@ + #include <linux/bitops.h> + #include <linux/spinlock.h> + #include <linux/completion.h> ++#include <linux/vserver/inode.h> + #include <asm/uaccess.h> + + #include "internal.h" +@@ -424,11 +425,15 @@ struct dentry *proc_lookup_de(struct pro + for (de = de->subdir; de ; de = de->next) { + if (de->namelen != dentry->d_name.len) + continue; ++ if (!vx_hide_check(0, de->vx_flags)) ++ continue; + if (!memcmp(dentry->d_name.name, de->name, de->namelen)) { + pde_get(de); + spin_unlock(&proc_subdir_lock); + error = -EINVAL; + inode = proc_get_inode(dir->i_sb, de); ++ /* generic proc entries belong to the host */ ++ inode->i_tag = 0; + goto out_unlock; + } + } +@@ -506,6 +511,8 @@ int proc_readdir_de(struct proc_dir_entr + + /* filldir passes info to user space */ + pde_get(de); ++ if (!vx_hide_check(0, de->vx_flags)) ++ goto skip; + spin_unlock(&proc_subdir_lock); + if (filldir(dirent, de->name, de->namelen, filp->f_pos, + de->low_ino, de->mode >> 12) < 0) { +@@ -513,6 +520,7 @@ int proc_readdir_de(struct proc_dir_entr + goto out; + } + spin_lock(&proc_subdir_lock); ++ skip: + filp->f_pos++; + next = de->next; + pde_put(de); +@@ -627,6 +635,7 @@ static struct proc_dir_entry *__proc_cre + ent->nlink = nlink; + atomic_set(&ent->count, 1); + ent->pde_users = 0; ++ ent->vx_flags = IATTR_PROC_DEFAULT; + spin_lock_init(&ent->pde_unload_lock); + ent->pde_unload_completion = NULL; + INIT_LIST_HEAD(&ent->pde_openers); +@@ -650,7 +659,8 @@ struct proc_dir_entry *proc_symlink(cons + kfree(ent->data); + kfree(ent); + ent = NULL; +- } ++ } else ++ ent->vx_flags = IATTR_PROC_SYMLINK; + } else { + kfree(ent); + ent = NULL; +diff -NurpP --minimal linux-3.0.9/fs/proc/inode.c linux-3.0.9-vs2.3.2.1/fs/proc/inode.c +--- linux-3.0.9/fs/proc/inode.c 2011-07-22 11:18:06.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/proc/inode.c 2011-06-10 22:11:24.000000000 +0200 +@@ -442,6 +442,8 @@ struct inode *proc_get_inode(struct supe + inode->i_uid = de->uid; + inode->i_gid = de->gid; + } ++ if (de->vx_flags) ++ PROC_I(inode)->vx_flags = de->vx_flags; + if (de->size) + inode->i_size = de->size; + if (de->nlink) +diff -NurpP --minimal linux-3.0.9/fs/proc/internal.h linux-3.0.9-vs2.3.2.1/fs/proc/internal.h +--- linux-3.0.9/fs/proc/internal.h 2011-07-22 11:18:06.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/proc/internal.h 2011-06-10 22:11:24.000000000 +0200 +@@ -10,6 +10,7 @@ + */ + + #include <linux/proc_fs.h> ++#include <linux/vs_pid.h> + + extern struct proc_dir_entry proc_root; + #ifdef CONFIG_PROC_SYSCTL +@@ -51,6 +52,9 @@ extern int proc_pid_status(struct seq_fi + struct pid *pid, struct task_struct *task); + extern int proc_pid_statm(struct seq_file *m, struct pid_namespace *ns, + struct pid *pid, struct task_struct *task); ++extern int proc_pid_nsproxy(struct seq_file *m, struct pid_namespace *ns, ++ struct pid *pid, struct task_struct *task); ++ + extern loff_t mem_lseek(struct file *file, loff_t offset, int orig); + + extern const struct file_operations proc_maps_operations; +@@ -76,11 +80,16 @@ static inline struct pid *proc_pid(struc + return PROC_I(inode)->pid; + } + +-static inline struct task_struct *get_proc_task(struct inode *inode) ++static inline struct task_struct *get_proc_task_real(struct inode *inode) + { + return get_pid_task(proc_pid(inode), PIDTYPE_PID); + } + ++static inline struct task_struct *get_proc_task(struct inode *inode) ++{ ++ return vx_get_proc_task(inode, proc_pid(inode)); ++} ++ + static inline int proc_fd(struct inode *inode) + { + return PROC_I(inode)->fd; +diff -NurpP --minimal linux-3.0.9/fs/proc/loadavg.c linux-3.0.9-vs2.3.2.1/fs/proc/loadavg.c +--- linux-3.0.9/fs/proc/loadavg.c 2009-09-10 15:26:23.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/proc/loadavg.c 2011-06-10 22:11:24.000000000 +0200 +@@ -12,15 +12,27 @@ + + static int loadavg_proc_show(struct seq_file *m, void *v) + { ++ unsigned long running; ++ unsigned int threads; + unsigned long avnrun[3]; + + get_avenrun(avnrun, FIXED_1/200, 0); + ++ if (vx_flags(VXF_VIRT_LOAD, 0)) { ++ struct vx_info *vxi = current_vx_info(); ++ ++ running = atomic_read(&vxi->cvirt.nr_running); ++ threads = atomic_read(&vxi->cvirt.nr_threads); ++ } else { ++ running = nr_running(); ++ threads = nr_threads; ++ } ++ + seq_printf(m, "%lu.%02lu %lu.%02lu %lu.%02lu %ld/%d %d\n", + LOAD_INT(avnrun[0]), LOAD_FRAC(avnrun[0]), + LOAD_INT(avnrun[1]), LOAD_FRAC(avnrun[1]), + LOAD_INT(avnrun[2]), LOAD_FRAC(avnrun[2]), +- nr_running(), nr_threads, ++ running, threads, + task_active_pid_ns(current)->last_pid); + return 0; + } +diff -NurpP --minimal linux-3.0.9/fs/proc/meminfo.c linux-3.0.9-vs2.3.2.1/fs/proc/meminfo.c +--- linux-3.0.9/fs/proc/meminfo.c 2011-03-15 18:07:33.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/proc/meminfo.c 2011-06-10 22:11:24.000000000 +0200 +@@ -39,7 +39,8 @@ static int meminfo_proc_show(struct seq_ + allowed = ((totalram_pages - hugetlb_total_pages()) + * sysctl_overcommit_ratio / 100) + total_swap_pages; + +- cached = global_page_state(NR_FILE_PAGES) - ++ cached = vx_flags(VXF_VIRT_MEM, 0) ? ++ vx_vsi_cached(&i) : global_page_state(NR_FILE_PAGES) - + total_swapcache_pages - i.bufferram; + if (cached < 0) + cached = 0; +diff -NurpP --minimal linux-3.0.9/fs/proc/root.c linux-3.0.9-vs2.3.2.1/fs/proc/root.c +--- linux-3.0.9/fs/proc/root.c 2011-07-22 11:18:06.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/proc/root.c 2011-06-22 12:39:15.000000000 +0200 +@@ -18,9 +18,14 @@ + #include <linux/bitops.h> + #include <linux/mount.h> + #include <linux/pid_namespace.h> ++#include <linux/vserver/inode.h> + + #include "internal.h" + ++struct proc_dir_entry *proc_virtual; ++ ++extern void proc_vx_init(void); ++ + static int proc_test_super(struct super_block *sb, void *data) + { + return sb->s_fs_info == data; +@@ -125,6 +130,7 @@ void __init proc_root_init(void) + #endif + proc_mkdir("bus", NULL); + proc_sys_init(); ++ proc_vx_init(); + } + + static int proc_root_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat +@@ -193,6 +199,7 @@ struct proc_dir_entry proc_root = { + .proc_iops = &proc_root_inode_operations, + .proc_fops = &proc_root_operations, + .parent = &proc_root, ++ .vx_flags = IATTR_ADMIN | IATTR_WATCH, + }; + + int pid_ns_prepare_proc(struct pid_namespace *ns) +diff -NurpP --minimal linux-3.0.9/fs/proc/uptime.c linux-3.0.9-vs2.3.2.1/fs/proc/uptime.c +--- linux-3.0.9/fs/proc/uptime.c 2009-12-03 20:02:53.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/proc/uptime.c 2011-06-10 22:11:24.000000000 +0200 +@@ -4,22 +4,22 @@ + #include <linux/sched.h> + #include <linux/seq_file.h> + #include <linux/time.h> +-#include <linux/kernel_stat.h> ++#include <linux/vserver/cvirt.h> + #include <asm/cputime.h> + + static int uptime_proc_show(struct seq_file *m, void *v) + { + struct timespec uptime; + struct timespec idle; +- int i; +- cputime_t idletime = cputime_zero; +- +- for_each_possible_cpu(i) +- idletime = cputime64_add(idletime, kstat_cpu(i).cpustat.idle); ++ cputime_t idletime = cputime_add(init_task.utime, init_task.stime); + + do_posix_clock_monotonic_gettime(&uptime); + monotonic_to_bootbased(&uptime); + cputime_to_timespec(idletime, &idle); ++ ++ if (vx_flags(VXF_VIRT_UPTIME, 0)) ++ vx_vsi_uptime(&uptime, &idle); ++ + seq_printf(m, "%lu.%02lu %lu.%02lu\n", + (unsigned long) uptime.tv_sec, + (uptime.tv_nsec / (NSEC_PER_SEC / 100)), +diff -NurpP --minimal linux-3.0.9/fs/quota/dquot.c linux-3.0.9-vs2.3.2.1/fs/quota/dquot.c +--- linux-3.0.9/fs/quota/dquot.c 2011-07-22 11:18:06.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/quota/dquot.c 2011-06-10 22:11:24.000000000 +0200 +@@ -1548,6 +1548,9 @@ int __dquot_alloc_space(struct inode *in + int reserve = flags & DQUOT_SPACE_RESERVE; + int nofail = flags & DQUOT_SPACE_NOFAIL; + ++ if ((ret = dl_alloc_space(inode, number))) ++ return ret; ++ + /* + * First test before acquiring mutex - solves deadlocks when we + * re-enter the quota code and are already holding the mutex +@@ -1602,6 +1605,9 @@ int dquot_alloc_inode(const struct inode + int cnt, ret = 0; + char warntype[MAXQUOTAS]; + ++ if ((ret = dl_alloc_inode(inode))) ++ return ret; ++ + /* First test before acquiring mutex - solves deadlocks when we + * re-enter the quota code and are already holding the mutex */ + if (!dquot_active(inode)) +@@ -1672,6 +1678,8 @@ void __dquot_free_space(struct inode *in + char warntype[MAXQUOTAS]; + int reserve = flags & DQUOT_SPACE_RESERVE; + ++ dl_free_space(inode, number); ++ + /* First test before acquiring mutex - solves deadlocks when we + * re-enter the quota code and are already holding the mutex */ + if (!dquot_active(inode)) { +@@ -1710,6 +1718,8 @@ void dquot_free_inode(const struct inode + unsigned int cnt; + char warntype[MAXQUOTAS]; + ++ dl_free_inode(inode); ++ + /* First test before acquiring mutex - solves deadlocks when we + * re-enter the quota code and are already holding the mutex */ + if (!dquot_active(inode)) +diff -NurpP --minimal linux-3.0.9/fs/quota/quota.c linux-3.0.9-vs2.3.2.1/fs/quota/quota.c +--- linux-3.0.9/fs/quota/quota.c 2011-11-15 16:40:47.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/quota/quota.c 2011-11-15 17:37:07.000000000 +0100 +@@ -8,6 +8,7 @@ + #include <linux/fs.h> + #include <linux/namei.h> + #include <linux/slab.h> ++#include <linux/vs_context.h> + #include <asm/current.h> + #include <asm/uaccess.h> + #include <linux/kernel.h> +@@ -38,7 +39,7 @@ static int check_quotactl_permission(str + break; + /*FALLTHROUGH*/ + default: +- if (!capable(CAP_SYS_ADMIN)) ++ if (!vx_capable(CAP_SYS_ADMIN, VXC_QUOTA_CTL)) + return -EPERM; + } + +@@ -293,6 +294,46 @@ static int do_quotactl(struct super_bloc + } + } + ++#if defined(CONFIG_BLK_DEV_VROOT) || defined(CONFIG_BLK_DEV_VROOT_MODULE) ++ ++#include <linux/vroot.h> ++#include <linux/major.h> ++#include <linux/module.h> ++#include <linux/kallsyms.h> ++#include <linux/vserver/debug.h> ++ ++static vroot_grb_func *vroot_get_real_bdev = NULL; ++ ++static DEFINE_SPINLOCK(vroot_grb_lock); ++ ++int register_vroot_grb(vroot_grb_func *func) { ++ int ret = -EBUSY; ++ ++ spin_lock(&vroot_grb_lock); ++ if (!vroot_get_real_bdev) { ++ vroot_get_real_bdev = func; ++ ret = 0; ++ } ++ spin_unlock(&vroot_grb_lock); ++ return ret; ++} ++EXPORT_SYMBOL(register_vroot_grb); ++ ++int unregister_vroot_grb(vroot_grb_func *func) { ++ int ret = -EINVAL; ++ ++ spin_lock(&vroot_grb_lock); ++ if (vroot_get_real_bdev) { ++ vroot_get_real_bdev = NULL; ++ ret = 0; ++ } ++ spin_unlock(&vroot_grb_lock); ++ return ret; ++} ++EXPORT_SYMBOL(unregister_vroot_grb); ++ ++#endif ++ + /* + * look up a superblock on which quota ops will be performed + * - use the name of a block device to find the superblock thereon +@@ -310,6 +351,22 @@ static struct super_block *quotactl_bloc + putname(tmp); + if (IS_ERR(bdev)) + return ERR_CAST(bdev); ++#if defined(CONFIG_BLK_DEV_VROOT) || defined(CONFIG_BLK_DEV_VROOT_MODULE) ++ if (bdev && bdev->bd_inode && ++ imajor(bdev->bd_inode) == VROOT_MAJOR) { ++ struct block_device *bdnew = (void *)-EINVAL; ++ ++ if (vroot_get_real_bdev) ++ bdnew = vroot_get_real_bdev(bdev); ++ else ++ vxdprintk(VXD_CBIT(misc, 0), ++ "vroot_get_real_bdev not set"); ++ bdput(bdev); ++ if (IS_ERR(bdnew)) ++ return ERR_PTR(PTR_ERR(bdnew)); ++ bdev = bdnew; ++ } ++#endif + sb = get_super(bdev); + bdput(bdev); + if (!sb) +diff -NurpP --minimal linux-3.0.9/fs/reiserfs/file.c linux-3.0.9-vs2.3.2.1/fs/reiserfs/file.c +--- linux-3.0.9/fs/reiserfs/file.c 2011-01-05 21:50:26.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/reiserfs/file.c 2011-06-10 22:11:24.000000000 +0200 +@@ -312,4 +312,5 @@ const struct inode_operations reiserfs_f + .listxattr = reiserfs_listxattr, + .removexattr = reiserfs_removexattr, + .permission = reiserfs_permission, ++ .sync_flags = reiserfs_sync_flags, + }; +diff -NurpP --minimal linux-3.0.9/fs/reiserfs/inode.c linux-3.0.9-vs2.3.2.1/fs/reiserfs/inode.c +--- linux-3.0.9/fs/reiserfs/inode.c 2011-05-22 16:17:53.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/reiserfs/inode.c 2011-06-10 22:11:24.000000000 +0200 +@@ -18,6 +18,7 @@ + #include <linux/writeback.h> + #include <linux/quotaops.h> + #include <linux/swap.h> ++#include <linux/vs_tag.h> + + int reiserfs_commit_write(struct file *f, struct page *page, + unsigned from, unsigned to); +@@ -1131,6 +1132,8 @@ static void init_inode(struct inode *ino + struct buffer_head *bh; + struct item_head *ih; + __u32 rdev; ++ uid_t uid; ++ gid_t gid; + //int version = ITEM_VERSION_1; + + bh = PATH_PLAST_BUFFER(path); +@@ -1151,12 +1154,13 @@ static void init_inode(struct inode *ino + (struct stat_data_v1 *)B_I_PITEM(bh, ih); + unsigned long blocks; + ++ uid = sd_v1_uid(sd); ++ gid = sd_v1_gid(sd); ++ + set_inode_item_key_version(inode, KEY_FORMAT_3_5); + set_inode_sd_version(inode, STAT_DATA_V1); + inode->i_mode = sd_v1_mode(sd); + inode->i_nlink = sd_v1_nlink(sd); +- inode->i_uid = sd_v1_uid(sd); +- inode->i_gid = sd_v1_gid(sd); + inode->i_size = sd_v1_size(sd); + inode->i_atime.tv_sec = sd_v1_atime(sd); + inode->i_mtime.tv_sec = sd_v1_mtime(sd); +@@ -1198,11 +1202,12 @@ static void init_inode(struct inode *ino + // (directories and symlinks) + struct stat_data *sd = (struct stat_data *)B_I_PITEM(bh, ih); + ++ uid = sd_v2_uid(sd); ++ gid = sd_v2_gid(sd); ++ + inode->i_mode = sd_v2_mode(sd); + inode->i_nlink = sd_v2_nlink(sd); +- inode->i_uid = sd_v2_uid(sd); + inode->i_size = sd_v2_size(sd); +- inode->i_gid = sd_v2_gid(sd); + inode->i_mtime.tv_sec = sd_v2_mtime(sd); + inode->i_atime.tv_sec = sd_v2_atime(sd); + inode->i_ctime.tv_sec = sd_v2_ctime(sd); +@@ -1232,6 +1237,10 @@ static void init_inode(struct inode *ino + sd_attrs_to_i_attrs(sd_v2_attrs(sd), inode); + } + ++ inode->i_uid = INOTAG_UID(DX_TAG(inode), uid, gid); ++ inode->i_gid = INOTAG_GID(DX_TAG(inode), uid, gid); ++ inode->i_tag = INOTAG_TAG(DX_TAG(inode), uid, gid, 0); ++ + pathrelse(path); + if (S_ISREG(inode->i_mode)) { + inode->i_op = &reiserfs_file_inode_operations; +@@ -1254,13 +1263,15 @@ static void init_inode(struct inode *ino + static void inode2sd(void *sd, struct inode *inode, loff_t size) + { + struct stat_data *sd_v2 = (struct stat_data *)sd; ++ uid_t uid = TAGINO_UID(DX_TAG(inode), inode->i_uid, inode->i_tag); ++ gid_t gid = TAGINO_GID(DX_TAG(inode), inode->i_gid, inode->i_tag); + __u16 flags; + ++ set_sd_v2_uid(sd_v2, uid); ++ set_sd_v2_gid(sd_v2, gid); + set_sd_v2_mode(sd_v2, inode->i_mode); + set_sd_v2_nlink(sd_v2, inode->i_nlink); +- set_sd_v2_uid(sd_v2, inode->i_uid); + set_sd_v2_size(sd_v2, size); +- set_sd_v2_gid(sd_v2, inode->i_gid); + set_sd_v2_mtime(sd_v2, inode->i_mtime.tv_sec); + set_sd_v2_atime(sd_v2, inode->i_atime.tv_sec); + set_sd_v2_ctime(sd_v2, inode->i_ctime.tv_sec); +@@ -2863,14 +2874,19 @@ int reiserfs_commit_write(struct file *f + void sd_attrs_to_i_attrs(__u16 sd_attrs, struct inode *inode) + { + if (reiserfs_attrs(inode->i_sb)) { +- if (sd_attrs & REISERFS_SYNC_FL) +- inode->i_flags |= S_SYNC; +- else +- inode->i_flags &= ~S_SYNC; + if (sd_attrs & REISERFS_IMMUTABLE_FL) + inode->i_flags |= S_IMMUTABLE; + else + inode->i_flags &= ~S_IMMUTABLE; ++ if (sd_attrs & REISERFS_IXUNLINK_FL) ++ inode->i_flags |= S_IXUNLINK; ++ else ++ inode->i_flags &= ~S_IXUNLINK; ++ ++ if (sd_attrs & REISERFS_SYNC_FL) ++ inode->i_flags |= S_SYNC; ++ else ++ inode->i_flags &= ~S_SYNC; + if (sd_attrs & REISERFS_APPEND_FL) + inode->i_flags |= S_APPEND; + else +@@ -2883,6 +2899,15 @@ void sd_attrs_to_i_attrs(__u16 sd_attrs, + REISERFS_I(inode)->i_flags |= i_nopack_mask; + else + REISERFS_I(inode)->i_flags &= ~i_nopack_mask; ++ ++ if (sd_attrs & REISERFS_BARRIER_FL) ++ inode->i_vflags |= V_BARRIER; ++ else ++ inode->i_vflags &= ~V_BARRIER; ++ if (sd_attrs & REISERFS_COW_FL) ++ inode->i_vflags |= V_COW; ++ else ++ inode->i_vflags &= ~V_COW; + } + } + +@@ -2893,6 +2918,11 @@ void i_attrs_to_sd_attrs(struct inode *i + *sd_attrs |= REISERFS_IMMUTABLE_FL; + else + *sd_attrs &= ~REISERFS_IMMUTABLE_FL; ++ if (inode->i_flags & S_IXUNLINK) ++ *sd_attrs |= REISERFS_IXUNLINK_FL; ++ else ++ *sd_attrs &= ~REISERFS_IXUNLINK_FL; ++ + if (inode->i_flags & S_SYNC) + *sd_attrs |= REISERFS_SYNC_FL; + else +@@ -2905,6 +2935,15 @@ void i_attrs_to_sd_attrs(struct inode *i + *sd_attrs |= REISERFS_NOTAIL_FL; + else + *sd_attrs &= ~REISERFS_NOTAIL_FL; ++ ++ if (inode->i_vflags & V_BARRIER) ++ *sd_attrs |= REISERFS_BARRIER_FL; ++ else ++ *sd_attrs &= ~REISERFS_BARRIER_FL; ++ if (inode->i_vflags & V_COW) ++ *sd_attrs |= REISERFS_COW_FL; ++ else ++ *sd_attrs &= ~REISERFS_COW_FL; + } + } + +@@ -3148,7 +3187,8 @@ int reiserfs_setattr(struct dentry *dent + } + + if ((ia_valid & ATTR_UID && attr->ia_uid != inode->i_uid) || +- (ia_valid & ATTR_GID && attr->ia_gid != inode->i_gid)) { ++ (ia_valid & ATTR_GID && attr->ia_gid != inode->i_gid) || ++ (ia_valid & ATTR_TAG && attr->ia_tag != inode->i_tag)) { + struct reiserfs_transaction_handle th; + int jbegin_count = + 2 * +@@ -3177,6 +3217,9 @@ int reiserfs_setattr(struct dentry *dent + inode->i_uid = attr->ia_uid; + if (attr->ia_valid & ATTR_GID) + inode->i_gid = attr->ia_gid; ++ if ((attr->ia_valid & ATTR_TAG) && ++ IS_TAGGED(inode)) ++ inode->i_tag = attr->ia_tag; + mark_inode_dirty(inode); + error = journal_end(&th, inode->i_sb, jbegin_count); + if (error) +diff -NurpP --minimal linux-3.0.9/fs/reiserfs/ioctl.c linux-3.0.9-vs2.3.2.1/fs/reiserfs/ioctl.c +--- linux-3.0.9/fs/reiserfs/ioctl.c 2011-05-22 16:17:53.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/reiserfs/ioctl.c 2011-06-10 22:11:24.000000000 +0200 +@@ -11,6 +11,21 @@ + #include <linux/pagemap.h> + #include <linux/compat.h> + ++ ++int reiserfs_sync_flags(struct inode *inode, int flags, int vflags) ++{ ++ __u16 sd_attrs = 0; ++ ++ inode->i_flags = flags; ++ inode->i_vflags = vflags; ++ ++ i_attrs_to_sd_attrs(inode, &sd_attrs); ++ REISERFS_I(inode)->i_attrs = sd_attrs; ++ inode->i_ctime = CURRENT_TIME_SEC; ++ mark_inode_dirty(inode); ++ return 0; ++} ++ + /* + * reiserfs_ioctl - handler for ioctl for inode + * supported commands: +@@ -22,7 +37,7 @@ + long reiserfs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) + { + struct inode *inode = filp->f_path.dentry->d_inode; +- unsigned int flags; ++ unsigned int flags, oldflags; + int err = 0; + + reiserfs_write_lock(inode->i_sb); +@@ -47,6 +62,7 @@ long reiserfs_ioctl(struct file *filp, u + + flags = REISERFS_I(inode)->i_attrs; + i_attrs_to_sd_attrs(inode, (__u16 *) & flags); ++ flags &= REISERFS_FL_USER_VISIBLE; + err = put_user(flags, (int __user *)arg); + break; + case REISERFS_IOC_SETFLAGS:{ +@@ -67,6 +83,10 @@ long reiserfs_ioctl(struct file *filp, u + err = -EFAULT; + goto setflags_out; + } ++ if (IS_BARRIER(inode)) { ++ vxwprintk_task(1, "messing with the barrier."); ++ return -EACCES; ++ } + /* + * Is it quota file? Do not allow user to mess with it + */ +@@ -91,6 +111,10 @@ long reiserfs_ioctl(struct file *filp, u + goto setflags_out; + } + } ++ ++ oldflags = REISERFS_I(inode)->i_attrs; ++ flags &= REISERFS_FL_USER_MODIFIABLE; ++ flags |= oldflags & ~REISERFS_FL_USER_MODIFIABLE; + sd_attrs_to_i_attrs(flags, inode); + REISERFS_I(inode)->i_attrs = flags; + inode->i_ctime = CURRENT_TIME_SEC; +diff -NurpP --minimal linux-3.0.9/fs/reiserfs/namei.c linux-3.0.9-vs2.3.2.1/fs/reiserfs/namei.c +--- linux-3.0.9/fs/reiserfs/namei.c 2011-05-22 16:17:53.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/reiserfs/namei.c 2011-06-10 22:11:24.000000000 +0200 +@@ -18,6 +18,7 @@ + #include <linux/reiserfs_acl.h> + #include <linux/reiserfs_xattr.h> + #include <linux/quotaops.h> ++#include <linux/vs_tag.h> + + #define INC_DIR_INODE_NLINK(i) if (i->i_nlink != 1) { inc_nlink(i); if (i->i_nlink >= REISERFS_LINK_MAX) i->i_nlink=1; } + #define DEC_DIR_INODE_NLINK(i) if (i->i_nlink != 1) drop_nlink(i); +@@ -362,6 +363,7 @@ static struct dentry *reiserfs_lookup(st + if (retval == IO_ERROR) { + return ERR_PTR(-EIO); + } ++ dx_propagate_tag(nd, inode); + + return d_splice_alias(inode, dentry); + } +@@ -1529,6 +1531,7 @@ const struct inode_operations reiserfs_d + .listxattr = reiserfs_listxattr, + .removexattr = reiserfs_removexattr, + .permission = reiserfs_permission, ++ .sync_flags = reiserfs_sync_flags, + }; + + /* +diff -NurpP --minimal linux-3.0.9/fs/reiserfs/super.c linux-3.0.9-vs2.3.2.1/fs/reiserfs/super.c +--- linux-3.0.9/fs/reiserfs/super.c 2011-07-22 11:18:06.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/reiserfs/super.c 2011-06-10 22:11:24.000000000 +0200 +@@ -899,6 +899,14 @@ static int reiserfs_parse_options(struct + {"user_xattr",.setmask = 1 << REISERFS_UNSUPPORTED_OPT}, + {"nouser_xattr",.clrmask = 1 << REISERFS_UNSUPPORTED_OPT}, + #endif ++#ifndef CONFIG_TAGGING_NONE ++ {"tagxid",.setmask = 1 << REISERFS_TAGGED}, ++ {"tag",.setmask = 1 << REISERFS_TAGGED}, ++ {"notag",.clrmask = 1 << REISERFS_TAGGED}, ++#endif ++#ifdef CONFIG_PROPAGATE ++ {"tag",.arg_required = 'T',.values = NULL}, ++#endif + #ifdef CONFIG_REISERFS_FS_POSIX_ACL + {"acl",.setmask = 1 << REISERFS_POSIXACL}, + {"noacl",.clrmask = 1 << REISERFS_POSIXACL}, +@@ -1208,6 +1216,14 @@ static int reiserfs_remount(struct super + handle_quota_files(s, qf_names, &qfmt); + #endif + ++ if ((mount_options & (1 << REISERFS_TAGGED)) && ++ !(s->s_flags & MS_TAGGED)) { ++ reiserfs_warning(s, "super-vs01", ++ "reiserfs: tagging not permitted on remount."); ++ err = -EINVAL; ++ goto out_err; ++ } ++ + handle_attrs(s); + + /* Add options that are safe here */ +@@ -1690,6 +1706,10 @@ static int reiserfs_fill_super(struct su + goto error; + } + ++ /* map mount option tagxid */ ++ if (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_TAGGED)) ++ s->s_flags |= MS_TAGGED; ++ + rs = SB_DISK_SUPER_BLOCK(s); + /* Let's do basic sanity check to verify that underlying device is not + smaller than the filesystem. If the check fails then abort and scream, +diff -NurpP --minimal linux-3.0.9/fs/reiserfs/xattr.c linux-3.0.9-vs2.3.2.1/fs/reiserfs/xattr.c +--- linux-3.0.9/fs/reiserfs/xattr.c 2011-07-22 11:18:06.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/reiserfs/xattr.c 2011-06-22 12:39:15.000000000 +0200 +@@ -40,6 +40,7 @@ + #include <linux/errno.h> + #include <linux/gfp.h> + #include <linux/fs.h> ++#include <linux/mount.h> + #include <linux/file.h> + #include <linux/pagemap.h> + #include <linux/xattr.h> +diff -NurpP --minimal linux-3.0.9/fs/stat.c linux-3.0.9-vs2.3.2.1/fs/stat.c +--- linux-3.0.9/fs/stat.c 2011-11-15 16:40:47.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/stat.c 2011-11-15 17:37:07.000000000 +0100 +@@ -26,6 +26,7 @@ void generic_fillattr(struct inode *inod + stat->nlink = inode->i_nlink; + stat->uid = inode->i_uid; + stat->gid = inode->i_gid; ++ stat->tag = inode->i_tag; + stat->rdev = inode->i_rdev; + stat->atime = inode->i_atime; + stat->mtime = inode->i_mtime; +diff -NurpP --minimal linux-3.0.9/fs/statfs.c linux-3.0.9-vs2.3.2.1/fs/statfs.c +--- linux-3.0.9/fs/statfs.c 2011-11-15 16:40:47.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/statfs.c 2011-11-15 17:37:07.000000000 +0100 +@@ -7,6 +7,8 @@ + #include <linux/statfs.h> + #include <linux/security.h> + #include <linux/uaccess.h> ++#include <linux/vs_base.h> ++#include <linux/vs_dlimit.h> + + static int flags_by_mnt(int mnt_flags) + { +@@ -59,6 +61,8 @@ int statfs_by_dentry(struct dentry *dent + retval = dentry->d_sb->s_op->statfs(dentry, buf); + if (retval == 0 && buf->f_frsize == 0) + buf->f_frsize = buf->f_bsize; ++ if (!vx_check(0, VS_ADMIN|VS_WATCH)) ++ vx_vsi_statfs(dentry->d_sb, buf); + return retval; + } + +diff -NurpP --minimal linux-3.0.9/fs/super.c linux-3.0.9-vs2.3.2.1/fs/super.c +--- linux-3.0.9/fs/super.c 2011-07-22 11:18:06.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/super.c 2011-07-23 16:49:09.000000000 +0200 +@@ -32,6 +32,9 @@ + #include <linux/backing-dev.h> + #include <linux/rculist_bl.h> + #include <linux/cleancache.h> ++#include <linux/devpts_fs.h> ++#include <linux/proc_fs.h> ++#include <linux/vs_context.h> + #include "internal.h" + + +@@ -943,6 +946,13 @@ mount_fs(struct file_system_type *type, + WARN_ON(sb->s_bdi == &default_backing_dev_info); + sb->s_flags |= MS_BORN; + ++ error = -EPERM; ++ if (!vx_capable(CAP_SYS_ADMIN, VXC_BINARY_MOUNT) && ++ !sb->s_bdev && ++ (sb->s_magic != PROC_SUPER_MAGIC) && ++ (sb->s_magic != DEVPTS_SUPER_MAGIC)) ++ goto out_sb; ++ + error = security_sb_kern_mount(sb, flags, secdata); + if (error) + goto out_sb; +diff -NurpP --minimal linux-3.0.9/fs/sysfs/mount.c linux-3.0.9-vs2.3.2.1/fs/sysfs/mount.c +--- linux-3.0.9/fs/sysfs/mount.c 2011-07-22 11:18:06.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/sysfs/mount.c 2011-06-22 12:39:15.000000000 +0200 +@@ -47,7 +47,7 @@ static int sysfs_fill_super(struct super + + sb->s_blocksize = PAGE_CACHE_SIZE; + sb->s_blocksize_bits = PAGE_CACHE_SHIFT; +- sb->s_magic = SYSFS_MAGIC; ++ sb->s_magic = SYSFS_SUPER_MAGIC; + sb->s_op = &sysfs_ops; + sb->s_time_gran = 1; + +diff -NurpP --minimal linux-3.0.9/fs/utimes.c linux-3.0.9-vs2.3.2.1/fs/utimes.c +--- linux-3.0.9/fs/utimes.c 2011-05-22 16:17:54.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/utimes.c 2011-06-10 22:11:24.000000000 +0200 +@@ -8,6 +8,8 @@ + #include <linux/stat.h> + #include <linux/utime.h> + #include <linux/syscalls.h> ++#include <linux/mount.h> ++#include <linux/vs_cowbl.h> + #include <asm/uaccess.h> + #include <asm/unistd.h> + +@@ -52,12 +54,18 @@ static int utimes_common(struct path *pa + { + int error; + struct iattr newattrs; +- struct inode *inode = path->dentry->d_inode; ++ struct inode *inode; + + error = mnt_want_write(path->mnt); + if (error) + goto out; + ++ error = cow_check_and_break(path); ++ if (error) ++ goto mnt_drop_write_and_out; ++ ++ inode = path->dentry->d_inode; ++ + if (times && times[0].tv_nsec == UTIME_NOW && + times[1].tv_nsec == UTIME_NOW) + times = NULL; +diff -NurpP --minimal linux-3.0.9/fs/xattr.c linux-3.0.9-vs2.3.2.1/fs/xattr.c +--- linux-3.0.9/fs/xattr.c 2011-07-22 11:18:09.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/xattr.c 2011-06-10 23:10:19.000000000 +0200 +@@ -18,6 +18,7 @@ + #include <linux/module.h> + #include <linux/fsnotify.h> + #include <linux/audit.h> ++#include <linux/mount.h> + #include <asm/uaccess.h> + + +@@ -49,7 +50,7 @@ xattr_permission(struct inode *inode, co + * The trusted.* namespace can only be accessed by privileged users. + */ + if (!strncmp(name, XATTR_TRUSTED_PREFIX, XATTR_TRUSTED_PREFIX_LEN)) { +- if (!capable(CAP_SYS_ADMIN)) ++ if (!vx_capable(CAP_SYS_ADMIN, VXC_FS_TRUSTED)) + return (mask & MAY_WRITE) ? -EPERM : -ENODATA; + return 0; + } +diff -NurpP --minimal linux-3.0.9/fs/xfs/linux-2.6/xfs_ioctl.c linux-3.0.9-vs2.3.2.1/fs/xfs/linux-2.6/xfs_ioctl.c +--- linux-3.0.9/fs/xfs/linux-2.6/xfs_ioctl.c 2011-05-22 16:17:54.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/xfs/linux-2.6/xfs_ioctl.c 2011-06-10 22:11:24.000000000 +0200 +@@ -28,7 +28,7 @@ + #include "xfs_bmap_btree.h" + #include "xfs_dinode.h" + #include "xfs_inode.h" +-#include "xfs_ioctl.h" ++// #include "xfs_ioctl.h" + #include "xfs_rtalloc.h" + #include "xfs_itable.h" + #include "xfs_error.h" +@@ -748,6 +748,10 @@ xfs_merge_ioc_xflags( + xflags |= XFS_XFLAG_IMMUTABLE; + else + xflags &= ~XFS_XFLAG_IMMUTABLE; ++ if (flags & FS_IXUNLINK_FL) ++ xflags |= XFS_XFLAG_IXUNLINK; ++ else ++ xflags &= ~XFS_XFLAG_IXUNLINK; + if (flags & FS_APPEND_FL) + xflags |= XFS_XFLAG_APPEND; + else +@@ -776,6 +780,8 @@ xfs_di2lxflags( + + if (di_flags & XFS_DIFLAG_IMMUTABLE) + flags |= FS_IMMUTABLE_FL; ++ if (di_flags & XFS_DIFLAG_IXUNLINK) ++ flags |= FS_IXUNLINK_FL; + if (di_flags & XFS_DIFLAG_APPEND) + flags |= FS_APPEND_FL; + if (di_flags & XFS_DIFLAG_SYNC) +@@ -836,6 +842,8 @@ xfs_set_diflags( + di_flags = (ip->i_d.di_flags & XFS_DIFLAG_PREALLOC); + if (xflags & XFS_XFLAG_IMMUTABLE) + di_flags |= XFS_DIFLAG_IMMUTABLE; ++ if (xflags & XFS_XFLAG_IXUNLINK) ++ di_flags |= XFS_DIFLAG_IXUNLINK; + if (xflags & XFS_XFLAG_APPEND) + di_flags |= XFS_DIFLAG_APPEND; + if (xflags & XFS_XFLAG_SYNC) +@@ -878,6 +886,10 @@ xfs_diflags_to_linux( + inode->i_flags |= S_IMMUTABLE; + else + inode->i_flags &= ~S_IMMUTABLE; ++ if (xflags & XFS_XFLAG_IXUNLINK) ++ inode->i_flags |= S_IXUNLINK; ++ else ++ inode->i_flags &= ~S_IXUNLINK; + if (xflags & XFS_XFLAG_APPEND) + inode->i_flags |= S_APPEND; + else +@@ -1370,10 +1382,18 @@ xfs_file_ioctl( + case XFS_IOC_FSGETXATTRA: + return xfs_ioc_fsgetxattr(ip, 1, arg); + case XFS_IOC_FSSETXATTR: ++ if (IS_BARRIER(inode)) { ++ vxwprintk_task(1, "messing with the barrier."); ++ return -XFS_ERROR(EACCES); ++ } + return xfs_ioc_fssetxattr(ip, filp, arg); + case XFS_IOC_GETXFLAGS: + return xfs_ioc_getxflags(ip, arg); + case XFS_IOC_SETXFLAGS: ++ if (IS_BARRIER(inode)) { ++ vxwprintk_task(1, "messing with the barrier."); ++ return -XFS_ERROR(EACCES); ++ } + return xfs_ioc_setxflags(ip, filp, arg); + + case XFS_IOC_FSSETDM: { +diff -NurpP --minimal linux-3.0.9/fs/xfs/linux-2.6/xfs_ioctl.h linux-3.0.9-vs2.3.2.1/fs/xfs/linux-2.6/xfs_ioctl.h +--- linux-3.0.9/fs/xfs/linux-2.6/xfs_ioctl.h 2010-07-07 18:31:54.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/xfs/linux-2.6/xfs_ioctl.h 2011-06-10 22:11:24.000000000 +0200 +@@ -70,6 +70,12 @@ xfs_handle_to_dentry( + void __user *uhandle, + u32 hlen); + ++extern int ++xfs_sync_flags( ++ struct inode *inode, ++ int flags, ++ int vflags); ++ + extern long + xfs_file_ioctl( + struct file *filp, +diff -NurpP --minimal linux-3.0.9/fs/xfs/linux-2.6/xfs_iops.c linux-3.0.9-vs2.3.2.1/fs/xfs/linux-2.6/xfs_iops.c +--- linux-3.0.9/fs/xfs/linux-2.6/xfs_iops.c 2011-07-22 11:18:09.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/xfs/linux-2.6/xfs_iops.c 2011-06-22 12:39:15.000000000 +0200 +@@ -30,6 +30,7 @@ + #include "xfs_bmap_btree.h" + #include "xfs_dinode.h" + #include "xfs_inode.h" ++#include "xfs_ioctl.h" + #include "xfs_bmap.h" + #include "xfs_rtalloc.h" + #include "xfs_error.h" +@@ -48,6 +49,7 @@ + #include <linux/security.h> + #include <linux/fiemap.h> + #include <linux/slab.h> ++#include <linux/vs_tag.h> + + /* + * Bring the timestamps in the XFS inode uptodate. +@@ -464,6 +466,7 @@ xfs_vn_getattr( + stat->nlink = ip->i_d.di_nlink; + stat->uid = ip->i_d.di_uid; + stat->gid = ip->i_d.di_gid; ++ stat->tag = ip->i_d.di_tag; + stat->ino = ip->i_ino; + stat->atime = inode->i_atime; + stat->mtime = inode->i_mtime; +@@ -599,6 +602,7 @@ static const struct inode_operations xfs + .removexattr = generic_removexattr, + .listxattr = xfs_vn_listxattr, + .fiemap = xfs_vn_fiemap, ++ .sync_flags = xfs_sync_flags, + }; + + static const struct inode_operations xfs_dir_inode_operations = { +@@ -624,6 +628,7 @@ static const struct inode_operations xfs + .getxattr = generic_getxattr, + .removexattr = generic_removexattr, + .listxattr = xfs_vn_listxattr, ++ .sync_flags = xfs_sync_flags, + }; + + static const struct inode_operations xfs_dir_ci_inode_operations = { +@@ -673,6 +678,10 @@ xfs_diflags_to_iflags( + inode->i_flags |= S_IMMUTABLE; + else + inode->i_flags &= ~S_IMMUTABLE; ++ if (ip->i_d.di_flags & XFS_DIFLAG_IXUNLINK) ++ inode->i_flags |= S_IXUNLINK; ++ else ++ inode->i_flags &= ~S_IXUNLINK; + if (ip->i_d.di_flags & XFS_DIFLAG_APPEND) + inode->i_flags |= S_APPEND; + else +@@ -685,6 +694,15 @@ xfs_diflags_to_iflags( + inode->i_flags |= S_NOATIME; + else + inode->i_flags &= ~S_NOATIME; ++ ++ if (ip->i_d.di_vflags & XFS_DIVFLAG_BARRIER) ++ inode->i_vflags |= V_BARRIER; ++ else ++ inode->i_vflags &= ~V_BARRIER; ++ if (ip->i_d.di_vflags & XFS_DIVFLAG_COW) ++ inode->i_vflags |= V_COW; ++ else ++ inode->i_vflags &= ~V_COW; + } + + /* +@@ -716,6 +734,7 @@ xfs_setup_inode( + inode->i_nlink = ip->i_d.di_nlink; + inode->i_uid = ip->i_d.di_uid; + inode->i_gid = ip->i_d.di_gid; ++ inode->i_tag = ip->i_d.di_tag; + + switch (inode->i_mode & S_IFMT) { + case S_IFBLK: +diff -NurpP --minimal linux-3.0.9/fs/xfs/linux-2.6/xfs_linux.h linux-3.0.9-vs2.3.2.1/fs/xfs/linux-2.6/xfs_linux.h +--- linux-3.0.9/fs/xfs/linux-2.6/xfs_linux.h 2011-11-15 16:40:47.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/xfs/linux-2.6/xfs_linux.h 2011-11-15 17:37:07.000000000 +0100 +@@ -117,6 +117,7 @@ + + #define current_cpu() (raw_smp_processor_id()) + #define current_pid() (current->pid) ++#define current_fstag(vp) (dx_current_fstag((vp)->i_sb)) + #define current_test_flags(f) (current->flags & (f)) + #define current_set_flags_nested(sp, f) \ + (*(sp) = current->flags, current->flags |= (f)) +diff -NurpP --minimal linux-3.0.9/fs/xfs/linux-2.6/xfs_super.c linux-3.0.9-vs2.3.2.1/fs/xfs/linux-2.6/xfs_super.c +--- linux-3.0.9/fs/xfs/linux-2.6/xfs_super.c 2011-11-15 16:40:47.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/xfs/linux-2.6/xfs_super.c 2011-11-15 17:37:07.000000000 +0100 +@@ -114,6 +114,9 @@ mempool_t *xfs_ioend_pool; + #define MNTOPT_NODELAYLOG "nodelaylog" /* Delayed logging disabled */ + #define MNTOPT_DISCARD "discard" /* Discard unused blocks */ + #define MNTOPT_NODISCARD "nodiscard" /* Do not discard unused blocks */ ++#define MNTOPT_TAGXID "tagxid" /* context tagging for inodes */ ++#define MNTOPT_TAGGED "tag" /* context tagging for inodes */ ++#define MNTOPT_NOTAGTAG "notag" /* do not use context tagging */ + + /* + * Table driven mount option parser. +@@ -122,10 +125,14 @@ mempool_t *xfs_ioend_pool; + * in the future, too. + */ + enum { ++ Opt_tag, Opt_notag, + Opt_barrier, Opt_nobarrier, Opt_err + }; + + static const match_table_t tokens = { ++ {Opt_tag, "tagxid"}, ++ {Opt_tag, "tag"}, ++ {Opt_notag, "notag"}, + {Opt_barrier, "barrier"}, + {Opt_nobarrier, "nobarrier"}, + {Opt_err, NULL} +@@ -373,6 +380,19 @@ xfs_parseargs( + } else if (!strcmp(this_char, "irixsgid")) { + xfs_warn(mp, + "irixsgid is now a sysctl(2) variable, option is deprecated."); ++#ifndef CONFIG_TAGGING_NONE ++ } else if (!strcmp(this_char, MNTOPT_TAGGED)) { ++ mp->m_flags |= XFS_MOUNT_TAGGED; ++ } else if (!strcmp(this_char, MNTOPT_NOTAGTAG)) { ++ mp->m_flags &= ~XFS_MOUNT_TAGGED; ++ } else if (!strcmp(this_char, MNTOPT_TAGXID)) { ++ mp->m_flags |= XFS_MOUNT_TAGGED; ++#endif ++#ifdef CONFIG_PROPAGATE ++ } else if (!strcmp(this_char, MNTOPT_TAGGED)) { ++ /* use value */ ++ mp->m_flags |= XFS_MOUNT_TAGGED; ++#endif + } else { + xfs_warn(mp, "unknown mount option [%s].", this_char); + return EINVAL; +@@ -1182,6 +1202,16 @@ xfs_fs_remount( + case Opt_nobarrier: + mp->m_flags &= ~XFS_MOUNT_BARRIER; + break; ++ case Opt_tag: ++ if (!(sb->s_flags & MS_TAGGED)) { ++ printk(KERN_INFO ++ "XFS: %s: tagging not permitted on remount.\n", ++ sb->s_id); ++ return -EINVAL; ++ } ++ break; ++ case Opt_notag: ++ break; + default: + /* + * Logically we would return an error here to prevent +@@ -1397,6 +1427,9 @@ xfs_fs_fill_super( + if (error) + goto out_free_sb; + ++ if (mp->m_flags & XFS_MOUNT_TAGGED) ++ sb->s_flags |= MS_TAGGED; ++ + /* + * we must configure the block size in the superblock before we run the + * full mount process as the mount process can lookup and cache inodes. +diff -NurpP --minimal linux-3.0.9/fs/xfs/xfs_dinode.h linux-3.0.9-vs2.3.2.1/fs/xfs/xfs_dinode.h +--- linux-3.0.9/fs/xfs/xfs_dinode.h 2011-01-05 21:50:28.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/xfs/xfs_dinode.h 2011-06-10 22:11:24.000000000 +0200 +@@ -51,7 +51,9 @@ typedef struct xfs_dinode { + __be32 di_nlink; /* number of links to file */ + __be16 di_projid_lo; /* lower part of owner's project id */ + __be16 di_projid_hi; /* higher part owner's project id */ +- __u8 di_pad[6]; /* unused, zeroed space */ ++ __u8 di_pad[2]; /* unused, zeroed space */ ++ __be16 di_tag; /* context tagging */ ++ __be16 di_vflags; /* vserver specific flags */ + __be16 di_flushiter; /* incremented on flush */ + xfs_timestamp_t di_atime; /* time last accessed */ + xfs_timestamp_t di_mtime; /* time last modified */ +@@ -184,6 +186,8 @@ static inline void xfs_dinode_put_rdev(s + #define XFS_DIFLAG_EXTSZINHERIT_BIT 12 /* inherit inode extent size */ + #define XFS_DIFLAG_NODEFRAG_BIT 13 /* do not reorganize/defragment */ + #define XFS_DIFLAG_FILESTREAM_BIT 14 /* use filestream allocator */ ++#define XFS_DIFLAG_IXUNLINK_BIT 15 /* Immutable inver on unlink */ ++ + #define XFS_DIFLAG_REALTIME (1 << XFS_DIFLAG_REALTIME_BIT) + #define XFS_DIFLAG_PREALLOC (1 << XFS_DIFLAG_PREALLOC_BIT) + #define XFS_DIFLAG_NEWRTBM (1 << XFS_DIFLAG_NEWRTBM_BIT) +@@ -199,6 +203,7 @@ static inline void xfs_dinode_put_rdev(s + #define XFS_DIFLAG_EXTSZINHERIT (1 << XFS_DIFLAG_EXTSZINHERIT_BIT) + #define XFS_DIFLAG_NODEFRAG (1 << XFS_DIFLAG_NODEFRAG_BIT) + #define XFS_DIFLAG_FILESTREAM (1 << XFS_DIFLAG_FILESTREAM_BIT) ++#define XFS_DIFLAG_IXUNLINK (1 << XFS_DIFLAG_IXUNLINK_BIT) + + #ifdef CONFIG_XFS_RT + #define XFS_IS_REALTIME_INODE(ip) ((ip)->i_d.di_flags & XFS_DIFLAG_REALTIME) +@@ -211,6 +216,10 @@ static inline void xfs_dinode_put_rdev(s + XFS_DIFLAG_IMMUTABLE | XFS_DIFLAG_APPEND | XFS_DIFLAG_SYNC | \ + XFS_DIFLAG_NOATIME | XFS_DIFLAG_NODUMP | XFS_DIFLAG_RTINHERIT | \ + XFS_DIFLAG_PROJINHERIT | XFS_DIFLAG_NOSYMLINKS | XFS_DIFLAG_EXTSIZE | \ +- XFS_DIFLAG_EXTSZINHERIT | XFS_DIFLAG_NODEFRAG | XFS_DIFLAG_FILESTREAM) ++ XFS_DIFLAG_EXTSZINHERIT | XFS_DIFLAG_NODEFRAG | XFS_DIFLAG_FILESTREAM | \ ++ XFS_DIFLAG_IXUNLINK) ++ ++#define XFS_DIVFLAG_BARRIER 0x01 ++#define XFS_DIVFLAG_COW 0x02 + + #endif /* __XFS_DINODE_H__ */ +diff -NurpP --minimal linux-3.0.9/fs/xfs/xfs_fs.h linux-3.0.9-vs2.3.2.1/fs/xfs/xfs_fs.h +--- linux-3.0.9/fs/xfs/xfs_fs.h 2011-01-05 21:50:28.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/fs/xfs/xfs_fs.h 2011-06-10 22:11:24.000000000 +0200 +@@ -67,6 +67,9 @@ struct fsxattr { + #define XFS_XFLAG_EXTSZINHERIT 0x00001000 /* inherit inode extent size */ + #define XFS_XFLAG_NODEFRAG 0x00002000 /* do not defragment */ + #define XFS_XFLAG_FILESTREAM 0x00004000 /* use filestream allocator */ ++#define XFS_XFLAG_IXUNLINK 0x00008000 /* immutable invert on unlink */ ++#define XFS_XFLAG_BARRIER 0x10000000 /* chroot() barrier */ ++#define XFS_XFLAG_COW 0x20000000 /* copy on write mark */ + #define XFS_XFLAG_HASATTR 0x80000000 /* no DIFLAG for this */ + + /* +@@ -297,7 +300,8 @@ typedef struct xfs_bstat { + #define bs_projid bs_projid_lo /* (previously just bs_projid) */ + __u16 bs_forkoff; /* inode fork offset in bytes */ + __u16 bs_projid_hi; /* higher part of project id */ +- unsigned char bs_pad[10]; /* pad space, unused */ ++ unsigned char bs_pad[8]; /* pad space, unused */ ++ __u16 bs_tag; /* context tagging */ + __u32 bs_dmevmask; /* DMIG event mask */ + __u16 bs_dmstate; /* DMIG state info */ + __u16 bs_aextents; /* attribute number of extents */ +diff -NurpP --minimal linux-3.0.9/fs/xfs/xfs_ialloc.c linux-3.0.9-vs2.3.2.1/fs/xfs/xfs_ialloc.c +--- linux-3.0.9/fs/xfs/xfs_ialloc.c 2011-05-22 16:17:54.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/xfs/xfs_ialloc.c 2011-06-10 22:11:24.000000000 +0200 +@@ -37,7 +37,6 @@ + #include "xfs_error.h" + #include "xfs_bmap.h" + +- + /* + * Allocation group level functions. + */ +diff -NurpP --minimal linux-3.0.9/fs/xfs/xfs_inode.c linux-3.0.9-vs2.3.2.1/fs/xfs/xfs_inode.c +--- linux-3.0.9/fs/xfs/xfs_inode.c 2011-07-22 11:18:10.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/xfs/xfs_inode.c 2011-06-10 22:11:24.000000000 +0200 +@@ -243,6 +243,7 @@ xfs_inotobp( + return 0; + } + ++#include <linux/vs_tag.h> + + /* + * This routine is called to map an inode to the buffer containing +@@ -641,15 +642,25 @@ xfs_iformat_btree( + STATIC void + xfs_dinode_from_disk( + xfs_icdinode_t *to, +- xfs_dinode_t *from) ++ xfs_dinode_t *from, ++ int tagged) + { ++ uint32_t uid, gid, tag; ++ + to->di_magic = be16_to_cpu(from->di_magic); + to->di_mode = be16_to_cpu(from->di_mode); + to->di_version = from ->di_version; + to->di_format = from->di_format; + to->di_onlink = be16_to_cpu(from->di_onlink); +- to->di_uid = be32_to_cpu(from->di_uid); +- to->di_gid = be32_to_cpu(from->di_gid); ++ ++ uid = be32_to_cpu(from->di_uid); ++ gid = be32_to_cpu(from->di_gid); ++ tag = be16_to_cpu(from->di_tag); ++ ++ to->di_uid = INOTAG_UID(tagged, uid, gid); ++ to->di_gid = INOTAG_GID(tagged, uid, gid); ++ to->di_tag = INOTAG_TAG(tagged, uid, gid, tag); ++ + to->di_nlink = be32_to_cpu(from->di_nlink); + to->di_projid_lo = be16_to_cpu(from->di_projid_lo); + to->di_projid_hi = be16_to_cpu(from->di_projid_hi); +@@ -671,21 +682,26 @@ xfs_dinode_from_disk( + to->di_dmevmask = be32_to_cpu(from->di_dmevmask); + to->di_dmstate = be16_to_cpu(from->di_dmstate); + to->di_flags = be16_to_cpu(from->di_flags); ++ to->di_vflags = be16_to_cpu(from->di_vflags); + to->di_gen = be32_to_cpu(from->di_gen); + } + + void + xfs_dinode_to_disk( + xfs_dinode_t *to, +- xfs_icdinode_t *from) ++ xfs_icdinode_t *from, ++ int tagged) + { + to->di_magic = cpu_to_be16(from->di_magic); + to->di_mode = cpu_to_be16(from->di_mode); + to->di_version = from ->di_version; + to->di_format = from->di_format; + to->di_onlink = cpu_to_be16(from->di_onlink); +- to->di_uid = cpu_to_be32(from->di_uid); +- to->di_gid = cpu_to_be32(from->di_gid); ++ ++ to->di_uid = cpu_to_be32(TAGINO_UID(tagged, from->di_uid, from->di_tag)); ++ to->di_gid = cpu_to_be32(TAGINO_GID(tagged, from->di_gid, from->di_tag)); ++ to->di_tag = cpu_to_be16(TAGINO_TAG(tagged, from->di_tag)); ++ + to->di_nlink = cpu_to_be32(from->di_nlink); + to->di_projid_lo = cpu_to_be16(from->di_projid_lo); + to->di_projid_hi = cpu_to_be16(from->di_projid_hi); +@@ -707,12 +723,14 @@ xfs_dinode_to_disk( + to->di_dmevmask = cpu_to_be32(from->di_dmevmask); + to->di_dmstate = cpu_to_be16(from->di_dmstate); + to->di_flags = cpu_to_be16(from->di_flags); ++ to->di_vflags = cpu_to_be16(from->di_vflags); + to->di_gen = cpu_to_be32(from->di_gen); + } + + STATIC uint + _xfs_dic2xflags( +- __uint16_t di_flags) ++ __uint16_t di_flags, ++ __uint16_t di_vflags) + { + uint flags = 0; + +@@ -723,6 +741,8 @@ _xfs_dic2xflags( + flags |= XFS_XFLAG_PREALLOC; + if (di_flags & XFS_DIFLAG_IMMUTABLE) + flags |= XFS_XFLAG_IMMUTABLE; ++ if (di_flags & XFS_DIFLAG_IXUNLINK) ++ flags |= XFS_XFLAG_IXUNLINK; + if (di_flags & XFS_DIFLAG_APPEND) + flags |= XFS_XFLAG_APPEND; + if (di_flags & XFS_DIFLAG_SYNC) +@@ -747,6 +767,10 @@ _xfs_dic2xflags( + flags |= XFS_XFLAG_FILESTREAM; + } + ++ if (di_vflags & XFS_DIVFLAG_BARRIER) ++ flags |= FS_BARRIER_FL; ++ if (di_vflags & XFS_DIVFLAG_COW) ++ flags |= FS_COW_FL; + return flags; + } + +@@ -756,7 +780,7 @@ xfs_ip2xflags( + { + xfs_icdinode_t *dic = &ip->i_d; + +- return _xfs_dic2xflags(dic->di_flags) | ++ return _xfs_dic2xflags(dic->di_flags, dic->di_vflags) | + (XFS_IFORK_Q(ip) ? XFS_XFLAG_HASATTR : 0); + } + +@@ -764,7 +788,8 @@ uint + xfs_dic2xflags( + xfs_dinode_t *dip) + { +- return _xfs_dic2xflags(be16_to_cpu(dip->di_flags)) | ++ return _xfs_dic2xflags(be16_to_cpu(dip->di_flags), ++ be16_to_cpu(dip->di_vflags)) | + (XFS_DFORK_Q(dip) ? XFS_XFLAG_HASATTR : 0); + } + +@@ -797,7 +822,6 @@ xfs_iread( + if (error) + return error; + dip = (xfs_dinode_t *)xfs_buf_offset(bp, ip->i_imap.im_boffset); +- + /* + * If we got something that isn't an inode it means someone + * (nfs or dmi) has a stale handle. +@@ -820,7 +844,8 @@ xfs_iread( + * Otherwise, just get the truly permanent information. + */ + if (dip->di_mode) { +- xfs_dinode_from_disk(&ip->i_d, dip); ++ xfs_dinode_from_disk(&ip->i_d, dip, ++ mp->m_flags & XFS_MOUNT_TAGGED); + error = xfs_iformat(ip, dip); + if (error) { + #ifdef DEBUG +@@ -1015,6 +1040,7 @@ xfs_ialloc( + ASSERT(ip->i_d.di_nlink == nlink); + ip->i_d.di_uid = current_fsuid(); + ip->i_d.di_gid = current_fsgid(); ++ ip->i_d.di_tag = current_fstag(&ip->i_vnode); + xfs_set_projid(ip, prid); + memset(&(ip->i_d.di_pad[0]), 0, sizeof(ip->i_d.di_pad)); + +@@ -1075,6 +1101,7 @@ xfs_ialloc( + ip->i_d.di_dmevmask = 0; + ip->i_d.di_dmstate = 0; + ip->i_d.di_flags = 0; ++ ip->i_d.di_vflags = 0; + flags = XFS_ILOG_CORE; + switch (mode & S_IFMT) { + case S_IFIFO: +@@ -2108,6 +2135,7 @@ xfs_ifree( + } + ip->i_d.di_mode = 0; /* mark incore inode as free */ + ip->i_d.di_flags = 0; ++ ip->i_d.di_vflags = 0; + ip->i_d.di_dmevmask = 0; + ip->i_d.di_forkoff = 0; /* mark the attr fork not in use */ + ip->i_df.if_ext_max = +@@ -2987,7 +3015,8 @@ xfs_iflush_int( + * because if the inode is dirty at all the core must + * be. + */ +- xfs_dinode_to_disk(dip, &ip->i_d); ++ xfs_dinode_to_disk(dip, &ip->i_d, ++ mp->m_flags & XFS_MOUNT_TAGGED); + + /* Wrap, we never let the log put out DI_MAX_FLUSH */ + if (ip->i_d.di_flushiter == DI_MAX_FLUSH) +diff -NurpP --minimal linux-3.0.9/fs/xfs/xfs_inode.h linux-3.0.9-vs2.3.2.1/fs/xfs/xfs_inode.h +--- linux-3.0.9/fs/xfs/xfs_inode.h 2011-07-22 11:18:10.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/xfs/xfs_inode.h 2011-07-01 11:35:35.000000000 +0200 +@@ -135,7 +135,9 @@ typedef struct xfs_icdinode { + __uint32_t di_nlink; /* number of links to file */ + __uint16_t di_projid_lo; /* lower part of owner's project id */ + __uint16_t di_projid_hi; /* higher part of owner's project id */ +- __uint8_t di_pad[6]; /* unused, zeroed space */ ++ __uint8_t di_pad[2]; /* unused, zeroed space */ ++ __uint16_t di_tag; /* context tagging */ ++ __uint16_t di_vflags; /* vserver specific flags */ + __uint16_t di_flushiter; /* incremented on flush */ + xfs_ictimestamp_t di_atime; /* time last accessed */ + xfs_ictimestamp_t di_mtime; /* time last modified */ +@@ -546,7 +548,7 @@ int xfs_itobp(struct xfs_mount *, struc + int xfs_iread(struct xfs_mount *, struct xfs_trans *, + struct xfs_inode *, uint); + void xfs_dinode_to_disk(struct xfs_dinode *, +- struct xfs_icdinode *); ++ struct xfs_icdinode *, int); + void xfs_idestroy_fork(struct xfs_inode *, int); + void xfs_idata_realloc(struct xfs_inode *, int, int); + void xfs_iroot_realloc(struct xfs_inode *, int, int); +diff -NurpP --minimal linux-3.0.9/fs/xfs/xfs_itable.c linux-3.0.9-vs2.3.2.1/fs/xfs/xfs_itable.c +--- linux-3.0.9/fs/xfs/xfs_itable.c 2011-05-22 16:17:54.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/xfs/xfs_itable.c 2011-06-10 22:11:24.000000000 +0200 +@@ -98,6 +98,7 @@ xfs_bulkstat_one_int( + buf->bs_mode = dic->di_mode; + buf->bs_uid = dic->di_uid; + buf->bs_gid = dic->di_gid; ++ buf->bs_tag = dic->di_tag; + buf->bs_size = dic->di_size; + + /* +diff -NurpP --minimal linux-3.0.9/fs/xfs/xfs_log_recover.c linux-3.0.9-vs2.3.2.1/fs/xfs/xfs_log_recover.c +--- linux-3.0.9/fs/xfs/xfs_log_recover.c 2011-07-22 11:18:10.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/xfs/xfs_log_recover.c 2011-06-10 22:11:24.000000000 +0200 +@@ -2343,7 +2343,8 @@ xlog_recover_inode_pass2( + } + + /* The core is in in-core format */ +- xfs_dinode_to_disk(dip, item->ri_buf[1].i_addr); ++ xfs_dinode_to_disk(dip, item->ri_buf[1].i_addr, ++ mp->m_flags & XFS_MOUNT_TAGGED); + + /* the rest is in on-disk format */ + if (item->ri_buf[1].i_len > sizeof(struct xfs_icdinode)) { +diff -NurpP --minimal linux-3.0.9/fs/xfs/xfs_mount.h linux-3.0.9-vs2.3.2.1/fs/xfs/xfs_mount.h +--- linux-3.0.9/fs/xfs/xfs_mount.h 2011-07-22 11:18:10.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/xfs/xfs_mount.h 2011-06-10 22:11:24.000000000 +0200 +@@ -249,6 +249,7 @@ typedef struct xfs_mount { + allocator */ + #define XFS_MOUNT_NOATTR2 (1ULL << 25) /* disable use of attr2 format */ + ++#define XFS_MOUNT_TAGGED (1ULL << 31) /* context tagging */ + + /* + * Default minimum read and write sizes. +diff -NurpP --minimal linux-3.0.9/fs/xfs/xfs_vnodeops.c linux-3.0.9-vs2.3.2.1/fs/xfs/xfs_vnodeops.c +--- linux-3.0.9/fs/xfs/xfs_vnodeops.c 2011-07-22 11:18:10.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/xfs/xfs_vnodeops.c 2011-07-01 11:35:35.000000000 +0200 +@@ -50,6 +50,78 @@ + #include "xfs_vnodeops.h" + #include "xfs_trace.h" + ++ ++STATIC void ++xfs_get_inode_flags( ++ xfs_inode_t *ip) ++{ ++ struct inode *inode = VFS_I(ip); ++ unsigned int flags = inode->i_flags; ++ unsigned int vflags = inode->i_vflags; ++ ++ if (flags & S_IMMUTABLE) ++ ip->i_d.di_flags |= XFS_DIFLAG_IMMUTABLE; ++ else ++ ip->i_d.di_flags &= ~XFS_DIFLAG_IMMUTABLE; ++ if (flags & S_IXUNLINK) ++ ip->i_d.di_flags |= XFS_DIFLAG_IXUNLINK; ++ else ++ ip->i_d.di_flags &= ~XFS_DIFLAG_IXUNLINK; ++ ++ if (vflags & V_BARRIER) ++ ip->i_d.di_vflags |= XFS_DIVFLAG_BARRIER; ++ else ++ ip->i_d.di_vflags &= ~XFS_DIVFLAG_BARRIER; ++ if (vflags & V_COW) ++ ip->i_d.di_vflags |= XFS_DIVFLAG_COW; ++ else ++ ip->i_d.di_vflags &= ~XFS_DIVFLAG_COW; ++} ++ ++int ++xfs_sync_flags( ++ struct inode *inode, ++ int flags, ++ int vflags) ++{ ++ struct xfs_inode *ip = XFS_I(inode); ++ struct xfs_mount *mp = ip->i_mount; ++ struct xfs_trans *tp; ++ unsigned int lock_flags = 0; ++ int code; ++ ++ tp = xfs_trans_alloc(mp, XFS_TRANS_SETATTR_NOT_SIZE); ++ code = xfs_trans_reserve(tp, 0, XFS_ICHANGE_LOG_RES(mp), 0, 0, 0); ++ if (code) ++ goto error_out; ++ ++ xfs_ilock(ip, XFS_ILOCK_EXCL); ++ ++ xfs_trans_ijoin(tp, ip); ++ ++ inode->i_flags = flags; ++ inode->i_vflags = vflags; ++ xfs_get_inode_flags(ip); ++ ++ xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); ++ xfs_trans_ichgtime(tp, ip, XFS_ICHGTIME_CHG); ++ ++ XFS_STATS_INC(xs_ig_attrchg); ++ ++ if (mp->m_flags & XFS_MOUNT_WSYNC) ++ xfs_trans_set_sync(tp); ++ code = xfs_trans_commit(tp, 0); ++ xfs_iunlock(ip, XFS_ILOCK_EXCL); ++ return code; ++ ++error_out: ++ xfs_trans_cancel(tp, 0); ++ if (lock_flags) ++ xfs_iunlock(ip, XFS_ILOCK_EXCL); ++ return code; ++} ++ ++ + int + xfs_setattr( + struct xfs_inode *ip, +@@ -65,6 +137,7 @@ xfs_setattr( + uint commit_flags=0; + uid_t uid=0, iuid=0; + gid_t gid=0, igid=0; ++ tag_t tag=0, itag=0; + struct xfs_dquot *udqp, *gdqp, *olddquot1, *olddquot2; + int need_iolock = 1; + +@@ -147,7 +220,7 @@ xfs_setattr( + /* + * Change file ownership. Must be the owner or privileged. + */ +- if (mask & (ATTR_UID|ATTR_GID)) { ++ if (mask & (ATTR_UID|ATTR_GID|ATTR_TAG)) { + /* + * These IDs could have changed since we last looked at them. + * But, we're assured that if the ownership did change +@@ -156,8 +229,10 @@ xfs_setattr( + */ + iuid = ip->i_d.di_uid; + igid = ip->i_d.di_gid; ++ itag = ip->i_d.di_tag; + gid = (mask & ATTR_GID) ? iattr->ia_gid : igid; + uid = (mask & ATTR_UID) ? iattr->ia_uid : iuid; ++ tag = (mask & ATTR_TAG) ? iattr->ia_tag : itag; + + /* + * Do a quota reservation only if uid/gid is actually +@@ -165,7 +240,8 @@ xfs_setattr( + */ + if (XFS_IS_QUOTA_RUNNING(mp) && + ((XFS_IS_UQUOTA_ON(mp) && iuid != uid) || +- (XFS_IS_GQUOTA_ON(mp) && igid != gid))) { ++ (XFS_IS_GQUOTA_ON(mp) && igid != gid) || ++ (XFS_IS_GQUOTA_ON(mp) && itag != tag))) { + ASSERT(tp); + code = xfs_qm_vop_chown_reserve(tp, ip, udqp, gdqp, + capable(CAP_FOWNER) ? +@@ -329,7 +405,7 @@ xfs_setattr( + /* + * Change file ownership. Must be the owner or privileged. + */ +- if (mask & (ATTR_UID|ATTR_GID)) { ++ if (mask & (ATTR_UID|ATTR_GID|ATTR_TAG)) { + /* + * CAP_FSETID overrides the following restrictions: + * +@@ -345,6 +421,10 @@ xfs_setattr( + * Change the ownerships and register quota modifications + * in the transaction. + */ ++ if (itag != tag) { ++ ip->i_d.di_tag = tag; ++ inode->i_tag = tag; ++ } + if (iuid != uid) { + if (XFS_IS_QUOTA_RUNNING(mp) && XFS_IS_UQUOTA_ON(mp)) { + ASSERT(mask & ATTR_UID); +diff -NurpP --minimal linux-3.0.9/fs/xfs/xfs_vnodeops.h linux-3.0.9-vs2.3.2.1/fs/xfs/xfs_vnodeops.h +--- linux-3.0.9/fs/xfs/xfs_vnodeops.h 2011-05-22 16:17:54.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/fs/xfs/xfs_vnodeops.h 2011-06-10 22:11:24.000000000 +0200 +@@ -13,6 +13,7 @@ struct xfs_inode; + struct xfs_iomap; + + ++int xfs_sync_xflags(struct xfs_inode *ip); + int xfs_setattr(struct xfs_inode *ip, struct iattr *vap, int flags); + #define XFS_ATTR_DMI 0x01 /* invocation from a DMI function */ + #define XFS_ATTR_NONBLOCK 0x02 /* return EAGAIN if operation would block */ +diff -NurpP --minimal linux-3.0.9/include/asm-generic/tlb.h linux-3.0.9-vs2.3.2.1/include/asm-generic/tlb.h +--- linux-3.0.9/include/asm-generic/tlb.h 2011-07-22 11:18:10.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/asm-generic/tlb.h 2011-06-10 22:11:24.000000000 +0200 +@@ -16,6 +16,7 @@ + #define _ASM_GENERIC__TLB_H + + #include <linux/swap.h> ++#include <linux/vs_memory.h> + #include <asm/pgalloc.h> + #include <asm/tlbflush.h> + +diff -NurpP --minimal linux-3.0.9/include/linux/Kbuild linux-3.0.9-vs2.3.2.1/include/linux/Kbuild +--- linux-3.0.9/include/linux/Kbuild 2011-07-22 11:18:10.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/linux/Kbuild 2011-06-10 22:11:24.000000000 +0200 +@@ -17,6 +17,7 @@ header-y += netfilter_bridge/ + header-y += netfilter_ipv4/ + header-y += netfilter_ipv6/ + header-y += usb/ ++header-y += vserver/ + header-y += wimax/ + + objhdr-y += version.h +diff -NurpP --minimal linux-3.0.9/include/linux/capability.h linux-3.0.9-vs2.3.2.1/include/linux/capability.h +--- linux-3.0.9/include/linux/capability.h 2011-07-22 11:18:10.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/linux/capability.h 2011-06-10 22:11:24.000000000 +0200 +@@ -279,6 +279,7 @@ struct cpu_vfs_cap_data { + arbitrary SCSI commands */ + /* Allow setting encryption key on loopback filesystem */ + /* Allow setting zone reclaim policy */ ++/* Allow the selection of a security context */ + + #define CAP_SYS_ADMIN 21 + +@@ -362,7 +363,12 @@ struct cpu_vfs_cap_data { + + #define CAP_LAST_CAP CAP_WAKE_ALARM + +-#define cap_valid(x) ((x) >= 0 && (x) <= CAP_LAST_CAP) ++/* Allow context manipulations */ ++/* Allow changing context info on files */ ++ ++#define CAP_CONTEXT 63 ++ ++#define cap_valid(x) ((x) >= 0 && ((x) <= CAP_LAST_CAP || (x) == CAP_CONTEXT)) + + /* + * Bit location of each capability (used by user-space library and kernel) +diff -NurpP --minimal linux-3.0.9/include/linux/cred.h linux-3.0.9-vs2.3.2.1/include/linux/cred.h +--- linux-3.0.9/include/linux/cred.h 2011-07-22 11:18:10.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/linux/cred.h 2011-06-10 22:11:24.000000000 +0200 +@@ -156,6 +156,7 @@ extern void exit_creds(struct task_struc + extern int copy_creds(struct task_struct *, unsigned long); + extern const struct cred *get_task_cred(struct task_struct *); + extern struct cred *cred_alloc_blank(void); ++extern struct cred *__prepare_creds(const struct cred *); + extern struct cred *prepare_creds(void); + extern struct cred *prepare_exec_creds(void); + extern int commit_creds(struct cred *); +@@ -209,6 +210,31 @@ static inline void validate_process_cred + } + #endif + ++static inline void set_cred_subscribers(struct cred *cred, int n) ++{ ++#ifdef CONFIG_DEBUG_CREDENTIALS ++ atomic_set(&cred->subscribers, n); ++#endif ++} ++ ++static inline int read_cred_subscribers(const struct cred *cred) ++{ ++#ifdef CONFIG_DEBUG_CREDENTIALS ++ return atomic_read(&cred->subscribers); ++#else ++ return 0; ++#endif ++} ++ ++static inline void alter_cred_subscribers(const struct cred *_cred, int n) ++{ ++#ifdef CONFIG_DEBUG_CREDENTIALS ++ struct cred *cred = (struct cred *) _cred; ++ ++ atomic_add(n, &cred->subscribers); ++#endif ++} ++ + /** + * get_new_cred - Get a reference on a new set of credentials + * @cred: The new credentials to reference +diff -NurpP --minimal linux-3.0.9/include/linux/devpts_fs.h linux-3.0.9-vs2.3.2.1/include/linux/devpts_fs.h +--- linux-3.0.9/include/linux/devpts_fs.h 2008-12-25 00:26:37.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/devpts_fs.h 2011-06-10 22:11:24.000000000 +0200 +@@ -45,5 +45,4 @@ static inline void devpts_pty_kill(struc + + #endif + +- + #endif /* _LINUX_DEVPTS_FS_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/ext2_fs.h linux-3.0.9-vs2.3.2.1/include/linux/ext2_fs.h +--- linux-3.0.9/include/linux/ext2_fs.h 2011-11-15 16:40:47.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/ext2_fs.h 2011-11-15 17:37:07.000000000 +0100 +@@ -189,8 +189,12 @@ struct ext2_group_desc + #define EXT2_NOTAIL_FL FS_NOTAIL_FL /* file tail should not be merged */ + #define EXT2_DIRSYNC_FL FS_DIRSYNC_FL /* dirsync behaviour (directories only) */ + #define EXT2_TOPDIR_FL FS_TOPDIR_FL /* Top of directory hierarchies*/ ++#define EXT2_IXUNLINK_FL FS_IXUNLINK_FL /* Immutable invert on unlink */ + #define EXT2_RESERVED_FL FS_RESERVED_FL /* reserved for ext2 lib */ + ++#define EXT2_BARRIER_FL FS_BARRIER_FL /* Barrier for chroot() */ ++#define EXT2_COW_FL FS_COW_FL /* Copy on Write marker */ ++ + #define EXT2_FL_USER_VISIBLE FS_FL_USER_VISIBLE /* User visible flags */ + #define EXT2_FL_USER_MODIFIABLE FS_FL_USER_MODIFIABLE /* User modifiable flags */ + +@@ -274,7 +278,8 @@ struct ext2_inode { + __u16 i_pad1; + __le16 l_i_uid_high; /* these 2 fields */ + __le16 l_i_gid_high; /* were reserved2[0] */ +- __u32 l_i_reserved2; ++ __le16 l_i_tag; /* Context Tag */ ++ __u16 l_i_reserved2; + } linux2; + struct { + __u8 h_i_frag; /* Fragment number */ +@@ -303,6 +308,7 @@ struct ext2_inode { + #define i_gid_low i_gid + #define i_uid_high osd2.linux2.l_i_uid_high + #define i_gid_high osd2.linux2.l_i_gid_high ++#define i_raw_tag osd2.linux2.l_i_tag + #define i_reserved2 osd2.linux2.l_i_reserved2 + #endif + +@@ -347,6 +353,7 @@ struct ext2_inode { + #define EXT2_MOUNT_USRQUOTA 0x020000 /* user quota */ + #define EXT2_MOUNT_GRPQUOTA 0x040000 /* group quota */ + #define EXT2_MOUNT_RESERVATION 0x080000 /* Preallocation */ ++#define EXT2_MOUNT_TAGGED (1<<24) /* Enable Context Tags */ + + + #define clear_opt(o, opt) o &= ~EXT2_MOUNT_##opt +diff -NurpP --minimal linux-3.0.9/include/linux/ext3_fs.h linux-3.0.9-vs2.3.2.1/include/linux/ext3_fs.h +--- linux-3.0.9/include/linux/ext3_fs.h 2011-11-15 16:40:47.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/ext3_fs.h 2011-11-15 17:37:07.000000000 +0100 +@@ -173,10 +173,14 @@ struct ext3_group_desc + #define EXT3_NOTAIL_FL 0x00008000 /* file tail should not be merged */ + #define EXT3_DIRSYNC_FL 0x00010000 /* dirsync behaviour (directories only) */ + #define EXT3_TOPDIR_FL 0x00020000 /* Top of directory hierarchies*/ ++#define EXT3_IXUNLINK_FL 0x08000000 /* Immutable invert on unlink */ + #define EXT3_RESERVED_FL 0x80000000 /* reserved for ext3 lib */ + +-#define EXT3_FL_USER_VISIBLE 0x0003DFFF /* User visible flags */ +-#define EXT3_FL_USER_MODIFIABLE 0x000380FF /* User modifiable flags */ ++#define EXT3_BARRIER_FL 0x04000000 /* Barrier for chroot() */ ++#define EXT3_COW_FL 0x20000000 /* Copy on Write marker */ ++ ++#define EXT3_FL_USER_VISIBLE 0x0103DFFF /* User visible flags */ ++#define EXT3_FL_USER_MODIFIABLE 0x010380FF /* User modifiable flags */ + + /* Flags that should be inherited by new inodes from their parent. */ + #define EXT3_FL_INHERITED (EXT3_SECRM_FL | EXT3_UNRM_FL | EXT3_COMPR_FL |\ +@@ -312,7 +316,8 @@ struct ext3_inode { + __u16 i_pad1; + __le16 l_i_uid_high; /* these 2 fields */ + __le16 l_i_gid_high; /* were reserved2[0] */ +- __u32 l_i_reserved2; ++ __le16 l_i_tag; /* Context Tag */ ++ __u16 l_i_reserved2; + } linux2; + struct { + __u8 h_i_frag; /* Fragment number */ +@@ -343,6 +348,7 @@ struct ext3_inode { + #define i_gid_low i_gid + #define i_uid_high osd2.linux2.l_i_uid_high + #define i_gid_high osd2.linux2.l_i_gid_high ++#define i_raw_tag osd2.linux2.l_i_tag + #define i_reserved2 osd2.linux2.l_i_reserved2 + + #elif defined(__GNU__) +@@ -405,6 +411,7 @@ struct ext3_inode { + #define EXT3_MOUNT_GRPQUOTA 0x200000 /* "old" group quota */ + #define EXT3_MOUNT_DATA_ERR_ABORT 0x400000 /* Abort on file data write + * error in ordered mode */ ++#define EXT3_MOUNT_TAGGED (1<<24) /* Enable Context Tags */ + + /* Compatibility, for having both ext2_fs.h and ext3_fs.h included at once */ + #ifndef _LINUX_EXT2_FS_H +@@ -919,6 +926,7 @@ extern void ext3_get_inode_flags(struct + extern void ext3_set_aops(struct inode *inode); + extern int ext3_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, + u64 start, u64 len); ++extern int ext3_sync_flags(struct inode *, int, int); + + /* ioctl.c */ + extern long ext3_ioctl(struct file *, unsigned int, unsigned long); +diff -NurpP --minimal linux-3.0.9/include/linux/fs.h linux-3.0.9-vs2.3.2.1/include/linux/fs.h +--- linux-3.0.9/include/linux/fs.h 2011-07-22 11:18:10.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/linux/fs.h 2011-07-01 11:35:35.000000000 +0200 +@@ -208,6 +208,9 @@ struct inodes_stat_t { + #define MS_KERNMOUNT (1<<22) /* this is a kern_mount call */ + #define MS_I_VERSION (1<<23) /* Update inode I_version field */ + #define MS_STRICTATIME (1<<24) /* Always perform atime updates */ ++#define MS_TAGGED (1<<25) /* use generic inode tagging */ ++#define MS_TAGID (1<<26) /* use specific tag for this mount */ ++#define MS_NOTAGCHECK (1<<27) /* don't check tags */ + #define MS_NOSEC (1<<28) + #define MS_BORN (1<<29) + #define MS_ACTIVE (1<<30) +@@ -239,6 +242,14 @@ struct inodes_stat_t { + #define S_IMA 1024 /* Inode has an associated IMA struct */ + #define S_AUTOMOUNT 2048 /* Automount/referral quasi-directory */ + #define S_NOSEC 4096 /* no suid or xattr security attributes */ ++#define S_IXUNLINK 8192 /* Immutable Invert on unlink */ ++ ++/* Linux-VServer related Inode flags */ ++ ++#define V_VALID 1 ++#define V_XATTR 2 ++#define V_BARRIER 4 /* Barrier for chroot() */ ++#define V_COW 8 /* Copy on Write */ + + /* + * Note that nosuid etc flags are inode-specific: setting some file-system +@@ -261,12 +272,15 @@ struct inodes_stat_t { + #define IS_DIRSYNC(inode) (__IS_FLG(inode, MS_SYNCHRONOUS|MS_DIRSYNC) || \ + ((inode)->i_flags & (S_SYNC|S_DIRSYNC))) + #define IS_MANDLOCK(inode) __IS_FLG(inode, MS_MANDLOCK) +-#define IS_NOATIME(inode) __IS_FLG(inode, MS_RDONLY|MS_NOATIME) +-#define IS_I_VERSION(inode) __IS_FLG(inode, MS_I_VERSION) ++#define IS_NOATIME(inode) __IS_FLG(inode, MS_RDONLY|MS_NOATIME) ++#define IS_I_VERSION(inode) __IS_FLG(inode, MS_I_VERSION) ++#define IS_TAGGED(inode) __IS_FLG(inode, MS_TAGGED) + + #define IS_NOQUOTA(inode) ((inode)->i_flags & S_NOQUOTA) + #define IS_APPEND(inode) ((inode)->i_flags & S_APPEND) + #define IS_IMMUTABLE(inode) ((inode)->i_flags & S_IMMUTABLE) ++#define IS_IXUNLINK(inode) ((inode)->i_flags & S_IXUNLINK) ++#define IS_IXORUNLINK(inode) ((IS_IXUNLINK(inode) ? S_IMMUTABLE : 0) ^ IS_IMMUTABLE(inode)) + #define IS_POSIXACL(inode) __IS_FLG(inode, MS_POSIXACL) + + #define IS_DEADDIR(inode) ((inode)->i_flags & S_DEAD) +@@ -277,6 +291,16 @@ struct inodes_stat_t { + #define IS_AUTOMOUNT(inode) ((inode)->i_flags & S_AUTOMOUNT) + #define IS_NOSEC(inode) ((inode)->i_flags & S_NOSEC) + ++#define IS_BARRIER(inode) (S_ISDIR((inode)->i_mode) && ((inode)->i_vflags & V_BARRIER)) ++ ++#ifdef CONFIG_VSERVER_COWBL ++# define IS_COW(inode) (IS_IXUNLINK(inode) && IS_IMMUTABLE(inode)) ++# define IS_COW_LINK(inode) (S_ISREG((inode)->i_mode) && ((inode)->i_nlink > 1)) ++#else ++# define IS_COW(inode) (0) ++# define IS_COW_LINK(inode) (0) ++#endif ++ + /* the read-only stuff doesn't really belong here, but any other place is + probably as bad and I don't want to create yet another include file. */ + +@@ -362,11 +386,14 @@ struct inodes_stat_t { + #define FS_EXTENT_FL 0x00080000 /* Extents */ + #define FS_DIRECTIO_FL 0x00100000 /* Use direct i/o */ + #define FS_NOCOW_FL 0x00800000 /* Do not cow file */ ++#define FS_IXUNLINK_FL 0x08000000 /* Immutable invert on unlink */ + #define FS_RESERVED_FL 0x80000000 /* reserved for ext2 lib */ + +-#define FS_FL_USER_VISIBLE 0x0003DFFF /* User visible flags */ +-#define FS_FL_USER_MODIFIABLE 0x000380FF /* User modifiable flags */ +- ++#define FS_BARRIER_FL 0x04000000 /* Barrier for chroot() */ ++#define FS_COW_FL 0x20000000 /* Copy on Write marker */ ++ ++#define FS_FL_USER_VISIBLE 0x0103DFFF /* User visible flags */ ++#define FS_FL_USER_MODIFIABLE 0x010380FF /* User modifiable flags */ + + #define SYNC_FILE_RANGE_WAIT_BEFORE 1 + #define SYNC_FILE_RANGE_WRITE 2 +@@ -447,6 +474,7 @@ typedef void (dio_iodone_t)(struct kiocb + #define ATTR_KILL_PRIV (1 << 14) + #define ATTR_OPEN (1 << 15) /* Truncating from open(O_TRUNC) */ + #define ATTR_TIMES_SET (1 << 16) ++#define ATTR_TAG (1 << 17) + + /* + * This is the Inode Attributes structure, used for notify_change(). It +@@ -462,6 +490,7 @@ struct iattr { + umode_t ia_mode; + uid_t ia_uid; + gid_t ia_gid; ++ tag_t ia_tag; + loff_t ia_size; + struct timespec ia_atime; + struct timespec ia_mtime; +@@ -475,6 +504,9 @@ struct iattr { + struct file *ia_file; + }; + ++#define ATTR_FLAG_BARRIER 512 /* Barrier for chroot() */ ++#define ATTR_FLAG_IXUNLINK 1024 /* Immutable invert on unlink */ ++ + /* + * Includes for diskquotas. + */ +@@ -740,11 +772,13 @@ struct inode { + umode_t i_mode; + uid_t i_uid; + gid_t i_gid; ++ tag_t i_tag; + const struct inode_operations *i_op; + struct super_block *i_sb; + + spinlock_t i_lock; /* i_blocks, i_bytes, maybe i_size */ +- unsigned int i_flags; ++ unsigned short i_flags; ++ unsigned short i_vflags; + unsigned long i_state; + #ifdef CONFIG_SECURITY + void *i_security; +@@ -766,6 +800,7 @@ struct inode { + atomic_t i_count; + unsigned int i_nlink; + dev_t i_rdev; ++ dev_t i_mdev; + unsigned int i_blkbits; + u64 i_version; + loff_t i_size; +@@ -890,12 +925,12 @@ static inline void i_size_write(struct i + + static inline unsigned iminor(const struct inode *inode) + { +- return MINOR(inode->i_rdev); ++ return MINOR(inode->i_mdev); + } + + static inline unsigned imajor(const struct inode *inode) + { +- return MAJOR(inode->i_rdev); ++ return MAJOR(inode->i_mdev); + } + + extern struct block_device *I_BDEV(struct inode *inode); +@@ -957,6 +992,7 @@ struct file { + loff_t f_pos; + struct fown_struct f_owner; + const struct cred *f_cred; ++ xid_t f_xid; + struct file_ra_state f_ra; + + u64 f_version; +@@ -1101,6 +1137,7 @@ struct file_lock { + struct file *fl_file; + loff_t fl_start; + loff_t fl_end; ++ xid_t fl_xid; + + struct fasync_struct * fl_fasync; /* for lease break notifications */ + unsigned long fl_break_time; /* for nonblocking lease breaks */ +@@ -1600,6 +1637,7 @@ struct inode_operations { + ssize_t (*getxattr) (struct dentry *, const char *, void *, size_t); + ssize_t (*listxattr) (struct dentry *, char *, size_t); + int (*removexattr) (struct dentry *, const char *); ++ int (*sync_flags) (struct inode *, int, int); + void (*truncate_range)(struct inode *, loff_t, loff_t); + int (*fiemap)(struct inode *, struct fiemap_extent_info *, u64 start, + u64 len); +@@ -1618,6 +1656,7 @@ extern ssize_t vfs_readv(struct file *, + unsigned long, loff_t *); + extern ssize_t vfs_writev(struct file *, const struct iovec __user *, + unsigned long, loff_t *); ++ssize_t vfs_sendfile(struct file *, struct file *, loff_t *, size_t, loff_t); + + struct super_operations { + struct inode *(*alloc_inode)(struct super_block *sb); +@@ -2437,6 +2476,7 @@ extern int dcache_dir_open(struct inode + extern int dcache_dir_close(struct inode *, struct file *); + extern loff_t dcache_dir_lseek(struct file *, loff_t, int); + extern int dcache_readdir(struct file *, void *, filldir_t); ++extern int dcache_readdir_filter(struct file *, void *, filldir_t, int (*)(struct dentry *)); + extern int simple_setattr(struct dentry *, struct iattr *); + extern int simple_getattr(struct vfsmount *, struct dentry *, struct kstat *); + extern int simple_statfs(struct dentry *, struct kstatfs *); +diff -NurpP --minimal linux-3.0.9/include/linux/gfs2_ondisk.h linux-3.0.9-vs2.3.2.1/include/linux/gfs2_ondisk.h +--- linux-3.0.9/include/linux/gfs2_ondisk.h 2010-07-07 18:31:55.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/linux/gfs2_ondisk.h 2011-06-10 22:11:24.000000000 +0200 +@@ -211,6 +211,9 @@ enum { + gfs2fl_NoAtime = 7, + gfs2fl_Sync = 8, + gfs2fl_System = 9, ++ gfs2fl_IXUnlink = 16, ++ gfs2fl_Barrier = 17, ++ gfs2fl_Cow = 18, + gfs2fl_TruncInProg = 29, + gfs2fl_InheritDirectio = 30, + gfs2fl_InheritJdata = 31, +@@ -227,6 +230,9 @@ enum { + #define GFS2_DIF_NOATIME 0x00000080 + #define GFS2_DIF_SYNC 0x00000100 + #define GFS2_DIF_SYSTEM 0x00000200 /* New in gfs2 */ ++#define GFS2_DIF_IXUNLINK 0x00010000 ++#define GFS2_DIF_BARRIER 0x00020000 ++#define GFS2_DIF_COW 0x00040000 + #define GFS2_DIF_TRUNC_IN_PROG 0x20000000 /* New in gfs2 */ + #define GFS2_DIF_INHERIT_DIRECTIO 0x40000000 + #define GFS2_DIF_INHERIT_JDATA 0x80000000 +diff -NurpP --minimal linux-3.0.9/include/linux/if_tun.h linux-3.0.9-vs2.3.2.1/include/linux/if_tun.h +--- linux-3.0.9/include/linux/if_tun.h 2010-08-02 16:52:54.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/linux/if_tun.h 2011-06-10 22:11:24.000000000 +0200 +@@ -53,6 +53,7 @@ + #define TUNDETACHFILTER _IOW('T', 214, struct sock_fprog) + #define TUNGETVNETHDRSZ _IOR('T', 215, int) + #define TUNSETVNETHDRSZ _IOW('T', 216, int) ++#define TUNSETNID _IOW('T', 217, int) + + /* TUNSETIFF ifr flags */ + #define IFF_TUN 0x0001 +diff -NurpP --minimal linux-3.0.9/include/linux/init_task.h linux-3.0.9-vs2.3.2.1/include/linux/init_task.h +--- linux-3.0.9/include/linux/init_task.h 2011-07-22 11:18:10.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/linux/init_task.h 2011-06-10 22:11:24.000000000 +0200 +@@ -193,6 +193,10 @@ extern struct cred init_cred; + INIT_FTRACE_GRAPH \ + INIT_TRACE_RECURSION \ + INIT_TASK_RCU_PREEMPT(tsk) \ ++ .xid = 0, \ ++ .vx_info = NULL, \ ++ .nid = 0, \ ++ .nx_info = NULL, \ + } + + +diff -NurpP --minimal linux-3.0.9/include/linux/ipc.h linux-3.0.9-vs2.3.2.1/include/linux/ipc.h +--- linux-3.0.9/include/linux/ipc.h 2009-12-03 20:02:55.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/ipc.h 2011-06-10 22:11:24.000000000 +0200 +@@ -91,6 +91,7 @@ struct kern_ipc_perm + key_t key; + uid_t uid; + gid_t gid; ++ xid_t xid; + uid_t cuid; + gid_t cgid; + mode_t mode; +diff -NurpP --minimal linux-3.0.9/include/linux/ipc_namespace.h linux-3.0.9-vs2.3.2.1/include/linux/ipc_namespace.h +--- linux-3.0.9/include/linux/ipc_namespace.h 2011-05-22 16:17:55.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/linux/ipc_namespace.h 2011-06-13 14:09:44.000000000 +0200 +@@ -94,7 +94,8 @@ static inline int mq_init_ns(struct ipc_ + + #if defined(CONFIG_IPC_NS) + extern struct ipc_namespace *copy_ipcs(unsigned long flags, +- struct task_struct *tsk); ++ struct ipc_namespace *old_ns, ++ struct user_namespace *user_ns); + static inline struct ipc_namespace *get_ipc_ns(struct ipc_namespace *ns) + { + if (ns) +@@ -105,12 +106,13 @@ static inline struct ipc_namespace *get_ + extern void put_ipc_ns(struct ipc_namespace *ns); + #else + static inline struct ipc_namespace *copy_ipcs(unsigned long flags, +- struct task_struct *tsk) ++ struct ipc_namespace *old_ns, ++ struct user_namespace *user_ns) + { + if (flags & CLONE_NEWIPC) + return ERR_PTR(-EINVAL); + +- return tsk->nsproxy->ipc_ns; ++ return old_ns; + } + + static inline struct ipc_namespace *get_ipc_ns(struct ipc_namespace *ns) +diff -NurpP --minimal linux-3.0.9/include/linux/loop.h linux-3.0.9-vs2.3.2.1/include/linux/loop.h +--- linux-3.0.9/include/linux/loop.h 2009-09-10 15:26:25.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/linux/loop.h 2011-06-10 22:11:24.000000000 +0200 +@@ -45,6 +45,7 @@ struct loop_device { + struct loop_func_table *lo_encryption; + __u32 lo_init[2]; + uid_t lo_key_owner; /* Who set the key */ ++ xid_t lo_xid; + int (*ioctl)(struct loop_device *, int cmd, + unsigned long arg); + +diff -NurpP --minimal linux-3.0.9/include/linux/magic.h linux-3.0.9-vs2.3.2.1/include/linux/magic.h +--- linux-3.0.9/include/linux/magic.h 2011-05-22 16:17:55.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/linux/magic.h 2011-06-10 22:11:24.000000000 +0200 +@@ -3,7 +3,7 @@ + + #define ADFS_SUPER_MAGIC 0xadf5 + #define AFFS_SUPER_MAGIC 0xadff +-#define AFS_SUPER_MAGIC 0x5346414F ++#define AFS_SUPER_MAGIC 0x5346414F + #define AUTOFS_SUPER_MAGIC 0x0187 + #define CODA_SUPER_MAGIC 0x73757245 + #define CRAMFS_MAGIC 0x28cd3d45 /* some random number */ +@@ -41,6 +41,7 @@ + #define NFS_SUPER_MAGIC 0x6969 + #define OPENPROM_SUPER_MAGIC 0x9fa1 + #define PROC_SUPER_MAGIC 0x9fa0 ++#define DEVPTS_SUPER_MAGIC 0x1cd1 + #define QNX4_SUPER_MAGIC 0x002f /* qnx4 fs detection */ + + #define REISERFS_SUPER_MAGIC 0x52654973 /* used by gcc */ +diff -NurpP --minimal linux-3.0.9/include/linux/major.h linux-3.0.9-vs2.3.2.1/include/linux/major.h +--- linux-3.0.9/include/linux/major.h 2009-09-10 15:26:25.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/linux/major.h 2011-06-10 22:11:24.000000000 +0200 +@@ -15,6 +15,7 @@ + #define HD_MAJOR IDE0_MAJOR + #define PTY_SLAVE_MAJOR 3 + #define TTY_MAJOR 4 ++#define VROOT_MAJOR 4 + #define TTYAUX_MAJOR 5 + #define LP_MAJOR 6 + #define VCS_MAJOR 7 +diff -NurpP --minimal linux-3.0.9/include/linux/memcontrol.h linux-3.0.9-vs2.3.2.1/include/linux/memcontrol.h +--- linux-3.0.9/include/linux/memcontrol.h 2011-07-22 11:18:10.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/linux/memcontrol.h 2011-06-22 12:39:15.000000000 +0200 +@@ -86,6 +86,13 @@ extern struct mem_cgroup *try_get_mem_cg + extern struct mem_cgroup *mem_cgroup_from_task(struct task_struct *p); + extern struct mem_cgroup *try_get_mem_cgroup_from_mm(struct mm_struct *mm); + ++extern u64 mem_cgroup_res_read_u64(struct mem_cgroup *mem, int member); ++extern u64 mem_cgroup_memsw_read_u64(struct mem_cgroup *mem, int member); ++ ++extern s64 mem_cgroup_stat_read_cache(struct mem_cgroup *mem); ++extern s64 mem_cgroup_stat_read_anon(struct mem_cgroup *mem); ++extern s64 mem_cgroup_stat_read_mapped(struct mem_cgroup *mem); ++ + static inline + int mm_match_cgroup(const struct mm_struct *mm, const struct mem_cgroup *cgroup) + { +diff -NurpP --minimal linux-3.0.9/include/linux/mm_types.h linux-3.0.9-vs2.3.2.1/include/linux/mm_types.h +--- linux-3.0.9/include/linux/mm_types.h 2011-11-15 16:40:47.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/mm_types.h 2011-11-15 17:37:07.000000000 +0100 +@@ -282,6 +282,7 @@ struct mm_struct { + + /* Architecture-specific MM context */ + mm_context_t context; ++ struct vx_info *mm_vx_info; + + /* Swap token stuff */ + /* +diff -NurpP --minimal linux-3.0.9/include/linux/mmzone.h linux-3.0.9-vs2.3.2.1/include/linux/mmzone.h +--- linux-3.0.9/include/linux/mmzone.h 2011-07-22 11:18:11.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/linux/mmzone.h 2011-07-01 11:35:35.000000000 +0200 +@@ -654,6 +654,13 @@ typedef struct pglist_data { + __pgdat->node_start_pfn + __pgdat->node_spanned_pages;\ + }) + ++#define node_start_pfn(nid) (NODE_DATA(nid)->node_start_pfn) ++ ++#define node_end_pfn(nid) ({\ ++ pg_data_t *__pgdat = NODE_DATA(nid);\ ++ __pgdat->node_start_pfn + __pgdat->node_spanned_pages;\ ++}) ++ + #include <linux/memory_hotplug.h> + + extern struct mutex zonelists_mutex; +diff -NurpP --minimal linux-3.0.9/include/linux/mount.h linux-3.0.9-vs2.3.2.1/include/linux/mount.h +--- linux-3.0.9/include/linux/mount.h 2011-03-15 18:07:39.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/mount.h 2011-06-10 22:11:24.000000000 +0200 +@@ -52,6 +52,9 @@ struct mnt_pcp { + int mnt_writers; + }; + ++#define MNT_TAGID 0x10000 ++#define MNT_NOTAG 0x20000 ++ + struct vfsmount { + struct list_head mnt_hash; + struct vfsmount *mnt_parent; /* fs we are mounted on */ +@@ -86,6 +89,7 @@ struct vfsmount { + int mnt_expiry_mark; /* true if marked for expiry */ + int mnt_pinned; + int mnt_ghosts; ++ tag_t mnt_tag; /* tagging used for vfsmount */ + }; + + struct file; /* forward dec */ +diff -NurpP --minimal linux-3.0.9/include/linux/net.h linux-3.0.9-vs2.3.2.1/include/linux/net.h +--- linux-3.0.9/include/linux/net.h 2011-07-22 11:18:11.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/linux/net.h 2011-06-10 22:11:24.000000000 +0200 +@@ -72,6 +72,7 @@ struct net; + #define SOCK_NOSPACE 2 + #define SOCK_PASSCRED 3 + #define SOCK_PASSSEC 4 ++#define SOCK_USER_SOCKET 5 + + #ifndef ARCH_HAS_SOCKET_TYPES + /** +diff -NurpP --minimal linux-3.0.9/include/linux/netdevice.h linux-3.0.9-vs2.3.2.1/include/linux/netdevice.h +--- linux-3.0.9/include/linux/netdevice.h 2011-11-15 16:40:47.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/netdevice.h 2011-09-17 19:22:49.000000000 +0200 +@@ -1650,6 +1650,7 @@ extern void netdev_resync_ops(struct ne + extern int call_netdevice_notifiers(unsigned long val, struct net_device *dev); + extern struct net_device *dev_get_by_index(struct net *net, int ifindex); + extern struct net_device *__dev_get_by_index(struct net *net, int ifindex); ++extern struct net_device *dev_get_by_index_real_rcu(struct net *net, int ifindex); + extern struct net_device *dev_get_by_index_rcu(struct net *net, int ifindex); + extern int dev_restart(struct net_device *dev); + #ifdef CONFIG_NETPOLL_TRAP +diff -NurpP --minimal linux-3.0.9/include/linux/nfs_mount.h linux-3.0.9-vs2.3.2.1/include/linux/nfs_mount.h +--- linux-3.0.9/include/linux/nfs_mount.h 2011-01-05 21:50:31.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/nfs_mount.h 2011-06-10 22:11:24.000000000 +0200 +@@ -63,7 +63,8 @@ struct nfs_mount_data { + #define NFS_MOUNT_SECFLAVOUR 0x2000 /* 5 */ + #define NFS_MOUNT_NORDIRPLUS 0x4000 /* 5 */ + #define NFS_MOUNT_UNSHARED 0x8000 /* 5 */ +-#define NFS_MOUNT_FLAGMASK 0xFFFF ++#define NFS_MOUNT_TAGGED 0x10000 /* context tagging */ ++#define NFS_MOUNT_FLAGMASK 0x1FFFF + + /* The following are for internal use only */ + #define NFS_MOUNT_LOOKUP_CACHE_NONEG 0x10000 +diff -NurpP --minimal linux-3.0.9/include/linux/nsproxy.h linux-3.0.9-vs2.3.2.1/include/linux/nsproxy.h +--- linux-3.0.9/include/linux/nsproxy.h 2011-07-22 11:18:11.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/linux/nsproxy.h 2011-06-10 22:11:24.000000000 +0200 +@@ -3,6 +3,7 @@ + + #include <linux/spinlock.h> + #include <linux/sched.h> ++#include <linux/vserver/debug.h> + + struct mnt_namespace; + struct uts_namespace; +@@ -63,22 +64,33 @@ static inline struct nsproxy *task_nspro + } + + int copy_namespaces(unsigned long flags, struct task_struct *tsk); ++struct nsproxy *copy_nsproxy(struct nsproxy *orig); + void exit_task_namespaces(struct task_struct *tsk); + void switch_task_namespaces(struct task_struct *tsk, struct nsproxy *new); + void free_nsproxy(struct nsproxy *ns); + int unshare_nsproxy_namespaces(unsigned long, struct nsproxy **, + struct fs_struct *); + +-static inline void put_nsproxy(struct nsproxy *ns) ++#define get_nsproxy(n) __get_nsproxy(n, __FILE__, __LINE__) ++ ++static inline void __get_nsproxy(struct nsproxy *ns, ++ const char *_file, int _line) + { +- if (atomic_dec_and_test(&ns->count)) { +- free_nsproxy(ns); +- } ++ vxlprintk(VXD_CBIT(space, 0), "get_nsproxy(%p[%u])", ++ ns, atomic_read(&ns->count), _file, _line); ++ atomic_inc(&ns->count); + } + +-static inline void get_nsproxy(struct nsproxy *ns) ++#define put_nsproxy(n) __put_nsproxy(n, __FILE__, __LINE__) ++ ++static inline void __put_nsproxy(struct nsproxy *ns, ++ const char *_file, int _line) + { +- atomic_inc(&ns->count); ++ vxlprintk(VXD_CBIT(space, 0), "put_nsproxy(%p[%u])", ++ ns, atomic_read(&ns->count), _file, _line); ++ if (atomic_dec_and_test(&ns->count)) { ++ free_nsproxy(ns); ++ } + } + + #endif +diff -NurpP --minimal linux-3.0.9/include/linux/pid.h linux-3.0.9-vs2.3.2.1/include/linux/pid.h +--- linux-3.0.9/include/linux/pid.h 2011-07-22 11:18:11.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/linux/pid.h 2011-06-10 22:11:24.000000000 +0200 +@@ -8,7 +8,8 @@ enum pid_type + PIDTYPE_PID, + PIDTYPE_PGID, + PIDTYPE_SID, +- PIDTYPE_MAX ++ PIDTYPE_MAX, ++ PIDTYPE_REALPID + }; + + /* +@@ -171,6 +172,7 @@ static inline pid_t pid_nr(struct pid *p + } + + pid_t pid_nr_ns(struct pid *pid, struct pid_namespace *ns); ++pid_t pid_unmapped_nr_ns(struct pid *pid, struct pid_namespace *ns); + pid_t pid_vnr(struct pid *pid); + + #define do_each_pid_task(pid, type, task) \ +diff -NurpP --minimal linux-3.0.9/include/linux/proc_fs.h linux-3.0.9-vs2.3.2.1/include/linux/proc_fs.h +--- linux-3.0.9/include/linux/proc_fs.h 2011-07-22 11:18:11.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/linux/proc_fs.h 2011-06-10 22:11:24.000000000 +0200 +@@ -56,6 +56,7 @@ struct proc_dir_entry { + nlink_t nlink; + uid_t uid; + gid_t gid; ++ int vx_flags; + loff_t size; + const struct inode_operations *proc_iops; + /* +@@ -252,12 +253,18 @@ extern const struct proc_ns_operations n + extern const struct proc_ns_operations utsns_operations; + extern const struct proc_ns_operations ipcns_operations; + ++struct vx_info; ++struct nx_info; ++ + union proc_op { + int (*proc_get_link)(struct inode *, struct path *); + int (*proc_read)(struct task_struct *task, char *page); + int (*proc_show)(struct seq_file *m, + struct pid_namespace *ns, struct pid *pid, + struct task_struct *task); ++ int (*proc_vs_read)(char *page); ++ int (*proc_vxi_read)(struct vx_info *vxi, char *page); ++ int (*proc_nxi_read)(struct nx_info *nxi, char *page); + }; + + struct ctl_table_header; +@@ -265,6 +272,7 @@ struct ctl_table; + + struct proc_inode { + struct pid *pid; ++ int vx_flags; + int fd; + union proc_op op; + struct proc_dir_entry *pde; +diff -NurpP --minimal linux-3.0.9/include/linux/quotaops.h linux-3.0.9-vs2.3.2.1/include/linux/quotaops.h +--- linux-3.0.9/include/linux/quotaops.h 2011-05-22 16:17:57.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/linux/quotaops.h 2011-06-10 22:11:24.000000000 +0200 +@@ -8,6 +8,7 @@ + #define _LINUX_QUOTAOPS_ + + #include <linux/fs.h> ++#include <linux/vs_dlimit.h> + + #define DQUOT_SPACE_WARN 0x1 + #define DQUOT_SPACE_RESERVE 0x2 +@@ -204,11 +205,12 @@ static inline void dquot_drop(struct ino + + static inline int dquot_alloc_inode(const struct inode *inode) + { +- return 0; ++ return dl_alloc_inode(inode); + } + + static inline void dquot_free_inode(const struct inode *inode) + { ++ dl_free_inode(inode); + } + + static inline int dquot_transfer(struct inode *inode, struct iattr *iattr) +@@ -219,6 +221,10 @@ static inline int dquot_transfer(struct + static inline int __dquot_alloc_space(struct inode *inode, qsize_t number, + int flags) + { ++ int ret = 0; ++ ++ if ((ret = dl_alloc_space(inode, number))) ++ return ret; + if (!(flags & DQUOT_SPACE_RESERVE)) + inode_add_bytes(inode, number); + return 0; +@@ -229,6 +235,7 @@ static inline void __dquot_free_space(st + { + if (!(flags & DQUOT_SPACE_RESERVE)) + inode_sub_bytes(inode, number); ++ dl_free_space(inode, number); + } + + static inline int dquot_claim_space_nodirty(struct inode *inode, qsize_t number) +diff -NurpP --minimal linux-3.0.9/include/linux/reboot.h linux-3.0.9-vs2.3.2.1/include/linux/reboot.h +--- linux-3.0.9/include/linux/reboot.h 2010-07-07 18:31:56.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/linux/reboot.h 2011-06-10 22:11:24.000000000 +0200 +@@ -33,6 +33,7 @@ + #define LINUX_REBOOT_CMD_RESTART2 0xA1B2C3D4 + #define LINUX_REBOOT_CMD_SW_SUSPEND 0xD000FCE2 + #define LINUX_REBOOT_CMD_KEXEC 0x45584543 ++#define LINUX_REBOOT_CMD_OOM 0xDEADBEEF + + + #ifdef __KERNEL__ +diff -NurpP --minimal linux-3.0.9/include/linux/reiserfs_fs.h linux-3.0.9-vs2.3.2.1/include/linux/reiserfs_fs.h +--- linux-3.0.9/include/linux/reiserfs_fs.h 2011-05-22 16:17:58.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/linux/reiserfs_fs.h 2011-06-10 22:11:24.000000000 +0200 +@@ -976,6 +976,11 @@ struct stat_data_v1 { + #define REISERFS_COMPR_FL FS_COMPR_FL + #define REISERFS_NOTAIL_FL FS_NOTAIL_FL + ++/* unfortunately reiserfs sdattr is only 16 bit */ ++#define REISERFS_IXUNLINK_FL (FS_IXUNLINK_FL >> 16) ++#define REISERFS_BARRIER_FL (FS_BARRIER_FL >> 16) ++#define REISERFS_COW_FL (FS_COW_FL >> 16) ++ + /* persistent flags that file inherits from the parent directory */ + #define REISERFS_INHERIT_MASK ( REISERFS_IMMUTABLE_FL | \ + REISERFS_SYNC_FL | \ +@@ -985,6 +990,9 @@ struct stat_data_v1 { + REISERFS_COMPR_FL | \ + REISERFS_NOTAIL_FL ) + ++#define REISERFS_FL_USER_VISIBLE 0x80FF ++#define REISERFS_FL_USER_MODIFIABLE 0x80FF ++ + /* Stat Data on disk (reiserfs version of UFS disk inode minus the + address blocks) */ + struct stat_data { +@@ -2073,6 +2081,7 @@ static inline void reiserfs_update_sd(st + void sd_attrs_to_i_attrs(__u16 sd_attrs, struct inode *inode); + void i_attrs_to_sd_attrs(struct inode *inode, __u16 * sd_attrs); + int reiserfs_setattr(struct dentry *dentry, struct iattr *attr); ++int reiserfs_sync_flags(struct inode *inode, int, int); + + int __reiserfs_write_begin(struct page *page, unsigned from, unsigned len); + +diff -NurpP --minimal linux-3.0.9/include/linux/reiserfs_fs_sb.h linux-3.0.9-vs2.3.2.1/include/linux/reiserfs_fs_sb.h +--- linux-3.0.9/include/linux/reiserfs_fs_sb.h 2010-02-25 11:52:07.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/reiserfs_fs_sb.h 2011-06-10 22:11:24.000000000 +0200 +@@ -476,6 +476,7 @@ enum reiserfs_mount_options { + REISERFS_EXPOSE_PRIVROOT, + REISERFS_BARRIER_NONE, + REISERFS_BARRIER_FLUSH, ++ REISERFS_TAGGED, + + /* Actions on error */ + REISERFS_ERROR_PANIC, +diff -NurpP --minimal linux-3.0.9/include/linux/sched.h linux-3.0.9-vs2.3.2.1/include/linux/sched.h +--- linux-3.0.9/include/linux/sched.h 2011-11-15 16:40:47.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/sched.h 2011-10-18 13:51:13.000000000 +0200 +@@ -1406,6 +1406,14 @@ struct task_struct { + #endif + seccomp_t seccomp; + ++/* vserver context data */ ++ struct vx_info *vx_info; ++ struct nx_info *nx_info; ++ ++ xid_t xid; ++ nid_t nid; ++ tag_t tag; ++ + /* Thread group tracking */ + u32 parent_exec_id; + u32 self_exec_id; +@@ -1649,6 +1657,11 @@ struct pid_namespace; + pid_t __task_pid_nr_ns(struct task_struct *task, enum pid_type type, + struct pid_namespace *ns); + ++#include <linux/vserver/base.h> ++#include <linux/vserver/context.h> ++#include <linux/vserver/debug.h> ++#include <linux/vserver/pid.h> ++ + static inline pid_t task_pid_nr(struct task_struct *tsk) + { + return tsk->pid; +@@ -1662,7 +1675,8 @@ static inline pid_t task_pid_nr_ns(struc + + static inline pid_t task_pid_vnr(struct task_struct *tsk) + { +- return __task_pid_nr_ns(tsk, PIDTYPE_PID, NULL); ++ // return __task_pid_nr_ns(tsk, PIDTYPE_PID, NULL); ++ return vx_map_pid(__task_pid_nr_ns(tsk, PIDTYPE_PID, NULL)); + } + + +@@ -1675,7 +1689,7 @@ pid_t task_tgid_nr_ns(struct task_struct + + static inline pid_t task_tgid_vnr(struct task_struct *tsk) + { +- return pid_vnr(task_tgid(tsk)); ++ return vx_map_tgid(pid_vnr(task_tgid(tsk))); + } + + +diff -NurpP --minimal linux-3.0.9/include/linux/shmem_fs.h linux-3.0.9-vs2.3.2.1/include/linux/shmem_fs.h +--- linux-3.0.9/include/linux/shmem_fs.h 2011-07-22 11:18:11.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/linux/shmem_fs.h 2011-07-01 11:35:35.000000000 +0200 +@@ -12,6 +12,9 @@ + + #define SHMEM_SYMLINK_INLINE_LEN (SHMEM_NR_DIRECT * sizeof(swp_entry_t)) + ++#define TMPFS_SUPER_MAGIC 0x01021994 ++ ++ + struct shmem_inode_info { + spinlock_t lock; + unsigned long flags; +diff -NurpP --minimal linux-3.0.9/include/linux/stat.h linux-3.0.9-vs2.3.2.1/include/linux/stat.h +--- linux-3.0.9/include/linux/stat.h 2008-12-25 00:26:37.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/stat.h 2011-06-10 22:11:24.000000000 +0200 +@@ -66,6 +66,7 @@ struct kstat { + unsigned int nlink; + uid_t uid; + gid_t gid; ++ tag_t tag; + dev_t rdev; + loff_t size; + struct timespec atime; +diff -NurpP --minimal linux-3.0.9/include/linux/sunrpc/auth.h linux-3.0.9-vs2.3.2.1/include/linux/sunrpc/auth.h +--- linux-3.0.9/include/linux/sunrpc/auth.h 2011-03-15 18:07:39.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/sunrpc/auth.h 2011-06-10 22:11:24.000000000 +0200 +@@ -25,6 +25,7 @@ + struct auth_cred { + uid_t uid; + gid_t gid; ++ tag_t tag; + struct group_info *group_info; + unsigned char machine_cred : 1; + }; +diff -NurpP --minimal linux-3.0.9/include/linux/sunrpc/clnt.h linux-3.0.9-vs2.3.2.1/include/linux/sunrpc/clnt.h +--- linux-3.0.9/include/linux/sunrpc/clnt.h 2011-05-22 16:17:58.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/linux/sunrpc/clnt.h 2011-06-10 22:11:24.000000000 +0200 +@@ -49,7 +49,8 @@ struct rpc_clnt { + unsigned int cl_softrtry : 1,/* soft timeouts */ + cl_discrtry : 1,/* disconnect before retry */ + cl_autobind : 1,/* use getport() */ +- cl_chatty : 1;/* be verbose */ ++ cl_chatty : 1,/* be verbose */ ++ cl_tag : 1;/* context tagging */ + + struct rpc_rtt * cl_rtt; /* RTO estimator data */ + const struct rpc_timeout *cl_timeout; /* Timeout strategy */ +diff -NurpP --minimal linux-3.0.9/include/linux/syscalls.h linux-3.0.9-vs2.3.2.1/include/linux/syscalls.h +--- linux-3.0.9/include/linux/syscalls.h 2011-07-22 11:18:11.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/linux/syscalls.h 2011-06-10 22:11:24.000000000 +0200 +@@ -483,6 +483,8 @@ asmlinkage long sys_symlink(const char _ + asmlinkage long sys_unlink(const char __user *pathname); + asmlinkage long sys_rename(const char __user *oldname, + const char __user *newname); ++asmlinkage long sys_copyfile(const char __user *from, const char __user *to, ++ umode_t mode); + asmlinkage long sys_chmod(const char __user *filename, mode_t mode); + asmlinkage long sys_fchmod(unsigned int fd, mode_t mode); + +diff -NurpP --minimal linux-3.0.9/include/linux/sysctl.h linux-3.0.9-vs2.3.2.1/include/linux/sysctl.h +--- linux-3.0.9/include/linux/sysctl.h 2011-03-15 18:07:40.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/sysctl.h 2011-06-10 22:11:24.000000000 +0200 +@@ -60,6 +60,7 @@ enum + CTL_ABI=9, /* Binary emulation */ + CTL_CPU=10, /* CPU stuff (speed scaling, etc) */ + CTL_ARLAN=254, /* arlan wireless driver */ ++ CTL_VSERVER=4242, /* Linux-VServer debug */ + CTL_S390DBF=5677, /* s390 debug */ + CTL_SUNRPC=7249, /* sunrpc debug */ + CTL_PM=9899, /* frv power management */ +@@ -94,6 +95,7 @@ enum + + KERN_PANIC=15, /* int: panic timeout */ + KERN_REALROOTDEV=16, /* real root device to mount after initrd */ ++ KERN_VSHELPER=17, /* string: path to vshelper policy agent */ + + KERN_SPARC_REBOOT=21, /* reboot command on Sparc */ + KERN_CTLALTDEL=22, /* int: allow ctl-alt-del to reboot */ +diff -NurpP --minimal linux-3.0.9/include/linux/sysfs.h linux-3.0.9-vs2.3.2.1/include/linux/sysfs.h +--- linux-3.0.9/include/linux/sysfs.h 2011-07-22 11:18:11.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/linux/sysfs.h 2011-06-22 12:39:15.000000000 +0200 +@@ -19,6 +19,8 @@ + #include <linux/kobject_ns.h> + #include <asm/atomic.h> + ++#define SYSFS_SUPER_MAGIC 0x62656572 ++ + struct kobject; + struct module; + enum kobj_ns_type; +diff -NurpP --minimal linux-3.0.9/include/linux/time.h linux-3.0.9-vs2.3.2.1/include/linux/time.h +--- linux-3.0.9/include/linux/time.h 2011-07-22 11:18:11.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/linux/time.h 2011-06-10 22:11:24.000000000 +0200 +@@ -256,6 +256,9 @@ static __always_inline void timespec_add + a->tv_sec += __iter_div_u64_rem(a->tv_nsec + ns, NSEC_PER_SEC, &ns); + a->tv_nsec = ns; + } ++ ++#include <linux/vs_time.h> ++ + #endif /* __KERNEL__ */ + + #define NFDBITS __NFDBITS +diff -NurpP --minimal linux-3.0.9/include/linux/types.h linux-3.0.9-vs2.3.2.1/include/linux/types.h +--- linux-3.0.9/include/linux/types.h 2011-05-22 16:17:58.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/linux/types.h 2011-06-10 22:11:24.000000000 +0200 +@@ -40,6 +40,9 @@ typedef __kernel_uid32_t uid_t; + typedef __kernel_gid32_t gid_t; + typedef __kernel_uid16_t uid16_t; + typedef __kernel_gid16_t gid16_t; ++typedef unsigned int xid_t; ++typedef unsigned int nid_t; ++typedef unsigned int tag_t; + + typedef unsigned long uintptr_t; + +diff -NurpP --minimal linux-3.0.9/include/linux/utsname.h linux-3.0.9-vs2.3.2.1/include/linux/utsname.h +--- linux-3.0.9/include/linux/utsname.h 2011-05-22 16:17:58.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/linux/utsname.h 2011-06-13 14:36:48.000000000 +0200 +@@ -54,7 +54,8 @@ static inline void get_uts_ns(struct uts + } + + extern struct uts_namespace *copy_utsname(unsigned long flags, +- struct task_struct *tsk); ++ struct uts_namespace *old_ns, ++ struct user_namespace *user_ns); + extern void free_uts_ns(struct kref *kref); + + static inline void put_uts_ns(struct uts_namespace *ns) +@@ -71,12 +72,13 @@ static inline void put_uts_ns(struct uts + } + + static inline struct uts_namespace *copy_utsname(unsigned long flags, +- struct task_struct *tsk) ++ struct uts_namespace *old_ns, ++ struct user_namespace *user_ns) + { + if (flags & CLONE_NEWUTS) + return ERR_PTR(-EINVAL); + +- return tsk->nsproxy->uts_ns; ++ return old_ns; + } + #endif + +diff -NurpP --minimal linux-3.0.9/include/linux/vroot.h linux-3.0.9-vs2.3.2.1/include/linux/vroot.h +--- linux-3.0.9/include/linux/vroot.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vroot.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,51 @@ ++ ++/* ++ * include/linux/vroot.h ++ * ++ * written by Herbert Pötzl, 9/11/2002 ++ * ported to 2.6 by Herbert Pötzl, 30/12/2004 ++ * ++ * Copyright (C) 2002-2007 by Herbert Pötzl. ++ * Redistribution of this file is permitted under the ++ * GNU General Public License. ++ */ ++ ++#ifndef _LINUX_VROOT_H ++#define _LINUX_VROOT_H ++ ++ ++#ifdef __KERNEL__ ++ ++/* Possible states of device */ ++enum { ++ Vr_unbound, ++ Vr_bound, ++}; ++ ++struct vroot_device { ++ int vr_number; ++ int vr_refcnt; ++ ++ struct semaphore vr_ctl_mutex; ++ struct block_device *vr_device; ++ int vr_state; ++}; ++ ++ ++typedef struct block_device *(vroot_grb_func)(struct block_device *); ++ ++extern int register_vroot_grb(vroot_grb_func *); ++extern int unregister_vroot_grb(vroot_grb_func *); ++ ++#endif /* __KERNEL__ */ ++ ++#define MAX_VROOT_DEFAULT 8 ++ ++/* ++ * IOCTL commands --- we will commandeer 0x56 ('V') ++ */ ++ ++#define VROOT_SET_DEV 0x5600 ++#define VROOT_CLR_DEV 0x5601 ++ ++#endif /* _LINUX_VROOT_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vs_base.h linux-3.0.9-vs2.3.2.1/include/linux/vs_base.h +--- linux-3.0.9/include/linux/vs_base.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vs_base.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,10 @@ ++#ifndef _VS_BASE_H ++#define _VS_BASE_H ++ ++#include "vserver/base.h" ++#include "vserver/check.h" ++#include "vserver/debug.h" ++ ++#else ++#warning duplicate inclusion ++#endif +diff -NurpP --minimal linux-3.0.9/include/linux/vs_context.h linux-3.0.9-vs2.3.2.1/include/linux/vs_context.h +--- linux-3.0.9/include/linux/vs_context.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vs_context.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,242 @@ ++#ifndef _VS_CONTEXT_H ++#define _VS_CONTEXT_H ++ ++#include "vserver/base.h" ++#include "vserver/check.h" ++#include "vserver/context.h" ++#include "vserver/history.h" ++#include "vserver/debug.h" ++ ++#include <linux/sched.h> ++ ++ ++#define get_vx_info(i) __get_vx_info(i, __FILE__, __LINE__, __HERE__) ++ ++static inline struct vx_info *__get_vx_info(struct vx_info *vxi, ++ const char *_file, int _line, void *_here) ++{ ++ if (!vxi) ++ return NULL; ++ ++ vxlprintk(VXD_CBIT(xid, 2), "get_vx_info(%p[#%d.%d])", ++ vxi, vxi ? vxi->vx_id : 0, ++ vxi ? atomic_read(&vxi->vx_usecnt) : 0, ++ _file, _line); ++ __vxh_get_vx_info(vxi, _here); ++ ++ atomic_inc(&vxi->vx_usecnt); ++ return vxi; ++} ++ ++ ++extern void free_vx_info(struct vx_info *); ++ ++#define put_vx_info(i) __put_vx_info(i, __FILE__, __LINE__, __HERE__) ++ ++static inline void __put_vx_info(struct vx_info *vxi, ++ const char *_file, int _line, void *_here) ++{ ++ if (!vxi) ++ return; ++ ++ vxlprintk(VXD_CBIT(xid, 2), "put_vx_info(%p[#%d.%d])", ++ vxi, vxi ? vxi->vx_id : 0, ++ vxi ? atomic_read(&vxi->vx_usecnt) : 0, ++ _file, _line); ++ __vxh_put_vx_info(vxi, _here); ++ ++ if (atomic_dec_and_test(&vxi->vx_usecnt)) ++ free_vx_info(vxi); ++} ++ ++ ++#define init_vx_info(p, i) \ ++ __init_vx_info(p, i, __FILE__, __LINE__, __HERE__) ++ ++static inline void __init_vx_info(struct vx_info **vxp, struct vx_info *vxi, ++ const char *_file, int _line, void *_here) ++{ ++ if (vxi) { ++ vxlprintk(VXD_CBIT(xid, 3), ++ "init_vx_info(%p[#%d.%d])", ++ vxi, vxi ? vxi->vx_id : 0, ++ vxi ? atomic_read(&vxi->vx_usecnt) : 0, ++ _file, _line); ++ __vxh_init_vx_info(vxi, vxp, _here); ++ ++ atomic_inc(&vxi->vx_usecnt); ++ } ++ *vxp = vxi; ++} ++ ++ ++#define set_vx_info(p, i) \ ++ __set_vx_info(p, i, __FILE__, __LINE__, __HERE__) ++ ++static inline void __set_vx_info(struct vx_info **vxp, struct vx_info *vxi, ++ const char *_file, int _line, void *_here) ++{ ++ struct vx_info *vxo; ++ ++ if (!vxi) ++ return; ++ ++ vxlprintk(VXD_CBIT(xid, 3), "set_vx_info(%p[#%d.%d])", ++ vxi, vxi ? vxi->vx_id : 0, ++ vxi ? atomic_read(&vxi->vx_usecnt) : 0, ++ _file, _line); ++ __vxh_set_vx_info(vxi, vxp, _here); ++ ++ atomic_inc(&vxi->vx_usecnt); ++ vxo = xchg(vxp, vxi); ++ BUG_ON(vxo); ++} ++ ++ ++#define clr_vx_info(p) __clr_vx_info(p, __FILE__, __LINE__, __HERE__) ++ ++static inline void __clr_vx_info(struct vx_info **vxp, ++ const char *_file, int _line, void *_here) ++{ ++ struct vx_info *vxo; ++ ++ vxo = xchg(vxp, NULL); ++ if (!vxo) ++ return; ++ ++ vxlprintk(VXD_CBIT(xid, 3), "clr_vx_info(%p[#%d.%d])", ++ vxo, vxo ? vxo->vx_id : 0, ++ vxo ? atomic_read(&vxo->vx_usecnt) : 0, ++ _file, _line); ++ __vxh_clr_vx_info(vxo, vxp, _here); ++ ++ if (atomic_dec_and_test(&vxo->vx_usecnt)) ++ free_vx_info(vxo); ++} ++ ++ ++#define claim_vx_info(v, p) \ ++ __claim_vx_info(v, p, __FILE__, __LINE__, __HERE__) ++ ++static inline void __claim_vx_info(struct vx_info *vxi, ++ struct task_struct *task, ++ const char *_file, int _line, void *_here) ++{ ++ vxlprintk(VXD_CBIT(xid, 3), "claim_vx_info(%p[#%d.%d.%d]) %p", ++ vxi, vxi ? vxi->vx_id : 0, ++ vxi ? atomic_read(&vxi->vx_usecnt) : 0, ++ vxi ? atomic_read(&vxi->vx_tasks) : 0, ++ task, _file, _line); ++ __vxh_claim_vx_info(vxi, task, _here); ++ ++ atomic_inc(&vxi->vx_tasks); ++} ++ ++ ++extern void unhash_vx_info(struct vx_info *); ++ ++#define release_vx_info(v, p) \ ++ __release_vx_info(v, p, __FILE__, __LINE__, __HERE__) ++ ++static inline void __release_vx_info(struct vx_info *vxi, ++ struct task_struct *task, ++ const char *_file, int _line, void *_here) ++{ ++ vxlprintk(VXD_CBIT(xid, 3), "release_vx_info(%p[#%d.%d.%d]) %p", ++ vxi, vxi ? vxi->vx_id : 0, ++ vxi ? atomic_read(&vxi->vx_usecnt) : 0, ++ vxi ? atomic_read(&vxi->vx_tasks) : 0, ++ task, _file, _line); ++ __vxh_release_vx_info(vxi, task, _here); ++ ++ might_sleep(); ++ ++ if (atomic_dec_and_test(&vxi->vx_tasks)) ++ unhash_vx_info(vxi); ++} ++ ++ ++#define task_get_vx_info(p) \ ++ __task_get_vx_info(p, __FILE__, __LINE__, __HERE__) ++ ++static inline struct vx_info *__task_get_vx_info(struct task_struct *p, ++ const char *_file, int _line, void *_here) ++{ ++ struct vx_info *vxi; ++ ++ task_lock(p); ++ vxlprintk(VXD_CBIT(xid, 5), "task_get_vx_info(%p)", ++ p, _file, _line); ++ vxi = __get_vx_info(p->vx_info, _file, _line, _here); ++ task_unlock(p); ++ return vxi; ++} ++ ++ ++static inline void __wakeup_vx_info(struct vx_info *vxi) ++{ ++ if (waitqueue_active(&vxi->vx_wait)) ++ wake_up_interruptible(&vxi->vx_wait); ++} ++ ++ ++#define enter_vx_info(v, s) __enter_vx_info(v, s, __FILE__, __LINE__) ++ ++static inline void __enter_vx_info(struct vx_info *vxi, ++ struct vx_info_save *vxis, const char *_file, int _line) ++{ ++ vxlprintk(VXD_CBIT(xid, 5), "enter_vx_info(%p[#%d],%p) %p[#%d,%p]", ++ vxi, vxi ? vxi->vx_id : 0, vxis, current, ++ current->xid, current->vx_info, _file, _line); ++ vxis->vxi = xchg(¤t->vx_info, vxi); ++ vxis->xid = current->xid; ++ current->xid = vxi ? vxi->vx_id : 0; ++} ++ ++#define leave_vx_info(s) __leave_vx_info(s, __FILE__, __LINE__) ++ ++static inline void __leave_vx_info(struct vx_info_save *vxis, ++ const char *_file, int _line) ++{ ++ vxlprintk(VXD_CBIT(xid, 5), "leave_vx_info(%p[#%d,%p]) %p[#%d,%p]", ++ vxis, vxis->xid, vxis->vxi, current, ++ current->xid, current->vx_info, _file, _line); ++ (void)xchg(¤t->vx_info, vxis->vxi); ++ current->xid = vxis->xid; ++} ++ ++ ++static inline void __enter_vx_admin(struct vx_info_save *vxis) ++{ ++ vxis->vxi = xchg(¤t->vx_info, NULL); ++ vxis->xid = xchg(¤t->xid, (xid_t)0); ++} ++ ++static inline void __leave_vx_admin(struct vx_info_save *vxis) ++{ ++ (void)xchg(¤t->xid, vxis->xid); ++ (void)xchg(¤t->vx_info, vxis->vxi); ++} ++ ++#define task_is_init(p) \ ++ __task_is_init(p, __FILE__, __LINE__, __HERE__) ++ ++static inline int __task_is_init(struct task_struct *p, ++ const char *_file, int _line, void *_here) ++{ ++ int is_init = is_global_init(p); ++ ++ task_lock(p); ++ if (p->vx_info) ++ is_init = p->vx_info->vx_initpid == p->pid; ++ task_unlock(p); ++ return is_init; ++} ++ ++extern void exit_vx_info(struct task_struct *, int); ++extern void exit_vx_info_early(struct task_struct *, int); ++ ++ ++#else ++#warning duplicate inclusion ++#endif +diff -NurpP --minimal linux-3.0.9/include/linux/vs_cowbl.h linux-3.0.9-vs2.3.2.1/include/linux/vs_cowbl.h +--- linux-3.0.9/include/linux/vs_cowbl.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vs_cowbl.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,48 @@ ++#ifndef _VS_COWBL_H ++#define _VS_COWBL_H ++ ++#include <linux/fs.h> ++#include <linux/dcache.h> ++#include <linux/namei.h> ++#include <linux/slab.h> ++ ++extern struct dentry *cow_break_link(const char *pathname); ++ ++static inline int cow_check_and_break(struct path *path) ++{ ++ struct inode *inode = path->dentry->d_inode; ++ int error = 0; ++ ++ /* do we need this check? */ ++ if (IS_RDONLY(inode)) ++ return -EROFS; ++ ++ if (IS_COW(inode)) { ++ if (IS_COW_LINK(inode)) { ++ struct dentry *new_dentry, *old_dentry = path->dentry; ++ char *pp, *buf; ++ ++ buf = kmalloc(PATH_MAX, GFP_KERNEL); ++ if (!buf) { ++ return -ENOMEM; ++ } ++ pp = d_path(path, buf, PATH_MAX); ++ new_dentry = cow_break_link(pp); ++ kfree(buf); ++ if (!IS_ERR(new_dentry)) { ++ path->dentry = new_dentry; ++ dput(old_dentry); ++ } else ++ error = PTR_ERR(new_dentry); ++ } else { ++ inode->i_flags &= ~(S_IXUNLINK | S_IMMUTABLE); ++ inode->i_ctime = CURRENT_TIME; ++ mark_inode_dirty(inode); ++ } ++ } ++ return error; ++} ++ ++#else ++#warning duplicate inclusion ++#endif +diff -NurpP --minimal linux-3.0.9/include/linux/vs_cvirt.h linux-3.0.9-vs2.3.2.1/include/linux/vs_cvirt.h +--- linux-3.0.9/include/linux/vs_cvirt.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vs_cvirt.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,50 @@ ++#ifndef _VS_CVIRT_H ++#define _VS_CVIRT_H ++ ++#include "vserver/cvirt.h" ++#include "vserver/context.h" ++#include "vserver/base.h" ++#include "vserver/check.h" ++#include "vserver/debug.h" ++ ++ ++static inline void vx_activate_task(struct task_struct *p) ++{ ++ struct vx_info *vxi; ++ ++ if ((vxi = p->vx_info)) { ++ vx_update_load(vxi); ++ atomic_inc(&vxi->cvirt.nr_running); ++ } ++} ++ ++static inline void vx_deactivate_task(struct task_struct *p) ++{ ++ struct vx_info *vxi; ++ ++ if ((vxi = p->vx_info)) { ++ vx_update_load(vxi); ++ atomic_dec(&vxi->cvirt.nr_running); ++ } ++} ++ ++static inline void vx_uninterruptible_inc(struct task_struct *p) ++{ ++ struct vx_info *vxi; ++ ++ if ((vxi = p->vx_info)) ++ atomic_inc(&vxi->cvirt.nr_uninterruptible); ++} ++ ++static inline void vx_uninterruptible_dec(struct task_struct *p) ++{ ++ struct vx_info *vxi; ++ ++ if ((vxi = p->vx_info)) ++ atomic_dec(&vxi->cvirt.nr_uninterruptible); ++} ++ ++ ++#else ++#warning duplicate inclusion ++#endif +diff -NurpP --minimal linux-3.0.9/include/linux/vs_device.h linux-3.0.9-vs2.3.2.1/include/linux/vs_device.h +--- linux-3.0.9/include/linux/vs_device.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vs_device.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,45 @@ ++#ifndef _VS_DEVICE_H ++#define _VS_DEVICE_H ++ ++#include "vserver/base.h" ++#include "vserver/device.h" ++#include "vserver/debug.h" ++ ++ ++#ifdef CONFIG_VSERVER_DEVICE ++ ++int vs_map_device(struct vx_info *, dev_t, dev_t *, umode_t); ++ ++#define vs_device_perm(v, d, m, p) \ ++ ((vs_map_device(current_vx_info(), d, NULL, m) & (p)) == (p)) ++ ++#else ++ ++static inline ++int vs_map_device(struct vx_info *vxi, ++ dev_t device, dev_t *target, umode_t mode) ++{ ++ if (target) ++ *target = device; ++ return ~0; ++} ++ ++#define vs_device_perm(v, d, m, p) ((p) == (p)) ++ ++#endif ++ ++ ++#define vs_map_chrdev(d, t, p) \ ++ ((vs_map_device(current_vx_info(), d, t, S_IFCHR) & (p)) == (p)) ++#define vs_map_blkdev(d, t, p) \ ++ ((vs_map_device(current_vx_info(), d, t, S_IFBLK) & (p)) == (p)) ++ ++#define vs_chrdev_perm(d, p) \ ++ vs_device_perm(current_vx_info(), d, S_IFCHR, p) ++#define vs_blkdev_perm(d, p) \ ++ vs_device_perm(current_vx_info(), d, S_IFBLK, p) ++ ++ ++#else ++#warning duplicate inclusion ++#endif +diff -NurpP --minimal linux-3.0.9/include/linux/vs_dlimit.h linux-3.0.9-vs2.3.2.1/include/linux/vs_dlimit.h +--- linux-3.0.9/include/linux/vs_dlimit.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vs_dlimit.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,215 @@ ++#ifndef _VS_DLIMIT_H ++#define _VS_DLIMIT_H ++ ++#include <linux/fs.h> ++ ++#include "vserver/dlimit.h" ++#include "vserver/base.h" ++#include "vserver/debug.h" ++ ++ ++#define get_dl_info(i) __get_dl_info(i, __FILE__, __LINE__) ++ ++static inline struct dl_info *__get_dl_info(struct dl_info *dli, ++ const char *_file, int _line) ++{ ++ if (!dli) ++ return NULL; ++ vxlprintk(VXD_CBIT(dlim, 4), "get_dl_info(%p[#%d.%d])", ++ dli, dli ? dli->dl_tag : 0, ++ dli ? atomic_read(&dli->dl_usecnt) : 0, ++ _file, _line); ++ atomic_inc(&dli->dl_usecnt); ++ return dli; ++} ++ ++ ++#define free_dl_info(i) \ ++ call_rcu(&(i)->dl_rcu, rcu_free_dl_info) ++ ++#define put_dl_info(i) __put_dl_info(i, __FILE__, __LINE__) ++ ++static inline void __put_dl_info(struct dl_info *dli, ++ const char *_file, int _line) ++{ ++ if (!dli) ++ return; ++ vxlprintk(VXD_CBIT(dlim, 4), "put_dl_info(%p[#%d.%d])", ++ dli, dli ? dli->dl_tag : 0, ++ dli ? atomic_read(&dli->dl_usecnt) : 0, ++ _file, _line); ++ if (atomic_dec_and_test(&dli->dl_usecnt)) ++ free_dl_info(dli); ++} ++ ++ ++#define __dlimit_char(d) ((d) ? '*' : ' ') ++ ++static inline int __dl_alloc_space(struct super_block *sb, ++ tag_t tag, dlsize_t nr, const char *file, int line) ++{ ++ struct dl_info *dli = NULL; ++ int ret = 0; ++ ++ if (nr == 0) ++ goto out; ++ dli = locate_dl_info(sb, tag); ++ if (!dli) ++ goto out; ++ ++ spin_lock(&dli->dl_lock); ++ ret = (dli->dl_space_used + nr > dli->dl_space_total); ++ if (!ret) ++ dli->dl_space_used += nr; ++ spin_unlock(&dli->dl_lock); ++ put_dl_info(dli); ++out: ++ vxlprintk(VXD_CBIT(dlim, 1), ++ "ALLOC (%p,#%d)%c %lld bytes (%d)", ++ sb, tag, __dlimit_char(dli), (long long)nr, ++ ret, file, line); ++ return ret ? -ENOSPC : 0; ++} ++ ++static inline void __dl_free_space(struct super_block *sb, ++ tag_t tag, dlsize_t nr, const char *_file, int _line) ++{ ++ struct dl_info *dli = NULL; ++ ++ if (nr == 0) ++ goto out; ++ dli = locate_dl_info(sb, tag); ++ if (!dli) ++ goto out; ++ ++ spin_lock(&dli->dl_lock); ++ if (dli->dl_space_used > nr) ++ dli->dl_space_used -= nr; ++ else ++ dli->dl_space_used = 0; ++ spin_unlock(&dli->dl_lock); ++ put_dl_info(dli); ++out: ++ vxlprintk(VXD_CBIT(dlim, 1), ++ "FREE (%p,#%d)%c %lld bytes", ++ sb, tag, __dlimit_char(dli), (long long)nr, ++ _file, _line); ++} ++ ++static inline int __dl_alloc_inode(struct super_block *sb, ++ tag_t tag, const char *_file, int _line) ++{ ++ struct dl_info *dli; ++ int ret = 0; ++ ++ dli = locate_dl_info(sb, tag); ++ if (!dli) ++ goto out; ++ ++ spin_lock(&dli->dl_lock); ++ dli->dl_inodes_used++; ++ ret = (dli->dl_inodes_used > dli->dl_inodes_total); ++ spin_unlock(&dli->dl_lock); ++ put_dl_info(dli); ++out: ++ vxlprintk(VXD_CBIT(dlim, 0), ++ "ALLOC (%p,#%d)%c inode (%d)", ++ sb, tag, __dlimit_char(dli), ret, _file, _line); ++ return ret ? -ENOSPC : 0; ++} ++ ++static inline void __dl_free_inode(struct super_block *sb, ++ tag_t tag, const char *_file, int _line) ++{ ++ struct dl_info *dli; ++ ++ dli = locate_dl_info(sb, tag); ++ if (!dli) ++ goto out; ++ ++ spin_lock(&dli->dl_lock); ++ if (dli->dl_inodes_used > 1) ++ dli->dl_inodes_used--; ++ else ++ dli->dl_inodes_used = 0; ++ spin_unlock(&dli->dl_lock); ++ put_dl_info(dli); ++out: ++ vxlprintk(VXD_CBIT(dlim, 0), ++ "FREE (%p,#%d)%c inode", ++ sb, tag, __dlimit_char(dli), _file, _line); ++} ++ ++static inline void __dl_adjust_block(struct super_block *sb, tag_t tag, ++ unsigned long long *free_blocks, unsigned long long *root_blocks, ++ const char *_file, int _line) ++{ ++ struct dl_info *dli; ++ uint64_t broot, bfree; ++ ++ dli = locate_dl_info(sb, tag); ++ if (!dli) ++ return; ++ ++ spin_lock(&dli->dl_lock); ++ broot = (dli->dl_space_total - ++ (dli->dl_space_total >> 10) * dli->dl_nrlmult) ++ >> sb->s_blocksize_bits; ++ bfree = (dli->dl_space_total - dli->dl_space_used) ++ >> sb->s_blocksize_bits; ++ spin_unlock(&dli->dl_lock); ++ ++ vxlprintk(VXD_CBIT(dlim, 2), ++ "ADJUST: %lld,%lld on %lld,%lld [mult=%d]", ++ (long long)bfree, (long long)broot, ++ *free_blocks, *root_blocks, dli->dl_nrlmult, ++ _file, _line); ++ if (free_blocks) { ++ if (*free_blocks > bfree) ++ *free_blocks = bfree; ++ } ++ if (root_blocks) { ++ if (*root_blocks > broot) ++ *root_blocks = broot; ++ } ++ put_dl_info(dli); ++} ++ ++#define dl_prealloc_space(in, bytes) \ ++ __dl_alloc_space((in)->i_sb, (in)->i_tag, (dlsize_t)(bytes), \ ++ __FILE__, __LINE__ ) ++ ++#define dl_alloc_space(in, bytes) \ ++ __dl_alloc_space((in)->i_sb, (in)->i_tag, (dlsize_t)(bytes), \ ++ __FILE__, __LINE__ ) ++ ++#define dl_reserve_space(in, bytes) \ ++ __dl_alloc_space((in)->i_sb, (in)->i_tag, (dlsize_t)(bytes), \ ++ __FILE__, __LINE__ ) ++ ++#define dl_claim_space(in, bytes) (0) ++ ++#define dl_release_space(in, bytes) \ ++ __dl_free_space((in)->i_sb, (in)->i_tag, (dlsize_t)(bytes), \ ++ __FILE__, __LINE__ ) ++ ++#define dl_free_space(in, bytes) \ ++ __dl_free_space((in)->i_sb, (in)->i_tag, (dlsize_t)(bytes), \ ++ __FILE__, __LINE__ ) ++ ++ ++ ++#define dl_alloc_inode(in) \ ++ __dl_alloc_inode((in)->i_sb, (in)->i_tag, __FILE__, __LINE__ ) ++ ++#define dl_free_inode(in) \ ++ __dl_free_inode((in)->i_sb, (in)->i_tag, __FILE__, __LINE__ ) ++ ++ ++#define dl_adjust_block(sb, tag, fb, rb) \ ++ __dl_adjust_block(sb, tag, fb, rb, __FILE__, __LINE__ ) ++ ++ ++#else ++#warning duplicate inclusion ++#endif +diff -NurpP --minimal linux-3.0.9/include/linux/vs_inet.h linux-3.0.9-vs2.3.2.1/include/linux/vs_inet.h +--- linux-3.0.9/include/linux/vs_inet.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vs_inet.h 2011-08-18 17:01:00.000000000 +0200 +@@ -0,0 +1,353 @@ ++#ifndef _VS_INET_H ++#define _VS_INET_H ++ ++#include "vserver/base.h" ++#include "vserver/network.h" ++#include "vserver/debug.h" ++ ++#define IPI_LOOPBACK htonl(INADDR_LOOPBACK) ++ ++#define NXAV4(a) NIPQUAD((a)->ip[0]), NIPQUAD((a)->ip[1]), \ ++ NIPQUAD((a)->mask), (a)->type ++#define NXAV4_FMT "[" NIPQUAD_FMT "-" NIPQUAD_FMT "/" NIPQUAD_FMT ":%04x]" ++ ++#define NIPQUAD(addr) \ ++ ((unsigned char *)&addr)[0], \ ++ ((unsigned char *)&addr)[1], \ ++ ((unsigned char *)&addr)[2], \ ++ ((unsigned char *)&addr)[3] ++ ++#define NIPQUAD_FMT "%u.%u.%u.%u" ++ ++ ++static inline ++int v4_addr_match(struct nx_addr_v4 *nxa, __be32 addr, uint16_t tmask) ++{ ++ __be32 ip = nxa->ip[0].s_addr; ++ __be32 mask = nxa->mask.s_addr; ++ __be32 bcast = ip | ~mask; ++ int ret = 0; ++ ++ switch (nxa->type & tmask) { ++ case NXA_TYPE_MASK: ++ ret = (ip == (addr & mask)); ++ break; ++ case NXA_TYPE_ADDR: ++ ret = 3; ++ if (addr == ip) ++ break; ++ /* fall through to broadcast */ ++ case NXA_MOD_BCAST: ++ ret = ((tmask & NXA_MOD_BCAST) && (addr == bcast)); ++ break; ++ case NXA_TYPE_RANGE: ++ ret = ((nxa->ip[0].s_addr <= addr) && ++ (nxa->ip[1].s_addr > addr)); ++ break; ++ case NXA_TYPE_ANY: ++ ret = 2; ++ break; ++ } ++ ++ vxdprintk(VXD_CBIT(net, 0), ++ "v4_addr_match(%p" NXAV4_FMT "," NIPQUAD_FMT ",%04x) = %d", ++ nxa, NXAV4(nxa), NIPQUAD(addr), tmask, ret); ++ return ret; ++} ++ ++static inline ++int v4_addr_in_nx_info(struct nx_info *nxi, __be32 addr, uint16_t tmask) ++{ ++ struct nx_addr_v4 *nxa; ++ int ret = 1; ++ ++ if (!nxi) ++ goto out; ++ ++ ret = 2; ++ /* allow 127.0.0.1 when remapping lback */ ++ if ((tmask & NXA_LOOPBACK) && ++ (addr == IPI_LOOPBACK) && ++ nx_info_flags(nxi, NXF_LBACK_REMAP, 0)) ++ goto out; ++ ret = 3; ++ /* check for lback address */ ++ if ((tmask & NXA_MOD_LBACK) && ++ (nxi->v4_lback.s_addr == addr)) ++ goto out; ++ ret = 4; ++ /* check for broadcast address */ ++ if ((tmask & NXA_MOD_BCAST) && ++ (nxi->v4_bcast.s_addr == addr)) ++ goto out; ++ ret = 5; ++ /* check for v4 addresses */ ++ for (nxa = &nxi->v4; nxa; nxa = nxa->next) ++ if (v4_addr_match(nxa, addr, tmask)) ++ goto out; ++ ret = 0; ++out: ++ vxdprintk(VXD_CBIT(net, 0), ++ "v4_addr_in_nx_info(%p[#%u]," NIPQUAD_FMT ",%04x) = %d", ++ nxi, nxi ? nxi->nx_id : 0, NIPQUAD(addr), tmask, ret); ++ return ret; ++} ++ ++static inline ++int v4_nx_addr_match(struct nx_addr_v4 *nxa, struct nx_addr_v4 *addr, uint16_t mask) ++{ ++ /* FIXME: needs full range checks */ ++ return v4_addr_match(nxa, addr->ip[0].s_addr, mask); ++} ++ ++static inline ++int v4_nx_addr_in_nx_info(struct nx_info *nxi, struct nx_addr_v4 *nxa, uint16_t mask) ++{ ++ struct nx_addr_v4 *ptr; ++ ++ for (ptr = &nxi->v4; ptr; ptr = ptr->next) ++ if (v4_nx_addr_match(ptr, nxa, mask)) ++ return 1; ++ return 0; ++} ++ ++#include <net/inet_sock.h> ++ ++/* ++ * Check if a given address matches for a socket ++ * ++ * nxi: the socket's nx_info if any ++ * addr: to be verified address ++ */ ++static inline ++int v4_sock_addr_match ( ++ struct nx_info *nxi, ++ struct inet_sock *inet, ++ __be32 addr) ++{ ++ __be32 saddr = inet->inet_rcv_saddr; ++ __be32 bcast = nxi ? nxi->v4_bcast.s_addr : INADDR_BROADCAST; ++ ++ if (addr && (saddr == addr || bcast == addr)) ++ return 1; ++ if (!saddr) ++ return v4_addr_in_nx_info(nxi, addr, NXA_MASK_BIND); ++ return 0; ++} ++ ++ ++/* inet related checks and helpers */ ++ ++ ++struct in_ifaddr; ++struct net_device; ++struct sock; ++ ++#ifdef CONFIG_INET ++ ++#include <linux/netdevice.h> ++#include <linux/inetdevice.h> ++#include <net/inet_sock.h> ++#include <net/inet_timewait_sock.h> ++ ++ ++int dev_in_nx_info(struct net_device *, struct nx_info *); ++int v4_dev_in_nx_info(struct net_device *, struct nx_info *); ++int nx_v4_addr_conflict(struct nx_info *, struct nx_info *); ++ ++ ++/* ++ * check if address is covered by socket ++ * ++ * sk: the socket to check against ++ * addr: the address in question (must be != 0) ++ */ ++ ++static inline ++int __v4_addr_match_socket(const struct sock *sk, struct nx_addr_v4 *nxa) ++{ ++ struct nx_info *nxi = sk->sk_nx_info; ++ __be32 saddr = sk_rcv_saddr(sk); ++ ++ vxdprintk(VXD_CBIT(net, 5), ++ "__v4_addr_in_socket(%p," NXAV4_FMT ") %p:" NIPQUAD_FMT " %p;%lx", ++ sk, NXAV4(nxa), nxi, NIPQUAD(saddr), sk->sk_socket, ++ (sk->sk_socket?sk->sk_socket->flags:0)); ++ ++ if (saddr) { /* direct address match */ ++ return v4_addr_match(nxa, saddr, -1); ++ } else if (nxi) { /* match against nx_info */ ++ return v4_nx_addr_in_nx_info(nxi, nxa, -1); ++ } else { /* unrestricted any socket */ ++ return 1; ++ } ++} ++ ++ ++ ++static inline ++int nx_dev_visible(struct nx_info *nxi, struct net_device *dev) ++{ ++ vxdprintk(VXD_CBIT(net, 1), ++ "nx_dev_visible(%p[#%u],%p " VS_Q("%s") ") %d", ++ nxi, nxi ? nxi->nx_id : 0, dev, dev->name, ++ nxi ? dev_in_nx_info(dev, nxi) : 0); ++ ++ if (!nx_info_flags(nxi, NXF_HIDE_NETIF, 0)) ++ return 1; ++ if (dev_in_nx_info(dev, nxi)) ++ return 1; ++ return 0; ++} ++ ++ ++static inline ++int v4_ifa_in_nx_info(struct in_ifaddr *ifa, struct nx_info *nxi) ++{ ++ if (!nxi) ++ return 1; ++ if (!ifa) ++ return 0; ++ return v4_addr_in_nx_info(nxi, ifa->ifa_local, NXA_MASK_SHOW); ++} ++ ++static inline ++int nx_v4_ifa_visible(struct nx_info *nxi, struct in_ifaddr *ifa) ++{ ++ vxdprintk(VXD_CBIT(net, 1), "nx_v4_ifa_visible(%p[#%u],%p) %d", ++ nxi, nxi ? nxi->nx_id : 0, ifa, ++ nxi ? v4_ifa_in_nx_info(ifa, nxi) : 0); ++ ++ if (!nx_info_flags(nxi, NXF_HIDE_NETIF, 0)) ++ return 1; ++ if (v4_ifa_in_nx_info(ifa, nxi)) ++ return 1; ++ return 0; ++} ++ ++ ++struct nx_v4_sock_addr { ++ __be32 saddr; /* Address used for validation */ ++ __be32 baddr; /* Address used for socket bind */ ++}; ++ ++static inline ++int v4_map_sock_addr(struct inet_sock *inet, struct sockaddr_in *addr, ++ struct nx_v4_sock_addr *nsa) ++{ ++ struct sock *sk = &inet->sk; ++ struct nx_info *nxi = sk->sk_nx_info; ++ __be32 saddr = addr->sin_addr.s_addr; ++ __be32 baddr = saddr; ++ ++ vxdprintk(VXD_CBIT(net, 3), ++ "inet_bind(%p)* %p,%p;%lx " NIPQUAD_FMT, ++ sk, sk->sk_nx_info, sk->sk_socket, ++ (sk->sk_socket ? sk->sk_socket->flags : 0), ++ NIPQUAD(saddr)); ++ ++ if (nxi) { ++ if (saddr == INADDR_ANY) { ++ if (nx_info_flags(nxi, NXF_SINGLE_IP, 0)) ++ baddr = nxi->v4.ip[0].s_addr; ++ } else if (saddr == IPI_LOOPBACK) { ++ if (nx_info_flags(nxi, NXF_LBACK_REMAP, 0)) ++ baddr = nxi->v4_lback.s_addr; ++ } else if (!ipv4_is_multicast(saddr) || ++ !nx_info_ncaps(nxi, NXC_MULTICAST)) { ++ /* normal address bind */ ++ if (!v4_addr_in_nx_info(nxi, saddr, NXA_MASK_BIND)) ++ return -EADDRNOTAVAIL; ++ } ++ } ++ ++ vxdprintk(VXD_CBIT(net, 3), ++ "inet_bind(%p) " NIPQUAD_FMT ", " NIPQUAD_FMT, ++ sk, NIPQUAD(saddr), NIPQUAD(baddr)); ++ ++ nsa->saddr = saddr; ++ nsa->baddr = baddr; ++ return 0; ++} ++ ++static inline ++void v4_set_sock_addr(struct inet_sock *inet, struct nx_v4_sock_addr *nsa) ++{ ++ inet->inet_saddr = nsa->baddr; ++ inet->inet_rcv_saddr = nsa->baddr; ++} ++ ++ ++/* ++ * helper to simplify inet_lookup_listener ++ * ++ * nxi: the socket's nx_info if any ++ * addr: to be verified address ++ * saddr: socket address ++ */ ++static inline int v4_inet_addr_match ( ++ struct nx_info *nxi, ++ __be32 addr, ++ __be32 saddr) ++{ ++ if (addr && (saddr == addr)) ++ return 1; ++ if (!saddr) ++ return nxi ? v4_addr_in_nx_info(nxi, addr, NXA_MASK_BIND) : 1; ++ return 0; ++} ++ ++static inline __be32 nx_map_sock_lback(struct nx_info *nxi, __be32 addr) ++{ ++ if (nx_info_flags(nxi, NXF_HIDE_LBACK, 0) && ++ (addr == nxi->v4_lback.s_addr)) ++ return IPI_LOOPBACK; ++ return addr; ++} ++ ++static inline ++int nx_info_has_v4(struct nx_info *nxi) ++{ ++ if (!nxi) ++ return 1; ++ if (NX_IPV4(nxi)) ++ return 1; ++ if (nx_info_flags(nxi, NXF_LBACK_REMAP, 0)) ++ return 1; ++ return 0; ++} ++ ++#else /* CONFIG_INET */ ++ ++static inline ++int nx_dev_visible(struct nx_info *n, struct net_device *d) ++{ ++ return 1; ++} ++ ++static inline ++int nx_v4_addr_conflict(struct nx_info *n, uint32_t a, const struct sock *s) ++{ ++ return 1; ++} ++ ++static inline ++int v4_ifa_in_nx_info(struct in_ifaddr *a, struct nx_info *n) ++{ ++ return 1; ++} ++ ++static inline ++int nx_info_has_v4(struct nx_info *nxi) ++{ ++ return 0; ++} ++ ++#endif /* CONFIG_INET */ ++ ++#define current_nx_info_has_v4() \ ++ nx_info_has_v4(current_nx_info()) ++ ++#else ++// #warning duplicate inclusion ++#endif +diff -NurpP --minimal linux-3.0.9/include/linux/vs_inet6.h linux-3.0.9-vs2.3.2.1/include/linux/vs_inet6.h +--- linux-3.0.9/include/linux/vs_inet6.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vs_inet6.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,246 @@ ++#ifndef _VS_INET6_H ++#define _VS_INET6_H ++ ++#include "vserver/base.h" ++#include "vserver/network.h" ++#include "vserver/debug.h" ++ ++#include <net/ipv6.h> ++ ++#define NXAV6(a) &(a)->ip, &(a)->mask, (a)->prefix, (a)->type ++#define NXAV6_FMT "[%pI6/%pI6/%d:%04x]" ++ ++ ++#ifdef CONFIG_IPV6 ++ ++static inline ++int v6_addr_match(struct nx_addr_v6 *nxa, ++ const struct in6_addr *addr, uint16_t mask) ++{ ++ int ret = 0; ++ ++ switch (nxa->type & mask) { ++ case NXA_TYPE_MASK: ++ ret = ipv6_masked_addr_cmp(&nxa->ip, &nxa->mask, addr); ++ break; ++ case NXA_TYPE_ADDR: ++ ret = ipv6_addr_equal(&nxa->ip, addr); ++ break; ++ case NXA_TYPE_ANY: ++ ret = 1; ++ break; ++ } ++ vxdprintk(VXD_CBIT(net, 0), ++ "v6_addr_match(%p" NXAV6_FMT ",%pI6,%04x) = %d", ++ nxa, NXAV6(nxa), addr, mask, ret); ++ return ret; ++} ++ ++static inline ++int v6_addr_in_nx_info(struct nx_info *nxi, ++ const struct in6_addr *addr, uint16_t mask) ++{ ++ struct nx_addr_v6 *nxa; ++ int ret = 1; ++ ++ if (!nxi) ++ goto out; ++ for (nxa = &nxi->v6; nxa; nxa = nxa->next) ++ if (v6_addr_match(nxa, addr, mask)) ++ goto out; ++ ret = 0; ++out: ++ vxdprintk(VXD_CBIT(net, 0), ++ "v6_addr_in_nx_info(%p[#%u],%pI6,%04x) = %d", ++ nxi, nxi ? nxi->nx_id : 0, addr, mask, ret); ++ return ret; ++} ++ ++static inline ++int v6_nx_addr_match(struct nx_addr_v6 *nxa, struct nx_addr_v6 *addr, uint16_t mask) ++{ ++ /* FIXME: needs full range checks */ ++ return v6_addr_match(nxa, &addr->ip, mask); ++} ++ ++static inline ++int v6_nx_addr_in_nx_info(struct nx_info *nxi, struct nx_addr_v6 *nxa, uint16_t mask) ++{ ++ struct nx_addr_v6 *ptr; ++ ++ for (ptr = &nxi->v6; ptr; ptr = ptr->next) ++ if (v6_nx_addr_match(ptr, nxa, mask)) ++ return 1; ++ return 0; ++} ++ ++ ++/* ++ * Check if a given address matches for a socket ++ * ++ * nxi: the socket's nx_info if any ++ * addr: to be verified address ++ */ ++static inline ++int v6_sock_addr_match ( ++ struct nx_info *nxi, ++ struct inet_sock *inet, ++ struct in6_addr *addr) ++{ ++ struct sock *sk = &inet->sk; ++ struct in6_addr *saddr = inet6_rcv_saddr(sk); ++ ++ if (!ipv6_addr_any(addr) && ++ ipv6_addr_equal(saddr, addr)) ++ return 1; ++ if (ipv6_addr_any(saddr)) ++ return v6_addr_in_nx_info(nxi, addr, -1); ++ return 0; ++} ++ ++/* ++ * check if address is covered by socket ++ * ++ * sk: the socket to check against ++ * addr: the address in question (must be != 0) ++ */ ++ ++static inline ++int __v6_addr_match_socket(const struct sock *sk, struct nx_addr_v6 *nxa) ++{ ++ struct nx_info *nxi = sk->sk_nx_info; ++ struct in6_addr *saddr = inet6_rcv_saddr(sk); ++ ++ vxdprintk(VXD_CBIT(net, 5), ++ "__v6_addr_in_socket(%p," NXAV6_FMT ") %p:%pI6 %p;%lx", ++ sk, NXAV6(nxa), nxi, saddr, sk->sk_socket, ++ (sk->sk_socket?sk->sk_socket->flags:0)); ++ ++ if (!ipv6_addr_any(saddr)) { /* direct address match */ ++ return v6_addr_match(nxa, saddr, -1); ++ } else if (nxi) { /* match against nx_info */ ++ return v6_nx_addr_in_nx_info(nxi, nxa, -1); ++ } else { /* unrestricted any socket */ ++ return 1; ++ } ++} ++ ++ ++/* inet related checks and helpers */ ++ ++ ++struct in_ifaddr; ++struct net_device; ++struct sock; ++ ++ ++#include <linux/netdevice.h> ++#include <linux/inetdevice.h> ++#include <net/inet_timewait_sock.h> ++ ++ ++int dev_in_nx_info(struct net_device *, struct nx_info *); ++int v6_dev_in_nx_info(struct net_device *, struct nx_info *); ++int nx_v6_addr_conflict(struct nx_info *, struct nx_info *); ++ ++ ++ ++static inline ++int v6_ifa_in_nx_info(struct inet6_ifaddr *ifa, struct nx_info *nxi) ++{ ++ if (!nxi) ++ return 1; ++ if (!ifa) ++ return 0; ++ return v6_addr_in_nx_info(nxi, &ifa->addr, -1); ++} ++ ++static inline ++int nx_v6_ifa_visible(struct nx_info *nxi, struct inet6_ifaddr *ifa) ++{ ++ vxdprintk(VXD_CBIT(net, 1), "nx_v6_ifa_visible(%p[#%u],%p) %d", ++ nxi, nxi ? nxi->nx_id : 0, ifa, ++ nxi ? v6_ifa_in_nx_info(ifa, nxi) : 0); ++ ++ if (!nx_info_flags(nxi, NXF_HIDE_NETIF, 0)) ++ return 1; ++ if (v6_ifa_in_nx_info(ifa, nxi)) ++ return 1; ++ return 0; ++} ++ ++ ++struct nx_v6_sock_addr { ++ struct in6_addr saddr; /* Address used for validation */ ++ struct in6_addr baddr; /* Address used for socket bind */ ++}; ++ ++static inline ++int v6_map_sock_addr(struct inet_sock *inet, struct sockaddr_in6 *addr, ++ struct nx_v6_sock_addr *nsa) ++{ ++ // struct sock *sk = &inet->sk; ++ // struct nx_info *nxi = sk->sk_nx_info; ++ struct in6_addr saddr = addr->sin6_addr; ++ struct in6_addr baddr = saddr; ++ ++ nsa->saddr = saddr; ++ nsa->baddr = baddr; ++ return 0; ++} ++ ++static inline ++void v6_set_sock_addr(struct inet_sock *inet, struct nx_v6_sock_addr *nsa) ++{ ++ // struct sock *sk = &inet->sk; ++ // struct in6_addr *saddr = inet6_rcv_saddr(sk); ++ ++ // *saddr = nsa->baddr; ++ // inet->inet_saddr = nsa->baddr; ++} ++ ++static inline ++int nx_info_has_v6(struct nx_info *nxi) ++{ ++ if (!nxi) ++ return 1; ++ if (NX_IPV6(nxi)) ++ return 1; ++ return 0; ++} ++ ++#else /* CONFIG_IPV6 */ ++ ++static inline ++int nx_v6_dev_visible(struct nx_info *n, struct net_device *d) ++{ ++ return 1; ++} ++ ++ ++static inline ++int nx_v6_addr_conflict(struct nx_info *n, uint32_t a, const struct sock *s) ++{ ++ return 1; ++} ++ ++static inline ++int v6_ifa_in_nx_info(struct in_ifaddr *a, struct nx_info *n) ++{ ++ return 1; ++} ++ ++static inline ++int nx_info_has_v6(struct nx_info *nxi) ++{ ++ return 0; ++} ++ ++#endif /* CONFIG_IPV6 */ ++ ++#define current_nx_info_has_v6() \ ++ nx_info_has_v6(current_nx_info()) ++ ++#else ++#warning duplicate inclusion ++#endif +diff -NurpP --minimal linux-3.0.9/include/linux/vs_limit.h linux-3.0.9-vs2.3.2.1/include/linux/vs_limit.h +--- linux-3.0.9/include/linux/vs_limit.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vs_limit.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,140 @@ ++#ifndef _VS_LIMIT_H ++#define _VS_LIMIT_H ++ ++#include "vserver/limit.h" ++#include "vserver/base.h" ++#include "vserver/context.h" ++#include "vserver/debug.h" ++#include "vserver/context.h" ++#include "vserver/limit_int.h" ++ ++ ++#define vx_acc_cres(v, d, p, r) \ ++ __vx_acc_cres(v, r, d, p, __FILE__, __LINE__) ++ ++#define vx_acc_cres_cond(x, d, p, r) \ ++ __vx_acc_cres(((x) == vx_current_xid()) ? current_vx_info() : 0, \ ++ r, d, p, __FILE__, __LINE__) ++ ++ ++#define vx_add_cres(v, a, p, r) \ ++ __vx_add_cres(v, r, a, p, __FILE__, __LINE__) ++#define vx_sub_cres(v, a, p, r) vx_add_cres(v, -(a), p, r) ++ ++#define vx_add_cres_cond(x, a, p, r) \ ++ __vx_add_cres(((x) == vx_current_xid()) ? current_vx_info() : 0, \ ++ r, a, p, __FILE__, __LINE__) ++#define vx_sub_cres_cond(x, a, p, r) vx_add_cres_cond(x, -(a), p, r) ++ ++ ++/* process and file limits */ ++ ++#define vx_nproc_inc(p) \ ++ vx_acc_cres((p)->vx_info, 1, p, RLIMIT_NPROC) ++ ++#define vx_nproc_dec(p) \ ++ vx_acc_cres((p)->vx_info,-1, p, RLIMIT_NPROC) ++ ++#define vx_files_inc(f) \ ++ vx_acc_cres_cond((f)->f_xid, 1, f, RLIMIT_NOFILE) ++ ++#define vx_files_dec(f) \ ++ vx_acc_cres_cond((f)->f_xid,-1, f, RLIMIT_NOFILE) ++ ++#define vx_locks_inc(l) \ ++ vx_acc_cres_cond((l)->fl_xid, 1, l, RLIMIT_LOCKS) ++ ++#define vx_locks_dec(l) \ ++ vx_acc_cres_cond((l)->fl_xid,-1, l, RLIMIT_LOCKS) ++ ++#define vx_openfd_inc(f) \ ++ vx_acc_cres(current_vx_info(), 1, (void *)(long)(f), VLIMIT_OPENFD) ++ ++#define vx_openfd_dec(f) \ ++ vx_acc_cres(current_vx_info(),-1, (void *)(long)(f), VLIMIT_OPENFD) ++ ++ ++#define vx_cres_avail(v, n, r) \ ++ __vx_cres_avail(v, r, n, __FILE__, __LINE__) ++ ++ ++#define vx_nproc_avail(n) \ ++ vx_cres_avail(current_vx_info(), n, RLIMIT_NPROC) ++ ++#define vx_files_avail(n) \ ++ vx_cres_avail(current_vx_info(), n, RLIMIT_NOFILE) ++ ++#define vx_locks_avail(n) \ ++ vx_cres_avail(current_vx_info(), n, RLIMIT_LOCKS) ++ ++#define vx_openfd_avail(n) \ ++ vx_cres_avail(current_vx_info(), n, VLIMIT_OPENFD) ++ ++ ++/* dentry limits */ ++ ++#define vx_dentry_inc(d) do { \ ++ if ((d)->d_count == 1) \ ++ vx_acc_cres(current_vx_info(), 1, d, VLIMIT_DENTRY); \ ++ } while (0) ++ ++#define vx_dentry_dec(d) do { \ ++ if ((d)->d_count == 0) \ ++ vx_acc_cres(current_vx_info(),-1, d, VLIMIT_DENTRY); \ ++ } while (0) ++ ++#define vx_dentry_avail(n) \ ++ vx_cres_avail(current_vx_info(), n, VLIMIT_DENTRY) ++ ++ ++/* socket limits */ ++ ++#define vx_sock_inc(s) \ ++ vx_acc_cres((s)->sk_vx_info, 1, s, VLIMIT_NSOCK) ++ ++#define vx_sock_dec(s) \ ++ vx_acc_cres((s)->sk_vx_info,-1, s, VLIMIT_NSOCK) ++ ++#define vx_sock_avail(n) \ ++ vx_cres_avail(current_vx_info(), n, VLIMIT_NSOCK) ++ ++ ++/* ipc resource limits */ ++ ++#define vx_ipcmsg_add(v, u, a) \ ++ vx_add_cres(v, a, u, RLIMIT_MSGQUEUE) ++ ++#define vx_ipcmsg_sub(v, u, a) \ ++ vx_sub_cres(v, a, u, RLIMIT_MSGQUEUE) ++ ++#define vx_ipcmsg_avail(v, a) \ ++ vx_cres_avail(v, a, RLIMIT_MSGQUEUE) ++ ++ ++#define vx_ipcshm_add(v, k, a) \ ++ vx_add_cres(v, a, (void *)(long)(k), VLIMIT_SHMEM) ++ ++#define vx_ipcshm_sub(v, k, a) \ ++ vx_sub_cres(v, a, (void *)(long)(k), VLIMIT_SHMEM) ++ ++#define vx_ipcshm_avail(v, a) \ ++ vx_cres_avail(v, a, VLIMIT_SHMEM) ++ ++ ++#define vx_semary_inc(a) \ ++ vx_acc_cres(current_vx_info(), 1, a, VLIMIT_SEMARY) ++ ++#define vx_semary_dec(a) \ ++ vx_acc_cres(current_vx_info(), -1, a, VLIMIT_SEMARY) ++ ++ ++#define vx_nsems_add(a,n) \ ++ vx_add_cres(current_vx_info(), n, a, VLIMIT_NSEMS) ++ ++#define vx_nsems_sub(a,n) \ ++ vx_sub_cres(current_vx_info(), n, a, VLIMIT_NSEMS) ++ ++ ++#else ++#warning duplicate inclusion ++#endif +diff -NurpP --minimal linux-3.0.9/include/linux/vs_memory.h linux-3.0.9-vs2.3.2.1/include/linux/vs_memory.h +--- linux-3.0.9/include/linux/vs_memory.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vs_memory.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,58 @@ ++#ifndef _VS_MEMORY_H ++#define _VS_MEMORY_H ++ ++#include "vserver/limit.h" ++#include "vserver/base.h" ++#include "vserver/context.h" ++#include "vserver/debug.h" ++#include "vserver/context.h" ++#include "vserver/limit_int.h" ++ ++enum { ++ VXPT_UNKNOWN = 0, ++ VXPT_ANON, ++ VXPT_NONE, ++ VXPT_FILE, ++ VXPT_SWAP, ++ VXPT_WRITE ++}; ++ ++#if 0 ++#define vx_page_fault(mm, vma, type, ret) ++#else ++ ++static inline ++void __vx_page_fault(struct mm_struct *mm, ++ struct vm_area_struct *vma, int type, int ret) ++{ ++ struct vx_info *vxi = mm->mm_vx_info; ++ int what; ++/* ++ static char *page_type[6] = ++ { "UNKNOWN", "ANON", "NONE", "FILE", "SWAP", "WRITE" }; ++ static char *page_what[4] = ++ { "FAULT_OOM", "FAULT_SIGBUS", "FAULT_MINOR", "FAULT_MAJOR" }; ++*/ ++ ++ if (!vxi) ++ return; ++ ++ what = (ret & 0x3); ++ ++/* printk("[%d] page[%d][%d] %2x %s %s\n", vxi->vx_id, ++ type, what, ret, page_type[type], page_what[what]); ++*/ ++ if (ret & VM_FAULT_WRITE) ++ what |= 0x4; ++ atomic_inc(&vxi->cacct.page[type][what]); ++} ++ ++#define vx_page_fault(mm, vma, type, ret) __vx_page_fault(mm, vma, type, ret) ++#endif ++ ++ ++extern unsigned long vx_badness(struct task_struct *task, struct mm_struct *mm); ++ ++#else ++#warning duplicate inclusion ++#endif +diff -NurpP --minimal linux-3.0.9/include/linux/vs_network.h linux-3.0.9-vs2.3.2.1/include/linux/vs_network.h +--- linux-3.0.9/include/linux/vs_network.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vs_network.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,169 @@ ++#ifndef _NX_VS_NETWORK_H ++#define _NX_VS_NETWORK_H ++ ++#include "vserver/context.h" ++#include "vserver/network.h" ++#include "vserver/base.h" ++#include "vserver/check.h" ++#include "vserver/debug.h" ++ ++#include <linux/sched.h> ++ ++ ++#define get_nx_info(i) __get_nx_info(i, __FILE__, __LINE__) ++ ++static inline struct nx_info *__get_nx_info(struct nx_info *nxi, ++ const char *_file, int _line) ++{ ++ if (!nxi) ++ return NULL; ++ ++ vxlprintk(VXD_CBIT(nid, 2), "get_nx_info(%p[#%d.%d])", ++ nxi, nxi ? nxi->nx_id : 0, ++ nxi ? atomic_read(&nxi->nx_usecnt) : 0, ++ _file, _line); ++ ++ atomic_inc(&nxi->nx_usecnt); ++ return nxi; ++} ++ ++ ++extern void free_nx_info(struct nx_info *); ++ ++#define put_nx_info(i) __put_nx_info(i, __FILE__, __LINE__) ++ ++static inline void __put_nx_info(struct nx_info *nxi, const char *_file, int _line) ++{ ++ if (!nxi) ++ return; ++ ++ vxlprintk(VXD_CBIT(nid, 2), "put_nx_info(%p[#%d.%d])", ++ nxi, nxi ? nxi->nx_id : 0, ++ nxi ? atomic_read(&nxi->nx_usecnt) : 0, ++ _file, _line); ++ ++ if (atomic_dec_and_test(&nxi->nx_usecnt)) ++ free_nx_info(nxi); ++} ++ ++ ++#define init_nx_info(p, i) __init_nx_info(p, i, __FILE__, __LINE__) ++ ++static inline void __init_nx_info(struct nx_info **nxp, struct nx_info *nxi, ++ const char *_file, int _line) ++{ ++ if (nxi) { ++ vxlprintk(VXD_CBIT(nid, 3), ++ "init_nx_info(%p[#%d.%d])", ++ nxi, nxi ? nxi->nx_id : 0, ++ nxi ? atomic_read(&nxi->nx_usecnt) : 0, ++ _file, _line); ++ ++ atomic_inc(&nxi->nx_usecnt); ++ } ++ *nxp = nxi; ++} ++ ++ ++#define set_nx_info(p, i) __set_nx_info(p, i, __FILE__, __LINE__) ++ ++static inline void __set_nx_info(struct nx_info **nxp, struct nx_info *nxi, ++ const char *_file, int _line) ++{ ++ struct nx_info *nxo; ++ ++ if (!nxi) ++ return; ++ ++ vxlprintk(VXD_CBIT(nid, 3), "set_nx_info(%p[#%d.%d])", ++ nxi, nxi ? nxi->nx_id : 0, ++ nxi ? atomic_read(&nxi->nx_usecnt) : 0, ++ _file, _line); ++ ++ atomic_inc(&nxi->nx_usecnt); ++ nxo = xchg(nxp, nxi); ++ BUG_ON(nxo); ++} ++ ++#define clr_nx_info(p) __clr_nx_info(p, __FILE__, __LINE__) ++ ++static inline void __clr_nx_info(struct nx_info **nxp, ++ const char *_file, int _line) ++{ ++ struct nx_info *nxo; ++ ++ nxo = xchg(nxp, NULL); ++ if (!nxo) ++ return; ++ ++ vxlprintk(VXD_CBIT(nid, 3), "clr_nx_info(%p[#%d.%d])", ++ nxo, nxo ? nxo->nx_id : 0, ++ nxo ? atomic_read(&nxo->nx_usecnt) : 0, ++ _file, _line); ++ ++ if (atomic_dec_and_test(&nxo->nx_usecnt)) ++ free_nx_info(nxo); ++} ++ ++ ++#define claim_nx_info(v, p) __claim_nx_info(v, p, __FILE__, __LINE__) ++ ++static inline void __claim_nx_info(struct nx_info *nxi, ++ struct task_struct *task, const char *_file, int _line) ++{ ++ vxlprintk(VXD_CBIT(nid, 3), "claim_nx_info(%p[#%d.%d.%d]) %p", ++ nxi, nxi ? nxi->nx_id : 0, ++ nxi?atomic_read(&nxi->nx_usecnt):0, ++ nxi?atomic_read(&nxi->nx_tasks):0, ++ task, _file, _line); ++ ++ atomic_inc(&nxi->nx_tasks); ++} ++ ++ ++extern void unhash_nx_info(struct nx_info *); ++ ++#define release_nx_info(v, p) __release_nx_info(v, p, __FILE__, __LINE__) ++ ++static inline void __release_nx_info(struct nx_info *nxi, ++ struct task_struct *task, const char *_file, int _line) ++{ ++ vxlprintk(VXD_CBIT(nid, 3), "release_nx_info(%p[#%d.%d.%d]) %p", ++ nxi, nxi ? nxi->nx_id : 0, ++ nxi ? atomic_read(&nxi->nx_usecnt) : 0, ++ nxi ? atomic_read(&nxi->nx_tasks) : 0, ++ task, _file, _line); ++ ++ might_sleep(); ++ ++ if (atomic_dec_and_test(&nxi->nx_tasks)) ++ unhash_nx_info(nxi); ++} ++ ++ ++#define task_get_nx_info(i) __task_get_nx_info(i, __FILE__, __LINE__) ++ ++static __inline__ struct nx_info *__task_get_nx_info(struct task_struct *p, ++ const char *_file, int _line) ++{ ++ struct nx_info *nxi; ++ ++ task_lock(p); ++ vxlprintk(VXD_CBIT(nid, 5), "task_get_nx_info(%p)", ++ p, _file, _line); ++ nxi = __get_nx_info(p->nx_info, _file, _line); ++ task_unlock(p); ++ return nxi; ++} ++ ++ ++static inline void exit_nx_info(struct task_struct *p) ++{ ++ if (p->nx_info) ++ release_nx_info(p->nx_info, p); ++} ++ ++ ++#else ++#warning duplicate inclusion ++#endif +diff -NurpP --minimal linux-3.0.9/include/linux/vs_pid.h linux-3.0.9-vs2.3.2.1/include/linux/vs_pid.h +--- linux-3.0.9/include/linux/vs_pid.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vs_pid.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,50 @@ ++#ifndef _VS_PID_H ++#define _VS_PID_H ++ ++#include "vserver/base.h" ++#include "vserver/check.h" ++#include "vserver/context.h" ++#include "vserver/debug.h" ++#include "vserver/pid.h" ++#include <linux/pid_namespace.h> ++ ++ ++#define VXF_FAKE_INIT (VXF_INFO_INIT | VXF_STATE_INIT) ++ ++static inline ++int vx_proc_task_visible(struct task_struct *task) ++{ ++ if ((task->pid == 1) && ++ !vx_flags(VXF_FAKE_INIT, VXF_FAKE_INIT)) ++ /* show a blend through init */ ++ goto visible; ++ if (vx_check(vx_task_xid(task), VS_WATCH | VS_IDENT)) ++ goto visible; ++ return 0; ++visible: ++ return 1; ++} ++ ++#define find_task_by_real_pid(pid) find_task_by_pid_ns(pid, &init_pid_ns) ++ ++ ++static inline ++struct task_struct *vx_get_proc_task(struct inode *inode, struct pid *pid) ++{ ++ struct task_struct *task = get_pid_task(pid, PIDTYPE_PID); ++ ++ if (task && !vx_proc_task_visible(task)) { ++ vxdprintk(VXD_CBIT(misc, 6), ++ "dropping task (get) %p[#%u,%u] for %p[#%u,%u]", ++ task, task->xid, task->pid, ++ current, current->xid, current->pid); ++ put_task_struct(task); ++ task = NULL; ++ } ++ return task; ++} ++ ++ ++#else ++#warning duplicate inclusion ++#endif +diff -NurpP --minimal linux-3.0.9/include/linux/vs_sched.h linux-3.0.9-vs2.3.2.1/include/linux/vs_sched.h +--- linux-3.0.9/include/linux/vs_sched.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vs_sched.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,40 @@ ++#ifndef _VS_SCHED_H ++#define _VS_SCHED_H ++ ++#include "vserver/base.h" ++#include "vserver/context.h" ++#include "vserver/sched.h" ++ ++ ++#define MAX_PRIO_BIAS 20 ++#define MIN_PRIO_BIAS -20 ++ ++static inline ++int vx_adjust_prio(struct task_struct *p, int prio, int max_user) ++{ ++ struct vx_info *vxi = p->vx_info; ++ ++ if (vxi) ++ prio += vx_cpu(vxi, sched_pc).prio_bias; ++ return prio; ++} ++ ++static inline void vx_account_user(struct vx_info *vxi, ++ cputime_t cputime, int nice) ++{ ++ if (!vxi) ++ return; ++ vx_cpu(vxi, sched_pc).user_ticks += cputime; ++} ++ ++static inline void vx_account_system(struct vx_info *vxi, ++ cputime_t cputime, int idle) ++{ ++ if (!vxi) ++ return; ++ vx_cpu(vxi, sched_pc).sys_ticks += cputime; ++} ++ ++#else ++#warning duplicate inclusion ++#endif +diff -NurpP --minimal linux-3.0.9/include/linux/vs_socket.h linux-3.0.9-vs2.3.2.1/include/linux/vs_socket.h +--- linux-3.0.9/include/linux/vs_socket.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vs_socket.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,67 @@ ++#ifndef _VS_SOCKET_H ++#define _VS_SOCKET_H ++ ++#include "vserver/debug.h" ++#include "vserver/base.h" ++#include "vserver/cacct.h" ++#include "vserver/context.h" ++#include "vserver/tag.h" ++ ++ ++/* socket accounting */ ++ ++#include <linux/socket.h> ++ ++static inline int vx_sock_type(int family) ++{ ++ switch (family) { ++ case PF_UNSPEC: ++ return VXA_SOCK_UNSPEC; ++ case PF_UNIX: ++ return VXA_SOCK_UNIX; ++ case PF_INET: ++ return VXA_SOCK_INET; ++ case PF_INET6: ++ return VXA_SOCK_INET6; ++ case PF_PACKET: ++ return VXA_SOCK_PACKET; ++ default: ++ return VXA_SOCK_OTHER; ++ } ++} ++ ++#define vx_acc_sock(v, f, p, s) \ ++ __vx_acc_sock(v, f, p, s, __FILE__, __LINE__) ++ ++static inline void __vx_acc_sock(struct vx_info *vxi, ++ int family, int pos, int size, char *file, int line) ++{ ++ if (vxi) { ++ int type = vx_sock_type(family); ++ ++ atomic_long_inc(&vxi->cacct.sock[type][pos].count); ++ atomic_long_add(size, &vxi->cacct.sock[type][pos].total); ++ } ++} ++ ++#define vx_sock_recv(sk, s) \ ++ vx_acc_sock((sk)->sk_vx_info, (sk)->sk_family, 0, s) ++#define vx_sock_send(sk, s) \ ++ vx_acc_sock((sk)->sk_vx_info, (sk)->sk_family, 1, s) ++#define vx_sock_fail(sk, s) \ ++ vx_acc_sock((sk)->sk_vx_info, (sk)->sk_family, 2, s) ++ ++ ++#define sock_vx_init(s) do { \ ++ (s)->sk_xid = 0; \ ++ (s)->sk_vx_info = NULL; \ ++ } while (0) ++ ++#define sock_nx_init(s) do { \ ++ (s)->sk_nid = 0; \ ++ (s)->sk_nx_info = NULL; \ ++ } while (0) ++ ++#else ++#warning duplicate inclusion ++#endif +diff -NurpP --minimal linux-3.0.9/include/linux/vs_tag.h linux-3.0.9-vs2.3.2.1/include/linux/vs_tag.h +--- linux-3.0.9/include/linux/vs_tag.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vs_tag.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,47 @@ ++#ifndef _VS_TAG_H ++#define _VS_TAG_H ++ ++#include <linux/vserver/tag.h> ++ ++/* check conditions */ ++ ++#define DX_ADMIN 0x0001 ++#define DX_WATCH 0x0002 ++#define DX_HOSTID 0x0008 ++ ++#define DX_IDENT 0x0010 ++ ++#define DX_ARG_MASK 0x0010 ++ ++ ++#define dx_task_tag(t) ((t)->tag) ++ ++#define dx_current_tag() dx_task_tag(current) ++ ++#define dx_check(c, m) __dx_check(dx_current_tag(), c, m) ++ ++#define dx_weak_check(c, m) ((m) ? dx_check(c, m) : 1) ++ ++ ++/* ++ * check current context for ADMIN/WATCH and ++ * optionally against supplied argument ++ */ ++static inline int __dx_check(tag_t cid, tag_t id, unsigned int mode) ++{ ++ if (mode & DX_ARG_MASK) { ++ if ((mode & DX_IDENT) && (id == cid)) ++ return 1; ++ } ++ return (((mode & DX_ADMIN) && (cid == 0)) || ++ ((mode & DX_WATCH) && (cid == 1)) || ++ ((mode & DX_HOSTID) && (id == 0))); ++} ++ ++struct inode; ++int dx_permission(const struct inode *inode, int mask); ++ ++ ++#else ++#warning duplicate inclusion ++#endif +diff -NurpP --minimal linux-3.0.9/include/linux/vs_time.h linux-3.0.9-vs2.3.2.1/include/linux/vs_time.h +--- linux-3.0.9/include/linux/vs_time.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vs_time.h 2011-06-13 14:57:45.000000000 +0200 +@@ -0,0 +1,19 @@ ++#ifndef _VS_TIME_H ++#define _VS_TIME_H ++ ++ ++/* time faking stuff */ ++ ++#ifdef CONFIG_VSERVER_VTIME ++ ++extern void vx_adjust_timespec(struct timespec *ts); ++extern int vx_settimeofday(const struct timespec *ts); ++ ++#else ++#define vx_adjust_timespec(t) do { } while (0) ++#define vx_settimeofday(t) do_settimeofday(t) ++#endif ++ ++#else ++#warning duplicate inclusion ++#endif +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/Kbuild linux-3.0.9-vs2.3.2.1/include/linux/vserver/Kbuild +--- linux-3.0.9/include/linux/vserver/Kbuild 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/Kbuild 2011-08-08 18:02:46.000000000 +0200 +@@ -0,0 +1,8 @@ ++ ++header-y += context_cmd.h network_cmd.h space_cmd.h \ ++ cacct_cmd.h cvirt_cmd.h limit_cmd.h dlimit_cmd.h \ ++ inode_cmd.h tag_cmd.h sched_cmd.h signal_cmd.h \ ++ debug_cmd.h device_cmd.h ++ ++header-y += switch.h network.h monitor.h inode.h device.h ++ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/base.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/base.h +--- linux-3.0.9/include/linux/vserver/base.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/base.h 2011-08-01 18:26:07.000000000 +0200 +@@ -0,0 +1,178 @@ ++#ifndef _VX_BASE_H ++#define _VX_BASE_H ++ ++ ++/* context state changes */ ++ ++enum { ++ VSC_STARTUP = 1, ++ VSC_SHUTDOWN, ++ ++ VSC_NETUP, ++ VSC_NETDOWN, ++}; ++ ++ ++ ++#define vx_task_xid(t) ((t)->xid) ++ ++#define vx_current_xid() vx_task_xid(current) ++ ++#define current_vx_info() (current->vx_info) ++ ++ ++#define nx_task_nid(t) ((t)->nid) ++ ++#define nx_current_nid() nx_task_nid(current) ++ ++#define current_nx_info() (current->nx_info) ++ ++ ++/* generic flag merging */ ++ ++#define vs_check_flags(v, m, f) (((v) & (m)) ^ (f)) ++ ++#define vs_mask_flags(v, f, m) (((v) & ~(m)) | ((f) & (m))) ++ ++#define vs_mask_mask(v, f, m) (((v) & ~(m)) | ((v) & (f) & (m))) ++ ++#define vs_check_bit(v, n) ((v) & (1LL << (n))) ++ ++ ++/* context flags */ ++ ++#define __vx_flags(v) ((v) ? (v)->vx_flags : 0) ++ ++#define vx_current_flags() __vx_flags(current_vx_info()) ++ ++#define vx_info_flags(v, m, f) \ ++ vs_check_flags(__vx_flags(v), m, f) ++ ++#define task_vx_flags(t, m, f) \ ++ ((t) && vx_info_flags((t)->vx_info, m, f)) ++ ++#define vx_flags(m, f) vx_info_flags(current_vx_info(), m, f) ++ ++ ++/* context caps */ ++ ++#define __vx_ccaps(v) ((v) ? (v)->vx_ccaps : 0) ++ ++#define vx_current_ccaps() __vx_ccaps(current_vx_info()) ++ ++#define vx_info_ccaps(v, c) (__vx_ccaps(v) & (c)) ++ ++#define vx_ccaps(c) vx_info_ccaps(current_vx_info(), (c)) ++ ++ ++ ++/* network flags */ ++ ++#define __nx_flags(n) ((n) ? (n)->nx_flags : 0) ++ ++#define nx_current_flags() __nx_flags(current_nx_info()) ++ ++#define nx_info_flags(n, m, f) \ ++ vs_check_flags(__nx_flags(n), m, f) ++ ++#define task_nx_flags(t, m, f) \ ++ ((t) && nx_info_flags((t)->nx_info, m, f)) ++ ++#define nx_flags(m, f) nx_info_flags(current_nx_info(), m, f) ++ ++ ++/* network caps */ ++ ++#define __nx_ncaps(n) ((n) ? (n)->nx_ncaps : 0) ++ ++#define nx_current_ncaps() __nx_ncaps(current_nx_info()) ++ ++#define nx_info_ncaps(n, c) (__nx_ncaps(n) & (c)) ++ ++#define nx_ncaps(c) nx_info_ncaps(current_nx_info(), c) ++ ++ ++/* context mask capabilities */ ++ ++#define __vx_mcaps(v) ((v) ? (v)->vx_ccaps >> 32UL : ~0 ) ++ ++#define vx_info_mcaps(v, c) (__vx_mcaps(v) & (c)) ++ ++#define vx_mcaps(c) vx_info_mcaps(current_vx_info(), c) ++ ++ ++/* context bcap mask */ ++ ++#define __vx_bcaps(v) ((v)->vx_bcaps) ++ ++#define vx_current_bcaps() __vx_bcaps(current_vx_info()) ++ ++ ++/* mask given bcaps */ ++ ++#define vx_info_mbcaps(v, c) ((v) ? cap_intersect(__vx_bcaps(v), c) : c) ++ ++#define vx_mbcaps(c) vx_info_mbcaps(current_vx_info(), c) ++ ++ ++/* masked cap_bset */ ++ ++#define vx_info_cap_bset(v) vx_info_mbcaps(v, current->cap_bset) ++ ++#define vx_current_cap_bset() vx_info_cap_bset(current_vx_info()) ++ ++#if 0 ++#define vx_info_mbcap(v, b) \ ++ (!vx_info_flags(v, VXF_STATE_SETUP, 0) ? \ ++ vx_info_bcaps(v, b) : (b)) ++ ++#define task_vx_mbcap(t, b) \ ++ vx_info_mbcap((t)->vx_info, (t)->b) ++ ++#define vx_mbcap(b) task_vx_mbcap(current, b) ++#endif ++ ++#define vx_cap_raised(v, c, f) cap_raised(vx_info_mbcaps(v, c), f) ++ ++#define vx_capable(b, c) (capable(b) || \ ++ (cap_raised(current_cap(), b) && vx_ccaps(c))) ++ ++#define vx_ns_capable(n, b, c) (ns_capable(n, b) || \ ++ (cap_raised(current_cap(), b) && vx_ccaps(c))) ++ ++#define nx_capable(b, c) (capable(b) || \ ++ (cap_raised(current_cap(), b) && nx_ncaps(c))) ++ ++#define vx_task_initpid(t, n) \ ++ ((t)->vx_info && \ ++ ((t)->vx_info->vx_initpid == (n))) ++ ++#define vx_current_initpid(n) vx_task_initpid(current, n) ++ ++ ++/* context unshare mask */ ++ ++#define __vx_umask(v) ((v)->vx_umask) ++ ++#define vx_current_umask() __vx_umask(current_vx_info()) ++ ++#define vx_can_unshare(b, f) (capable(b) || \ ++ (cap_raised(current_cap(), b) && \ ++ !((f) & ~vx_current_umask()))) ++ ++ ++#define __vx_wmask(v) ((v)->vx_wmask) ++ ++#define vx_current_wmask() __vx_wmask(current_vx_info()) ++ ++ ++#define __vx_state(v) ((v) ? ((v)->vx_state) : 0) ++ ++#define vx_info_state(v, m) (__vx_state(v) & (m)) ++ ++ ++#define __nx_state(n) ((n) ? ((n)->nx_state) : 0) ++ ++#define nx_info_state(n, m) (__nx_state(n) & (m)) ++ ++#endif +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/cacct.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/cacct.h +--- linux-3.0.9/include/linux/vserver/cacct.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/cacct.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,15 @@ ++#ifndef _VX_CACCT_H ++#define _VX_CACCT_H ++ ++ ++enum sock_acc_field { ++ VXA_SOCK_UNSPEC = 0, ++ VXA_SOCK_UNIX, ++ VXA_SOCK_INET, ++ VXA_SOCK_INET6, ++ VXA_SOCK_PACKET, ++ VXA_SOCK_OTHER, ++ VXA_SOCK_SIZE /* array size */ ++}; ++ ++#endif /* _VX_CACCT_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/cacct_cmd.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/cacct_cmd.h +--- linux-3.0.9/include/linux/vserver/cacct_cmd.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/cacct_cmd.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,23 @@ ++#ifndef _VX_CACCT_CMD_H ++#define _VX_CACCT_CMD_H ++ ++ ++/* virtual host info name commands */ ++ ++#define VCMD_sock_stat VC_CMD(VSTAT, 5, 0) ++ ++struct vcmd_sock_stat_v0 { ++ uint32_t field; ++ uint32_t count[3]; ++ uint64_t total[3]; ++}; ++ ++ ++#ifdef __KERNEL__ ++ ++#include <linux/compiler.h> ++ ++extern int vc_sock_stat(struct vx_info *, void __user *); ++ ++#endif /* __KERNEL__ */ ++#endif /* _VX_CACCT_CMD_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/cacct_def.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/cacct_def.h +--- linux-3.0.9/include/linux/vserver/cacct_def.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/cacct_def.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,43 @@ ++#ifndef _VX_CACCT_DEF_H ++#define _VX_CACCT_DEF_H ++ ++#include <asm/atomic.h> ++#include <linux/vserver/cacct.h> ++ ++ ++struct _vx_sock_acc { ++ atomic_long_t count; ++ atomic_long_t total; ++}; ++ ++/* context sub struct */ ++ ++struct _vx_cacct { ++ struct _vx_sock_acc sock[VXA_SOCK_SIZE][3]; ++ atomic_t slab[8]; ++ atomic_t page[6][8]; ++}; ++ ++#ifdef CONFIG_VSERVER_DEBUG ++ ++static inline void __dump_vx_cacct(struct _vx_cacct *cacct) ++{ ++ int i, j; ++ ++ printk("\t_vx_cacct:"); ++ for (i = 0; i < 6; i++) { ++ struct _vx_sock_acc *ptr = cacct->sock[i]; ++ ++ printk("\t [%d] =", i); ++ for (j = 0; j < 3; j++) { ++ printk(" [%d] = %8lu, %8lu", j, ++ atomic_long_read(&ptr[j].count), ++ atomic_long_read(&ptr[j].total)); ++ } ++ printk("\n"); ++ } ++} ++ ++#endif ++ ++#endif /* _VX_CACCT_DEF_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/cacct_int.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/cacct_int.h +--- linux-3.0.9/include/linux/vserver/cacct_int.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/cacct_int.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,21 @@ ++#ifndef _VX_CACCT_INT_H ++#define _VX_CACCT_INT_H ++ ++ ++#ifdef __KERNEL__ ++ ++static inline ++unsigned long vx_sock_count(struct _vx_cacct *cacct, int type, int pos) ++{ ++ return atomic_long_read(&cacct->sock[type][pos].count); ++} ++ ++ ++static inline ++unsigned long vx_sock_total(struct _vx_cacct *cacct, int type, int pos) ++{ ++ return atomic_long_read(&cacct->sock[type][pos].total); ++} ++ ++#endif /* __KERNEL__ */ ++#endif /* _VX_CACCT_INT_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/check.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/check.h +--- linux-3.0.9/include/linux/vserver/check.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/check.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,89 @@ ++#ifndef _VS_CHECK_H ++#define _VS_CHECK_H ++ ++ ++#define MAX_S_CONTEXT 65535 /* Arbitrary limit */ ++ ++#ifdef CONFIG_VSERVER_DYNAMIC_IDS ++#define MIN_D_CONTEXT 49152 /* dynamic contexts start here */ ++#else ++#define MIN_D_CONTEXT 65536 ++#endif ++ ++/* check conditions */ ++ ++#define VS_ADMIN 0x0001 ++#define VS_WATCH 0x0002 ++#define VS_HIDE 0x0004 ++#define VS_HOSTID 0x0008 ++ ++#define VS_IDENT 0x0010 ++#define VS_EQUIV 0x0020 ++#define VS_PARENT 0x0040 ++#define VS_CHILD 0x0080 ++ ++#define VS_ARG_MASK 0x00F0 ++ ++#define VS_DYNAMIC 0x0100 ++#define VS_STATIC 0x0200 ++ ++#define VS_ATR_MASK 0x0F00 ++ ++#ifdef CONFIG_VSERVER_PRIVACY ++#define VS_ADMIN_P (0) ++#define VS_WATCH_P (0) ++#else ++#define VS_ADMIN_P VS_ADMIN ++#define VS_WATCH_P VS_WATCH ++#endif ++ ++#define VS_HARDIRQ 0x1000 ++#define VS_SOFTIRQ 0x2000 ++#define VS_IRQ 0x4000 ++ ++#define VS_IRQ_MASK 0xF000 ++ ++#include <linux/hardirq.h> ++ ++/* ++ * check current context for ADMIN/WATCH and ++ * optionally against supplied argument ++ */ ++static inline int __vs_check(int cid, int id, unsigned int mode) ++{ ++ if (mode & VS_ARG_MASK) { ++ if ((mode & VS_IDENT) && (id == cid)) ++ return 1; ++ } ++ if (mode & VS_ATR_MASK) { ++ if ((mode & VS_DYNAMIC) && ++ (id >= MIN_D_CONTEXT) && ++ (id <= MAX_S_CONTEXT)) ++ return 1; ++ if ((mode & VS_STATIC) && ++ (id > 1) && (id < MIN_D_CONTEXT)) ++ return 1; ++ } ++ if (mode & VS_IRQ_MASK) { ++ if ((mode & VS_IRQ) && unlikely(in_interrupt())) ++ return 1; ++ if ((mode & VS_HARDIRQ) && unlikely(in_irq())) ++ return 1; ++ if ((mode & VS_SOFTIRQ) && unlikely(in_softirq())) ++ return 1; ++ } ++ return (((mode & VS_ADMIN) && (cid == 0)) || ++ ((mode & VS_WATCH) && (cid == 1)) || ++ ((mode & VS_HOSTID) && (id == 0))); ++} ++ ++#define vx_check(c, m) __vs_check(vx_current_xid(), c, (m) | VS_IRQ) ++ ++#define vx_weak_check(c, m) ((m) ? vx_check(c, m) : 1) ++ ++ ++#define nx_check(c, m) __vs_check(nx_current_nid(), c, m) ++ ++#define nx_weak_check(c, m) ((m) ? nx_check(c, m) : 1) ++ ++#endif +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/context.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/context.h +--- linux-3.0.9/include/linux/vserver/context.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/context.h 2011-10-27 13:59:29.000000000 +0200 +@@ -0,0 +1,188 @@ ++#ifndef _VX_CONTEXT_H ++#define _VX_CONTEXT_H ++ ++#include <linux/types.h> ++#include <linux/capability.h> ++ ++ ++/* context flags */ ++ ++#define VXF_INFO_SCHED 0x00000002 ++#define VXF_INFO_NPROC 0x00000004 ++#define VXF_INFO_PRIVATE 0x00000008 ++ ++#define VXF_INFO_INIT 0x00000010 ++#define VXF_INFO_HIDE 0x00000020 ++#define VXF_INFO_ULIMIT 0x00000040 ++#define VXF_INFO_NSPACE 0x00000080 ++ ++#define VXF_SCHED_HARD 0x00000100 ++#define VXF_SCHED_PRIO 0x00000200 ++#define VXF_SCHED_PAUSE 0x00000400 ++ ++#define VXF_VIRT_MEM 0x00010000 ++#define VXF_VIRT_UPTIME 0x00020000 ++#define VXF_VIRT_CPU 0x00040000 ++#define VXF_VIRT_LOAD 0x00080000 ++#define VXF_VIRT_TIME 0x00100000 ++ ++#define VXF_HIDE_MOUNT 0x01000000 ++/* was VXF_HIDE_NETIF 0x02000000 */ ++#define VXF_HIDE_VINFO 0x04000000 ++ ++#define VXF_STATE_SETUP (1ULL << 32) ++#define VXF_STATE_INIT (1ULL << 33) ++#define VXF_STATE_ADMIN (1ULL << 34) ++ ++#define VXF_SC_HELPER (1ULL << 36) ++#define VXF_REBOOT_KILL (1ULL << 37) ++#define VXF_PERSISTENT (1ULL << 38) ++ ++#define VXF_FORK_RSS (1ULL << 48) ++#define VXF_PROLIFIC (1ULL << 49) ++ ++#define VXF_IGNEG_NICE (1ULL << 52) ++ ++#define VXF_ONE_TIME (0x0007ULL << 32) ++ ++#define VXF_INIT_SET (VXF_STATE_SETUP | VXF_STATE_INIT | VXF_STATE_ADMIN) ++ ++ ++/* context migration */ ++ ++#define VXM_SET_INIT 0x00000001 ++#define VXM_SET_REAPER 0x00000002 ++ ++/* context caps */ ++ ++#define VXC_SET_UTSNAME 0x00000001 ++#define VXC_SET_RLIMIT 0x00000002 ++#define VXC_FS_SECURITY 0x00000004 ++#define VXC_FS_TRUSTED 0x00000008 ++#define VXC_TIOCSTI 0x00000010 ++ ++/* was VXC_RAW_ICMP 0x00000100 */ ++#define VXC_SYSLOG 0x00001000 ++#define VXC_OOM_ADJUST 0x00002000 ++#define VXC_AUDIT_CONTROL 0x00004000 ++ ++#define VXC_SECURE_MOUNT 0x00010000 ++#define VXC_SECURE_REMOUNT 0x00020000 ++#define VXC_BINARY_MOUNT 0x00040000 ++ ++#define VXC_QUOTA_CTL 0x00100000 ++#define VXC_ADMIN_MAPPER 0x00200000 ++#define VXC_ADMIN_CLOOP 0x00400000 ++ ++#define VXC_KTHREAD 0x01000000 ++#define VXC_NAMESPACE 0x02000000 ++ ++ ++#ifdef __KERNEL__ ++ ++#include <linux/list.h> ++#include <linux/spinlock.h> ++#include <linux/rcupdate.h> ++ ++#include "limit_def.h" ++#include "sched_def.h" ++#include "cvirt_def.h" ++#include "cacct_def.h" ++#include "device_def.h" ++ ++#define VX_SPACES 2 ++ ++struct _vx_info_pc { ++ struct _vx_sched_pc sched_pc; ++ struct _vx_cvirt_pc cvirt_pc; ++}; ++ ++struct _vx_space { ++ unsigned long vx_nsmask; /* assignment mask */ ++ struct nsproxy *vx_nsproxy; /* private namespaces */ ++ struct fs_struct *vx_fs; /* private namespace fs */ ++ const struct cred *vx_cred; /* task credentials */ ++}; ++ ++struct vx_info { ++ struct hlist_node vx_hlist; /* linked list of contexts */ ++ xid_t vx_id; /* context id */ ++ atomic_t vx_usecnt; /* usage count */ ++ atomic_t vx_tasks; /* tasks count */ ++ struct vx_info *vx_parent; /* parent context */ ++ int vx_state; /* context state */ ++ ++ struct _vx_space space[VX_SPACES]; /* namespace store */ ++ ++ uint64_t vx_flags; /* context flags */ ++ uint64_t vx_ccaps; /* context caps (vserver) */ ++ uint64_t vx_umask; /* unshare mask (guest) */ ++ uint64_t vx_wmask; /* warn mask (guest) */ ++ kernel_cap_t vx_bcaps; /* bounding caps (system) */ ++ ++ struct task_struct *vx_reaper; /* guest reaper process */ ++ pid_t vx_initpid; /* PID of guest init */ ++ int64_t vx_badness_bias; /* OOM points bias */ ++ ++ struct _vx_limit limit; /* vserver limits */ ++ struct _vx_sched sched; /* vserver scheduler */ ++ struct _vx_cvirt cvirt; /* virtual/bias stuff */ ++ struct _vx_cacct cacct; /* context accounting */ ++ ++ struct _vx_device dmap; /* default device map targets */ ++ ++#ifndef CONFIG_SMP ++ struct _vx_info_pc info_pc; /* per cpu data */ ++#else ++ struct _vx_info_pc *ptr_pc; /* per cpu array */ ++#endif ++ ++ wait_queue_head_t vx_wait; /* context exit waitqueue */ ++ int reboot_cmd; /* last sys_reboot() cmd */ ++ int exit_code; /* last process exit code */ ++ ++ char vx_name[65]; /* vserver name */ ++}; ++ ++#ifndef CONFIG_SMP ++#define vx_ptr_pc(vxi) (&(vxi)->info_pc) ++#define vx_per_cpu(vxi, v, id) vx_ptr_pc(vxi)->v ++#else ++#define vx_ptr_pc(vxi) ((vxi)->ptr_pc) ++#define vx_per_cpu(vxi, v, id) per_cpu_ptr(vx_ptr_pc(vxi), id)->v ++#endif ++ ++#define vx_cpu(vxi, v) vx_per_cpu(vxi, v, smp_processor_id()) ++ ++ ++struct vx_info_save { ++ struct vx_info *vxi; ++ xid_t xid; ++}; ++ ++ ++/* status flags */ ++ ++#define VXS_HASHED 0x0001 ++#define VXS_PAUSED 0x0010 ++#define VXS_SHUTDOWN 0x0100 ++#define VXS_HELPER 0x1000 ++#define VXS_RELEASED 0x8000 ++ ++ ++extern void claim_vx_info(struct vx_info *, struct task_struct *); ++extern void release_vx_info(struct vx_info *, struct task_struct *); ++ ++extern struct vx_info *lookup_vx_info(int); ++extern struct vx_info *lookup_or_create_vx_info(int); ++ ++extern int get_xid_list(int, unsigned int *, int); ++extern int xid_is_hashed(xid_t); ++ ++extern int vx_migrate_task(struct task_struct *, struct vx_info *, int); ++ ++extern long vs_state_change(struct vx_info *, unsigned int); ++ ++ ++#endif /* __KERNEL__ */ ++#endif /* _VX_CONTEXT_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/context_cmd.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/context_cmd.h +--- linux-3.0.9/include/linux/vserver/context_cmd.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/context_cmd.h 2011-08-01 18:25:07.000000000 +0200 +@@ -0,0 +1,162 @@ ++#ifndef _VX_CONTEXT_CMD_H ++#define _VX_CONTEXT_CMD_H ++ ++ ++/* vinfo commands */ ++ ++#define VCMD_task_xid VC_CMD(VINFO, 1, 0) ++ ++#ifdef __KERNEL__ ++extern int vc_task_xid(uint32_t); ++ ++#endif /* __KERNEL__ */ ++ ++#define VCMD_vx_info VC_CMD(VINFO, 5, 0) ++ ++struct vcmd_vx_info_v0 { ++ uint32_t xid; ++ uint32_t initpid; ++ /* more to come */ ++}; ++ ++#ifdef __KERNEL__ ++extern int vc_vx_info(struct vx_info *, void __user *); ++ ++#endif /* __KERNEL__ */ ++ ++#define VCMD_ctx_stat VC_CMD(VSTAT, 0, 0) ++ ++struct vcmd_ctx_stat_v0 { ++ uint32_t usecnt; ++ uint32_t tasks; ++ /* more to come */ ++}; ++ ++#ifdef __KERNEL__ ++extern int vc_ctx_stat(struct vx_info *, void __user *); ++ ++#endif /* __KERNEL__ */ ++ ++/* context commands */ ++ ++#define VCMD_ctx_create_v0 VC_CMD(VPROC, 1, 0) ++#define VCMD_ctx_create VC_CMD(VPROC, 1, 1) ++ ++struct vcmd_ctx_create { ++ uint64_t flagword; ++}; ++ ++#define VCMD_ctx_migrate_v0 VC_CMD(PROCMIG, 1, 0) ++#define VCMD_ctx_migrate VC_CMD(PROCMIG, 1, 1) ++ ++struct vcmd_ctx_migrate { ++ uint64_t flagword; ++}; ++ ++#ifdef __KERNEL__ ++extern int vc_ctx_create(uint32_t, void __user *); ++extern int vc_ctx_migrate(struct vx_info *, void __user *); ++ ++#endif /* __KERNEL__ */ ++ ++ ++/* flag commands */ ++ ++#define VCMD_get_cflags VC_CMD(FLAGS, 1, 0) ++#define VCMD_set_cflags VC_CMD(FLAGS, 2, 0) ++ ++struct vcmd_ctx_flags_v0 { ++ uint64_t flagword; ++ uint64_t mask; ++}; ++ ++#ifdef __KERNEL__ ++extern int vc_get_cflags(struct vx_info *, void __user *); ++extern int vc_set_cflags(struct vx_info *, void __user *); ++ ++#endif /* __KERNEL__ */ ++ ++ ++/* context caps commands */ ++ ++#define VCMD_get_ccaps VC_CMD(FLAGS, 3, 1) ++#define VCMD_set_ccaps VC_CMD(FLAGS, 4, 1) ++ ++struct vcmd_ctx_caps_v1 { ++ uint64_t ccaps; ++ uint64_t cmask; ++}; ++ ++#ifdef __KERNEL__ ++extern int vc_get_ccaps(struct vx_info *, void __user *); ++extern int vc_set_ccaps(struct vx_info *, void __user *); ++ ++#endif /* __KERNEL__ */ ++ ++ ++/* bcaps commands */ ++ ++#define VCMD_get_bcaps VC_CMD(FLAGS, 9, 0) ++#define VCMD_set_bcaps VC_CMD(FLAGS, 10, 0) ++ ++struct vcmd_bcaps { ++ uint64_t bcaps; ++ uint64_t bmask; ++}; ++ ++#ifdef __KERNEL__ ++extern int vc_get_bcaps(struct vx_info *, void __user *); ++extern int vc_set_bcaps(struct vx_info *, void __user *); ++ ++#endif /* __KERNEL__ */ ++ ++ ++/* umask commands */ ++ ++#define VCMD_get_umask VC_CMD(FLAGS, 13, 0) ++#define VCMD_set_umask VC_CMD(FLAGS, 14, 0) ++ ++struct vcmd_umask { ++ uint64_t umask; ++ uint64_t mask; ++}; ++ ++#ifdef __KERNEL__ ++extern int vc_get_umask(struct vx_info *, void __user *); ++extern int vc_set_umask(struct vx_info *, void __user *); ++ ++#endif /* __KERNEL__ */ ++ ++ ++/* wmask commands */ ++ ++#define VCMD_get_wmask VC_CMD(FLAGS, 15, 0) ++#define VCMD_set_wmask VC_CMD(FLAGS, 16, 0) ++ ++struct vcmd_wmask { ++ uint64_t wmask; ++ uint64_t mask; ++}; ++ ++#ifdef __KERNEL__ ++extern int vc_get_wmask(struct vx_info *, void __user *); ++extern int vc_set_wmask(struct vx_info *, void __user *); ++ ++#endif /* __KERNEL__ */ ++ ++ ++/* OOM badness */ ++ ++#define VCMD_get_badness VC_CMD(MEMCTRL, 5, 0) ++#define VCMD_set_badness VC_CMD(MEMCTRL, 6, 0) ++ ++struct vcmd_badness_v0 { ++ int64_t bias; ++}; ++ ++#ifdef __KERNEL__ ++extern int vc_get_badness(struct vx_info *, void __user *); ++extern int vc_set_badness(struct vx_info *, void __user *); ++ ++#endif /* __KERNEL__ */ ++#endif /* _VX_CONTEXT_CMD_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/cvirt.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/cvirt.h +--- linux-3.0.9/include/linux/vserver/cvirt.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/cvirt.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,20 @@ ++#ifndef _VX_CVIRT_H ++#define _VX_CVIRT_H ++ ++ ++#ifdef __KERNEL__ ++ ++struct timespec; ++ ++void vx_vsi_uptime(struct timespec *, struct timespec *); ++ ++ ++struct vx_info; ++ ++void vx_update_load(struct vx_info *); ++ ++ ++int vx_do_syslog(int, char __user *, int); ++ ++#endif /* __KERNEL__ */ ++#endif /* _VX_CVIRT_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/cvirt_cmd.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/cvirt_cmd.h +--- linux-3.0.9/include/linux/vserver/cvirt_cmd.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/cvirt_cmd.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,53 @@ ++#ifndef _VX_CVIRT_CMD_H ++#define _VX_CVIRT_CMD_H ++ ++ ++/* virtual host info name commands */ ++ ++#define VCMD_set_vhi_name VC_CMD(VHOST, 1, 0) ++#define VCMD_get_vhi_name VC_CMD(VHOST, 2, 0) ++ ++struct vcmd_vhi_name_v0 { ++ uint32_t field; ++ char name[65]; ++}; ++ ++ ++enum vhi_name_field { ++ VHIN_CONTEXT = 0, ++ VHIN_SYSNAME, ++ VHIN_NODENAME, ++ VHIN_RELEASE, ++ VHIN_VERSION, ++ VHIN_MACHINE, ++ VHIN_DOMAINNAME, ++}; ++ ++ ++#ifdef __KERNEL__ ++ ++#include <linux/compiler.h> ++ ++extern int vc_set_vhi_name(struct vx_info *, void __user *); ++extern int vc_get_vhi_name(struct vx_info *, void __user *); ++ ++#endif /* __KERNEL__ */ ++ ++#define VCMD_virt_stat VC_CMD(VSTAT, 3, 0) ++ ++struct vcmd_virt_stat_v0 { ++ uint64_t offset; ++ uint64_t uptime; ++ uint32_t nr_threads; ++ uint32_t nr_running; ++ uint32_t nr_uninterruptible; ++ uint32_t nr_onhold; ++ uint32_t nr_forks; ++ uint32_t load[3]; ++}; ++ ++#ifdef __KERNEL__ ++extern int vc_virt_stat(struct vx_info *, void __user *); ++ ++#endif /* __KERNEL__ */ ++#endif /* _VX_CVIRT_CMD_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/cvirt_def.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/cvirt_def.h +--- linux-3.0.9/include/linux/vserver/cvirt_def.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/cvirt_def.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,80 @@ ++#ifndef _VX_CVIRT_DEF_H ++#define _VX_CVIRT_DEF_H ++ ++#include <linux/jiffies.h> ++#include <linux/spinlock.h> ++#include <linux/wait.h> ++#include <linux/time.h> ++#include <asm/atomic.h> ++ ++ ++struct _vx_usage_stat { ++ uint64_t user; ++ uint64_t nice; ++ uint64_t system; ++ uint64_t softirq; ++ uint64_t irq; ++ uint64_t idle; ++ uint64_t iowait; ++}; ++ ++struct _vx_syslog { ++ wait_queue_head_t log_wait; ++ spinlock_t logbuf_lock; /* lock for the log buffer */ ++ ++ unsigned long log_start; /* next char to be read by syslog() */ ++ unsigned long con_start; /* next char to be sent to consoles */ ++ unsigned long log_end; /* most-recently-written-char + 1 */ ++ unsigned long logged_chars; /* #chars since last read+clear operation */ ++ ++ char log_buf[1024]; ++}; ++ ++ ++/* context sub struct */ ++ ++struct _vx_cvirt { ++ atomic_t nr_threads; /* number of current threads */ ++ atomic_t nr_running; /* number of running threads */ ++ atomic_t nr_uninterruptible; /* number of uninterruptible threads */ ++ ++ atomic_t nr_onhold; /* processes on hold */ ++ uint32_t onhold_last; /* jiffies when put on hold */ ++ ++ struct timespec bias_ts; /* time offset to the host */ ++ struct timespec bias_idle; ++ struct timespec bias_uptime; /* context creation point */ ++ uint64_t bias_clock; /* offset in clock_t */ ++ ++ spinlock_t load_lock; /* lock for the load averages */ ++ atomic_t load_updates; /* nr of load updates done so far */ ++ uint32_t load_last; /* last time load was calculated */ ++ uint32_t load[3]; /* load averages 1,5,15 */ ++ ++ atomic_t total_forks; /* number of forks so far */ ++ ++ struct _vx_syslog syslog; ++}; ++ ++struct _vx_cvirt_pc { ++ struct _vx_usage_stat cpustat; ++}; ++ ++ ++#ifdef CONFIG_VSERVER_DEBUG ++ ++static inline void __dump_vx_cvirt(struct _vx_cvirt *cvirt) ++{ ++ printk("\t_vx_cvirt:\n"); ++ printk("\t threads: %4d, %4d, %4d, %4d\n", ++ atomic_read(&cvirt->nr_threads), ++ atomic_read(&cvirt->nr_running), ++ atomic_read(&cvirt->nr_uninterruptible), ++ atomic_read(&cvirt->nr_onhold)); ++ /* add rest here */ ++ printk("\t total_forks = %d\n", atomic_read(&cvirt->total_forks)); ++} ++ ++#endif ++ ++#endif /* _VX_CVIRT_DEF_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/debug.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/debug.h +--- linux-3.0.9/include/linux/vserver/debug.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/debug.h 2011-11-17 21:25:04.000000000 +0100 +@@ -0,0 +1,145 @@ ++#ifndef _VX_DEBUG_H ++#define _VX_DEBUG_H ++ ++ ++#define VXD_CBIT(n, m) (vs_debug_ ## n & (1 << (m))) ++#define VXD_CMIN(n, m) (vs_debug_ ## n > (m)) ++#define VXD_MASK(n, m) (vs_debug_ ## n & (m)) ++ ++#define VXD_DEV(d) (d), (d)->bd_inode->i_ino, \ ++ imajor((d)->bd_inode), iminor((d)->bd_inode) ++#define VXF_DEV "%p[%lu,%d:%d]" ++ ++#if defined(CONFIG_QUOTES_UTF8) ++#define VS_Q_LQM "\xc2\xbb" ++#define VS_Q_RQM "\xc2\xab" ++#elif defined(CONFIG_QUOTES_ASCII) ++#define VS_Q_LQM "\x27" ++#define VS_Q_RQM "\x27" ++#else ++#define VS_Q_LQM "\xbb" ++#define VS_Q_RQM "\xab" ++#endif ++ ++#define VS_Q(f) VS_Q_LQM f VS_Q_RQM ++ ++ ++#define vxd_path(p) \ ++ ({ static char _buffer[PATH_MAX]; \ ++ d_path(p, _buffer, sizeof(_buffer)); }) ++ ++#define vxd_cond_path(n) \ ++ ((n) ? vxd_path(&(n)->path) : "<null>" ) ++ ++ ++#ifdef CONFIG_VSERVER_DEBUG ++ ++extern unsigned int vs_debug_switch; ++extern unsigned int vs_debug_xid; ++extern unsigned int vs_debug_nid; ++extern unsigned int vs_debug_tag; ++extern unsigned int vs_debug_net; ++extern unsigned int vs_debug_limit; ++extern unsigned int vs_debug_cres; ++extern unsigned int vs_debug_dlim; ++extern unsigned int vs_debug_quota; ++extern unsigned int vs_debug_cvirt; ++extern unsigned int vs_debug_space; ++extern unsigned int vs_debug_perm; ++extern unsigned int vs_debug_misc; ++ ++ ++#define VX_LOGLEVEL "vxD: " ++#define VX_PROC_FMT "%p: " ++#define VX_PROCESS current ++ ++#define vxdprintk(c, f, x...) \ ++ do { \ ++ if (c) \ ++ printk(VX_LOGLEVEL VX_PROC_FMT f "\n", \ ++ VX_PROCESS , ##x); \ ++ } while (0) ++ ++#define vxlprintk(c, f, x...) \ ++ do { \ ++ if (c) \ ++ printk(VX_LOGLEVEL f " @%s:%d\n", x); \ ++ } while (0) ++ ++#define vxfprintk(c, f, x...) \ ++ do { \ ++ if (c) \ ++ printk(VX_LOGLEVEL f " %s@%s:%d\n", x); \ ++ } while (0) ++ ++ ++struct vx_info; ++ ++void dump_vx_info(struct vx_info *, int); ++void dump_vx_info_inactive(int); ++ ++#else /* CONFIG_VSERVER_DEBUG */ ++ ++#define vs_debug_switch 0 ++#define vs_debug_xid 0 ++#define vs_debug_nid 0 ++#define vs_debug_tag 0 ++#define vs_debug_net 0 ++#define vs_debug_limit 0 ++#define vs_debug_cres 0 ++#define vs_debug_dlim 0 ++#define vs_debug_quota 0 ++#define vs_debug_cvirt 0 ++#define vs_debug_space 0 ++#define vs_debug_perm 0 ++#define vs_debug_misc 0 ++ ++#define vxdprintk(x...) do { } while (0) ++#define vxlprintk(x...) do { } while (0) ++#define vxfprintk(x...) do { } while (0) ++ ++#endif /* CONFIG_VSERVER_DEBUG */ ++ ++ ++#ifdef CONFIG_VSERVER_WARN ++ ++#define VX_WARNLEVEL KERN_WARNING "vxW: " ++#define VX_WARN_TASK "[" VS_Q("%s") ",%u:#%u|%u|%u] " ++#define VX_WARN_XID "[xid #%u] " ++#define VX_WARN_NID "[nid #%u] " ++#define VX_WARN_TAG "[tag #%u] " ++ ++#define vxwprintk(c, f, x...) \ ++ do { \ ++ if (c) \ ++ printk(VX_WARNLEVEL f "\n", ##x); \ ++ } while (0) ++ ++#else /* CONFIG_VSERVER_WARN */ ++ ++#define vxwprintk(x...) do { } while (0) ++ ++#endif /* CONFIG_VSERVER_WARN */ ++ ++#define vxwprintk_task(c, f, x...) \ ++ vxwprintk(c, VX_WARN_TASK f, \ ++ current->comm, current->pid, \ ++ current->xid, current->nid, current->tag, ##x) ++#define vxwprintk_xid(c, f, x...) \ ++ vxwprintk(c, VX_WARN_XID f, current->xid, x) ++#define vxwprintk_nid(c, f, x...) \ ++ vxwprintk(c, VX_WARN_NID f, current->nid, x) ++#define vxwprintk_tag(c, f, x...) \ ++ vxwprintk(c, VX_WARN_TAG f, current->tag, x) ++ ++#ifdef CONFIG_VSERVER_DEBUG ++#define vxd_assert_lock(l) assert_spin_locked(l) ++#define vxd_assert(c, f, x...) vxlprintk(!(c), \ ++ "assertion [" f "] failed.", ##x, __FILE__, __LINE__) ++#else ++#define vxd_assert_lock(l) do { } while (0) ++#define vxd_assert(c, f, x...) do { } while (0) ++#endif ++ ++ ++#endif /* _VX_DEBUG_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/debug_cmd.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/debug_cmd.h +--- linux-3.0.9/include/linux/vserver/debug_cmd.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/debug_cmd.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,58 @@ ++#ifndef _VX_DEBUG_CMD_H ++#define _VX_DEBUG_CMD_H ++ ++ ++/* debug commands */ ++ ++#define VCMD_dump_history VC_CMD(DEBUG, 1, 0) ++ ++#define VCMD_read_history VC_CMD(DEBUG, 5, 0) ++#define VCMD_read_monitor VC_CMD(DEBUG, 6, 0) ++ ++struct vcmd_read_history_v0 { ++ uint32_t index; ++ uint32_t count; ++ char __user *data; ++}; ++ ++struct vcmd_read_monitor_v0 { ++ uint32_t index; ++ uint32_t count; ++ char __user *data; ++}; ++ ++ ++#ifdef __KERNEL__ ++ ++#ifdef CONFIG_COMPAT ++ ++#include <asm/compat.h> ++ ++struct vcmd_read_history_v0_x32 { ++ uint32_t index; ++ uint32_t count; ++ compat_uptr_t data_ptr; ++}; ++ ++struct vcmd_read_monitor_v0_x32 { ++ uint32_t index; ++ uint32_t count; ++ compat_uptr_t data_ptr; ++}; ++ ++#endif /* CONFIG_COMPAT */ ++ ++extern int vc_dump_history(uint32_t); ++ ++extern int vc_read_history(uint32_t, void __user *); ++extern int vc_read_monitor(uint32_t, void __user *); ++ ++#ifdef CONFIG_COMPAT ++ ++extern int vc_read_history_x32(uint32_t, void __user *); ++extern int vc_read_monitor_x32(uint32_t, void __user *); ++ ++#endif /* CONFIG_COMPAT */ ++ ++#endif /* __KERNEL__ */ ++#endif /* _VX_DEBUG_CMD_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/device.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/device.h +--- linux-3.0.9/include/linux/vserver/device.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/device.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,15 @@ ++#ifndef _VX_DEVICE_H ++#define _VX_DEVICE_H ++ ++ ++#define DATTR_CREATE 0x00000001 ++#define DATTR_OPEN 0x00000002 ++ ++#define DATTR_REMAP 0x00000010 ++ ++#define DATTR_MASK 0x00000013 ++ ++ ++#else /* _VX_DEVICE_H */ ++#warning duplicate inclusion ++#endif /* _VX_DEVICE_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/device_cmd.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/device_cmd.h +--- linux-3.0.9/include/linux/vserver/device_cmd.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/device_cmd.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,44 @@ ++#ifndef _VX_DEVICE_CMD_H ++#define _VX_DEVICE_CMD_H ++ ++ ++/* device vserver commands */ ++ ++#define VCMD_set_mapping VC_CMD(DEVICE, 1, 0) ++#define VCMD_unset_mapping VC_CMD(DEVICE, 2, 0) ++ ++struct vcmd_set_mapping_v0 { ++ const char __user *device; ++ const char __user *target; ++ uint32_t flags; ++}; ++ ++ ++#ifdef __KERNEL__ ++ ++#ifdef CONFIG_COMPAT ++ ++#include <asm/compat.h> ++ ++struct vcmd_set_mapping_v0_x32 { ++ compat_uptr_t device_ptr; ++ compat_uptr_t target_ptr; ++ uint32_t flags; ++}; ++ ++#endif /* CONFIG_COMPAT */ ++ ++#include <linux/compiler.h> ++ ++extern int vc_set_mapping(struct vx_info *, void __user *); ++extern int vc_unset_mapping(struct vx_info *, void __user *); ++ ++#ifdef CONFIG_COMPAT ++ ++extern int vc_set_mapping_x32(struct vx_info *, void __user *); ++extern int vc_unset_mapping_x32(struct vx_info *, void __user *); ++ ++#endif /* CONFIG_COMPAT */ ++ ++#endif /* __KERNEL__ */ ++#endif /* _VX_DEVICE_CMD_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/device_def.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/device_def.h +--- linux-3.0.9/include/linux/vserver/device_def.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/device_def.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,17 @@ ++#ifndef _VX_DEVICE_DEF_H ++#define _VX_DEVICE_DEF_H ++ ++#include <linux/types.h> ++ ++struct vx_dmap_target { ++ dev_t target; ++ uint32_t flags; ++}; ++ ++struct _vx_device { ++#ifdef CONFIG_VSERVER_DEVICE ++ struct vx_dmap_target targets[2]; ++#endif ++}; ++ ++#endif /* _VX_DEVICE_DEF_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/dlimit.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/dlimit.h +--- linux-3.0.9/include/linux/vserver/dlimit.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/dlimit.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,54 @@ ++#ifndef _VX_DLIMIT_H ++#define _VX_DLIMIT_H ++ ++#include "switch.h" ++ ++ ++#ifdef __KERNEL__ ++ ++/* keep in sync with CDLIM_INFINITY */ ++ ++#define DLIM_INFINITY (~0ULL) ++ ++#include <linux/spinlock.h> ++#include <linux/rcupdate.h> ++ ++struct super_block; ++ ++struct dl_info { ++ struct hlist_node dl_hlist; /* linked list of contexts */ ++ struct rcu_head dl_rcu; /* the rcu head */ ++ tag_t dl_tag; /* context tag */ ++ atomic_t dl_usecnt; /* usage count */ ++ atomic_t dl_refcnt; /* reference count */ ++ ++ struct super_block *dl_sb; /* associated superblock */ ++ ++ spinlock_t dl_lock; /* protect the values */ ++ ++ unsigned long long dl_space_used; /* used space in bytes */ ++ unsigned long long dl_space_total; /* maximum space in bytes */ ++ unsigned long dl_inodes_used; /* used inodes */ ++ unsigned long dl_inodes_total; /* maximum inodes */ ++ ++ unsigned int dl_nrlmult; /* non root limit mult */ ++}; ++ ++struct rcu_head; ++ ++extern void rcu_free_dl_info(struct rcu_head *); ++extern void unhash_dl_info(struct dl_info *); ++ ++extern struct dl_info *locate_dl_info(struct super_block *, tag_t); ++ ++ ++struct kstatfs; ++ ++extern void vx_vsi_statfs(struct super_block *, struct kstatfs *); ++ ++typedef uint64_t dlsize_t; ++ ++#endif /* __KERNEL__ */ ++#else /* _VX_DLIMIT_H */ ++#warning duplicate inclusion ++#endif /* _VX_DLIMIT_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/dlimit_cmd.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/dlimit_cmd.h +--- linux-3.0.9/include/linux/vserver/dlimit_cmd.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/dlimit_cmd.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,109 @@ ++#ifndef _VX_DLIMIT_CMD_H ++#define _VX_DLIMIT_CMD_H ++ ++ ++/* dlimit vserver commands */ ++ ++#define VCMD_add_dlimit VC_CMD(DLIMIT, 1, 0) ++#define VCMD_rem_dlimit VC_CMD(DLIMIT, 2, 0) ++ ++#define VCMD_set_dlimit VC_CMD(DLIMIT, 5, 0) ++#define VCMD_get_dlimit VC_CMD(DLIMIT, 6, 0) ++ ++struct vcmd_ctx_dlimit_base_v0 { ++ const char __user *name; ++ uint32_t flags; ++}; ++ ++struct vcmd_ctx_dlimit_v0 { ++ const char __user *name; ++ uint32_t space_used; /* used space in kbytes */ ++ uint32_t space_total; /* maximum space in kbytes */ ++ uint32_t inodes_used; /* used inodes */ ++ uint32_t inodes_total; /* maximum inodes */ ++ uint32_t reserved; /* reserved for root in % */ ++ uint32_t flags; ++}; ++ ++#define CDLIM_UNSET ((uint32_t)0UL) ++#define CDLIM_INFINITY ((uint32_t)~0UL) ++#define CDLIM_KEEP ((uint32_t)~1UL) ++ ++#define DLIME_UNIT 0 ++#define DLIME_KILO 1 ++#define DLIME_MEGA 2 ++#define DLIME_GIGA 3 ++ ++#define DLIMF_SHIFT 0x10 ++ ++#define DLIMS_USED 0 ++#define DLIMS_TOTAL 2 ++ ++static inline ++uint64_t dlimit_space_32to64(uint32_t val, uint32_t flags, int shift) ++{ ++ int exp = (flags & DLIMF_SHIFT) ? ++ (flags >> shift) & DLIME_GIGA : DLIME_KILO; ++ return ((uint64_t)val) << (10 * exp); ++} ++ ++static inline ++uint32_t dlimit_space_64to32(uint64_t val, uint32_t *flags, int shift) ++{ ++ int exp = 0; ++ ++ if (*flags & DLIMF_SHIFT) { ++ while (val > (1LL << 32) && (exp < 3)) { ++ val >>= 10; ++ exp++; ++ } ++ *flags &= ~(DLIME_GIGA << shift); ++ *flags |= exp << shift; ++ } else ++ val >>= 10; ++ return val; ++} ++ ++#ifdef __KERNEL__ ++ ++#ifdef CONFIG_COMPAT ++ ++#include <asm/compat.h> ++ ++struct vcmd_ctx_dlimit_base_v0_x32 { ++ compat_uptr_t name_ptr; ++ uint32_t flags; ++}; ++ ++struct vcmd_ctx_dlimit_v0_x32 { ++ compat_uptr_t name_ptr; ++ uint32_t space_used; /* used space in kbytes */ ++ uint32_t space_total; /* maximum space in kbytes */ ++ uint32_t inodes_used; /* used inodes */ ++ uint32_t inodes_total; /* maximum inodes */ ++ uint32_t reserved; /* reserved for root in % */ ++ uint32_t flags; ++}; ++ ++#endif /* CONFIG_COMPAT */ ++ ++#include <linux/compiler.h> ++ ++extern int vc_add_dlimit(uint32_t, void __user *); ++extern int vc_rem_dlimit(uint32_t, void __user *); ++ ++extern int vc_set_dlimit(uint32_t, void __user *); ++extern int vc_get_dlimit(uint32_t, void __user *); ++ ++#ifdef CONFIG_COMPAT ++ ++extern int vc_add_dlimit_x32(uint32_t, void __user *); ++extern int vc_rem_dlimit_x32(uint32_t, void __user *); ++ ++extern int vc_set_dlimit_x32(uint32_t, void __user *); ++extern int vc_get_dlimit_x32(uint32_t, void __user *); ++ ++#endif /* CONFIG_COMPAT */ ++ ++#endif /* __KERNEL__ */ ++#endif /* _VX_DLIMIT_CMD_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/global.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/global.h +--- linux-3.0.9/include/linux/vserver/global.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/global.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,19 @@ ++#ifndef _VX_GLOBAL_H ++#define _VX_GLOBAL_H ++ ++ ++extern atomic_t vx_global_ctotal; ++extern atomic_t vx_global_cactive; ++ ++extern atomic_t nx_global_ctotal; ++extern atomic_t nx_global_cactive; ++ ++extern atomic_t vs_global_nsproxy; ++extern atomic_t vs_global_fs; ++extern atomic_t vs_global_mnt_ns; ++extern atomic_t vs_global_uts_ns; ++extern atomic_t vs_global_user_ns; ++extern atomic_t vs_global_pid_ns; ++ ++ ++#endif /* _VX_GLOBAL_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/history.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/history.h +--- linux-3.0.9/include/linux/vserver/history.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/history.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,197 @@ ++#ifndef _VX_HISTORY_H ++#define _VX_HISTORY_H ++ ++ ++enum { ++ VXH_UNUSED = 0, ++ VXH_THROW_OOPS = 1, ++ ++ VXH_GET_VX_INFO, ++ VXH_PUT_VX_INFO, ++ VXH_INIT_VX_INFO, ++ VXH_SET_VX_INFO, ++ VXH_CLR_VX_INFO, ++ VXH_CLAIM_VX_INFO, ++ VXH_RELEASE_VX_INFO, ++ VXH_ALLOC_VX_INFO, ++ VXH_DEALLOC_VX_INFO, ++ VXH_HASH_VX_INFO, ++ VXH_UNHASH_VX_INFO, ++ VXH_LOC_VX_INFO, ++ VXH_LOOKUP_VX_INFO, ++ VXH_CREATE_VX_INFO, ++}; ++ ++struct _vxhe_vxi { ++ struct vx_info *ptr; ++ unsigned xid; ++ unsigned usecnt; ++ unsigned tasks; ++}; ++ ++struct _vxhe_set_clr { ++ void *data; ++}; ++ ++struct _vxhe_loc_lookup { ++ unsigned arg; ++}; ++ ++struct _vx_hist_entry { ++ void *loc; ++ unsigned short seq; ++ unsigned short type; ++ struct _vxhe_vxi vxi; ++ union { ++ struct _vxhe_set_clr sc; ++ struct _vxhe_loc_lookup ll; ++ }; ++}; ++ ++#ifdef CONFIG_VSERVER_HISTORY ++ ++extern unsigned volatile int vxh_active; ++ ++struct _vx_hist_entry *vxh_advance(void *loc); ++ ++ ++static inline ++void __vxh_copy_vxi(struct _vx_hist_entry *entry, struct vx_info *vxi) ++{ ++ entry->vxi.ptr = vxi; ++ if (vxi) { ++ entry->vxi.usecnt = atomic_read(&vxi->vx_usecnt); ++ entry->vxi.tasks = atomic_read(&vxi->vx_tasks); ++ entry->vxi.xid = vxi->vx_id; ++ } ++} ++ ++ ++#define __HERE__ current_text_addr() ++ ++#define __VXH_BODY(__type, __data, __here) \ ++ struct _vx_hist_entry *entry; \ ++ \ ++ preempt_disable(); \ ++ entry = vxh_advance(__here); \ ++ __data; \ ++ entry->type = __type; \ ++ preempt_enable(); ++ ++ ++ /* pass vxi only */ ++ ++#define __VXH_SMPL \ ++ __vxh_copy_vxi(entry, vxi) ++ ++static inline ++void __vxh_smpl(struct vx_info *vxi, int __type, void *__here) ++{ ++ __VXH_BODY(__type, __VXH_SMPL, __here) ++} ++ ++ /* pass vxi and data (void *) */ ++ ++#define __VXH_DATA \ ++ __vxh_copy_vxi(entry, vxi); \ ++ entry->sc.data = data ++ ++static inline ++void __vxh_data(struct vx_info *vxi, void *data, ++ int __type, void *__here) ++{ ++ __VXH_BODY(__type, __VXH_DATA, __here) ++} ++ ++ /* pass vxi and arg (long) */ ++ ++#define __VXH_LONG \ ++ __vxh_copy_vxi(entry, vxi); \ ++ entry->ll.arg = arg ++ ++static inline ++void __vxh_long(struct vx_info *vxi, long arg, ++ int __type, void *__here) ++{ ++ __VXH_BODY(__type, __VXH_LONG, __here) ++} ++ ++ ++static inline ++void __vxh_throw_oops(void *__here) ++{ ++ __VXH_BODY(VXH_THROW_OOPS, {}, __here); ++ /* prevent further acquisition */ ++ vxh_active = 0; ++} ++ ++ ++#define vxh_throw_oops() __vxh_throw_oops(__HERE__); ++ ++#define __vxh_get_vx_info(v, h) __vxh_smpl(v, VXH_GET_VX_INFO, h); ++#define __vxh_put_vx_info(v, h) __vxh_smpl(v, VXH_PUT_VX_INFO, h); ++ ++#define __vxh_init_vx_info(v, d, h) \ ++ __vxh_data(v, d, VXH_INIT_VX_INFO, h); ++#define __vxh_set_vx_info(v, d, h) \ ++ __vxh_data(v, d, VXH_SET_VX_INFO, h); ++#define __vxh_clr_vx_info(v, d, h) \ ++ __vxh_data(v, d, VXH_CLR_VX_INFO, h); ++ ++#define __vxh_claim_vx_info(v, d, h) \ ++ __vxh_data(v, d, VXH_CLAIM_VX_INFO, h); ++#define __vxh_release_vx_info(v, d, h) \ ++ __vxh_data(v, d, VXH_RELEASE_VX_INFO, h); ++ ++#define vxh_alloc_vx_info(v) \ ++ __vxh_smpl(v, VXH_ALLOC_VX_INFO, __HERE__); ++#define vxh_dealloc_vx_info(v) \ ++ __vxh_smpl(v, VXH_DEALLOC_VX_INFO, __HERE__); ++ ++#define vxh_hash_vx_info(v) \ ++ __vxh_smpl(v, VXH_HASH_VX_INFO, __HERE__); ++#define vxh_unhash_vx_info(v) \ ++ __vxh_smpl(v, VXH_UNHASH_VX_INFO, __HERE__); ++ ++#define vxh_loc_vx_info(v, l) \ ++ __vxh_long(v, l, VXH_LOC_VX_INFO, __HERE__); ++#define vxh_lookup_vx_info(v, l) \ ++ __vxh_long(v, l, VXH_LOOKUP_VX_INFO, __HERE__); ++#define vxh_create_vx_info(v, l) \ ++ __vxh_long(v, l, VXH_CREATE_VX_INFO, __HERE__); ++ ++extern void vxh_dump_history(void); ++ ++ ++#else /* CONFIG_VSERVER_HISTORY */ ++ ++#define __HERE__ 0 ++ ++#define vxh_throw_oops() do { } while (0) ++ ++#define __vxh_get_vx_info(v, h) do { } while (0) ++#define __vxh_put_vx_info(v, h) do { } while (0) ++ ++#define __vxh_init_vx_info(v, d, h) do { } while (0) ++#define __vxh_set_vx_info(v, d, h) do { } while (0) ++#define __vxh_clr_vx_info(v, d, h) do { } while (0) ++ ++#define __vxh_claim_vx_info(v, d, h) do { } while (0) ++#define __vxh_release_vx_info(v, d, h) do { } while (0) ++ ++#define vxh_alloc_vx_info(v) do { } while (0) ++#define vxh_dealloc_vx_info(v) do { } while (0) ++ ++#define vxh_hash_vx_info(v) do { } while (0) ++#define vxh_unhash_vx_info(v) do { } while (0) ++ ++#define vxh_loc_vx_info(v, l) do { } while (0) ++#define vxh_lookup_vx_info(v, l) do { } while (0) ++#define vxh_create_vx_info(v, l) do { } while (0) ++ ++#define vxh_dump_history() do { } while (0) ++ ++ ++#endif /* CONFIG_VSERVER_HISTORY */ ++ ++#endif /* _VX_HISTORY_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/inode.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/inode.h +--- linux-3.0.9/include/linux/vserver/inode.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/inode.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,39 @@ ++#ifndef _VX_INODE_H ++#define _VX_INODE_H ++ ++ ++#define IATTR_TAG 0x01000000 ++ ++#define IATTR_ADMIN 0x00000001 ++#define IATTR_WATCH 0x00000002 ++#define IATTR_HIDE 0x00000004 ++#define IATTR_FLAGS 0x00000007 ++ ++#define IATTR_BARRIER 0x00010000 ++#define IATTR_IXUNLINK 0x00020000 ++#define IATTR_IMMUTABLE 0x00040000 ++#define IATTR_COW 0x00080000 ++ ++#ifdef __KERNEL__ ++ ++ ++#ifdef CONFIG_VSERVER_PROC_SECURE ++#define IATTR_PROC_DEFAULT ( IATTR_ADMIN | IATTR_HIDE ) ++#define IATTR_PROC_SYMLINK ( IATTR_ADMIN ) ++#else ++#define IATTR_PROC_DEFAULT ( IATTR_ADMIN ) ++#define IATTR_PROC_SYMLINK ( IATTR_ADMIN ) ++#endif ++ ++#define vx_hide_check(c, m) (((m) & IATTR_HIDE) ? vx_check(c, m) : 1) ++ ++#endif /* __KERNEL__ */ ++ ++/* inode ioctls */ ++ ++#define FIOC_GETXFLG _IOR('x', 5, long) ++#define FIOC_SETXFLG _IOW('x', 6, long) ++ ++#else /* _VX_INODE_H */ ++#warning duplicate inclusion ++#endif /* _VX_INODE_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/inode_cmd.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/inode_cmd.h +--- linux-3.0.9/include/linux/vserver/inode_cmd.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/inode_cmd.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,59 @@ ++#ifndef _VX_INODE_CMD_H ++#define _VX_INODE_CMD_H ++ ++ ++/* inode vserver commands */ ++ ++#define VCMD_get_iattr VC_CMD(INODE, 1, 1) ++#define VCMD_set_iattr VC_CMD(INODE, 2, 1) ++ ++#define VCMD_fget_iattr VC_CMD(INODE, 3, 0) ++#define VCMD_fset_iattr VC_CMD(INODE, 4, 0) ++ ++struct vcmd_ctx_iattr_v1 { ++ const char __user *name; ++ uint32_t tag; ++ uint32_t flags; ++ uint32_t mask; ++}; ++ ++struct vcmd_ctx_fiattr_v0 { ++ uint32_t tag; ++ uint32_t flags; ++ uint32_t mask; ++}; ++ ++ ++#ifdef __KERNEL__ ++ ++ ++#ifdef CONFIG_COMPAT ++ ++#include <asm/compat.h> ++ ++struct vcmd_ctx_iattr_v1_x32 { ++ compat_uptr_t name_ptr; ++ uint32_t tag; ++ uint32_t flags; ++ uint32_t mask; ++}; ++ ++#endif /* CONFIG_COMPAT */ ++ ++#include <linux/compiler.h> ++ ++extern int vc_get_iattr(void __user *); ++extern int vc_set_iattr(void __user *); ++ ++extern int vc_fget_iattr(uint32_t, void __user *); ++extern int vc_fset_iattr(uint32_t, void __user *); ++ ++#ifdef CONFIG_COMPAT ++ ++extern int vc_get_iattr_x32(void __user *); ++extern int vc_set_iattr_x32(void __user *); ++ ++#endif /* CONFIG_COMPAT */ ++ ++#endif /* __KERNEL__ */ ++#endif /* _VX_INODE_CMD_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/limit.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/limit.h +--- linux-3.0.9/include/linux/vserver/limit.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/limit.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,71 @@ ++#ifndef _VX_LIMIT_H ++#define _VX_LIMIT_H ++ ++#define VLIMIT_NSOCK 16 ++#define VLIMIT_OPENFD 17 ++#define VLIMIT_ANON 18 ++#define VLIMIT_SHMEM 19 ++#define VLIMIT_SEMARY 20 ++#define VLIMIT_NSEMS 21 ++#define VLIMIT_DENTRY 22 ++#define VLIMIT_MAPPED 23 ++ ++ ++#ifdef __KERNEL__ ++ ++#define VLIM_NOCHECK ((1L << VLIMIT_DENTRY) | (1L << RLIMIT_RSS)) ++ ++/* keep in sync with CRLIM_INFINITY */ ++ ++#define VLIM_INFINITY (~0ULL) ++ ++#include <asm/atomic.h> ++#include <asm/resource.h> ++ ++#ifndef RLIM_INFINITY ++#warning RLIM_INFINITY is undefined ++#endif ++ ++#define __rlim_val(l, r, v) ((l)->res[r].v) ++ ++#define __rlim_soft(l, r) __rlim_val(l, r, soft) ++#define __rlim_hard(l, r) __rlim_val(l, r, hard) ++ ++#define __rlim_rcur(l, r) __rlim_val(l, r, rcur) ++#define __rlim_rmin(l, r) __rlim_val(l, r, rmin) ++#define __rlim_rmax(l, r) __rlim_val(l, r, rmax) ++ ++#define __rlim_lhit(l, r) __rlim_val(l, r, lhit) ++#define __rlim_hit(l, r) atomic_inc(&__rlim_lhit(l, r)) ++ ++typedef atomic_long_t rlim_atomic_t; ++typedef unsigned long rlim_t; ++ ++#define __rlim_get(l, r) atomic_long_read(&__rlim_rcur(l, r)) ++#define __rlim_set(l, r, v) atomic_long_set(&__rlim_rcur(l, r), v) ++#define __rlim_inc(l, r) atomic_long_inc(&__rlim_rcur(l, r)) ++#define __rlim_dec(l, r) atomic_long_dec(&__rlim_rcur(l, r)) ++#define __rlim_add(l, r, v) atomic_long_add(v, &__rlim_rcur(l, r)) ++#define __rlim_sub(l, r, v) atomic_long_sub(v, &__rlim_rcur(l, r)) ++ ++ ++#if (RLIM_INFINITY == VLIM_INFINITY) ++#define VX_VLIM(r) ((long long)(long)(r)) ++#define VX_RLIM(v) ((rlim_t)(v)) ++#else ++#define VX_VLIM(r) (((r) == RLIM_INFINITY) \ ++ ? VLIM_INFINITY : (long long)(r)) ++#define VX_RLIM(v) (((v) == VLIM_INFINITY) \ ++ ? RLIM_INFINITY : (rlim_t)(v)) ++#endif ++ ++struct sysinfo; ++ ++void vx_vsi_meminfo(struct sysinfo *); ++void vx_vsi_swapinfo(struct sysinfo *); ++long vx_vsi_cached(struct sysinfo *); ++ ++#define NUM_LIMITS 24 ++ ++#endif /* __KERNEL__ */ ++#endif /* _VX_LIMIT_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/limit_cmd.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/limit_cmd.h +--- linux-3.0.9/include/linux/vserver/limit_cmd.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/limit_cmd.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,71 @@ ++#ifndef _VX_LIMIT_CMD_H ++#define _VX_LIMIT_CMD_H ++ ++ ++/* rlimit vserver commands */ ++ ++#define VCMD_get_rlimit VC_CMD(RLIMIT, 1, 0) ++#define VCMD_set_rlimit VC_CMD(RLIMIT, 2, 0) ++#define VCMD_get_rlimit_mask VC_CMD(RLIMIT, 3, 0) ++#define VCMD_reset_hits VC_CMD(RLIMIT, 7, 0) ++#define VCMD_reset_minmax VC_CMD(RLIMIT, 9, 0) ++ ++struct vcmd_ctx_rlimit_v0 { ++ uint32_t id; ++ uint64_t minimum; ++ uint64_t softlimit; ++ uint64_t maximum; ++}; ++ ++struct vcmd_ctx_rlimit_mask_v0 { ++ uint32_t minimum; ++ uint32_t softlimit; ++ uint32_t maximum; ++}; ++ ++#define VCMD_rlimit_stat VC_CMD(VSTAT, 1, 0) ++ ++struct vcmd_rlimit_stat_v0 { ++ uint32_t id; ++ uint32_t hits; ++ uint64_t value; ++ uint64_t minimum; ++ uint64_t maximum; ++}; ++ ++#define CRLIM_UNSET (0ULL) ++#define CRLIM_INFINITY (~0ULL) ++#define CRLIM_KEEP (~1ULL) ++ ++#ifdef __KERNEL__ ++ ++#ifdef CONFIG_IA32_EMULATION ++ ++struct vcmd_ctx_rlimit_v0_x32 { ++ uint32_t id; ++ uint64_t minimum; ++ uint64_t softlimit; ++ uint64_t maximum; ++} __attribute__ ((packed)); ++ ++#endif /* CONFIG_IA32_EMULATION */ ++ ++#include <linux/compiler.h> ++ ++extern int vc_get_rlimit_mask(uint32_t, void __user *); ++extern int vc_get_rlimit(struct vx_info *, void __user *); ++extern int vc_set_rlimit(struct vx_info *, void __user *); ++extern int vc_reset_hits(struct vx_info *, void __user *); ++extern int vc_reset_minmax(struct vx_info *, void __user *); ++ ++extern int vc_rlimit_stat(struct vx_info *, void __user *); ++ ++#ifdef CONFIG_IA32_EMULATION ++ ++extern int vc_get_rlimit_x32(struct vx_info *, void __user *); ++extern int vc_set_rlimit_x32(struct vx_info *, void __user *); ++ ++#endif /* CONFIG_IA32_EMULATION */ ++ ++#endif /* __KERNEL__ */ ++#endif /* _VX_LIMIT_CMD_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/limit_def.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/limit_def.h +--- linux-3.0.9/include/linux/vserver/limit_def.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/limit_def.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,47 @@ ++#ifndef _VX_LIMIT_DEF_H ++#define _VX_LIMIT_DEF_H ++ ++#include <asm/atomic.h> ++#include <asm/resource.h> ++ ++#include "limit.h" ++ ++ ++struct _vx_res_limit { ++ rlim_t soft; /* Context soft limit */ ++ rlim_t hard; /* Context hard limit */ ++ ++ rlim_atomic_t rcur; /* Current value */ ++ rlim_t rmin; /* Context minimum */ ++ rlim_t rmax; /* Context maximum */ ++ ++ atomic_t lhit; /* Limit hits */ ++}; ++ ++/* context sub struct */ ++ ++struct _vx_limit { ++ struct _vx_res_limit res[NUM_LIMITS]; ++}; ++ ++#ifdef CONFIG_VSERVER_DEBUG ++ ++static inline void __dump_vx_limit(struct _vx_limit *limit) ++{ ++ int i; ++ ++ printk("\t_vx_limit:"); ++ for (i = 0; i < NUM_LIMITS; i++) { ++ printk("\t [%2d] = %8lu %8lu/%8lu, %8ld/%8ld, %8d\n", ++ i, (unsigned long)__rlim_get(limit, i), ++ (unsigned long)__rlim_rmin(limit, i), ++ (unsigned long)__rlim_rmax(limit, i), ++ (long)__rlim_soft(limit, i), ++ (long)__rlim_hard(limit, i), ++ atomic_read(&__rlim_lhit(limit, i))); ++ } ++} ++ ++#endif ++ ++#endif /* _VX_LIMIT_DEF_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/limit_int.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/limit_int.h +--- linux-3.0.9/include/linux/vserver/limit_int.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/limit_int.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,198 @@ ++#ifndef _VX_LIMIT_INT_H ++#define _VX_LIMIT_INT_H ++ ++#include "context.h" ++ ++#ifdef __KERNEL__ ++ ++#define VXD_RCRES_COND(r) VXD_CBIT(cres, r) ++#define VXD_RLIMIT_COND(r) VXD_CBIT(limit, r) ++ ++extern const char *vlimit_name[NUM_LIMITS]; ++ ++static inline void __vx_acc_cres(struct vx_info *vxi, ++ int res, int dir, void *_data, char *_file, int _line) ++{ ++ if (VXD_RCRES_COND(res)) ++ vxlprintk(1, "vx_acc_cres[%5d,%s,%2d]: %5ld%s (%p)", ++ (vxi ? vxi->vx_id : -1), vlimit_name[res], res, ++ (vxi ? (long)__rlim_get(&vxi->limit, res) : 0), ++ (dir > 0) ? "++" : "--", _data, _file, _line); ++ if (!vxi) ++ return; ++ ++ if (dir > 0) ++ __rlim_inc(&vxi->limit, res); ++ else ++ __rlim_dec(&vxi->limit, res); ++} ++ ++static inline void __vx_add_cres(struct vx_info *vxi, ++ int res, int amount, void *_data, char *_file, int _line) ++{ ++ if (VXD_RCRES_COND(res)) ++ vxlprintk(1, "vx_add_cres[%5d,%s,%2d]: %5ld += %5d (%p)", ++ (vxi ? vxi->vx_id : -1), vlimit_name[res], res, ++ (vxi ? (long)__rlim_get(&vxi->limit, res) : 0), ++ amount, _data, _file, _line); ++ if (amount == 0) ++ return; ++ if (!vxi) ++ return; ++ __rlim_add(&vxi->limit, res, amount); ++} ++ ++static inline ++int __vx_cres_adjust_max(struct _vx_limit *limit, int res, rlim_t value) ++{ ++ int cond = (value > __rlim_rmax(limit, res)); ++ ++ if (cond) ++ __rlim_rmax(limit, res) = value; ++ return cond; ++} ++ ++static inline ++int __vx_cres_adjust_min(struct _vx_limit *limit, int res, rlim_t value) ++{ ++ int cond = (value < __rlim_rmin(limit, res)); ++ ++ if (cond) ++ __rlim_rmin(limit, res) = value; ++ return cond; ++} ++ ++static inline ++void __vx_cres_fixup(struct _vx_limit *limit, int res, rlim_t value) ++{ ++ if (!__vx_cres_adjust_max(limit, res, value)) ++ __vx_cres_adjust_min(limit, res, value); ++} ++ ++ ++/* return values: ++ +1 ... no limit hit ++ -1 ... over soft limit ++ 0 ... over hard limit */ ++ ++static inline int __vx_cres_avail(struct vx_info *vxi, ++ int res, int num, char *_file, int _line) ++{ ++ struct _vx_limit *limit; ++ rlim_t value; ++ ++ if (VXD_RLIMIT_COND(res)) ++ vxlprintk(1, "vx_cres_avail[%5d,%s,%2d]: %5ld/%5ld > %5ld + %5d", ++ (vxi ? vxi->vx_id : -1), vlimit_name[res], res, ++ (vxi ? (long)__rlim_soft(&vxi->limit, res) : -1), ++ (vxi ? (long)__rlim_hard(&vxi->limit, res) : -1), ++ (vxi ? (long)__rlim_get(&vxi->limit, res) : 0), ++ num, _file, _line); ++ if (!vxi) ++ return 1; ++ ++ limit = &vxi->limit; ++ value = __rlim_get(limit, res); ++ ++ if (!__vx_cres_adjust_max(limit, res, value)) ++ __vx_cres_adjust_min(limit, res, value); ++ ++ if (num == 0) ++ return 1; ++ ++ if (__rlim_soft(limit, res) == RLIM_INFINITY) ++ return -1; ++ if (value + num <= __rlim_soft(limit, res)) ++ return -1; ++ ++ if (__rlim_hard(limit, res) == RLIM_INFINITY) ++ return 1; ++ if (value + num <= __rlim_hard(limit, res)) ++ return 1; ++ ++ __rlim_hit(limit, res); ++ return 0; ++} ++ ++ ++static const int VLA_RSS[] = { RLIMIT_RSS, VLIMIT_ANON, VLIMIT_MAPPED, 0 }; ++ ++static inline ++rlim_t __vx_cres_array_sum(struct _vx_limit *limit, const int *array) ++{ ++ rlim_t value, sum = 0; ++ int res; ++ ++ while ((res = *array++)) { ++ value = __rlim_get(limit, res); ++ __vx_cres_fixup(limit, res, value); ++ sum += value; ++ } ++ return sum; ++} ++ ++static inline ++rlim_t __vx_cres_array_fixup(struct _vx_limit *limit, const int *array) ++{ ++ rlim_t value = __vx_cres_array_sum(limit, array + 1); ++ int res = *array; ++ ++ if (value == __rlim_get(limit, res)) ++ return value; ++ ++ __rlim_set(limit, res, value); ++ /* now adjust min/max */ ++ if (!__vx_cres_adjust_max(limit, res, value)) ++ __vx_cres_adjust_min(limit, res, value); ++ ++ return value; ++} ++ ++static inline int __vx_cres_array_avail(struct vx_info *vxi, ++ const int *array, int num, char *_file, int _line) ++{ ++ struct _vx_limit *limit; ++ rlim_t value = 0; ++ int res; ++ ++ if (num == 0) ++ return 1; ++ if (!vxi) ++ return 1; ++ ++ limit = &vxi->limit; ++ res = *array; ++ value = __vx_cres_array_sum(limit, array + 1); ++ ++ __rlim_set(limit, res, value); ++ __vx_cres_fixup(limit, res, value); ++ ++ return __vx_cres_avail(vxi, res, num, _file, _line); ++} ++ ++ ++static inline void vx_limit_fixup(struct _vx_limit *limit, int id) ++{ ++ rlim_t value; ++ int res; ++ ++ /* complex resources first */ ++ if ((id < 0) || (id == RLIMIT_RSS)) ++ __vx_cres_array_fixup(limit, VLA_RSS); ++ ++ for (res = 0; res < NUM_LIMITS; res++) { ++ if ((id > 0) && (res != id)) ++ continue; ++ ++ value = __rlim_get(limit, res); ++ __vx_cres_fixup(limit, res, value); ++ ++ /* not supposed to happen, maybe warn? */ ++ if (__rlim_rmax(limit, res) > __rlim_hard(limit, res)) ++ __rlim_rmax(limit, res) = __rlim_hard(limit, res); ++ } ++} ++ ++ ++#endif /* __KERNEL__ */ ++#endif /* _VX_LIMIT_INT_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/monitor.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/monitor.h +--- linux-3.0.9/include/linux/vserver/monitor.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/monitor.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,96 @@ ++#ifndef _VX_MONITOR_H ++#define _VX_MONITOR_H ++ ++#include <linux/types.h> ++ ++enum { ++ VXM_UNUSED = 0, ++ ++ VXM_SYNC = 0x10, ++ ++ VXM_UPDATE = 0x20, ++ VXM_UPDATE_1, ++ VXM_UPDATE_2, ++ ++ VXM_RQINFO_1 = 0x24, ++ VXM_RQINFO_2, ++ ++ VXM_ACTIVATE = 0x40, ++ VXM_DEACTIVATE, ++ VXM_IDLE, ++ ++ VXM_HOLD = 0x44, ++ VXM_UNHOLD, ++ ++ VXM_MIGRATE = 0x48, ++ VXM_RESCHED, ++ ++ /* all other bits are flags */ ++ VXM_SCHED = 0x80, ++}; ++ ++struct _vxm_update_1 { ++ uint32_t tokens_max; ++ uint32_t fill_rate; ++ uint32_t interval; ++}; ++ ++struct _vxm_update_2 { ++ uint32_t tokens_min; ++ uint32_t fill_rate; ++ uint32_t interval; ++}; ++ ++struct _vxm_rqinfo_1 { ++ uint16_t running; ++ uint16_t onhold; ++ uint16_t iowait; ++ uint16_t uintr; ++ uint32_t idle_tokens; ++}; ++ ++struct _vxm_rqinfo_2 { ++ uint32_t norm_time; ++ uint32_t idle_time; ++ uint32_t idle_skip; ++}; ++ ++struct _vxm_sched { ++ uint32_t tokens; ++ uint32_t norm_time; ++ uint32_t idle_time; ++}; ++ ++struct _vxm_task { ++ uint16_t pid; ++ uint16_t state; ++}; ++ ++struct _vxm_event { ++ uint32_t jif; ++ union { ++ uint32_t seq; ++ uint32_t sec; ++ }; ++ union { ++ uint32_t tokens; ++ uint32_t nsec; ++ struct _vxm_task tsk; ++ }; ++}; ++ ++struct _vx_mon_entry { ++ uint16_t type; ++ uint16_t xid; ++ union { ++ struct _vxm_event ev; ++ struct _vxm_sched sd; ++ struct _vxm_update_1 u1; ++ struct _vxm_update_2 u2; ++ struct _vxm_rqinfo_1 q1; ++ struct _vxm_rqinfo_2 q2; ++ }; ++}; ++ ++ ++#endif /* _VX_MONITOR_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/network.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/network.h +--- linux-3.0.9/include/linux/vserver/network.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/network.h 2011-08-18 16:30:48.000000000 +0200 +@@ -0,0 +1,148 @@ ++#ifndef _VX_NETWORK_H ++#define _VX_NETWORK_H ++ ++#include <linux/types.h> ++ ++ ++#define MAX_N_CONTEXT 65535 /* Arbitrary limit */ ++ ++ ++/* network flags */ ++ ++#define NXF_INFO_PRIVATE 0x00000008 ++ ++#define NXF_SINGLE_IP 0x00000100 ++#define NXF_LBACK_REMAP 0x00000200 ++#define NXF_LBACK_ALLOW 0x00000400 ++ ++#define NXF_HIDE_NETIF 0x02000000 ++#define NXF_HIDE_LBACK 0x04000000 ++ ++#define NXF_STATE_SETUP (1ULL << 32) ++#define NXF_STATE_ADMIN (1ULL << 34) ++ ++#define NXF_SC_HELPER (1ULL << 36) ++#define NXF_PERSISTENT (1ULL << 38) ++ ++#define NXF_ONE_TIME (0x0005ULL << 32) ++ ++ ++#define NXF_INIT_SET (__nxf_init_set()) ++ ++static inline uint64_t __nxf_init_set(void) { ++ return NXF_STATE_ADMIN ++#ifdef CONFIG_VSERVER_AUTO_LBACK ++ | NXF_LBACK_REMAP ++ | NXF_HIDE_LBACK ++#endif ++#ifdef CONFIG_VSERVER_AUTO_SINGLE ++ | NXF_SINGLE_IP ++#endif ++ | NXF_HIDE_NETIF; ++} ++ ++ ++/* network caps */ ++ ++#define NXC_TUN_CREATE 0x00000001 ++ ++#define NXC_RAW_ICMP 0x00000100 ++ ++#define NXC_MULTICAST 0x00001000 ++ ++ ++/* address types */ ++ ++#define NXA_TYPE_IPV4 0x0001 ++#define NXA_TYPE_IPV6 0x0002 ++ ++#define NXA_TYPE_NONE 0x0000 ++#define NXA_TYPE_ANY 0x00FF ++ ++#define NXA_TYPE_ADDR 0x0010 ++#define NXA_TYPE_MASK 0x0020 ++#define NXA_TYPE_RANGE 0x0040 ++ ++#define NXA_MASK_ALL (NXA_TYPE_ADDR | NXA_TYPE_MASK | NXA_TYPE_RANGE) ++ ++#define NXA_MOD_BCAST 0x0100 ++#define NXA_MOD_LBACK 0x0200 ++ ++#define NXA_LOOPBACK 0x1000 ++ ++#define NXA_MASK_BIND (NXA_MASK_ALL | NXA_MOD_BCAST | NXA_MOD_LBACK) ++#define NXA_MASK_SHOW (NXA_MASK_ALL | NXA_LOOPBACK) ++ ++#ifdef __KERNEL__ ++ ++#include <linux/list.h> ++#include <linux/spinlock.h> ++#include <linux/rcupdate.h> ++#include <linux/in.h> ++#include <linux/in6.h> ++#include <asm/atomic.h> ++ ++struct nx_addr_v4 { ++ struct nx_addr_v4 *next; ++ struct in_addr ip[2]; ++ struct in_addr mask; ++ uint16_t type; ++ uint16_t flags; ++}; ++ ++struct nx_addr_v6 { ++ struct nx_addr_v6 *next; ++ struct in6_addr ip; ++ struct in6_addr mask; ++ uint32_t prefix; ++ uint16_t type; ++ uint16_t flags; ++}; ++ ++struct nx_info { ++ struct hlist_node nx_hlist; /* linked list of nxinfos */ ++ nid_t nx_id; /* vnet id */ ++ atomic_t nx_usecnt; /* usage count */ ++ atomic_t nx_tasks; /* tasks count */ ++ int nx_state; /* context state */ ++ ++ uint64_t nx_flags; /* network flag word */ ++ uint64_t nx_ncaps; /* network capabilities */ ++ ++ struct in_addr v4_lback; /* Loopback address */ ++ struct in_addr v4_bcast; /* Broadcast address */ ++ struct nx_addr_v4 v4; /* First/Single ipv4 address */ ++#ifdef CONFIG_IPV6 ++ struct nx_addr_v6 v6; /* First/Single ipv6 address */ ++#endif ++ char nx_name[65]; /* network context name */ ++}; ++ ++ ++/* status flags */ ++ ++#define NXS_HASHED 0x0001 ++#define NXS_SHUTDOWN 0x0100 ++#define NXS_RELEASED 0x8000 ++ ++extern struct nx_info *lookup_nx_info(int); ++ ++extern int get_nid_list(int, unsigned int *, int); ++extern int nid_is_hashed(nid_t); ++ ++extern int nx_migrate_task(struct task_struct *, struct nx_info *); ++ ++extern long vs_net_change(struct nx_info *, unsigned int); ++ ++struct sock; ++ ++ ++#define NX_IPV4(n) ((n)->v4.type != NXA_TYPE_NONE) ++#ifdef CONFIG_IPV6 ++#define NX_IPV6(n) ((n)->v6.type != NXA_TYPE_NONE) ++#else ++#define NX_IPV6(n) (0) ++#endif ++ ++#endif /* __KERNEL__ */ ++#endif /* _VX_NETWORK_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/network_cmd.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/network_cmd.h +--- linux-3.0.9/include/linux/vserver/network_cmd.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/network_cmd.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,164 @@ ++#ifndef _VX_NETWORK_CMD_H ++#define _VX_NETWORK_CMD_H ++ ++ ++/* vinfo commands */ ++ ++#define VCMD_task_nid VC_CMD(VINFO, 2, 0) ++ ++#ifdef __KERNEL__ ++extern int vc_task_nid(uint32_t); ++ ++#endif /* __KERNEL__ */ ++ ++#define VCMD_nx_info VC_CMD(VINFO, 6, 0) ++ ++struct vcmd_nx_info_v0 { ++ uint32_t nid; ++ /* more to come */ ++}; ++ ++#ifdef __KERNEL__ ++extern int vc_nx_info(struct nx_info *, void __user *); ++ ++#endif /* __KERNEL__ */ ++ ++#include <linux/in.h> ++#include <linux/in6.h> ++ ++#define VCMD_net_create_v0 VC_CMD(VNET, 1, 0) ++#define VCMD_net_create VC_CMD(VNET, 1, 1) ++ ++struct vcmd_net_create { ++ uint64_t flagword; ++}; ++ ++#define VCMD_net_migrate VC_CMD(NETMIG, 1, 0) ++ ++#define VCMD_net_add VC_CMD(NETALT, 1, 0) ++#define VCMD_net_remove VC_CMD(NETALT, 2, 0) ++ ++struct vcmd_net_addr_v0 { ++ uint16_t type; ++ uint16_t count; ++ struct in_addr ip[4]; ++ struct in_addr mask[4]; ++}; ++ ++#define VCMD_net_add_ipv4_v1 VC_CMD(NETALT, 1, 1) ++#define VCMD_net_rem_ipv4_v1 VC_CMD(NETALT, 2, 1) ++ ++struct vcmd_net_addr_ipv4_v1 { ++ uint16_t type; ++ uint16_t flags; ++ struct in_addr ip; ++ struct in_addr mask; ++}; ++ ++#define VCMD_net_add_ipv4 VC_CMD(NETALT, 1, 2) ++#define VCMD_net_rem_ipv4 VC_CMD(NETALT, 2, 2) ++ ++struct vcmd_net_addr_ipv4_v2 { ++ uint16_t type; ++ uint16_t flags; ++ struct in_addr ip; ++ struct in_addr ip2; ++ struct in_addr mask; ++}; ++ ++#define VCMD_net_add_ipv6 VC_CMD(NETALT, 3, 1) ++#define VCMD_net_remove_ipv6 VC_CMD(NETALT, 4, 1) ++ ++struct vcmd_net_addr_ipv6_v1 { ++ uint16_t type; ++ uint16_t flags; ++ uint32_t prefix; ++ struct in6_addr ip; ++ struct in6_addr mask; ++}; ++ ++#define VCMD_add_match_ipv4 VC_CMD(NETALT, 5, 0) ++#define VCMD_get_match_ipv4 VC_CMD(NETALT, 6, 0) ++ ++struct vcmd_match_ipv4_v0 { ++ uint16_t type; ++ uint16_t flags; ++ uint16_t parent; ++ uint16_t prefix; ++ struct in_addr ip; ++ struct in_addr ip2; ++ struct in_addr mask; ++}; ++ ++#define VCMD_add_match_ipv6 VC_CMD(NETALT, 7, 0) ++#define VCMD_get_match_ipv6 VC_CMD(NETALT, 8, 0) ++ ++struct vcmd_match_ipv6_v0 { ++ uint16_t type; ++ uint16_t flags; ++ uint16_t parent; ++ uint16_t prefix; ++ struct in6_addr ip; ++ struct in6_addr ip2; ++ struct in6_addr mask; ++}; ++ ++ ++#ifdef __KERNEL__ ++extern int vc_net_create(uint32_t, void __user *); ++extern int vc_net_migrate(struct nx_info *, void __user *); ++ ++extern int vc_net_add(struct nx_info *, void __user *); ++extern int vc_net_remove(struct nx_info *, void __user *); ++ ++extern int vc_net_add_ipv4_v1(struct nx_info *, void __user *); ++extern int vc_net_add_ipv4(struct nx_info *, void __user *); ++ ++extern int vc_net_rem_ipv4_v1(struct nx_info *, void __user *); ++extern int vc_net_rem_ipv4(struct nx_info *, void __user *); ++ ++extern int vc_net_add_ipv6(struct nx_info *, void __user *); ++extern int vc_net_remove_ipv6(struct nx_info *, void __user *); ++ ++extern int vc_add_match_ipv4(struct nx_info *, void __user *); ++extern int vc_get_match_ipv4(struct nx_info *, void __user *); ++ ++extern int vc_add_match_ipv6(struct nx_info *, void __user *); ++extern int vc_get_match_ipv6(struct nx_info *, void __user *); ++ ++#endif /* __KERNEL__ */ ++ ++ ++/* flag commands */ ++ ++#define VCMD_get_nflags VC_CMD(FLAGS, 5, 0) ++#define VCMD_set_nflags VC_CMD(FLAGS, 6, 0) ++ ++struct vcmd_net_flags_v0 { ++ uint64_t flagword; ++ uint64_t mask; ++}; ++ ++#ifdef __KERNEL__ ++extern int vc_get_nflags(struct nx_info *, void __user *); ++extern int vc_set_nflags(struct nx_info *, void __user *); ++ ++#endif /* __KERNEL__ */ ++ ++ ++/* network caps commands */ ++ ++#define VCMD_get_ncaps VC_CMD(FLAGS, 7, 0) ++#define VCMD_set_ncaps VC_CMD(FLAGS, 8, 0) ++ ++struct vcmd_net_caps_v0 { ++ uint64_t ncaps; ++ uint64_t cmask; ++}; ++ ++#ifdef __KERNEL__ ++extern int vc_get_ncaps(struct nx_info *, void __user *); ++extern int vc_set_ncaps(struct nx_info *, void __user *); ++ ++#endif /* __KERNEL__ */ ++#endif /* _VX_CONTEXT_CMD_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/percpu.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/percpu.h +--- linux-3.0.9/include/linux/vserver/percpu.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/percpu.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,14 @@ ++#ifndef _VX_PERCPU_H ++#define _VX_PERCPU_H ++ ++#include "cvirt_def.h" ++#include "sched_def.h" ++ ++struct _vx_percpu { ++ struct _vx_cvirt_pc cvirt; ++ struct _vx_sched_pc sched; ++}; ++ ++#define PERCPU_PERCTX (sizeof(struct _vx_percpu)) ++ ++#endif /* _VX_PERCPU_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/pid.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/pid.h +--- linux-3.0.9/include/linux/vserver/pid.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/pid.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,51 @@ ++#ifndef _VSERVER_PID_H ++#define _VSERVER_PID_H ++ ++/* pid faking stuff */ ++ ++#define vx_info_map_pid(v, p) \ ++ __vx_info_map_pid((v), (p), __func__, __FILE__, __LINE__) ++#define vx_info_map_tgid(v,p) vx_info_map_pid(v,p) ++#define vx_map_pid(p) vx_info_map_pid(current_vx_info(), p) ++#define vx_map_tgid(p) vx_map_pid(p) ++ ++static inline int __vx_info_map_pid(struct vx_info *vxi, int pid, ++ const char *func, const char *file, int line) ++{ ++ if (vx_info_flags(vxi, VXF_INFO_INIT, 0)) { ++ vxfprintk(VXD_CBIT(cvirt, 2), ++ "vx_map_tgid: %p/%llx: %d -> %d", ++ vxi, (long long)vxi->vx_flags, pid, ++ (pid && pid == vxi->vx_initpid) ? 1 : pid, ++ func, file, line); ++ if (pid == 0) ++ return 0; ++ if (pid == vxi->vx_initpid) ++ return 1; ++ } ++ return pid; ++} ++ ++#define vx_info_rmap_pid(v, p) \ ++ __vx_info_rmap_pid((v), (p), __func__, __FILE__, __LINE__) ++#define vx_rmap_pid(p) vx_info_rmap_pid(current_vx_info(), p) ++#define vx_rmap_tgid(p) vx_rmap_pid(p) ++ ++static inline int __vx_info_rmap_pid(struct vx_info *vxi, int pid, ++ const char *func, const char *file, int line) ++{ ++ if (vx_info_flags(vxi, VXF_INFO_INIT, 0)) { ++ vxfprintk(VXD_CBIT(cvirt, 2), ++ "vx_rmap_tgid: %p/%llx: %d -> %d", ++ vxi, (long long)vxi->vx_flags, pid, ++ (pid == 1) ? vxi->vx_initpid : pid, ++ func, file, line); ++ if ((pid == 1) && vxi->vx_initpid) ++ return vxi->vx_initpid; ++ if (pid == vxi->vx_initpid) ++ return ~0U; ++ } ++ return pid; ++} ++ ++#endif +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/sched.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/sched.h +--- linux-3.0.9/include/linux/vserver/sched.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/sched.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,23 @@ ++#ifndef _VX_SCHED_H ++#define _VX_SCHED_H ++ ++ ++#ifdef __KERNEL__ ++ ++struct timespec; ++ ++void vx_vsi_uptime(struct timespec *, struct timespec *); ++ ++ ++struct vx_info; ++ ++void vx_update_load(struct vx_info *); ++ ++ ++void vx_update_sched_param(struct _vx_sched *sched, ++ struct _vx_sched_pc *sched_pc); ++ ++#endif /* __KERNEL__ */ ++#else /* _VX_SCHED_H */ ++#warning duplicate inclusion ++#endif /* _VX_SCHED_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/sched_cmd.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/sched_cmd.h +--- linux-3.0.9/include/linux/vserver/sched_cmd.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/sched_cmd.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,21 @@ ++#ifndef _VX_SCHED_CMD_H ++#define _VX_SCHED_CMD_H ++ ++ ++struct vcmd_prio_bias { ++ int32_t cpu_id; ++ int32_t prio_bias; ++}; ++ ++#define VCMD_set_prio_bias VC_CMD(SCHED, 4, 0) ++#define VCMD_get_prio_bias VC_CMD(SCHED, 5, 0) ++ ++#ifdef __KERNEL__ ++ ++#include <linux/compiler.h> ++ ++extern int vc_set_prio_bias(struct vx_info *, void __user *); ++extern int vc_get_prio_bias(struct vx_info *, void __user *); ++ ++#endif /* __KERNEL__ */ ++#endif /* _VX_SCHED_CMD_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/sched_def.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/sched_def.h +--- linux-3.0.9/include/linux/vserver/sched_def.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/sched_def.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,38 @@ ++#ifndef _VX_SCHED_DEF_H ++#define _VX_SCHED_DEF_H ++ ++#include <linux/spinlock.h> ++#include <linux/jiffies.h> ++#include <linux/cpumask.h> ++#include <asm/atomic.h> ++#include <asm/param.h> ++ ++ ++/* context sub struct */ ++ ++struct _vx_sched { ++ int prio_bias; /* bias offset for priority */ ++ ++ cpumask_t update; /* CPUs which should update */ ++}; ++ ++struct _vx_sched_pc { ++ int prio_bias; /* bias offset for priority */ ++ ++ uint64_t user_ticks; /* token tick events */ ++ uint64_t sys_ticks; /* token tick events */ ++ uint64_t hold_ticks; /* token ticks paused */ ++}; ++ ++ ++#ifdef CONFIG_VSERVER_DEBUG ++ ++static inline void __dump_vx_sched(struct _vx_sched *sched) ++{ ++ printk("\t_vx_sched:\n"); ++ printk("\t priority = %4d\n", sched->prio_bias); ++} ++ ++#endif ++ ++#endif /* _VX_SCHED_DEF_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/signal.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/signal.h +--- linux-3.0.9/include/linux/vserver/signal.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/signal.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,14 @@ ++#ifndef _VX_SIGNAL_H ++#define _VX_SIGNAL_H ++ ++ ++#ifdef __KERNEL__ ++ ++struct vx_info; ++ ++int vx_info_kill(struct vx_info *, int, int); ++ ++#endif /* __KERNEL__ */ ++#else /* _VX_SIGNAL_H */ ++#warning duplicate inclusion ++#endif /* _VX_SIGNAL_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/signal_cmd.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/signal_cmd.h +--- linux-3.0.9/include/linux/vserver/signal_cmd.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/signal_cmd.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,43 @@ ++#ifndef _VX_SIGNAL_CMD_H ++#define _VX_SIGNAL_CMD_H ++ ++ ++/* signalling vserver commands */ ++ ++#define VCMD_ctx_kill VC_CMD(PROCTRL, 1, 0) ++#define VCMD_wait_exit VC_CMD(EVENT, 99, 0) ++ ++struct vcmd_ctx_kill_v0 { ++ int32_t pid; ++ int32_t sig; ++}; ++ ++struct vcmd_wait_exit_v0 { ++ int32_t reboot_cmd; ++ int32_t exit_code; ++}; ++ ++#ifdef __KERNEL__ ++ ++extern int vc_ctx_kill(struct vx_info *, void __user *); ++extern int vc_wait_exit(struct vx_info *, void __user *); ++ ++#endif /* __KERNEL__ */ ++ ++/* process alteration commands */ ++ ++#define VCMD_get_pflags VC_CMD(PROCALT, 5, 0) ++#define VCMD_set_pflags VC_CMD(PROCALT, 6, 0) ++ ++struct vcmd_pflags_v0 { ++ uint32_t flagword; ++ uint32_t mask; ++}; ++ ++#ifdef __KERNEL__ ++ ++extern int vc_get_pflags(uint32_t pid, void __user *); ++extern int vc_set_pflags(uint32_t pid, void __user *); ++ ++#endif /* __KERNEL__ */ ++#endif /* _VX_SIGNAL_CMD_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/space.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/space.h +--- linux-3.0.9/include/linux/vserver/space.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/space.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,12 @@ ++#ifndef _VX_SPACE_H ++#define _VX_SPACE_H ++ ++#include <linux/types.h> ++ ++struct vx_info; ++ ++int vx_set_space(struct vx_info *vxi, unsigned long mask, unsigned index); ++ ++#else /* _VX_SPACE_H */ ++#warning duplicate inclusion ++#endif /* _VX_SPACE_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/space_cmd.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/space_cmd.h +--- linux-3.0.9/include/linux/vserver/space_cmd.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/space_cmd.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,38 @@ ++#ifndef _VX_SPACE_CMD_H ++#define _VX_SPACE_CMD_H ++ ++ ++#define VCMD_enter_space_v0 VC_CMD(PROCALT, 1, 0) ++#define VCMD_enter_space_v1 VC_CMD(PROCALT, 1, 1) ++#define VCMD_enter_space VC_CMD(PROCALT, 1, 2) ++ ++#define VCMD_set_space_v0 VC_CMD(PROCALT, 3, 0) ++#define VCMD_set_space_v1 VC_CMD(PROCALT, 3, 1) ++#define VCMD_set_space VC_CMD(PROCALT, 3, 2) ++ ++#define VCMD_get_space_mask_v0 VC_CMD(PROCALT, 4, 0) ++ ++#define VCMD_get_space_mask VC_CMD(VSPACE, 0, 1) ++#define VCMD_get_space_default VC_CMD(VSPACE, 1, 0) ++ ++ ++struct vcmd_space_mask_v1 { ++ uint64_t mask; ++}; ++ ++struct vcmd_space_mask_v2 { ++ uint64_t mask; ++ uint32_t index; ++}; ++ ++ ++#ifdef __KERNEL__ ++ ++extern int vc_enter_space_v1(struct vx_info *, void __user *); ++extern int vc_set_space_v1(struct vx_info *, void __user *); ++extern int vc_enter_space(struct vx_info *, void __user *); ++extern int vc_set_space(struct vx_info *, void __user *); ++extern int vc_get_space_mask(void __user *, int); ++ ++#endif /* __KERNEL__ */ ++#endif /* _VX_SPACE_CMD_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/switch.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/switch.h +--- linux-3.0.9/include/linux/vserver/switch.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/switch.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,98 @@ ++#ifndef _VX_SWITCH_H ++#define _VX_SWITCH_H ++ ++#include <linux/types.h> ++ ++ ++#define VC_CATEGORY(c) (((c) >> 24) & 0x3F) ++#define VC_COMMAND(c) (((c) >> 16) & 0xFF) ++#define VC_VERSION(c) ((c) & 0xFFF) ++ ++#define VC_CMD(c, i, v) ((((VC_CAT_ ## c) & 0x3F) << 24) \ ++ | (((i) & 0xFF) << 16) | ((v) & 0xFFF)) ++ ++/* ++ ++ Syscall Matrix V2.8 ++ ++ |VERSION|CREATE |MODIFY |MIGRATE|CONTROL|EXPERIM| |SPECIAL|SPECIAL| ++ |STATS |DESTROY|ALTER |CHANGE |LIMIT |TEST | | | | ++ |INFO |SETUP | |MOVE | | | | | | ++ -------+-------+-------+-------+-------+-------+-------+ +-------+-------+ ++ SYSTEM |VERSION|VSETUP |VHOST | | | | |DEVICE | | ++ HOST | 00| 01| 02| 03| 04| 05| | 06| 07| ++ -------+-------+-------+-------+-------+-------+-------+ +-------+-------+ ++ CPU | |VPROC |PROCALT|PROCMIG|PROCTRL| | |SCHED. | | ++ PROCESS| 08| 09| 10| 11| 12| 13| | 14| 15| ++ -------+-------+-------+-------+-------+-------+-------+ +-------+-------+ ++ MEMORY | | | | |MEMCTRL| | |SWAP | | ++ | 16| 17| 18| 19| 20| 21| | 22| 23| ++ -------+-------+-------+-------+-------+-------+-------+ +-------+-------+ ++ NETWORK| |VNET |NETALT |NETMIG |NETCTL | | |SERIAL | | ++ | 24| 25| 26| 27| 28| 29| | 30| 31| ++ -------+-------+-------+-------+-------+-------+-------+ +-------+-------+ ++ DISK | | | |TAGMIG |DLIMIT | | |INODE | | ++ VFS | 32| 33| 34| 35| 36| 37| | 38| 39| ++ -------+-------+-------+-------+-------+-------+-------+ +-------+-------+ ++ OTHER |VSTAT | | | | | | |VINFO | | ++ | 40| 41| 42| 43| 44| 45| | 46| 47| ++ =======+=======+=======+=======+=======+=======+=======+ +=======+=======+ ++ SPECIAL|EVENT | | | |FLAGS | | |VSPACE | | ++ | 48| 49| 50| 51| 52| 53| | 54| 55| ++ -------+-------+-------+-------+-------+-------+-------+ +-------+-------+ ++ SPECIAL|DEBUG | | | |RLIMIT |SYSCALL| | |COMPAT | ++ | 56| 57| 58| 59| 60|TEST 61| | 62| 63| ++ -------+-------+-------+-------+-------+-------+-------+ +-------+-------+ ++ ++*/ ++ ++#define VC_CAT_VERSION 0 ++ ++#define VC_CAT_VSETUP 1 ++#define VC_CAT_VHOST 2 ++ ++#define VC_CAT_DEVICE 6 ++ ++#define VC_CAT_VPROC 9 ++#define VC_CAT_PROCALT 10 ++#define VC_CAT_PROCMIG 11 ++#define VC_CAT_PROCTRL 12 ++ ++#define VC_CAT_SCHED 14 ++#define VC_CAT_MEMCTRL 20 ++ ++#define VC_CAT_VNET 25 ++#define VC_CAT_NETALT 26 ++#define VC_CAT_NETMIG 27 ++#define VC_CAT_NETCTRL 28 ++ ++#define VC_CAT_TAGMIG 35 ++#define VC_CAT_DLIMIT 36 ++#define VC_CAT_INODE 38 ++ ++#define VC_CAT_VSTAT 40 ++#define VC_CAT_VINFO 46 ++#define VC_CAT_EVENT 48 ++ ++#define VC_CAT_FLAGS 52 ++#define VC_CAT_VSPACE 54 ++#define VC_CAT_DEBUG 56 ++#define VC_CAT_RLIMIT 60 ++ ++#define VC_CAT_SYSTEST 61 ++#define VC_CAT_COMPAT 63 ++ ++/* query version */ ++ ++#define VCMD_get_version VC_CMD(VERSION, 0, 0) ++#define VCMD_get_vci VC_CMD(VERSION, 1, 0) ++ ++ ++#ifdef __KERNEL__ ++ ++#include <linux/errno.h> ++ ++#endif /* __KERNEL__ */ ++ ++#endif /* _VX_SWITCH_H */ ++ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/tag.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/tag.h +--- linux-3.0.9/include/linux/vserver/tag.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/tag.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,143 @@ ++#ifndef _DX_TAG_H ++#define _DX_TAG_H ++ ++#include <linux/types.h> ++ ++ ++#define DX_TAG(in) (IS_TAGGED(in)) ++ ++ ++#ifdef CONFIG_TAG_NFSD ++#define DX_TAG_NFSD 1 ++#else ++#define DX_TAG_NFSD 0 ++#endif ++ ++ ++#ifdef CONFIG_TAGGING_NONE ++ ++#define MAX_UID 0xFFFFFFFF ++#define MAX_GID 0xFFFFFFFF ++ ++#define INOTAG_TAG(cond, uid, gid, tag) (0) ++ ++#define TAGINO_UID(cond, uid, tag) (uid) ++#define TAGINO_GID(cond, gid, tag) (gid) ++ ++#endif ++ ++ ++#ifdef CONFIG_TAGGING_GID16 ++ ++#define MAX_UID 0xFFFFFFFF ++#define MAX_GID 0x0000FFFF ++ ++#define INOTAG_TAG(cond, uid, gid, tag) \ ++ ((cond) ? (((gid) >> 16) & 0xFFFF) : 0) ++ ++#define TAGINO_UID(cond, uid, tag) (uid) ++#define TAGINO_GID(cond, gid, tag) \ ++ ((cond) ? (((gid) & 0xFFFF) | ((tag) << 16)) : (gid)) ++ ++#endif ++ ++ ++#ifdef CONFIG_TAGGING_ID24 ++ ++#define MAX_UID 0x00FFFFFF ++#define MAX_GID 0x00FFFFFF ++ ++#define INOTAG_TAG(cond, uid, gid, tag) \ ++ ((cond) ? ((((uid) >> 16) & 0xFF00) | (((gid) >> 24) & 0xFF)) : 0) ++ ++#define TAGINO_UID(cond, uid, tag) \ ++ ((cond) ? (((uid) & 0xFFFFFF) | (((tag) & 0xFF00) << 16)) : (uid)) ++#define TAGINO_GID(cond, gid, tag) \ ++ ((cond) ? (((gid) & 0xFFFFFF) | (((tag) & 0x00FF) << 24)) : (gid)) ++ ++#endif ++ ++ ++#ifdef CONFIG_TAGGING_UID16 ++ ++#define MAX_UID 0x0000FFFF ++#define MAX_GID 0xFFFFFFFF ++ ++#define INOTAG_TAG(cond, uid, gid, tag) \ ++ ((cond) ? (((uid) >> 16) & 0xFFFF) : 0) ++ ++#define TAGINO_UID(cond, uid, tag) \ ++ ((cond) ? (((uid) & 0xFFFF) | ((tag) << 16)) : (uid)) ++#define TAGINO_GID(cond, gid, tag) (gid) ++ ++#endif ++ ++ ++#ifdef CONFIG_TAGGING_INTERN ++ ++#define MAX_UID 0xFFFFFFFF ++#define MAX_GID 0xFFFFFFFF ++ ++#define INOTAG_TAG(cond, uid, gid, tag) \ ++ ((cond) ? (tag) : 0) ++ ++#define TAGINO_UID(cond, uid, tag) (uid) ++#define TAGINO_GID(cond, gid, tag) (gid) ++ ++#endif ++ ++ ++#ifndef CONFIG_TAGGING_NONE ++#define dx_current_fstag(sb) \ ++ ((sb)->s_flags & MS_TAGGED ? dx_current_tag() : 0) ++#else ++#define dx_current_fstag(sb) (0) ++#endif ++ ++#ifndef CONFIG_TAGGING_INTERN ++#define TAGINO_TAG(cond, tag) (0) ++#else ++#define TAGINO_TAG(cond, tag) ((cond) ? (tag) : 0) ++#endif ++ ++#define INOTAG_UID(cond, uid, gid) \ ++ ((cond) ? ((uid) & MAX_UID) : (uid)) ++#define INOTAG_GID(cond, uid, gid) \ ++ ((cond) ? ((gid) & MAX_GID) : (gid)) ++ ++ ++static inline uid_t dx_map_uid(uid_t uid) ++{ ++ if ((uid > MAX_UID) && (uid != -1)) ++ uid = -2; ++ return (uid & MAX_UID); ++} ++ ++static inline gid_t dx_map_gid(gid_t gid) ++{ ++ if ((gid > MAX_GID) && (gid != -1)) ++ gid = -2; ++ return (gid & MAX_GID); ++} ++ ++struct peer_tag { ++ int32_t xid; ++ int32_t nid; ++}; ++ ++#define dx_notagcheck(sb) ((sb) && ((sb)->s_flags & MS_NOTAGCHECK)) ++ ++int dx_parse_tag(char *string, tag_t *tag, int remove, int *mnt_flags, ++ unsigned long *flags); ++ ++#ifdef CONFIG_PROPAGATE ++ ++void __dx_propagate_tag(struct nameidata *nd, struct inode *inode); ++ ++#define dx_propagate_tag(n, i) __dx_propagate_tag(n, i) ++ ++#else ++#define dx_propagate_tag(n, i) do { } while (0) ++#endif ++ ++#endif /* _DX_TAG_H */ +diff -NurpP --minimal linux-3.0.9/include/linux/vserver/tag_cmd.h linux-3.0.9-vs2.3.2.1/include/linux/vserver/tag_cmd.h +--- linux-3.0.9/include/linux/vserver/tag_cmd.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/linux/vserver/tag_cmd.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,22 @@ ++#ifndef _VX_TAG_CMD_H ++#define _VX_TAG_CMD_H ++ ++ ++/* vinfo commands */ ++ ++#define VCMD_task_tag VC_CMD(VINFO, 3, 0) ++ ++#ifdef __KERNEL__ ++extern int vc_task_tag(uint32_t); ++ ++#endif /* __KERNEL__ */ ++ ++/* context commands */ ++ ++#define VCMD_tag_migrate VC_CMD(TAGMIG, 1, 0) ++ ++#ifdef __KERNEL__ ++extern int vc_tag_migrate(uint32_t); ++ ++#endif /* __KERNEL__ */ ++#endif /* _VX_TAG_CMD_H */ +diff -NurpP --minimal linux-3.0.9/include/net/addrconf.h linux-3.0.9-vs2.3.2.1/include/net/addrconf.h +--- linux-3.0.9/include/net/addrconf.h 2011-07-22 11:18:11.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/net/addrconf.h 2011-06-10 22:11:24.000000000 +0200 +@@ -80,7 +80,8 @@ extern int ipv6_dev_get_saddr(struct n + struct net_device *dev, + const struct in6_addr *daddr, + unsigned int srcprefs, +- struct in6_addr *saddr); ++ struct in6_addr *saddr, ++ struct nx_info *nxi); + extern int ipv6_get_lladdr(struct net_device *dev, + struct in6_addr *addr, + unsigned char banned_flags); +diff -NurpP --minimal linux-3.0.9/include/net/af_unix.h linux-3.0.9-vs2.3.2.1/include/net/af_unix.h +--- linux-3.0.9/include/net/af_unix.h 2011-07-22 11:18:11.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/net/af_unix.h 2011-06-10 22:11:24.000000000 +0200 +@@ -4,6 +4,7 @@ + #include <linux/socket.h> + #include <linux/un.h> + #include <linux/mutex.h> ++#include <linux/vs_base.h> + #include <net/sock.h> + + extern void unix_inflight(struct file *fp); +diff -NurpP --minimal linux-3.0.9/include/net/inet_timewait_sock.h linux-3.0.9-vs2.3.2.1/include/net/inet_timewait_sock.h +--- linux-3.0.9/include/net/inet_timewait_sock.h 2011-03-15 18:07:40.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/include/net/inet_timewait_sock.h 2011-06-10 22:11:24.000000000 +0200 +@@ -113,6 +113,10 @@ struct inet_timewait_sock { + #define tw_net __tw_common.skc_net + #define tw_daddr __tw_common.skc_daddr + #define tw_rcv_saddr __tw_common.skc_rcv_saddr ++#define tw_xid __tw_common.skc_xid ++#define tw_vx_info __tw_common.skc_vx_info ++#define tw_nid __tw_common.skc_nid ++#define tw_nx_info __tw_common.skc_nx_info + int tw_timeout; + volatile unsigned char tw_substate; + unsigned char tw_rcv_wscale; +diff -NurpP --minimal linux-3.0.9/include/net/ip6_route.h linux-3.0.9-vs2.3.2.1/include/net/ip6_route.h +--- linux-3.0.9/include/net/ip6_route.h 2011-07-22 11:18:11.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/net/ip6_route.h 2011-06-16 14:16:51.000000000 +0200 +@@ -86,7 +86,8 @@ extern int ip6_route_get_saddr(struct + struct rt6_info *rt, + const struct in6_addr *daddr, + unsigned int prefs, +- struct in6_addr *saddr); ++ struct in6_addr *saddr, ++ struct nx_info *nxi); + + extern struct rt6_info *rt6_lookup(struct net *net, + const struct in6_addr *daddr, +diff -NurpP --minimal linux-3.0.9/include/net/route.h linux-3.0.9-vs2.3.2.1/include/net/route.h +--- linux-3.0.9/include/net/route.h 2011-07-22 11:18:11.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/net/route.h 2011-07-27 19:42:59.000000000 +0200 +@@ -202,6 +202,9 @@ static inline void ip_rt_put(struct rtab + dst_release(&rt->dst); + } + ++#include <linux/vs_base.h> ++#include <linux/vs_inet.h> ++ + #define IPTOS_RT_MASK (IPTOS_TOS_MASK & ~3) + + extern const __u8 ip_tos2prio[16]; +@@ -253,6 +256,9 @@ static inline void ip_route_connect_init + protocol, flow_flags, dst, src, dport, sport); + } + ++extern struct rtable *ip_v4_find_src(struct net *net, struct nx_info *, ++ struct flowi4 *); ++ + static inline struct rtable *ip_route_connect(struct flowi4 *fl4, + __be32 dst, __be32 src, u32 tos, + int oif, u8 protocol, +@@ -261,11 +267,25 @@ static inline struct rtable *ip_route_co + { + struct net *net = sock_net(sk); + struct rtable *rt; ++ struct nx_info *nx_info = current_nx_info(); + + ip_route_connect_init(fl4, dst, src, tos, oif, protocol, + sport, dport, sk, can_sleep); + +- if (!dst || !src) { ++ if (sk) ++ nx_info = sk->sk_nx_info; ++ ++ vxdprintk(VXD_CBIT(net, 4), ++ "ip_route_connect(%p) %p,%p;%lx", ++ sk, nx_info, sk->sk_socket, ++ (sk->sk_socket?sk->sk_socket->flags:0)); ++ ++ rt = ip_v4_find_src(net, nx_info, fl4); ++ if (IS_ERR(rt)) ++ return rt; ++ ip_rt_put(rt); ++ ++ if (!fl4->daddr || !fl4->saddr) { + rt = __ip_route_output_key(net, fl4); + if (IS_ERR(rt)) + return rt; +diff -NurpP --minimal linux-3.0.9/include/net/sock.h linux-3.0.9-vs2.3.2.1/include/net/sock.h +--- linux-3.0.9/include/net/sock.h 2011-07-22 11:18:12.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/include/net/sock.h 2011-07-01 11:35:35.000000000 +0200 +@@ -149,6 +149,10 @@ struct sock_common { + #ifdef CONFIG_NET_NS + struct net *skc_net; + #endif ++ xid_t skc_xid; ++ struct vx_info *skc_vx_info; ++ nid_t skc_nid; ++ struct nx_info *skc_nx_info; + /* + * fields between dontcopy_begin/dontcopy_end + * are not copied in sock_copy() +@@ -256,6 +260,10 @@ struct sock { + #define sk_bind_node __sk_common.skc_bind_node + #define sk_prot __sk_common.skc_prot + #define sk_net __sk_common.skc_net ++#define sk_xid __sk_common.skc_xid ++#define sk_vx_info __sk_common.skc_vx_info ++#define sk_nid __sk_common.skc_nid ++#define sk_nx_info __sk_common.skc_nx_info + socket_lock_t sk_lock; + struct sk_buff_head sk_receive_queue; + /* +diff -NurpP --minimal linux-3.0.9/init/Kconfig linux-3.0.9-vs2.3.2.1/init/Kconfig +--- linux-3.0.9/init/Kconfig 2011-07-22 11:18:12.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/init/Kconfig 2011-08-08 18:08:57.000000000 +0200 +@@ -574,6 +574,7 @@ config HAVE_UNSTABLE_SCHED_CLOCK + menuconfig CGROUPS + boolean "Control Group support" + depends on EVENTFD ++ default y + help + This option adds support for grouping sets of processes together, for + use with process control subsystems such as Cpusets, CFS, memory +@@ -790,6 +791,7 @@ config IPC_NS + config USER_NS + bool "User namespace (EXPERIMENTAL)" + depends on EXPERIMENTAL ++ depends on VSERVER_DISABLED + default y + help + This allows containers, i.e. vservers, to use user namespaces +diff -NurpP --minimal linux-3.0.9/init/main.c linux-3.0.9-vs2.3.2.1/init/main.c +--- linux-3.0.9/init/main.c 2011-07-22 11:18:12.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/init/main.c 2011-06-22 12:39:15.000000000 +0200 +@@ -68,6 +68,7 @@ + #include <linux/shmem_fs.h> + #include <linux/slab.h> + #include <linux/perf_event.h> ++#include <linux/vserver/percpu.h> + + #include <asm/io.h> + #include <asm/bugs.h> +diff -NurpP --minimal linux-3.0.9/ipc/mqueue.c linux-3.0.9-vs2.3.2.1/ipc/mqueue.c +--- linux-3.0.9/ipc/mqueue.c 2011-11-15 16:40:47.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/ipc/mqueue.c 2011-10-18 14:03:46.000000000 +0200 +@@ -33,6 +33,8 @@ + #include <linux/pid.h> + #include <linux/ipc_namespace.h> + #include <linux/slab.h> ++#include <linux/vs_context.h> ++#include <linux/vs_limit.h> + + #include <net/sock.h> + #include "util.h" +@@ -66,6 +68,7 @@ struct mqueue_inode_info { + struct sigevent notify; + struct pid* notify_owner; + struct user_struct *user; /* user who created, for accounting */ ++ struct vx_info *vxi; + struct sock *notify_sock; + struct sk_buff *notify_cookie; + +@@ -128,6 +131,7 @@ static struct inode *mqueue_get_inode(st + if (S_ISREG(mode)) { + struct mqueue_inode_info *info; + struct task_struct *p = current; ++ struct vx_info *vxi = p->vx_info; + unsigned long mq_bytes, mq_msg_tblsz; + + inode->i_fop = &mqueue_file_operations; +@@ -141,6 +145,7 @@ static struct inode *mqueue_get_inode(st + info->notify_owner = NULL; + info->qsize = 0; + info->user = NULL; /* set when all is ok */ ++ info->vxi = NULL; + memset(&info->attr, 0, sizeof(info->attr)); + info->attr.mq_maxmsg = ipc_ns->mq_msg_max; + info->attr.mq_msgsize = ipc_ns->mq_msgsize_max; +@@ -158,17 +163,20 @@ static struct inode *mqueue_get_inode(st + + spin_lock(&mq_lock); + if (u->mq_bytes + mq_bytes < u->mq_bytes || +- u->mq_bytes + mq_bytes > task_rlimit(p, RLIMIT_MSGQUEUE)) { ++ u->mq_bytes + mq_bytes > task_rlimit(p, RLIMIT_MSGQUEUE) || ++ !vx_ipcmsg_avail(vxi, mq_bytes)) { + spin_unlock(&mq_lock); + /* mqueue_evict_inode() releases info->messages */ + ret = -EMFILE; + goto out_inode; + } + u->mq_bytes += mq_bytes; ++ vx_ipcmsg_add(vxi, u, mq_bytes); + spin_unlock(&mq_lock); + + /* all is ok */ + info->user = get_uid(u); ++ info->vxi = get_vx_info(vxi); + } else if (S_ISDIR(mode)) { + inc_nlink(inode); + /* Some things misbehave if size == 0 on a directory */ +@@ -278,8 +286,11 @@ static void mqueue_evict_inode(struct in + + info->attr.mq_msgsize); + user = info->user; + if (user) { ++ struct vx_info *vxi = info->vxi; ++ + spin_lock(&mq_lock); + user->mq_bytes -= mq_bytes; ++ vx_ipcmsg_sub(vxi, user, mq_bytes); + /* + * get_ns_from_inode() ensures that the + * (ipc_ns = sb->s_fs_info) is either a valid ipc_ns +@@ -289,6 +300,7 @@ static void mqueue_evict_inode(struct in + if (ipc_ns) + ipc_ns->mq_queues_count--; + spin_unlock(&mq_lock); ++ put_vx_info(vxi); + free_uid(user); + } + if (ipc_ns) +diff -NurpP --minimal linux-3.0.9/ipc/msg.c linux-3.0.9-vs2.3.2.1/ipc/msg.c +--- linux-3.0.9/ipc/msg.c 2011-05-22 16:17:59.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/ipc/msg.c 2011-06-10 22:11:24.000000000 +0200 +@@ -37,6 +37,7 @@ + #include <linux/rwsem.h> + #include <linux/nsproxy.h> + #include <linux/ipc_namespace.h> ++#include <linux/vs_base.h> + + #include <asm/current.h> + #include <asm/uaccess.h> +@@ -190,6 +191,7 @@ static int newque(struct ipc_namespace * + + msq->q_perm.mode = msgflg & S_IRWXUGO; + msq->q_perm.key = key; ++ msq->q_perm.xid = vx_current_xid(); + + msq->q_perm.security = NULL; + retval = security_msg_queue_alloc(msq); +diff -NurpP --minimal linux-3.0.9/ipc/namespace.c linux-3.0.9-vs2.3.2.1/ipc/namespace.c +--- linux-3.0.9/ipc/namespace.c 2011-07-22 11:18:12.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/ipc/namespace.c 2011-06-13 14:09:44.000000000 +0200 +@@ -13,11 +13,12 @@ + #include <linux/mount.h> + #include <linux/user_namespace.h> + #include <linux/proc_fs.h> ++#include <linux/vs_base.h> ++#include <linux/vserver/global.h> + + #include "util.h" + +-static struct ipc_namespace *create_ipc_ns(struct task_struct *tsk, +- struct ipc_namespace *old_ns) ++static struct ipc_namespace *create_ipc_ns(struct user_namespace *user_ns) + { + struct ipc_namespace *ns; + int err; +@@ -46,19 +47,18 @@ static struct ipc_namespace *create_ipc_ + ipcns_notify(IPCNS_CREATED); + register_ipcns_notifier(ns); + +- ns->user_ns = get_user_ns(task_cred_xxx(tsk, user)->user_ns); ++ ns->user_ns = get_user_ns(user_ns); + + return ns; + } + + struct ipc_namespace *copy_ipcs(unsigned long flags, +- struct task_struct *tsk) ++ struct ipc_namespace *old_ns, ++ struct user_namespace *user_ns) + { +- struct ipc_namespace *ns = tsk->nsproxy->ipc_ns; +- + if (!(flags & CLONE_NEWIPC)) +- return get_ipc_ns(ns); +- return create_ipc_ns(tsk, ns); ++ return get_ipc_ns(old_ns); ++ return create_ipc_ns(user_ns); + } + + /* +diff -NurpP --minimal linux-3.0.9/ipc/sem.c linux-3.0.9-vs2.3.2.1/ipc/sem.c +--- linux-3.0.9/ipc/sem.c 2011-11-15 16:40:47.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/ipc/sem.c 2011-08-08 23:04:47.000000000 +0200 +@@ -86,6 +86,8 @@ + #include <linux/rwsem.h> + #include <linux/nsproxy.h> + #include <linux/ipc_namespace.h> ++#include <linux/vs_base.h> ++#include <linux/vs_limit.h> + + #include <asm/uaccess.h> + #include "util.h" +@@ -260,6 +262,7 @@ static int newary(struct ipc_namespace * + + sma->sem_perm.mode = (semflg & S_IRWXUGO); + sma->sem_perm.key = key; ++ sma->sem_perm.xid = vx_current_xid(); + + sma->sem_perm.security = NULL; + retval = security_sem_alloc(sma); +@@ -275,6 +278,9 @@ static int newary(struct ipc_namespace * + return id; + } + ns->used_sems += nsems; ++ /* FIXME: obsoleted? */ ++ vx_semary_inc(sma); ++ vx_nsems_add(sma, nsems); + + sma->sem_base = (struct sem *) &sma[1]; + +@@ -730,6 +736,9 @@ static void freeary(struct ipc_namespace + + wake_up_sem_queue_do(&tasks); + ns->used_sems -= sma->sem_nsems; ++ /* FIXME: obsoleted? */ ++ vx_nsems_sub(sma, sma->sem_nsems); ++ vx_semary_dec(sma); + security_sem_free(sma); + ipc_rcu_putref(sma); + } +diff -NurpP --minimal linux-3.0.9/ipc/shm.c linux-3.0.9-vs2.3.2.1/ipc/shm.c +--- linux-3.0.9/ipc/shm.c 2011-07-22 11:18:12.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/ipc/shm.c 2011-06-10 22:11:24.000000000 +0200 +@@ -39,6 +39,8 @@ + #include <linux/nsproxy.h> + #include <linux/mount.h> + #include <linux/ipc_namespace.h> ++#include <linux/vs_context.h> ++#include <linux/vs_limit.h> + + #include <asm/uaccess.h> + +@@ -173,7 +175,12 @@ static void shm_open(struct vm_area_stru + */ + static void shm_destroy(struct ipc_namespace *ns, struct shmid_kernel *shp) + { +- ns->shm_tot -= (shp->shm_segsz + PAGE_SIZE - 1) >> PAGE_SHIFT; ++ struct vx_info *vxi = lookup_vx_info(shp->shm_perm.xid); ++ int numpages = (shp->shm_segsz + PAGE_SIZE - 1) >> PAGE_SHIFT; ++ ++ vx_ipcshm_sub(vxi, shp, numpages); ++ ns->shm_tot -= numpages; ++ + shm_rmid(ns, shp); + shm_unlock(shp); + if (!is_file_hugepages(shp->shm_file)) +@@ -183,6 +190,7 @@ static void shm_destroy(struct ipc_names + shp->mlock_user); + fput (shp->shm_file); + security_shm_free(shp); ++ put_vx_info(vxi); + ipc_rcu_putref(shp); + } + +@@ -355,11 +363,15 @@ static int newseg(struct ipc_namespace * + if (ns->shm_tot + numpages > ns->shm_ctlall) + return -ENOSPC; + ++ if (!vx_ipcshm_avail(current_vx_info(), numpages)) ++ return -ENOSPC; ++ + shp = ipc_rcu_alloc(sizeof(*shp)); + if (!shp) + return -ENOMEM; + + shp->shm_perm.key = key; ++ shp->shm_perm.xid = vx_current_xid(); + shp->shm_perm.mode = (shmflg & S_IRWXUGO); + shp->mlock_user = NULL; + +@@ -413,6 +425,7 @@ static int newseg(struct ipc_namespace * + ns->shm_tot += numpages; + error = shp->shm_perm.id; + shm_unlock(shp); ++ vx_ipcshm_add(current_vx_info(), key, numpages); + return error; + + no_id: +diff -NurpP --minimal linux-3.0.9/kernel/Makefile linux-3.0.9-vs2.3.2.1/kernel/Makefile +--- linux-3.0.9/kernel/Makefile 2011-07-22 11:18:12.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/kernel/Makefile 2011-06-10 22:11:24.000000000 +0200 +@@ -24,6 +24,7 @@ CFLAGS_REMOVE_sched_clock.o = -pg + CFLAGS_REMOVE_irq_work.o = -pg + endif + ++obj-y += vserver/ + obj-$(CONFIG_FREEZER) += freezer.o + obj-$(CONFIG_PROFILING) += profile.o + obj-$(CONFIG_SYSCTL_SYSCALL_CHECK) += sysctl_check.o +diff -NurpP --minimal linux-3.0.9/kernel/capability.c linux-3.0.9-vs2.3.2.1/kernel/capability.c +--- linux-3.0.9/kernel/capability.c 2011-07-22 11:18:12.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/kernel/capability.c 2011-10-27 13:59:20.000000000 +0200 +@@ -15,6 +15,7 @@ + #include <linux/syscalls.h> + #include <linux/pid_namespace.h> + #include <linux/user_namespace.h> ++#include <linux/vs_context.h> + #include <asm/uaccess.h> + + /* +@@ -116,6 +117,7 @@ static int cap_validate_magic(cap_user_h + return 0; + } + ++ + /* + * The only thing that can change the capabilities of the current + * process is the current process. As such, we can't be in this code +@@ -340,6 +342,8 @@ bool has_capability_noaudit(struct task_ + return (ret == 0); + } + ++#include <linux/vserver/base.h> ++ + /** + * capable - Determine if the current task has a superior capability in effect + * @cap: The capability to be tested for +diff -NurpP --minimal linux-3.0.9/kernel/compat.c linux-3.0.9-vs2.3.2.1/kernel/compat.c +--- linux-3.0.9/kernel/compat.c 2011-07-22 11:18:12.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/kernel/compat.c 2011-06-10 22:11:24.000000000 +0200 +@@ -970,7 +970,7 @@ asmlinkage long compat_sys_stime(compat_ + if (err) + return err; + +- do_settimeofday(&tv); ++ vx_settimeofday(&tv); + return 0; + } + +diff -NurpP --minimal linux-3.0.9/kernel/cred.c linux-3.0.9-vs2.3.2.1/kernel/cred.c +--- linux-3.0.9/kernel/cred.c 2011-07-22 11:18:12.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/kernel/cred.c 2011-06-10 22:11:24.000000000 +0200 +@@ -61,31 +61,6 @@ struct cred init_cred = { + #endif + }; + +-static inline void set_cred_subscribers(struct cred *cred, int n) +-{ +-#ifdef CONFIG_DEBUG_CREDENTIALS +- atomic_set(&cred->subscribers, n); +-#endif +-} +- +-static inline int read_cred_subscribers(const struct cred *cred) +-{ +-#ifdef CONFIG_DEBUG_CREDENTIALS +- return atomic_read(&cred->subscribers); +-#else +- return 0; +-#endif +-} +- +-static inline void alter_cred_subscribers(const struct cred *_cred, int n) +-{ +-#ifdef CONFIG_DEBUG_CREDENTIALS +- struct cred *cred = (struct cred *) _cred; +- +- atomic_add(n, &cred->subscribers); +-#endif +-} +- + /* + * Dispose of the shared task group credentials + */ +@@ -281,21 +256,16 @@ error: + * + * Call commit_creds() or abort_creds() to clean up. + */ +-struct cred *prepare_creds(void) ++struct cred *__prepare_creds(const struct cred *old) + { +- struct task_struct *task = current; +- const struct cred *old; + struct cred *new; + +- validate_process_creds(); +- + new = kmem_cache_alloc(cred_jar, GFP_KERNEL); + if (!new) + return NULL; + + kdebug("prepare_creds() alloc %p", new); + +- old = task->cred; + memcpy(new, old, sizeof(struct cred)); + + atomic_set(&new->usage, 1); +@@ -322,6 +292,13 @@ error: + abort_creds(new); + return NULL; + } ++ ++struct cred *prepare_creds(void) ++{ ++ validate_process_creds(); ++ ++ return __prepare_creds(current->cred); ++} + EXPORT_SYMBOL(prepare_creds); + + /* +diff -NurpP --minimal linux-3.0.9/kernel/exit.c linux-3.0.9-vs2.3.2.1/kernel/exit.c +--- linux-3.0.9/kernel/exit.c 2011-07-22 11:18:12.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/kernel/exit.c 2011-06-22 12:39:15.000000000 +0200 +@@ -48,6 +48,10 @@ + #include <linux/fs_struct.h> + #include <linux/init_task.h> + #include <linux/perf_event.h> ++#include <linux/vs_limit.h> ++#include <linux/vs_context.h> ++#include <linux/vs_network.h> ++#include <linux/vs_pid.h> + #include <trace/events/sched.h> + #include <linux/hw_breakpoint.h> + #include <linux/oom.h> +@@ -494,9 +498,11 @@ static void close_files(struct files_str + filp_close(file, files); + cond_resched(); + } ++ vx_openfd_dec(i); + } + i++; + set >>= 1; ++ cond_resched(); + } + } + } +@@ -1047,11 +1053,16 @@ NORET_TYPE void do_exit(long code) + + validate_creds_for_do_exit(tsk); + ++ /* needs to stay after exit_notify() */ ++ exit_vx_info(tsk, code); ++ exit_nx_info(tsk); ++ + preempt_disable(); + exit_rcu(); + /* causes final put_task_struct in finish_task_switch(). */ + tsk->state = TASK_DEAD; + schedule(); ++ printk("bad task: %p [%lx]\n", current, current->state); + BUG(); + /* Avoid "noreturn function does return". */ + for (;;) +diff -NurpP --minimal linux-3.0.9/kernel/fork.c linux-3.0.9-vs2.3.2.1/kernel/fork.c +--- linux-3.0.9/kernel/fork.c 2011-07-22 11:18:12.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/kernel/fork.c 2011-06-10 22:21:02.000000000 +0200 +@@ -67,6 +67,10 @@ + #include <linux/user-return-notifier.h> + #include <linux/oom.h> + #include <linux/khugepaged.h> ++#include <linux/vs_context.h> ++#include <linux/vs_network.h> ++#include <linux/vs_limit.h> ++#include <linux/vs_memory.h> + + #include <asm/pgtable.h> + #include <asm/pgalloc.h> +@@ -167,6 +171,8 @@ void free_task(struct task_struct *tsk) + account_kernel_stack(tsk->stack, -1); + free_thread_info(tsk->stack); + rt_mutex_debug_task_free(tsk); ++ clr_vx_info(&tsk->vx_info); ++ clr_nx_info(&tsk->nx_info); + ftrace_graph_exit_task(tsk); + free_task_struct(tsk); + } +@@ -505,6 +511,7 @@ static struct mm_struct * mm_init(struct + if (likely(!mm_alloc_pgd(mm))) { + mm->def_flags = 0; + mmu_notifier_mm_init(mm); ++ set_vx_info(&mm->mm_vx_info, p->vx_info); + return mm; + } + +@@ -542,6 +549,7 @@ void __mmdrop(struct mm_struct *mm) + #ifdef CONFIG_TRANSPARENT_HUGEPAGE + VM_BUG_ON(mm->pmd_huge_pte); + #endif ++ clr_vx_info(&mm->mm_vx_info); + free_mm(mm); + } + EXPORT_SYMBOL_GPL(__mmdrop); +@@ -729,6 +737,7 @@ struct mm_struct *dup_mm(struct task_str + goto fail_nomem; + + memcpy(mm, oldmm, sizeof(*mm)); ++ mm->mm_vx_info = NULL; + mm_init_cpumask(mm); + + /* Initializing for Swap token stuff */ +@@ -772,6 +781,7 @@ fail_nocontext: + * If init_new_context() failed, we cannot use mmput() to free the mm + * because it calls destroy_context() + */ ++ clr_vx_info(&mm->mm_vx_info); + mm_free_pgd(mm); + free_mm(mm); + return NULL; +@@ -1057,6 +1067,8 @@ static struct task_struct *copy_process( + int retval; + struct task_struct *p; + int cgroup_callbacks_done = 0; ++ struct vx_info *vxi; ++ struct nx_info *nxi; + + if ((clone_flags & (CLONE_NEWNS|CLONE_FS)) == (CLONE_NEWNS|CLONE_FS)) + return ERR_PTR(-EINVAL); +@@ -1103,7 +1115,12 @@ static struct task_struct *copy_process( + DEBUG_LOCKS_WARN_ON(!p->hardirqs_enabled); + DEBUG_LOCKS_WARN_ON(!p->softirqs_enabled); + #endif ++ init_vx_info(&p->vx_info, current_vx_info()); ++ init_nx_info(&p->nx_info, current_nx_info()); ++ + retval = -EAGAIN; ++ if (!vx_nproc_avail(1)) ++ goto bad_fork_free; + if (atomic_read(&p->real_cred->user->processes) >= + task_rlimit(p, RLIMIT_NPROC)) { + if (!capable(CAP_SYS_ADMIN) && !capable(CAP_SYS_RESOURCE) && +@@ -1360,6 +1377,18 @@ static struct task_struct *copy_process( + + total_forks++; + spin_unlock(¤t->sighand->siglock); ++ ++ /* p is copy of current */ ++ vxi = p->vx_info; ++ if (vxi) { ++ claim_vx_info(vxi, p); ++ atomic_inc(&vxi->cvirt.nr_threads); ++ atomic_inc(&vxi->cvirt.total_forks); ++ vx_nproc_inc(p); ++ } ++ nxi = p->nx_info; ++ if (nxi) ++ claim_nx_info(nxi, p); + write_unlock_irq(&tasklist_lock); + proc_fork_connector(p); + cgroup_post_fork(p); +diff -NurpP --minimal linux-3.0.9/kernel/kthread.c linux-3.0.9-vs2.3.2.1/kernel/kthread.c +--- linux-3.0.9/kernel/kthread.c 2011-07-22 11:18:12.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/kernel/kthread.c 2011-06-10 22:11:24.000000000 +0200 +@@ -16,6 +16,7 @@ + #include <linux/mutex.h> + #include <linux/slab.h> + #include <linux/freezer.h> ++#include <linux/vs_pid.h> + #include <trace/events/sched.h> + + static DEFINE_SPINLOCK(kthread_create_lock); +diff -NurpP --minimal linux-3.0.9/kernel/nsproxy.c linux-3.0.9-vs2.3.2.1/kernel/nsproxy.c +--- linux-3.0.9/kernel/nsproxy.c 2011-07-22 11:18:12.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/kernel/nsproxy.c 2011-06-13 14:09:44.000000000 +0200 +@@ -20,6 +20,8 @@ + #include <linux/mnt_namespace.h> + #include <linux/utsname.h> + #include <linux/pid_namespace.h> ++#include <linux/vserver/global.h> ++#include <linux/vserver/debug.h> + #include <net/net_namespace.h> + #include <linux/ipc_namespace.h> + #include <linux/proc_fs.h> +@@ -46,8 +48,11 @@ static inline struct nsproxy *create_nsp + struct nsproxy *nsproxy; + + nsproxy = kmem_cache_alloc(nsproxy_cachep, GFP_KERNEL); +- if (nsproxy) ++ if (nsproxy) { + atomic_set(&nsproxy->count, 1); ++ atomic_inc(&vs_global_nsproxy); ++ } ++ vxdprintk(VXD_CBIT(space, 2), "create_nsproxy = %p[1]", nsproxy); + return nsproxy; + } + +@@ -56,8 +61,11 @@ static inline struct nsproxy *create_nsp + * Return the newly created nsproxy. Do not attach this to the task, + * leave it to the caller to do proper locking and attach it to task. + */ +-static struct nsproxy *create_new_namespaces(unsigned long flags, +- struct task_struct *tsk, struct fs_struct *new_fs) ++static struct nsproxy *unshare_namespaces(unsigned long flags, ++ struct nsproxy *orig, ++ struct fs_struct *new_fs, ++ struct user_namespace *new_user, ++ struct pid_namespace *new_pid) + { + struct nsproxy *new_nsp; + int err; +@@ -66,31 +74,31 @@ static struct nsproxy *create_new_namesp + if (!new_nsp) + return ERR_PTR(-ENOMEM); + +- new_nsp->mnt_ns = copy_mnt_ns(flags, tsk->nsproxy->mnt_ns, new_fs); ++ new_nsp->mnt_ns = copy_mnt_ns(flags, orig->mnt_ns, new_fs); + if (IS_ERR(new_nsp->mnt_ns)) { + err = PTR_ERR(new_nsp->mnt_ns); + goto out_ns; + } + +- new_nsp->uts_ns = copy_utsname(flags, tsk); ++ new_nsp->uts_ns = copy_utsname(flags, orig->uts_ns, new_user); + if (IS_ERR(new_nsp->uts_ns)) { + err = PTR_ERR(new_nsp->uts_ns); + goto out_uts; + } + +- new_nsp->ipc_ns = copy_ipcs(flags, tsk); ++ new_nsp->ipc_ns = copy_ipcs(flags, orig->ipc_ns, new_user); + if (IS_ERR(new_nsp->ipc_ns)) { + err = PTR_ERR(new_nsp->ipc_ns); + goto out_ipc; + } + +- new_nsp->pid_ns = copy_pid_ns(flags, task_active_pid_ns(tsk)); ++ new_nsp->pid_ns = copy_pid_ns(flags, new_pid); + if (IS_ERR(new_nsp->pid_ns)) { + err = PTR_ERR(new_nsp->pid_ns); + goto out_pid; + } + +- new_nsp->net_ns = copy_net_ns(flags, tsk->nsproxy->net_ns); ++ new_nsp->net_ns = copy_net_ns(flags, orig->net_ns); + if (IS_ERR(new_nsp->net_ns)) { + err = PTR_ERR(new_nsp->net_ns); + goto out_net; +@@ -115,6 +123,40 @@ out_ns: + return ERR_PTR(err); + } + ++static struct nsproxy *create_new_namespaces(unsigned long flags, ++ struct task_struct *tsk, struct fs_struct *new_fs) ++{ ++ return unshare_namespaces(flags, tsk->nsproxy, ++ new_fs, task_cred_xxx(tsk, user)->user_ns, ++ task_active_pid_ns(tsk)); ++} ++ ++/* ++ * copies the nsproxy, setting refcount to 1, and grabbing a ++ * reference to all contained namespaces. ++ */ ++struct nsproxy *copy_nsproxy(struct nsproxy *orig) ++{ ++ struct nsproxy *ns = create_nsproxy(); ++ ++ if (ns) { ++ memcpy(ns, orig, sizeof(struct nsproxy)); ++ atomic_set(&ns->count, 1); ++ ++ if (ns->mnt_ns) ++ get_mnt_ns(ns->mnt_ns); ++ if (ns->uts_ns) ++ get_uts_ns(ns->uts_ns); ++ if (ns->ipc_ns) ++ get_ipc_ns(ns->ipc_ns); ++ if (ns->pid_ns) ++ get_pid_ns(ns->pid_ns); ++ if (ns->net_ns) ++ get_net(ns->net_ns); ++ } ++ return ns; ++} ++ + /* + * called from clone. This now handles copy for nsproxy and all + * namespaces therein. +@@ -122,9 +164,12 @@ out_ns: + int copy_namespaces(unsigned long flags, struct task_struct *tsk) + { + struct nsproxy *old_ns = tsk->nsproxy; +- struct nsproxy *new_ns; ++ struct nsproxy *new_ns = NULL; + int err = 0; + ++ vxdprintk(VXD_CBIT(space, 7), "copy_namespaces(0x%08lx,%p[%p])", ++ flags, tsk, old_ns); ++ + if (!old_ns) + return 0; + +@@ -134,7 +179,7 @@ int copy_namespaces(unsigned long flags, + CLONE_NEWPID | CLONE_NEWNET))) + return 0; + +- if (!capable(CAP_SYS_ADMIN)) { ++ if (!vx_can_unshare(CAP_SYS_ADMIN, flags)) { + err = -EPERM; + goto out; + } +@@ -161,6 +206,9 @@ int copy_namespaces(unsigned long flags, + + out: + put_nsproxy(old_ns); ++ vxdprintk(VXD_CBIT(space, 3), ++ "copy_namespaces(0x%08lx,%p[%p]) = %d [%p]", ++ flags, tsk, old_ns, err, new_ns); + return err; + } + +@@ -174,7 +222,9 @@ void free_nsproxy(struct nsproxy *ns) + put_ipc_ns(ns->ipc_ns); + if (ns->pid_ns) + put_pid_ns(ns->pid_ns); +- put_net(ns->net_ns); ++ if (ns->net_ns) ++ put_net(ns->net_ns); ++ atomic_dec(&vs_global_nsproxy); + kmem_cache_free(nsproxy_cachep, ns); + } + +@@ -187,11 +237,15 @@ int unshare_nsproxy_namespaces(unsigned + { + int err = 0; + ++ vxdprintk(VXD_CBIT(space, 4), ++ "unshare_nsproxy_namespaces(0x%08lx,[%p])", ++ unshare_flags, current->nsproxy); ++ + if (!(unshare_flags & (CLONE_NEWNS | CLONE_NEWUTS | CLONE_NEWIPC | + CLONE_NEWNET))) + return 0; + +- if (!capable(CAP_SYS_ADMIN)) ++ if (!vx_can_unshare(CAP_SYS_ADMIN, unshare_flags)) + return -EPERM; + + *new_nsp = create_new_namespaces(unshare_flags, current, +diff -NurpP --minimal linux-3.0.9/kernel/pid.c linux-3.0.9-vs2.3.2.1/kernel/pid.c +--- linux-3.0.9/kernel/pid.c 2011-05-22 16:17:59.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/kernel/pid.c 2011-06-10 22:11:24.000000000 +0200 +@@ -36,6 +36,7 @@ + #include <linux/pid_namespace.h> + #include <linux/init_task.h> + #include <linux/syscalls.h> ++#include <linux/vs_pid.h> + + #define pid_hashfn(nr, ns) \ + hash_long((unsigned long)nr + (unsigned long)ns, pidhash_shift) +@@ -342,7 +343,7 @@ EXPORT_SYMBOL_GPL(find_pid_ns); + + struct pid *find_vpid(int nr) + { +- return find_pid_ns(nr, current->nsproxy->pid_ns); ++ return find_pid_ns(vx_rmap_pid(nr), current->nsproxy->pid_ns); + } + EXPORT_SYMBOL_GPL(find_vpid); + +@@ -402,6 +403,9 @@ void transfer_pid(struct task_struct *ol + struct task_struct *pid_task(struct pid *pid, enum pid_type type) + { + struct task_struct *result = NULL; ++ ++ if (type == PIDTYPE_REALPID) ++ type = PIDTYPE_PID; + if (pid) { + struct hlist_node *first; + first = rcu_dereference_check(hlist_first_rcu(&pid->tasks[type]), +@@ -420,7 +424,7 @@ EXPORT_SYMBOL(pid_task); + struct task_struct *find_task_by_pid_ns(pid_t nr, struct pid_namespace *ns) + { + rcu_lockdep_assert(rcu_read_lock_held()); +- return pid_task(find_pid_ns(nr, ns), PIDTYPE_PID); ++ return pid_task(find_pid_ns(vx_rmap_pid(nr), ns), PIDTYPE_PID); + } + + struct task_struct *find_task_by_vpid(pid_t vnr) +@@ -464,7 +468,7 @@ struct pid *find_get_pid(pid_t nr) + } + EXPORT_SYMBOL_GPL(find_get_pid); + +-pid_t pid_nr_ns(struct pid *pid, struct pid_namespace *ns) ++pid_t pid_unmapped_nr_ns(struct pid *pid, struct pid_namespace *ns) + { + struct upid *upid; + pid_t nr = 0; +@@ -477,6 +481,11 @@ pid_t pid_nr_ns(struct pid *pid, struct + return nr; + } + ++pid_t pid_nr_ns(struct pid *pid, struct pid_namespace *ns) ++{ ++ return vx_map_pid(pid_unmapped_nr_ns(pid, ns)); ++} ++ + pid_t pid_vnr(struct pid *pid) + { + return pid_nr_ns(pid, current->nsproxy->pid_ns); +diff -NurpP --minimal linux-3.0.9/kernel/pid_namespace.c linux-3.0.9-vs2.3.2.1/kernel/pid_namespace.c +--- linux-3.0.9/kernel/pid_namespace.c 2011-05-22 16:17:59.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/kernel/pid_namespace.c 2011-06-10 22:17:45.000000000 +0200 +@@ -15,6 +15,7 @@ + #include <linux/acct.h> + #include <linux/slab.h> + #include <linux/proc_fs.h> ++#include <linux/vserver/global.h> + + #define BITS_PER_PAGE (PAGE_SIZE*8) + +@@ -88,6 +89,7 @@ static struct pid_namespace *create_pid_ + goto out_free_map; + + kref_init(&ns->kref); ++ atomic_inc(&vs_global_pid_ns); + ns->level = level; + ns->parent = get_pid_ns(parent_pid_ns); + +@@ -119,6 +121,7 @@ static void destroy_pid_namespace(struct + + for (i = 0; i < PIDMAP_ENTRIES; i++) + kfree(ns->pidmap[i].page); ++ atomic_dec(&vs_global_pid_ns); + kmem_cache_free(pid_ns_cachep, ns); + } + +diff -NurpP --minimal linux-3.0.9/kernel/posix-timers.c linux-3.0.9-vs2.3.2.1/kernel/posix-timers.c +--- linux-3.0.9/kernel/posix-timers.c 2011-07-22 11:18:12.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/kernel/posix-timers.c 2011-06-10 22:11:24.000000000 +0200 +@@ -47,6 +47,7 @@ + #include <linux/wait.h> + #include <linux/workqueue.h> + #include <linux/module.h> ++#include <linux/vs_context.h> + + /* + * Management arrays for POSIX timers. Timers are kept in slab memory +@@ -340,6 +341,7 @@ int posix_timer_event(struct k_itimer *t + { + struct task_struct *task; + int shared, ret = -1; ++ + /* + * FIXME: if ->sigq is queued we can race with + * dequeue_signal()->do_schedule_next_timer(). +@@ -356,10 +358,18 @@ int posix_timer_event(struct k_itimer *t + rcu_read_lock(); + task = pid_task(timr->it_pid, PIDTYPE_PID); + if (task) { ++ struct vx_info_save vxis; ++ struct vx_info *vxi; ++ ++ vxi = get_vx_info(task->vx_info); ++ enter_vx_info(vxi, &vxis); + shared = !(timr->it_sigev_notify & SIGEV_THREAD_ID); + ret = send_sigqueue(timr->sigq, task, shared); ++ leave_vx_info(&vxis); ++ put_vx_info(vxi); + } + rcu_read_unlock(); ++ + /* If we failed to send the signal the timer stops. */ + return ret > 0; + } +diff -NurpP --minimal linux-3.0.9/kernel/printk.c linux-3.0.9-vs2.3.2.1/kernel/printk.c +--- linux-3.0.9/kernel/printk.c 2011-11-15 16:40:47.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/kernel/printk.c 2011-10-18 13:51:13.000000000 +0200 +@@ -41,6 +41,7 @@ + #include <linux/cpu.h> + #include <linux/notifier.h> + #include <linux/rculist.h> ++#include <linux/vs_cvirt.h> + + #include <asm/uaccess.h> + +@@ -314,7 +315,7 @@ static int check_syslog_permissions(int + return 0; + + if (syslog_action_restricted(type)) { +- if (capable(CAP_SYSLOG)) ++ if (vx_capable(CAP_SYSLOG, VXC_SYSLOG)) + return 0; + /* For historical reasons, accept CAP_SYS_ADMIN too, with a warning */ + if (capable(CAP_SYS_ADMIN)) { +@@ -342,12 +343,9 @@ int do_syslog(int type, char __user *buf + if (error) + return error; + +- switch (type) { +- case SYSLOG_ACTION_CLOSE: /* Close log */ +- break; +- case SYSLOG_ACTION_OPEN: /* Open log */ +- break; +- case SYSLOG_ACTION_READ: /* Read from log */ ++ if ((type == SYSLOG_ACTION_READ) || ++ (type == SYSLOG_ACTION_READ_ALL) || ++ (type == SYSLOG_ACTION_READ_CLEAR)) { + error = -EINVAL; + if (!buf || len < 0) + goto out; +@@ -358,6 +356,16 @@ int do_syslog(int type, char __user *buf + error = -EFAULT; + goto out; + } ++ } ++ if (!vx_check(0, VS_ADMIN|VS_WATCH)) ++ return vx_do_syslog(type, buf, len); ++ ++ switch (type) { ++ case SYSLOG_ACTION_CLOSE: /* Close log */ ++ break; ++ case SYSLOG_ACTION_OPEN: /* Open log */ ++ break; ++ case SYSLOG_ACTION_READ: /* Read from log */ + error = wait_event_interruptible(log_wait, + (log_start - log_end)); + if (error) +@@ -384,16 +392,6 @@ int do_syslog(int type, char __user *buf + /* FALL THRU */ + /* Read last kernel messages */ + case SYSLOG_ACTION_READ_ALL: +- error = -EINVAL; +- if (!buf || len < 0) +- goto out; +- error = 0; +- if (!len) +- goto out; +- if (!access_ok(VERIFY_WRITE, buf, len)) { +- error = -EFAULT; +- goto out; +- } + count = len; + if (count > log_buf_len) + count = log_buf_len; +diff -NurpP --minimal linux-3.0.9/kernel/ptrace.c linux-3.0.9-vs2.3.2.1/kernel/ptrace.c +--- linux-3.0.9/kernel/ptrace.c 2011-07-22 11:18:12.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/kernel/ptrace.c 2011-10-27 16:12:46.000000000 +0200 +@@ -22,6 +22,7 @@ + #include <linux/syscalls.h> + #include <linux/uaccess.h> + #include <linux/regset.h> ++#include <linux/vs_context.h> + #include <linux/hw_breakpoint.h> + + +@@ -145,9 +146,15 @@ int __ptrace_may_access(struct task_stru + * or halting the specified task is impossible. + */ + int dumpable = 0; ++ + /* Don't let security modules deny introspection */ + if (task == current) + return 0; ++ ++ vxdprintk(VXD_CBIT(perm, 8), ++ "__ptrace_may_access(%p[#%d,%d,%d], %d)", ++ task, task->xid, task->pid, task->tgid, mode); ++ + rcu_read_lock(); + tcred = __task_cred(task); + if (cred->user->user_ns == tcred->user->user_ns && +@@ -161,6 +168,8 @@ int __ptrace_may_access(struct task_stru + if (ns_capable(tcred->user->user_ns, CAP_SYS_PTRACE)) + goto ok; + rcu_read_unlock(); ++ vxdprintk(VXD_CBIT(perm, 8), ++ "__ptrace_may_access(%p) cred/cap failed", task); + return -EPERM; + ok: + rcu_read_unlock(); +@@ -169,6 +178,24 @@ ok: + dumpable = get_dumpable(task->mm); + if (!dumpable && !task_ns_capable(task, CAP_SYS_PTRACE)) + return -EPERM; ++ vxdprintk(VXD_CBIT(perm, 8), ++ "__ptrace_may_access(%p) cap/dump ok", task); ++ ++ if (!vx_check(task->xid, VS_ADMIN_P|VS_WATCH_P|VS_IDENT)) ++ return -EPERM; ++ vxdprintk(VXD_CBIT(perm, 8), ++ "__ptrace_may_access(%p) check ok", task); ++ ++ printk("%d,%d %d,%d\n", ++ vx_check(task->xid, VS_IDENT), ++ task_vx_flags(task, VXF_STATE_ADMIN, 0), ++ current->xid, task->xid); ++ ++ if (!vx_check(task->xid, VS_IDENT) && ++ !task_vx_flags(task, VXF_STATE_ADMIN, 0)) ++ return -EACCES; ++ vxdprintk(VXD_CBIT(perm, 8), ++ "__ptrace_may_access(%p) admin ok", task); + + return security_ptrace_access_check(task, mode); + } +diff -NurpP --minimal linux-3.0.9/kernel/sched.c linux-3.0.9-vs2.3.2.1/kernel/sched.c +--- linux-3.0.9/kernel/sched.c 2011-11-15 16:40:47.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/kernel/sched.c 2011-10-18 13:51:13.000000000 +0200 +@@ -71,6 +71,8 @@ + #include <linux/ctype.h> + #include <linux/ftrace.h> + #include <linux/slab.h> ++#include <linux/vs_sched.h> ++#include <linux/vs_cvirt.h> + + #include <asm/tlb.h> + #include <asm/irq_regs.h> +@@ -3461,9 +3463,17 @@ static void calc_global_nohz(unsigned lo + */ + void get_avenrun(unsigned long *loads, unsigned long offset, int shift) + { +- loads[0] = (avenrun[0] + offset) << shift; +- loads[1] = (avenrun[1] + offset) << shift; +- loads[2] = (avenrun[2] + offset) << shift; ++ if (vx_flags(VXF_VIRT_LOAD, 0)) { ++ struct vx_info *vxi = current_vx_info(); ++ ++ loads[0] = (vxi->cvirt.load[0] + offset) << shift; ++ loads[1] = (vxi->cvirt.load[1] + offset) << shift; ++ loads[2] = (vxi->cvirt.load[2] + offset) << shift; ++ } else { ++ loads[0] = (avenrun[0] + offset) << shift; ++ loads[1] = (avenrun[1] + offset) << shift; ++ loads[2] = (avenrun[2] + offset) << shift; ++ } + } + + /* +@@ -3722,16 +3732,19 @@ void account_user_time(struct task_struc + cputime_t cputime_scaled) + { + struct cpu_usage_stat *cpustat = &kstat_this_cpu.cpustat; ++ struct vx_info *vxi = p->vx_info; /* p is _always_ current */ + cputime64_t tmp; ++ int nice = (TASK_NICE(p) > 0); + + /* Add user time to process. */ + p->utime = cputime_add(p->utime, cputime); + p->utimescaled = cputime_add(p->utimescaled, cputime_scaled); ++ vx_account_user(vxi, cputime, nice); + account_group_user_time(p, cputime); + + /* Add user time to cpustat. */ + tmp = cputime_to_cputime64(cputime); +- if (TASK_NICE(p) > 0) ++ if (nice) + cpustat->nice = cputime64_add(cpustat->nice, tmp); + else + cpustat->user = cputime64_add(cpustat->user, tmp); +@@ -3783,10 +3796,12 @@ void __account_system_time(struct task_s + cputime_t cputime_scaled, cputime64_t *target_cputime64) + { + cputime64_t tmp = cputime_to_cputime64(cputime); ++ struct vx_info *vxi = p->vx_info; /* p is _always_ current */ + + /* Add system time to process. */ + p->stime = cputime_add(p->stime, cputime); + p->stimescaled = cputime_add(p->stimescaled, cputime_scaled); ++ vx_account_system(vxi, cputime, 0 /* do we have idle time? */); + account_group_system_time(p, cputime); + + /* Add system time to cpustat. */ +@@ -4954,7 +4969,7 @@ SYSCALL_DEFINE1(nice, int, increment) + nice = 19; + + if (increment < 0 && !can_nice(current, nice)) +- return -EPERM; ++ return vx_flags(VXF_IGNEG_NICE, 0) ? 0 : -EPERM; + + retval = security_task_setnice(current, nice); + if (retval) +diff -NurpP --minimal linux-3.0.9/kernel/sched_fair.c linux-3.0.9-vs2.3.2.1/kernel/sched_fair.c +--- linux-3.0.9/kernel/sched_fair.c 2011-07-22 11:18:12.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/kernel/sched_fair.c 2011-07-22 11:20:39.000000000 +0200 +@@ -998,6 +998,8 @@ enqueue_entity(struct cfs_rq *cfs_rq, st + __enqueue_entity(cfs_rq, se); + se->on_rq = 1; + ++ if (entity_is_task(se)) ++ vx_activate_task(task_of(se)); + if (cfs_rq->nr_running == 1) + list_add_leaf_cfs_rq(cfs_rq); + } +@@ -1074,6 +1076,8 @@ dequeue_entity(struct cfs_rq *cfs_rq, st + if (se != cfs_rq->curr) + __dequeue_entity(cfs_rq, se); + se->on_rq = 0; ++ if (entity_is_task(se)) ++ vx_deactivate_task(task_of(se)); + update_cfs_load(cfs_rq, 0); + account_entity_dequeue(cfs_rq, se); + +diff -NurpP --minimal linux-3.0.9/kernel/signal.c linux-3.0.9-vs2.3.2.1/kernel/signal.c +--- linux-3.0.9/kernel/signal.c 2011-11-15 16:40:47.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/kernel/signal.c 2011-11-15 17:37:07.000000000 +0100 +@@ -28,6 +28,8 @@ + #include <linux/freezer.h> + #include <linux/pid_namespace.h> + #include <linux/nsproxy.h> ++#include <linux/vs_context.h> ++#include <linux/vs_pid.h> + #define CREATE_TRACE_POINTS + #include <trace/events/signal.h> + +@@ -744,9 +746,18 @@ static int check_kill_permission(int sig + struct pid *sid; + int error; + ++ vxdprintk(VXD_CBIT(misc, 7), ++ "check_kill_permission(%d,%p,%p[#%u,%u])", ++ sig, info, t, vx_task_xid(t), t->pid); ++ + if (!valid_signal(sig)) + return -EINVAL; + ++/* FIXME: needed? if so, why? ++ if ((info != SEND_SIG_NOINFO) && ++ (is_si_special(info) || !si_fromuser(info))) ++ goto skip; */ ++ + if (!si_fromuser(info)) + return 0; + +@@ -770,6 +781,20 @@ static int check_kill_permission(int sig + } + } + ++ error = -EPERM; ++ if (t->pid == 1 && current->xid) ++ return error; ++ ++ error = -ESRCH; ++ /* FIXME: we shouldn't return ESRCH ever, to avoid ++ loops, maybe ENOENT or EACCES? */ ++ if (!vx_check(vx_task_xid(t), VS_WATCH_P | VS_IDENT)) { ++ vxdprintk(current->xid || VXD_CBIT(misc, 7), ++ "signal %d[%p] xid mismatch %p[#%u,%u] xid=#%u", ++ sig, info, t, vx_task_xid(t), t->pid, current->xid); ++ return error; ++ } ++/* skip: */ + return security_task_kill(t, info, sig, 0); + } + +@@ -1246,7 +1271,7 @@ int kill_pid_info(int sig, struct siginf + rcu_read_lock(); + retry: + p = pid_task(pid, PIDTYPE_PID); +- if (p) { ++ if (p && vx_check(vx_task_xid(p), VS_IDENT)) { + error = group_send_sig_info(sig, info, p); + if (unlikely(error == -ESRCH)) + /* +@@ -1285,7 +1310,7 @@ int kill_pid_info_as_uid(int sig, struct + + rcu_read_lock(); + p = pid_task(pid, PIDTYPE_PID); +- if (!p) { ++ if (!p || !vx_check(vx_task_xid(p), VS_IDENT)) { + ret = -ESRCH; + goto out_unlock; + } +@@ -1340,8 +1365,10 @@ static int kill_something_info(int sig, + struct task_struct * p; + + for_each_process(p) { +- if (task_pid_vnr(p) > 1 && +- !same_thread_group(p, current)) { ++ if (vx_check(vx_task_xid(p), VS_ADMIN|VS_IDENT) && ++ task_pid_vnr(p) > 1 && ++ !same_thread_group(p, current) && ++ !vx_current_initpid(p->pid)) { + int err = group_send_sig_info(sig, info, p); + ++count; + if (err != -EPERM) +@@ -2138,6 +2165,11 @@ relock: + !sig_kernel_only(signr)) + continue; + ++ /* virtual init is protected against user signals */ ++ if ((info->si_code == SI_USER) && ++ vx_current_initpid(current->pid)) ++ continue; ++ + if (sig_kernel_stop(signr)) { + /* + * The default action is to stop all threads in +diff -NurpP --minimal linux-3.0.9/kernel/softirq.c linux-3.0.9-vs2.3.2.1/kernel/softirq.c +--- linux-3.0.9/kernel/softirq.c 2011-07-22 11:18:12.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/kernel/softirq.c 2011-07-22 11:20:39.000000000 +0200 +@@ -24,6 +24,7 @@ + #include <linux/ftrace.h> + #include <linux/smp.h> + #include <linux/tick.h> ++#include <linux/vs_context.h> + + #define CREATE_TRACE_POINTS + #include <trace/events/irq.h> +diff -NurpP --minimal linux-3.0.9/kernel/sys.c linux-3.0.9-vs2.3.2.1/kernel/sys.c +--- linux-3.0.9/kernel/sys.c 2011-11-15 16:40:47.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/kernel/sys.c 2011-11-15 17:37:07.000000000 +0100 +@@ -45,6 +45,7 @@ + #include <linux/syscalls.h> + #include <linux/kprobes.h> + #include <linux/user_namespace.h> ++#include <linux/vs_pid.h> + + #include <linux/kmsg_dump.h> + /* Move somewhere else to avoid recompiling? */ +@@ -155,7 +156,10 @@ static int set_one_prio(struct task_stru + goto out; + } + if (niceval < task_nice(p) && !can_nice(p, niceval)) { +- error = -EACCES; ++ if (vx_flags(VXF_IGNEG_NICE, 0)) ++ error = 0; ++ else ++ error = -EACCES; + goto out; + } + no_nice = security_task_setnice(p, niceval); +@@ -205,6 +209,8 @@ SYSCALL_DEFINE3(setpriority, int, which, + else + pgrp = task_pgrp(current); + do_each_pid_thread(pgrp, PIDTYPE_PGID, p) { ++ if (!vx_check(p->xid, VS_ADMIN_P | VS_IDENT)) ++ continue; + error = set_one_prio(p, niceval, error); + } while_each_pid_thread(pgrp, PIDTYPE_PGID, p); + break; +@@ -268,6 +274,8 @@ SYSCALL_DEFINE2(getpriority, int, which, + else + pgrp = task_pgrp(current); + do_each_pid_thread(pgrp, PIDTYPE_PGID, p) { ++ if (!vx_check(p->xid, VS_ADMIN_P | VS_IDENT)) ++ continue; + niceval = 20 - task_nice(p); + if (niceval > retval) + retval = niceval; +@@ -387,6 +395,8 @@ EXPORT_SYMBOL_GPL(kernel_power_off); + + static DEFINE_MUTEX(reboot_mutex); + ++long vs_reboot(unsigned int, void __user *); ++ + /* + * Reboot system call: for obvious reasons only root may call it, + * and even root needs to set up some magic numbers in the registers +@@ -419,6 +429,9 @@ SYSCALL_DEFINE4(reboot, int, magic1, int + if ((cmd == LINUX_REBOOT_CMD_POWER_OFF) && !pm_power_off) + cmd = LINUX_REBOOT_CMD_HALT; + ++ if (!vx_check(0, VS_ADMIN|VS_WATCH)) ++ return vs_reboot(cmd, arg); ++ + mutex_lock(&reboot_mutex); + switch (cmd) { + case LINUX_REBOOT_CMD_RESTART: +@@ -1235,7 +1248,8 @@ SYSCALL_DEFINE2(sethostname, char __user + int errno; + char tmp[__NEW_UTS_LEN]; + +- if (!ns_capable(current->nsproxy->uts_ns->user_ns, CAP_SYS_ADMIN)) ++ if (!vx_ns_capable(current->nsproxy->uts_ns->user_ns, ++ CAP_SYS_ADMIN, VXC_SET_UTSNAME)) + return -EPERM; + + if (len < 0 || len > __NEW_UTS_LEN) +@@ -1285,7 +1299,8 @@ SYSCALL_DEFINE2(setdomainname, char __us + int errno; + char tmp[__NEW_UTS_LEN]; + +- if (!ns_capable(current->nsproxy->uts_ns->user_ns, CAP_SYS_ADMIN)) ++ if (!vx_ns_capable(current->nsproxy->uts_ns->user_ns, ++ CAP_SYS_ADMIN, VXC_SET_UTSNAME)) + return -EPERM; + if (len < 0 || len > __NEW_UTS_LEN) + return -EINVAL; +@@ -1403,7 +1418,7 @@ int do_prlimit(struct task_struct *tsk, + /* Keep the capable check against init_user_ns until + cgroups can contain all limits */ + if (new_rlim->rlim_max > rlim->rlim_max && +- !capable(CAP_SYS_RESOURCE)) ++ !vx_capable(CAP_SYS_RESOURCE, VXC_SET_RLIMIT)) + retval = -EPERM; + if (!retval) + retval = security_task_setrlimit(tsk->group_leader, +@@ -1457,7 +1472,8 @@ static int check_prlimit_permission(stru + cred->gid == tcred->sgid && + cred->gid == tcred->gid)) + return 0; +- if (ns_capable(tcred->user->user_ns, CAP_SYS_RESOURCE)) ++ if (vx_ns_capable(tcred->user->user_ns, ++ CAP_SYS_RESOURCE, VXC_SET_RLIMIT)) + return 0; + + return -EPERM; +diff -NurpP --minimal linux-3.0.9/kernel/sysctl.c linux-3.0.9-vs2.3.2.1/kernel/sysctl.c +--- linux-3.0.9/kernel/sysctl.c 2011-07-22 11:18:12.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/kernel/sysctl.c 2011-06-15 02:40:14.000000000 +0200 +@@ -75,6 +75,7 @@ + #if defined(CONFIG_PROVE_LOCKING) || defined(CONFIG_LOCK_STAT) + #include <linux/lockdep.h> + #endif ++extern char vshelper_path[]; + #ifdef CONFIG_CHR_DEV_SG + #include <scsi/sg.h> + #endif +@@ -568,6 +569,13 @@ static struct ctl_table kern_table[] = { + .proc_handler = proc_dostring, + }, + #endif ++ { ++ .procname = "vshelper", ++ .data = &vshelper_path, ++ .maxlen = 256, ++ .mode = 0644, ++ .proc_handler = &proc_dostring, ++ }, + #ifdef CONFIG_CHR_DEV_SG + { + .procname = "sg-big-buff", +diff -NurpP --minimal linux-3.0.9/kernel/sysctl_binary.c linux-3.0.9-vs2.3.2.1/kernel/sysctl_binary.c +--- linux-3.0.9/kernel/sysctl_binary.c 2011-05-22 16:17:59.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/kernel/sysctl_binary.c 2011-06-10 22:11:24.000000000 +0200 +@@ -73,6 +73,7 @@ static const struct bin_table bin_kern_t + + { CTL_INT, KERN_PANIC, "panic" }, + { CTL_INT, KERN_REALROOTDEV, "real-root-dev" }, ++ { CTL_STR, KERN_VSHELPER, "vshelper" }, + + { CTL_STR, KERN_SPARC_REBOOT, "reboot-cmd" }, + { CTL_INT, KERN_CTLALTDEL, "ctrl-alt-del" }, +diff -NurpP --minimal linux-3.0.9/kernel/time/timekeeping.c linux-3.0.9-vs2.3.2.1/kernel/time/timekeeping.c +--- linux-3.0.9/kernel/time/timekeeping.c 2011-07-22 11:18:12.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/kernel/time/timekeeping.c 2011-06-10 22:11:24.000000000 +0200 +@@ -233,6 +233,7 @@ void getnstimeofday(struct timespec *ts) + } while (read_seqretry(&xtime_lock, seq)); + + timespec_add_ns(ts, nsecs); ++ vx_adjust_timespec(ts); + } + + EXPORT_SYMBOL(getnstimeofday); +diff -NurpP --minimal linux-3.0.9/kernel/time.c linux-3.0.9-vs2.3.2.1/kernel/time.c +--- linux-3.0.9/kernel/time.c 2011-11-15 16:40:47.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/kernel/time.c 2011-11-15 17:37:07.000000000 +0100 +@@ -92,7 +92,7 @@ SYSCALL_DEFINE1(stime, time_t __user *, + if (err) + return err; + +- do_settimeofday(&tv); ++ vx_settimeofday(&tv); + return 0; + } + +@@ -177,7 +177,7 @@ int do_sys_settimeofday(const struct tim + /* SMP safe, again the code in arch/foo/time.c should + * globally block out interrupts when it runs. + */ +- return do_settimeofday(tv); ++ return vx_settimeofday(tv); + } + return 0; + } +diff -NurpP --minimal linux-3.0.9/kernel/timer.c linux-3.0.9-vs2.3.2.1/kernel/timer.c +--- linux-3.0.9/kernel/timer.c 2011-07-22 11:18:12.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/kernel/timer.c 2011-06-15 02:40:14.000000000 +0200 +@@ -40,6 +40,10 @@ + #include <linux/irq_work.h> + #include <linux/sched.h> + #include <linux/slab.h> ++#include <linux/vs_base.h> ++#include <linux/vs_cvirt.h> ++#include <linux/vs_pid.h> ++#include <linux/vserver/sched.h> + + #include <asm/uaccess.h> + #include <asm/unistd.h> +@@ -1336,12 +1340,6 @@ SYSCALL_DEFINE1(alarm, unsigned int, sec + + #endif + +-#ifndef __alpha__ +- +-/* +- * The Alpha uses getxpid, getxuid, and getxgid instead. Maybe this +- * should be moved into arch/i386 instead? +- */ + + /** + * sys_getpid - return the thread group id of the current process +@@ -1370,10 +1368,23 @@ SYSCALL_DEFINE0(getppid) + rcu_read_lock(); + pid = task_tgid_vnr(current->real_parent); + rcu_read_unlock(); ++ return vx_map_pid(pid); ++} + +- return pid; ++#ifdef __alpha__ ++ ++/* ++ * The Alpha uses getxpid, getxuid, and getxgid instead. ++ */ ++ ++asmlinkage long do_getxpid(long *ppid) ++{ ++ *ppid = sys_getppid(); ++ return sys_getpid(); + } + ++#else /* _alpha_ */ ++ + SYSCALL_DEFINE0(getuid) + { + /* Only we change this so SMP safe */ +diff -NurpP --minimal linux-3.0.9/kernel/user_namespace.c linux-3.0.9-vs2.3.2.1/kernel/user_namespace.c +--- linux-3.0.9/kernel/user_namespace.c 2011-03-15 18:07:42.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/kernel/user_namespace.c 2011-06-10 22:11:24.000000000 +0200 +@@ -11,6 +11,7 @@ + #include <linux/user_namespace.h> + #include <linux/highuid.h> + #include <linux/cred.h> ++#include <linux/vserver/global.h> + + static struct kmem_cache *user_ns_cachep __read_mostly; + +@@ -33,6 +34,7 @@ int create_user_ns(struct cred *new) + return -ENOMEM; + + kref_init(&ns->kref); ++ atomic_inc(&vs_global_user_ns); + + for (n = 0; n < UIDHASH_SZ; ++n) + INIT_HLIST_HEAD(ns->uidhash_table + n); +@@ -81,6 +83,8 @@ void free_user_ns(struct kref *kref) + struct user_namespace *ns = + container_of(kref, struct user_namespace, kref); + ++ /* FIXME: maybe move into destroyer? */ ++ atomic_dec(&vs_global_user_ns); + INIT_WORK(&ns->destroyer, free_user_ns_work); + schedule_work(&ns->destroyer); + } +diff -NurpP --minimal linux-3.0.9/kernel/utsname.c linux-3.0.9-vs2.3.2.1/kernel/utsname.c +--- linux-3.0.9/kernel/utsname.c 2011-07-22 11:18:12.000000000 +0200 ++++ linux-3.0.9-vs2.3.2.1/kernel/utsname.c 2011-06-13 14:09:44.000000000 +0200 +@@ -16,14 +16,17 @@ + #include <linux/slab.h> + #include <linux/user_namespace.h> + #include <linux/proc_fs.h> ++#include <linux/vserver/global.h> + + static struct uts_namespace *create_uts_ns(void) + { + struct uts_namespace *uts_ns; + + uts_ns = kmalloc(sizeof(struct uts_namespace), GFP_KERNEL); +- if (uts_ns) ++ if (uts_ns) { + kref_init(&uts_ns->kref); ++ atomic_inc(&vs_global_uts_ns); ++ } + return uts_ns; + } + +@@ -32,8 +35,8 @@ static struct uts_namespace *create_uts_ + * @old_ns: namespace to clone + * Return NULL on error (failure to kmalloc), new ns otherwise + */ +-static struct uts_namespace *clone_uts_ns(struct task_struct *tsk, +- struct uts_namespace *old_ns) ++static struct uts_namespace *clone_uts_ns(struct uts_namespace *old_ns, ++ struct user_namespace *old_user) + { + struct uts_namespace *ns; + +@@ -43,7 +46,7 @@ static struct uts_namespace *clone_uts_n + + down_read(&uts_sem); + memcpy(&ns->name, &old_ns->name, sizeof(ns->name)); +- ns->user_ns = get_user_ns(task_cred_xxx(tsk, user)->user_ns); ++ ns->user_ns = get_user_ns(old_user); + up_read(&uts_sem); + return ns; + } +@@ -55,9 +58,9 @@ static struct uts_namespace *clone_uts_n + * versa. + */ + struct uts_namespace *copy_utsname(unsigned long flags, +- struct task_struct *tsk) ++ struct uts_namespace *old_ns, ++ struct user_namespace *user_ns) + { +- struct uts_namespace *old_ns = tsk->nsproxy->uts_ns; + struct uts_namespace *new_ns; + + BUG_ON(!old_ns); +@@ -66,7 +69,7 @@ struct uts_namespace *copy_utsname(unsig + if (!(flags & CLONE_NEWUTS)) + return old_ns; + +- new_ns = clone_uts_ns(tsk, old_ns); ++ new_ns = clone_uts_ns(old_ns, user_ns); + + put_uts_ns(old_ns); + return new_ns; +@@ -78,6 +81,7 @@ void free_uts_ns(struct kref *kref) + + ns = container_of(kref, struct uts_namespace, kref); + put_user_ns(ns->user_ns); ++ atomic_dec(&vs_global_uts_ns); + kfree(ns); + } + +diff -NurpP --minimal linux-3.0.9/kernel/vserver/Kconfig linux-3.0.9-vs2.3.2.1/kernel/vserver/Kconfig +--- linux-3.0.9/kernel/vserver/Kconfig 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/kernel/vserver/Kconfig 2011-08-08 18:06:22.000000000 +0200 +@@ -0,0 +1,232 @@ ++# ++# Linux VServer configuration ++# ++ ++menu "Linux VServer" ++ ++config VSERVER_AUTO_LBACK ++ bool "Automatically Assign Loopback IP" ++ default y ++ help ++ Automatically assign a guest specific loopback ++ IP and add it to the kernel network stack on ++ startup. ++ ++config VSERVER_AUTO_SINGLE ++ bool "Automatic Single IP Special Casing" ++ depends on EXPERIMENTAL ++ default y ++ help ++ This allows network contexts with a single IP to ++ automatically remap 0.0.0.0 bindings to that IP, ++ avoiding further network checks and improving ++ performance. ++ ++ (note: such guests do not allow to change the ip ++ on the fly and do not show loopback addresses) ++ ++config VSERVER_COWBL ++ bool "Enable COW Immutable Link Breaking" ++ default y ++ help ++ This enables the COW (Copy-On-Write) link break code. ++ It allows you to treat unified files like normal files ++ when writing to them (which will implicitely break the ++ link and create a copy of the unified file) ++ ++config VSERVER_VTIME ++ bool "Enable Virtualized Guest Time" ++ depends on EXPERIMENTAL ++ default n ++ help ++ This enables per guest time offsets to allow for ++ adjusting the system clock individually per guest. ++ this adds some overhead to the time functions and ++ therefore should not be enabled without good reason. ++ ++config VSERVER_DEVICE ++ bool "Enable Guest Device Mapping" ++ depends on EXPERIMENTAL ++ default n ++ help ++ This enables generic device remapping. ++ ++config VSERVER_PROC_SECURE ++ bool "Enable Proc Security" ++ depends on PROC_FS ++ default y ++ help ++ This configures ProcFS security to initially hide ++ non-process entries for all contexts except the main and ++ spectator context (i.e. for all guests), which is a secure ++ default. ++ ++ (note: on 1.2x the entries were visible by default) ++ ++choice ++ prompt "Persistent Inode Tagging" ++ default TAGGING_ID24 ++ help ++ This adds persistent context information to filesystems ++ mounted with the tagxid option. Tagging is a requirement ++ for per-context disk limits and per-context quota. ++ ++ ++config TAGGING_NONE ++ bool "Disabled" ++ help ++ do not store per-context information in inodes. ++ ++config TAGGING_UID16 ++ bool "UID16/GID32" ++ help ++ reduces UID to 16 bit, but leaves GID at 32 bit. ++ ++config TAGGING_GID16 ++ bool "UID32/GID16" ++ help ++ reduces GID to 16 bit, but leaves UID at 32 bit. ++ ++config TAGGING_ID24 ++ bool "UID24/GID24" ++ help ++ uses the upper 8bit from UID and GID for XID tagging ++ which leaves 24bit for UID/GID each, which should be ++ more than sufficient for normal use. ++ ++config TAGGING_INTERN ++ bool "UID32/GID32" ++ help ++ this uses otherwise reserved inode fields in the on ++ disk representation, which limits the use to a few ++ filesystems (currently ext2 and ext3) ++ ++endchoice ++ ++config TAG_NFSD ++ bool "Tag NFSD User Auth and Files" ++ default n ++ help ++ Enable this if you do want the in-kernel NFS ++ Server to use the tagging specified above. ++ (will require patched clients too) ++ ++config VSERVER_PRIVACY ++ bool "Honor Privacy Aspects of Guests" ++ default n ++ help ++ When enabled, most context checks will disallow ++ access to structures assigned to a specific context, ++ like ptys or loop devices. ++ ++config VSERVER_CONTEXTS ++ int "Maximum number of Contexts (1-65533)" if EMBEDDED ++ range 1 65533 ++ default "768" if 64BIT ++ default "256" ++ help ++ This setting will optimize certain data structures ++ and memory allocations according to the expected ++ maximum. ++ ++ note: this is not a strict upper limit. ++ ++config VSERVER_WARN ++ bool "VServer Warnings" ++ default y ++ help ++ This enables various runtime warnings, which will ++ notify about potential manipulation attempts or ++ resource shortage. It is generally considered to ++ be a good idea to have that enabled. ++ ++config VSERVER_WARN_DEVPTS ++ bool "VServer DevPTS Warnings" ++ depends on VSERVER_WARN ++ default y ++ help ++ This enables DevPTS related warnings, issued when a ++ process inside a context tries to lookup or access ++ a dynamic pts from the host or a different context. ++ ++config VSERVER_DEBUG ++ bool "VServer Debugging Code" ++ default n ++ help ++ Set this to yes if you want to be able to activate ++ debugging output at runtime. It adds a very small ++ overhead to all vserver related functions and ++ increases the kernel size by about 20k. ++ ++config VSERVER_HISTORY ++ bool "VServer History Tracing" ++ depends on VSERVER_DEBUG ++ default n ++ help ++ Set this to yes if you want to record the history of ++ linux-vserver activities, so they can be replayed in ++ the event of a kernel panic or oops. ++ ++config VSERVER_HISTORY_SIZE ++ int "Per-CPU History Size (32-65536)" ++ depends on VSERVER_HISTORY ++ range 32 65536 ++ default 64 ++ help ++ This allows you to specify the number of entries in ++ the per-CPU history buffer. ++ ++config VSERVER_LEGACY_MEM ++ bool "Legacy Memory Limits" ++ default n ++ help ++ This provides fake memory limits to keep ++ older tools happy in the face of memory ++ cgroups ++ ++choice ++ prompt "Quotes used in debug and warn messages" ++ default QUOTES_ISO8859 ++ ++config QUOTES_ISO8859 ++ bool "Extended ASCII (ISO 8859) angle quotes" ++ help ++ This uses the extended ASCII characters \xbb ++ and \xab for quoting file and process names. ++ ++config QUOTES_UTF8 ++ bool "UTF-8 angle quotes" ++ help ++ This uses the the UTF-8 sequences for angle ++ quotes to quote file and process names. ++ ++config QUOTES_ASCII ++ bool "ASCII single quotes" ++ help ++ This uses the ASCII single quote character ++ (\x27) to quote file and process names. ++ ++endchoice ++ ++endmenu ++ ++ ++config VSERVER ++ bool ++ default y ++ select NAMESPACES ++ select UTS_NS ++ select IPC_NS ++# select USER_NS ++ select SYSVIPC ++ ++config VSERVER_SECURITY ++ bool ++ depends on SECURITY ++ default y ++ select SECURITY_CAPABILITIES ++ ++config VSERVER_DISABLED ++ bool ++ default n ++ +diff -NurpP --minimal linux-3.0.9/kernel/vserver/Makefile linux-3.0.9-vs2.3.2.1/kernel/vserver/Makefile +--- linux-3.0.9/kernel/vserver/Makefile 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/kernel/vserver/Makefile 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,18 @@ ++# ++# Makefile for the Linux vserver routines. ++# ++ ++ ++obj-y += vserver.o ++ ++vserver-y := switch.o context.o space.o sched.o network.o inode.o \ ++ limit.o cvirt.o cacct.o signal.o helper.o init.o \ ++ dlimit.o tag.o ++ ++vserver-$(CONFIG_INET) += inet.o ++vserver-$(CONFIG_PROC_FS) += proc.o ++vserver-$(CONFIG_VSERVER_DEBUG) += sysctl.o debug.o ++vserver-$(CONFIG_VSERVER_HISTORY) += history.o ++vserver-$(CONFIG_VSERVER_MONITOR) += monitor.o ++vserver-$(CONFIG_VSERVER_DEVICE) += device.o ++ +diff -NurpP --minimal linux-3.0.9/kernel/vserver/cacct.c linux-3.0.9-vs2.3.2.1/kernel/vserver/cacct.c +--- linux-3.0.9/kernel/vserver/cacct.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/kernel/vserver/cacct.c 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,42 @@ ++/* ++ * linux/kernel/vserver/cacct.c ++ * ++ * Virtual Server: Context Accounting ++ * ++ * Copyright (C) 2006-2007 Herbert Pötzl ++ * ++ * V0.01 added accounting stats ++ * ++ */ ++ ++#include <linux/types.h> ++#include <linux/vs_context.h> ++#include <linux/vserver/cacct_cmd.h> ++#include <linux/vserver/cacct_int.h> ++ ++#include <asm/errno.h> ++#include <asm/uaccess.h> ++ ++ ++int vc_sock_stat(struct vx_info *vxi, void __user *data) ++{ ++ struct vcmd_sock_stat_v0 vc_data; ++ int j, field; ++ ++ if (copy_from_user(&vc_data, data, sizeof(vc_data))) ++ return -EFAULT; ++ ++ field = vc_data.field; ++ if ((field < 0) || (field >= VXA_SOCK_SIZE)) ++ return -EINVAL; ++ ++ for (j = 0; j < 3; j++) { ++ vc_data.count[j] = vx_sock_count(&vxi->cacct, field, j); ++ vc_data.total[j] = vx_sock_total(&vxi->cacct, field, j); ++ } ++ ++ if (copy_to_user(data, &vc_data, sizeof(vc_data))) ++ return -EFAULT; ++ return 0; ++} ++ +diff -NurpP --minimal linux-3.0.9/kernel/vserver/cacct_init.h linux-3.0.9-vs2.3.2.1/kernel/vserver/cacct_init.h +--- linux-3.0.9/kernel/vserver/cacct_init.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/kernel/vserver/cacct_init.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,25 @@ ++ ++ ++static inline void vx_info_init_cacct(struct _vx_cacct *cacct) ++{ ++ int i, j; ++ ++ ++ for (i = 0; i < VXA_SOCK_SIZE; i++) { ++ for (j = 0; j < 3; j++) { ++ atomic_long_set(&cacct->sock[i][j].count, 0); ++ atomic_long_set(&cacct->sock[i][j].total, 0); ++ } ++ } ++ for (i = 0; i < 8; i++) ++ atomic_set(&cacct->slab[i], 0); ++ for (i = 0; i < 5; i++) ++ for (j = 0; j < 4; j++) ++ atomic_set(&cacct->page[i][j], 0); ++} ++ ++static inline void vx_info_exit_cacct(struct _vx_cacct *cacct) ++{ ++ return; ++} ++ +diff -NurpP --minimal linux-3.0.9/kernel/vserver/cacct_proc.h linux-3.0.9-vs2.3.2.1/kernel/vserver/cacct_proc.h +--- linux-3.0.9/kernel/vserver/cacct_proc.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/kernel/vserver/cacct_proc.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,53 @@ ++#ifndef _VX_CACCT_PROC_H ++#define _VX_CACCT_PROC_H ++ ++#include <linux/vserver/cacct_int.h> ++ ++ ++#define VX_SOCKA_TOP \ ++ "Type\t recv #/bytes\t\t send #/bytes\t\t fail #/bytes\n" ++ ++static inline int vx_info_proc_cacct(struct _vx_cacct *cacct, char *buffer) ++{ ++ int i, j, length = 0; ++ static char *type[VXA_SOCK_SIZE] = { ++ "UNSPEC", "UNIX", "INET", "INET6", "PACKET", "OTHER" ++ }; ++ ++ length += sprintf(buffer + length, VX_SOCKA_TOP); ++ for (i = 0; i < VXA_SOCK_SIZE; i++) { ++ length += sprintf(buffer + length, "%s:", type[i]); ++ for (j = 0; j < 3; j++) { ++ length += sprintf(buffer + length, ++ "\t%10lu/%-10lu", ++ vx_sock_count(cacct, i, j), ++ vx_sock_total(cacct, i, j)); ++ } ++ buffer[length++] = '\n'; ++ } ++ ++ length += sprintf(buffer + length, "\n"); ++ length += sprintf(buffer + length, ++ "slab:\t %8u %8u %8u %8u\n", ++ atomic_read(&cacct->slab[1]), ++ atomic_read(&cacct->slab[4]), ++ atomic_read(&cacct->slab[0]), ++ atomic_read(&cacct->slab[2])); ++ ++ length += sprintf(buffer + length, "\n"); ++ for (i = 0; i < 5; i++) { ++ length += sprintf(buffer + length, ++ "page[%d]: %8u %8u %8u %8u\t %8u %8u %8u %8u\n", i, ++ atomic_read(&cacct->page[i][0]), ++ atomic_read(&cacct->page[i][1]), ++ atomic_read(&cacct->page[i][2]), ++ atomic_read(&cacct->page[i][3]), ++ atomic_read(&cacct->page[i][4]), ++ atomic_read(&cacct->page[i][5]), ++ atomic_read(&cacct->page[i][6]), ++ atomic_read(&cacct->page[i][7])); ++ } ++ return length; ++} ++ ++#endif /* _VX_CACCT_PROC_H */ +diff -NurpP --minimal linux-3.0.9/kernel/vserver/context.c linux-3.0.9-vs2.3.2.1/kernel/vserver/context.c +--- linux-3.0.9/kernel/vserver/context.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/kernel/vserver/context.c 2011-08-01 18:28:12.000000000 +0200 +@@ -0,0 +1,1107 @@ ++/* ++ * linux/kernel/vserver/context.c ++ * ++ * Virtual Server: Context Support ++ * ++ * Copyright (C) 2003-2011 Herbert Pötzl ++ * ++ * V0.01 context helper ++ * V0.02 vx_ctx_kill syscall command ++ * V0.03 replaced context_info calls ++ * V0.04 redesign of struct (de)alloc ++ * V0.05 rlimit basic implementation ++ * V0.06 task_xid and info commands ++ * V0.07 context flags and caps ++ * V0.08 switch to RCU based hash ++ * V0.09 revert to non RCU for now ++ * V0.10 and back to working RCU hash ++ * V0.11 and back to locking again ++ * V0.12 referenced context store ++ * V0.13 separate per cpu data ++ * V0.14 changed vcmds to vxi arg ++ * V0.15 added context stat ++ * V0.16 have __create claim() the vxi ++ * V0.17 removed older and legacy stuff ++ * V0.18 added user credentials ++ * V0.19 added warn mask ++ * ++ */ ++ ++#include <linux/slab.h> ++#include <linux/types.h> ++#include <linux/security.h> ++#include <linux/pid_namespace.h> ++#include <linux/capability.h> ++ ++#include <linux/vserver/context.h> ++#include <linux/vserver/network.h> ++#include <linux/vserver/debug.h> ++#include <linux/vserver/limit.h> ++#include <linux/vserver/limit_int.h> ++#include <linux/vserver/space.h> ++#include <linux/init_task.h> ++#include <linux/fs_struct.h> ++#include <linux/cred.h> ++ ++#include <linux/vs_context.h> ++#include <linux/vs_limit.h> ++#include <linux/vs_pid.h> ++#include <linux/vserver/context_cmd.h> ++ ++#include "cvirt_init.h" ++#include "cacct_init.h" ++#include "limit_init.h" ++#include "sched_init.h" ++ ++ ++atomic_t vx_global_ctotal = ATOMIC_INIT(0); ++atomic_t vx_global_cactive = ATOMIC_INIT(0); ++ ++ ++/* now inactive context structures */ ++ ++static struct hlist_head vx_info_inactive = HLIST_HEAD_INIT; ++ ++static DEFINE_SPINLOCK(vx_info_inactive_lock); ++ ++ ++/* __alloc_vx_info() ++ ++ * allocate an initialized vx_info struct ++ * doesn't make it visible (hash) */ ++ ++static struct vx_info *__alloc_vx_info(xid_t xid) ++{ ++ struct vx_info *new = NULL; ++ int cpu, index; ++ ++ vxdprintk(VXD_CBIT(xid, 0), "alloc_vx_info(%d)*", xid); ++ ++ /* would this benefit from a slab cache? */ ++ new = kmalloc(sizeof(struct vx_info), GFP_KERNEL); ++ if (!new) ++ return 0; ++ ++ memset(new, 0, sizeof(struct vx_info)); ++#ifdef CONFIG_SMP ++ new->ptr_pc = alloc_percpu(struct _vx_info_pc); ++ if (!new->ptr_pc) ++ goto error; ++#endif ++ new->vx_id = xid; ++ INIT_HLIST_NODE(&new->vx_hlist); ++ atomic_set(&new->vx_usecnt, 0); ++ atomic_set(&new->vx_tasks, 0); ++ new->vx_parent = NULL; ++ new->vx_state = 0; ++ init_waitqueue_head(&new->vx_wait); ++ ++ /* prepare reaper */ ++ get_task_struct(init_pid_ns.child_reaper); ++ new->vx_reaper = init_pid_ns.child_reaper; ++ new->vx_badness_bias = 0; ++ ++ /* rest of init goes here */ ++ vx_info_init_limit(&new->limit); ++ vx_info_init_sched(&new->sched); ++ vx_info_init_cvirt(&new->cvirt); ++ vx_info_init_cacct(&new->cacct); ++ ++ /* per cpu data structures */ ++ for_each_possible_cpu(cpu) { ++ vx_info_init_sched_pc( ++ &vx_per_cpu(new, sched_pc, cpu), cpu); ++ vx_info_init_cvirt_pc( ++ &vx_per_cpu(new, cvirt_pc, cpu), cpu); ++ } ++ ++ new->vx_flags = VXF_INIT_SET; ++ new->vx_bcaps = CAP_FULL_SET; // maybe ~CAP_SETPCAP ++ new->vx_ccaps = 0; ++ new->vx_umask = 0; ++ new->vx_wmask = 0; ++ ++ new->reboot_cmd = 0; ++ new->exit_code = 0; ++ ++ // preconfig spaces ++ for (index = 0; index < VX_SPACES; index++) { ++ struct _vx_space *space = &new->space[index]; ++ ++ // filesystem ++ spin_lock(&init_fs.lock); ++ init_fs.users++; ++ spin_unlock(&init_fs.lock); ++ space->vx_fs = &init_fs; ++ ++ /* FIXME: do we want defaults? */ ++ // space->vx_real_cred = 0; ++ // space->vx_cred = 0; ++ } ++ ++ ++ vxdprintk(VXD_CBIT(xid, 0), ++ "alloc_vx_info(%d) = %p", xid, new); ++ vxh_alloc_vx_info(new); ++ atomic_inc(&vx_global_ctotal); ++ return new; ++#ifdef CONFIG_SMP ++error: ++ kfree(new); ++ return 0; ++#endif ++} ++ ++/* __dealloc_vx_info() ++ ++ * final disposal of vx_info */ ++ ++static void __dealloc_vx_info(struct vx_info *vxi) ++{ ++#ifdef CONFIG_VSERVER_WARN ++ struct vx_info_save vxis; ++ int cpu; ++#endif ++ vxdprintk(VXD_CBIT(xid, 0), ++ "dealloc_vx_info(%p)", vxi); ++ vxh_dealloc_vx_info(vxi); ++ ++#ifdef CONFIG_VSERVER_WARN ++ enter_vx_info(vxi, &vxis); ++ vx_info_exit_limit(&vxi->limit); ++ vx_info_exit_sched(&vxi->sched); ++ vx_info_exit_cvirt(&vxi->cvirt); ++ vx_info_exit_cacct(&vxi->cacct); ++ ++ for_each_possible_cpu(cpu) { ++ vx_info_exit_sched_pc( ++ &vx_per_cpu(vxi, sched_pc, cpu), cpu); ++ vx_info_exit_cvirt_pc( ++ &vx_per_cpu(vxi, cvirt_pc, cpu), cpu); ++ } ++ leave_vx_info(&vxis); ++#endif ++ ++ vxi->vx_id = -1; ++ vxi->vx_state |= VXS_RELEASED; ++ ++#ifdef CONFIG_SMP ++ free_percpu(vxi->ptr_pc); ++#endif ++ kfree(vxi); ++ atomic_dec(&vx_global_ctotal); ++} ++ ++static void __shutdown_vx_info(struct vx_info *vxi) ++{ ++ struct nsproxy *nsproxy; ++ struct fs_struct *fs; ++ struct cred *cred; ++ int index, kill; ++ ++ might_sleep(); ++ ++ vxi->vx_state |= VXS_SHUTDOWN; ++ vs_state_change(vxi, VSC_SHUTDOWN); ++ ++ for (index = 0; index < VX_SPACES; index++) { ++ struct _vx_space *space = &vxi->space[index]; ++ ++ nsproxy = xchg(&space->vx_nsproxy, NULL); ++ if (nsproxy) ++ put_nsproxy(nsproxy); ++ ++ fs = xchg(&space->vx_fs, NULL); ++ spin_lock(&fs->lock); ++ kill = !--fs->users; ++ spin_unlock(&fs->lock); ++ if (kill) ++ free_fs_struct(fs); ++ ++ cred = (struct cred *)xchg(&space->vx_cred, NULL); ++ if (cred) ++ abort_creds(cred); ++ } ++} ++ ++/* exported stuff */ ++ ++void free_vx_info(struct vx_info *vxi) ++{ ++ unsigned long flags; ++ unsigned index; ++ ++ /* check for reference counts first */ ++ BUG_ON(atomic_read(&vxi->vx_usecnt)); ++ BUG_ON(atomic_read(&vxi->vx_tasks)); ++ ++ /* context must not be hashed */ ++ BUG_ON(vx_info_state(vxi, VXS_HASHED)); ++ ++ /* context shutdown is mandatory */ ++ BUG_ON(!vx_info_state(vxi, VXS_SHUTDOWN)); ++ ++ /* spaces check */ ++ for (index = 0; index < VX_SPACES; index++) { ++ struct _vx_space *space = &vxi->space[index]; ++ ++ BUG_ON(space->vx_nsproxy); ++ BUG_ON(space->vx_fs); ++ // BUG_ON(space->vx_real_cred); ++ // BUG_ON(space->vx_cred); ++ } ++ ++ spin_lock_irqsave(&vx_info_inactive_lock, flags); ++ hlist_del(&vxi->vx_hlist); ++ spin_unlock_irqrestore(&vx_info_inactive_lock, flags); ++ ++ __dealloc_vx_info(vxi); ++} ++ ++ ++/* hash table for vx_info hash */ ++ ++#define VX_HASH_SIZE 13 ++ ++static struct hlist_head vx_info_hash[VX_HASH_SIZE] = ++ { [0 ... VX_HASH_SIZE-1] = HLIST_HEAD_INIT }; ++ ++static DEFINE_SPINLOCK(vx_info_hash_lock); ++ ++ ++static inline unsigned int __hashval(xid_t xid) ++{ ++ return (xid % VX_HASH_SIZE); ++} ++ ++ ++ ++/* __hash_vx_info() ++ ++ * add the vxi to the global hash table ++ * requires the hash_lock to be held */ ++ ++static inline void __hash_vx_info(struct vx_info *vxi) ++{ ++ struct hlist_head *head; ++ ++ vxd_assert_lock(&vx_info_hash_lock); ++ vxdprintk(VXD_CBIT(xid, 4), ++ "__hash_vx_info: %p[#%d]", vxi, vxi->vx_id); ++ vxh_hash_vx_info(vxi); ++ ++ /* context must not be hashed */ ++ BUG_ON(vx_info_state(vxi, VXS_HASHED)); ++ ++ vxi->vx_state |= VXS_HASHED; ++ head = &vx_info_hash[__hashval(vxi->vx_id)]; ++ hlist_add_head(&vxi->vx_hlist, head); ++ atomic_inc(&vx_global_cactive); ++} ++ ++/* __unhash_vx_info() ++ ++ * remove the vxi from the global hash table ++ * requires the hash_lock to be held */ ++ ++static inline void __unhash_vx_info(struct vx_info *vxi) ++{ ++ unsigned long flags; ++ ++ vxd_assert_lock(&vx_info_hash_lock); ++ vxdprintk(VXD_CBIT(xid, 4), ++ "__unhash_vx_info: %p[#%d.%d.%d]", vxi, vxi->vx_id, ++ atomic_read(&vxi->vx_usecnt), atomic_read(&vxi->vx_tasks)); ++ vxh_unhash_vx_info(vxi); ++ ++ /* context must be hashed */ ++ BUG_ON(!vx_info_state(vxi, VXS_HASHED)); ++ /* but without tasks */ ++ BUG_ON(atomic_read(&vxi->vx_tasks)); ++ ++ vxi->vx_state &= ~VXS_HASHED; ++ hlist_del_init(&vxi->vx_hlist); ++ spin_lock_irqsave(&vx_info_inactive_lock, flags); ++ hlist_add_head(&vxi->vx_hlist, &vx_info_inactive); ++ spin_unlock_irqrestore(&vx_info_inactive_lock, flags); ++ atomic_dec(&vx_global_cactive); ++} ++ ++ ++/* __lookup_vx_info() ++ ++ * requires the hash_lock to be held ++ * doesn't increment the vx_refcnt */ ++ ++static inline struct vx_info *__lookup_vx_info(xid_t xid) ++{ ++ struct hlist_head *head = &vx_info_hash[__hashval(xid)]; ++ struct hlist_node *pos; ++ struct vx_info *vxi; ++ ++ vxd_assert_lock(&vx_info_hash_lock); ++ hlist_for_each(pos, head) { ++ vxi = hlist_entry(pos, struct vx_info, vx_hlist); ++ ++ if (vxi->vx_id == xid) ++ goto found; ++ } ++ vxi = NULL; ++found: ++ vxdprintk(VXD_CBIT(xid, 0), ++ "__lookup_vx_info(#%u): %p[#%u]", ++ xid, vxi, vxi ? vxi->vx_id : 0); ++ vxh_lookup_vx_info(vxi, xid); ++ return vxi; ++} ++ ++ ++/* __create_vx_info() ++ ++ * create the requested context ++ * get(), claim() and hash it */ ++ ++static struct vx_info *__create_vx_info(int id) ++{ ++ struct vx_info *new, *vxi = NULL; ++ ++ vxdprintk(VXD_CBIT(xid, 1), "create_vx_info(%d)*", id); ++ ++ if (!(new = __alloc_vx_info(id))) ++ return ERR_PTR(-ENOMEM); ++ ++ /* required to make dynamic xids unique */ ++ spin_lock(&vx_info_hash_lock); ++ ++ /* static context requested */ ++ if ((vxi = __lookup_vx_info(id))) { ++ vxdprintk(VXD_CBIT(xid, 0), ++ "create_vx_info(%d) = %p (already there)", id, vxi); ++ if (vx_info_flags(vxi, VXF_STATE_SETUP, 0)) ++ vxi = ERR_PTR(-EBUSY); ++ else ++ vxi = ERR_PTR(-EEXIST); ++ goto out_unlock; ++ } ++ /* new context */ ++ vxdprintk(VXD_CBIT(xid, 0), ++ "create_vx_info(%d) = %p (new)", id, new); ++ claim_vx_info(new, NULL); ++ __hash_vx_info(get_vx_info(new)); ++ vxi = new, new = NULL; ++ ++out_unlock: ++ spin_unlock(&vx_info_hash_lock); ++ vxh_create_vx_info(IS_ERR(vxi) ? NULL : vxi, id); ++ if (new) ++ __dealloc_vx_info(new); ++ return vxi; ++} ++ ++ ++/* exported stuff */ ++ ++ ++void unhash_vx_info(struct vx_info *vxi) ++{ ++ spin_lock(&vx_info_hash_lock); ++ __unhash_vx_info(vxi); ++ spin_unlock(&vx_info_hash_lock); ++ __shutdown_vx_info(vxi); ++ __wakeup_vx_info(vxi); ++} ++ ++ ++/* lookup_vx_info() ++ ++ * search for a vx_info and get() it ++ * negative id means current */ ++ ++struct vx_info *lookup_vx_info(int id) ++{ ++ struct vx_info *vxi = NULL; ++ ++ if (id < 0) { ++ vxi = get_vx_info(current_vx_info()); ++ } else if (id > 1) { ++ spin_lock(&vx_info_hash_lock); ++ vxi = get_vx_info(__lookup_vx_info(id)); ++ spin_unlock(&vx_info_hash_lock); ++ } ++ return vxi; ++} ++ ++/* xid_is_hashed() ++ ++ * verify that xid is still hashed */ ++ ++int xid_is_hashed(xid_t xid) ++{ ++ int hashed; ++ ++ spin_lock(&vx_info_hash_lock); ++ hashed = (__lookup_vx_info(xid) != NULL); ++ spin_unlock(&vx_info_hash_lock); ++ return hashed; ++} ++ ++#ifdef CONFIG_PROC_FS ++ ++/* get_xid_list() ++ ++ * get a subset of hashed xids for proc ++ * assumes size is at least one */ ++ ++int get_xid_list(int index, unsigned int *xids, int size) ++{ ++ int hindex, nr_xids = 0; ++ ++ /* only show current and children */ ++ if (!vx_check(0, VS_ADMIN | VS_WATCH)) { ++ if (index > 0) ++ return 0; ++ xids[nr_xids] = vx_current_xid(); ++ return 1; ++ } ++ ++ for (hindex = 0; hindex < VX_HASH_SIZE; hindex++) { ++ struct hlist_head *head = &vx_info_hash[hindex]; ++ struct hlist_node *pos; ++ ++ spin_lock(&vx_info_hash_lock); ++ hlist_for_each(pos, head) { ++ struct vx_info *vxi; ++ ++ if (--index > 0) ++ continue; ++ ++ vxi = hlist_entry(pos, struct vx_info, vx_hlist); ++ xids[nr_xids] = vxi->vx_id; ++ if (++nr_xids >= size) { ++ spin_unlock(&vx_info_hash_lock); ++ goto out; ++ } ++ } ++ /* keep the lock time short */ ++ spin_unlock(&vx_info_hash_lock); ++ } ++out: ++ return nr_xids; ++} ++#endif ++ ++#ifdef CONFIG_VSERVER_DEBUG ++ ++void dump_vx_info_inactive(int level) ++{ ++ struct hlist_node *entry, *next; ++ ++ hlist_for_each_safe(entry, next, &vx_info_inactive) { ++ struct vx_info *vxi = ++ list_entry(entry, struct vx_info, vx_hlist); ++ ++ dump_vx_info(vxi, level); ++ } ++} ++ ++#endif ++ ++#if 0 ++int vx_migrate_user(struct task_struct *p, struct vx_info *vxi) ++{ ++ struct user_struct *new_user, *old_user; ++ ++ if (!p || !vxi) ++ BUG(); ++ ++ if (vx_info_flags(vxi, VXF_INFO_PRIVATE, 0)) ++ return -EACCES; ++ ++ new_user = alloc_uid(vxi->vx_id, p->uid); ++ if (!new_user) ++ return -ENOMEM; ++ ++ old_user = p->user; ++ if (new_user != old_user) { ++ atomic_inc(&new_user->processes); ++ atomic_dec(&old_user->processes); ++ p->user = new_user; ++ } ++ free_uid(old_user); ++ return 0; ++} ++#endif ++ ++#if 0 ++void vx_mask_cap_bset(struct vx_info *vxi, struct task_struct *p) ++{ ++ // p->cap_effective &= vxi->vx_cap_bset; ++ p->cap_effective = ++ cap_intersect(p->cap_effective, vxi->cap_bset); ++ // p->cap_inheritable &= vxi->vx_cap_bset; ++ p->cap_inheritable = ++ cap_intersect(p->cap_inheritable, vxi->cap_bset); ++ // p->cap_permitted &= vxi->vx_cap_bset; ++ p->cap_permitted = ++ cap_intersect(p->cap_permitted, vxi->cap_bset); ++} ++#endif ++ ++ ++#include <linux/file.h> ++#include <linux/fdtable.h> ++ ++static int vx_openfd_task(struct task_struct *tsk) ++{ ++ struct files_struct *files = tsk->files; ++ struct fdtable *fdt; ++ const unsigned long *bptr; ++ int count, total; ++ ++ /* no rcu_read_lock() because of spin_lock() */ ++ spin_lock(&files->file_lock); ++ fdt = files_fdtable(files); ++ bptr = fdt->open_fds->fds_bits; ++ count = fdt->max_fds / (sizeof(unsigned long) * 8); ++ for (total = 0; count > 0; count--) { ++ if (*bptr) ++ total += hweight_long(*bptr); ++ bptr++; ++ } ++ spin_unlock(&files->file_lock); ++ return total; ++} ++ ++ ++/* for *space compatibility */ ++ ++asmlinkage long sys_unshare(unsigned long); ++ ++/* ++ * migrate task to new context ++ * gets vxi, puts old_vxi on change ++ * optionally unshares namespaces (hack) ++ */ ++ ++int vx_migrate_task(struct task_struct *p, struct vx_info *vxi, int unshare) ++{ ++ struct vx_info *old_vxi; ++ int ret = 0; ++ ++ if (!p || !vxi) ++ BUG(); ++ ++ vxdprintk(VXD_CBIT(xid, 5), ++ "vx_migrate_task(%p,%p[#%d.%d])", p, vxi, ++ vxi->vx_id, atomic_read(&vxi->vx_usecnt)); ++ ++ if (vx_info_flags(vxi, VXF_INFO_PRIVATE, 0) && ++ !vx_info_flags(vxi, VXF_STATE_SETUP, 0)) ++ return -EACCES; ++ ++ if (vx_info_state(vxi, VXS_SHUTDOWN)) ++ return -EFAULT; ++ ++ old_vxi = task_get_vx_info(p); ++ if (old_vxi == vxi) ++ goto out; ++ ++// if (!(ret = vx_migrate_user(p, vxi))) { ++ { ++ int openfd; ++ ++ task_lock(p); ++ openfd = vx_openfd_task(p); ++ ++ if (old_vxi) { ++ atomic_dec(&old_vxi->cvirt.nr_threads); ++ atomic_dec(&old_vxi->cvirt.nr_running); ++ __rlim_dec(&old_vxi->limit, RLIMIT_NPROC); ++ /* FIXME: what about the struct files here? */ ++ __rlim_sub(&old_vxi->limit, VLIMIT_OPENFD, openfd); ++ /* account for the executable */ ++ __rlim_dec(&old_vxi->limit, VLIMIT_DENTRY); ++ } ++ atomic_inc(&vxi->cvirt.nr_threads); ++ atomic_inc(&vxi->cvirt.nr_running); ++ __rlim_inc(&vxi->limit, RLIMIT_NPROC); ++ /* FIXME: what about the struct files here? */ ++ __rlim_add(&vxi->limit, VLIMIT_OPENFD, openfd); ++ /* account for the executable */ ++ __rlim_inc(&vxi->limit, VLIMIT_DENTRY); ++ ++ if (old_vxi) { ++ release_vx_info(old_vxi, p); ++ clr_vx_info(&p->vx_info); ++ } ++ claim_vx_info(vxi, p); ++ set_vx_info(&p->vx_info, vxi); ++ p->xid = vxi->vx_id; ++ ++ vxdprintk(VXD_CBIT(xid, 5), ++ "moved task %p into vxi:%p[#%d]", ++ p, vxi, vxi->vx_id); ++ ++ // vx_mask_cap_bset(vxi, p); ++ task_unlock(p); ++ ++ /* hack for *spaces to provide compatibility */ ++ if (unshare) { ++ struct nsproxy *old_nsp, *new_nsp; ++ ++ ret = unshare_nsproxy_namespaces( ++ CLONE_NEWUTS | CLONE_NEWIPC | CLONE_NEWUSER, ++ &new_nsp, NULL); ++ if (ret) ++ goto out; ++ ++ old_nsp = xchg(&p->nsproxy, new_nsp); ++ vx_set_space(vxi, ++ CLONE_NEWUTS | CLONE_NEWIPC | CLONE_NEWUSER, 0); ++ put_nsproxy(old_nsp); ++ } ++ } ++out: ++ put_vx_info(old_vxi); ++ return ret; ++} ++ ++int vx_set_reaper(struct vx_info *vxi, struct task_struct *p) ++{ ++ struct task_struct *old_reaper; ++ ++ if (!vxi) ++ return -EINVAL; ++ ++ vxdprintk(VXD_CBIT(xid, 6), ++ "vx_set_reaper(%p[#%d],%p[#%d,%d])", ++ vxi, vxi->vx_id, p, p->xid, p->pid); ++ ++ old_reaper = vxi->vx_reaper; ++ if (old_reaper == p) ++ return 0; ++ ++ /* set new child reaper */ ++ get_task_struct(p); ++ vxi->vx_reaper = p; ++ put_task_struct(old_reaper); ++ return 0; ++} ++ ++int vx_set_init(struct vx_info *vxi, struct task_struct *p) ++{ ++ if (!vxi) ++ return -EINVAL; ++ ++ vxdprintk(VXD_CBIT(xid, 6), ++ "vx_set_init(%p[#%d],%p[#%d,%d,%d])", ++ vxi, vxi->vx_id, p, p->xid, p->pid, p->tgid); ++ ++ vxi->vx_flags &= ~VXF_STATE_INIT; ++ // vxi->vx_initpid = p->tgid; ++ vxi->vx_initpid = p->pid; ++ return 0; ++} ++ ++void vx_exit_init(struct vx_info *vxi, struct task_struct *p, int code) ++{ ++ vxdprintk(VXD_CBIT(xid, 6), ++ "vx_exit_init(%p[#%d],%p[#%d,%d,%d])", ++ vxi, vxi->vx_id, p, p->xid, p->pid, p->tgid); ++ ++ vxi->exit_code = code; ++ vxi->vx_initpid = 0; ++} ++ ++ ++void vx_set_persistent(struct vx_info *vxi) ++{ ++ vxdprintk(VXD_CBIT(xid, 6), ++ "vx_set_persistent(%p[#%d])", vxi, vxi->vx_id); ++ ++ get_vx_info(vxi); ++ claim_vx_info(vxi, NULL); ++} ++ ++void vx_clear_persistent(struct vx_info *vxi) ++{ ++ vxdprintk(VXD_CBIT(xid, 6), ++ "vx_clear_persistent(%p[#%d])", vxi, vxi->vx_id); ++ ++ release_vx_info(vxi, NULL); ++ put_vx_info(vxi); ++} ++ ++void vx_update_persistent(struct vx_info *vxi) ++{ ++ if (vx_info_flags(vxi, VXF_PERSISTENT, 0)) ++ vx_set_persistent(vxi); ++ else ++ vx_clear_persistent(vxi); ++} ++ ++ ++/* task must be current or locked */ ++ ++void exit_vx_info(struct task_struct *p, int code) ++{ ++ struct vx_info *vxi = p->vx_info; ++ ++ if (vxi) { ++ atomic_dec(&vxi->cvirt.nr_threads); ++ vx_nproc_dec(p); ++ ++ vxi->exit_code = code; ++ release_vx_info(vxi, p); ++ } ++} ++ ++void exit_vx_info_early(struct task_struct *p, int code) ++{ ++ struct vx_info *vxi = p->vx_info; ++ ++ if (vxi) { ++ if (vxi->vx_initpid == p->pid) ++ vx_exit_init(vxi, p, code); ++ if (vxi->vx_reaper == p) ++ vx_set_reaper(vxi, init_pid_ns.child_reaper); ++ } ++} ++ ++ ++/* vserver syscall commands below here */ ++ ++/* taks xid and vx_info functions */ ++ ++#include <asm/uaccess.h> ++ ++ ++int vc_task_xid(uint32_t id) ++{ ++ xid_t xid; ++ ++ if (id) { ++ struct task_struct *tsk; ++ ++ rcu_read_lock(); ++ tsk = find_task_by_real_pid(id); ++ xid = (tsk) ? tsk->xid : -ESRCH; ++ rcu_read_unlock(); ++ } else ++ xid = vx_current_xid(); ++ return xid; ++} ++ ++ ++int vc_vx_info(struct vx_info *vxi, void __user *data) ++{ ++ struct vcmd_vx_info_v0 vc_data; ++ ++ vc_data.xid = vxi->vx_id; ++ vc_data.initpid = vxi->vx_initpid; ++ ++ if (copy_to_user(data, &vc_data, sizeof(vc_data))) ++ return -EFAULT; ++ return 0; ++} ++ ++ ++int vc_ctx_stat(struct vx_info *vxi, void __user *data) ++{ ++ struct vcmd_ctx_stat_v0 vc_data; ++ ++ vc_data.usecnt = atomic_read(&vxi->vx_usecnt); ++ vc_data.tasks = atomic_read(&vxi->vx_tasks); ++ ++ if (copy_to_user(data, &vc_data, sizeof(vc_data))) ++ return -EFAULT; ++ return 0; ++} ++ ++ ++/* context functions */ ++ ++int vc_ctx_create(uint32_t xid, void __user *data) ++{ ++ struct vcmd_ctx_create vc_data = { .flagword = VXF_INIT_SET }; ++ struct vx_info *new_vxi; ++ int ret; ++ ++ if (data && copy_from_user(&vc_data, data, sizeof(vc_data))) ++ return -EFAULT; ++ ++ if ((xid > MAX_S_CONTEXT) || (xid < 2)) ++ return -EINVAL; ++ ++ new_vxi = __create_vx_info(xid); ++ if (IS_ERR(new_vxi)) ++ return PTR_ERR(new_vxi); ++ ++ /* initial flags */ ++ new_vxi->vx_flags = vc_data.flagword; ++ ++ ret = -ENOEXEC; ++ if (vs_state_change(new_vxi, VSC_STARTUP)) ++ goto out; ++ ++ ret = vx_migrate_task(current, new_vxi, (!data)); ++ if (ret) ++ goto out; ++ ++ /* return context id on success */ ++ ret = new_vxi->vx_id; ++ ++ /* get a reference for persistent contexts */ ++ if ((vc_data.flagword & VXF_PERSISTENT)) ++ vx_set_persistent(new_vxi); ++out: ++ release_vx_info(new_vxi, NULL); ++ put_vx_info(new_vxi); ++ return ret; ++} ++ ++ ++int vc_ctx_migrate(struct vx_info *vxi, void __user *data) ++{ ++ struct vcmd_ctx_migrate vc_data = { .flagword = 0 }; ++ int ret; ++ ++ if (data && copy_from_user(&vc_data, data, sizeof(vc_data))) ++ return -EFAULT; ++ ++ ret = vx_migrate_task(current, vxi, 0); ++ if (ret) ++ return ret; ++ if (vc_data.flagword & VXM_SET_INIT) ++ ret = vx_set_init(vxi, current); ++ if (ret) ++ return ret; ++ if (vc_data.flagword & VXM_SET_REAPER) ++ ret = vx_set_reaper(vxi, current); ++ return ret; ++} ++ ++ ++int vc_get_cflags(struct vx_info *vxi, void __user *data) ++{ ++ struct vcmd_ctx_flags_v0 vc_data; ++ ++ vc_data.flagword = vxi->vx_flags; ++ ++ /* special STATE flag handling */ ++ vc_data.mask = vs_mask_flags(~0ULL, vxi->vx_flags, VXF_ONE_TIME); ++ ++ if (copy_to_user(data, &vc_data, sizeof(vc_data))) ++ return -EFAULT; ++ return 0; ++} ++ ++int vc_set_cflags(struct vx_info *vxi, void __user *data) ++{ ++ struct vcmd_ctx_flags_v0 vc_data; ++ uint64_t mask, trigger; ++ ++ if (copy_from_user(&vc_data, data, sizeof(vc_data))) ++ return -EFAULT; ++ ++ /* special STATE flag handling */ ++ mask = vs_mask_mask(vc_data.mask, vxi->vx_flags, VXF_ONE_TIME); ++ trigger = (mask & vxi->vx_flags) ^ (mask & vc_data.flagword); ++ ++ if (vxi == current_vx_info()) { ++ /* if (trigger & VXF_STATE_SETUP) ++ vx_mask_cap_bset(vxi, current); */ ++ if (trigger & VXF_STATE_INIT) { ++ int ret; ++ ++ ret = vx_set_init(vxi, current); ++ if (ret) ++ return ret; ++ ret = vx_set_reaper(vxi, current); ++ if (ret) ++ return ret; ++ } ++ } ++ ++ vxi->vx_flags = vs_mask_flags(vxi->vx_flags, ++ vc_data.flagword, mask); ++ if (trigger & VXF_PERSISTENT) ++ vx_update_persistent(vxi); ++ ++ return 0; ++} ++ ++ ++static inline uint64_t caps_from_cap_t(kernel_cap_t c) ++{ ++ uint64_t v = c.cap[0] | ((uint64_t)c.cap[1] << 32); ++ ++ // printk("caps_from_cap_t(%08x:%08x) = %016llx\n", c.cap[1], c.cap[0], v); ++ return v; ++} ++ ++static inline kernel_cap_t cap_t_from_caps(uint64_t v) ++{ ++ kernel_cap_t c = __cap_empty_set; ++ ++ c.cap[0] = v & 0xFFFFFFFF; ++ c.cap[1] = (v >> 32) & 0xFFFFFFFF; ++ ++ // printk("cap_t_from_caps(%016llx) = %08x:%08x\n", v, c.cap[1], c.cap[0]); ++ return c; ++} ++ ++ ++static int do_get_caps(struct vx_info *vxi, uint64_t *bcaps, uint64_t *ccaps) ++{ ++ if (bcaps) ++ *bcaps = caps_from_cap_t(vxi->vx_bcaps); ++ if (ccaps) ++ *ccaps = vxi->vx_ccaps; ++ ++ return 0; ++} ++ ++int vc_get_ccaps(struct vx_info *vxi, void __user *data) ++{ ++ struct vcmd_ctx_caps_v1 vc_data; ++ int ret; ++ ++ ret = do_get_caps(vxi, NULL, &vc_data.ccaps); ++ if (ret) ++ return ret; ++ vc_data.cmask = ~0ULL; ++ ++ if (copy_to_user(data, &vc_data, sizeof(vc_data))) ++ return -EFAULT; ++ return 0; ++} ++ ++static int do_set_caps(struct vx_info *vxi, ++ uint64_t bcaps, uint64_t bmask, uint64_t ccaps, uint64_t cmask) ++{ ++ uint64_t bcold = caps_from_cap_t(vxi->vx_bcaps); ++ ++#if 0 ++ printk("do_set_caps(%16llx, %16llx, %16llx, %16llx)\n", ++ bcaps, bmask, ccaps, cmask); ++#endif ++ vxi->vx_bcaps = cap_t_from_caps( ++ vs_mask_flags(bcold, bcaps, bmask)); ++ vxi->vx_ccaps = vs_mask_flags(vxi->vx_ccaps, ccaps, cmask); ++ ++ return 0; ++} ++ ++int vc_set_ccaps(struct vx_info *vxi, void __user *data) ++{ ++ struct vcmd_ctx_caps_v1 vc_data; ++ ++ if (copy_from_user(&vc_data, data, sizeof(vc_data))) ++ return -EFAULT; ++ ++ return do_set_caps(vxi, 0, 0, vc_data.ccaps, vc_data.cmask); ++} ++ ++int vc_get_bcaps(struct vx_info *vxi, void __user *data) ++{ ++ struct vcmd_bcaps vc_data; ++ int ret; ++ ++ ret = do_get_caps(vxi, &vc_data.bcaps, NULL); ++ if (ret) ++ return ret; ++ vc_data.bmask = ~0ULL; ++ ++ if (copy_to_user(data, &vc_data, sizeof(vc_data))) ++ return -EFAULT; ++ return 0; ++} ++ ++int vc_set_bcaps(struct vx_info *vxi, void __user *data) ++{ ++ struct vcmd_bcaps vc_data; ++ ++ if (copy_from_user(&vc_data, data, sizeof(vc_data))) ++ return -EFAULT; ++ ++ return do_set_caps(vxi, vc_data.bcaps, vc_data.bmask, 0, 0); ++} ++ ++ ++int vc_get_umask(struct vx_info *vxi, void __user *data) ++{ ++ struct vcmd_umask vc_data; ++ ++ vc_data.umask = vxi->vx_umask; ++ vc_data.mask = ~0ULL; ++ ++ if (copy_to_user(data, &vc_data, sizeof(vc_data))) ++ return -EFAULT; ++ return 0; ++} ++ ++int vc_set_umask(struct vx_info *vxi, void __user *data) ++{ ++ struct vcmd_umask vc_data; ++ ++ if (copy_from_user(&vc_data, data, sizeof(vc_data))) ++ return -EFAULT; ++ ++ vxi->vx_umask = vs_mask_flags(vxi->vx_umask, ++ vc_data.umask, vc_data.mask); ++ return 0; ++} ++ ++ ++int vc_get_wmask(struct vx_info *vxi, void __user *data) ++{ ++ struct vcmd_wmask vc_data; ++ ++ vc_data.wmask = vxi->vx_wmask; ++ vc_data.mask = ~0ULL; ++ ++ if (copy_to_user(data, &vc_data, sizeof(vc_data))) ++ return -EFAULT; ++ return 0; ++} ++ ++int vc_set_wmask(struct vx_info *vxi, void __user *data) ++{ ++ struct vcmd_wmask vc_data; ++ ++ if (copy_from_user(&vc_data, data, sizeof(vc_data))) ++ return -EFAULT; ++ ++ vxi->vx_wmask = vs_mask_flags(vxi->vx_wmask, ++ vc_data.wmask, vc_data.mask); ++ return 0; ++} ++ ++ ++int vc_get_badness(struct vx_info *vxi, void __user *data) ++{ ++ struct vcmd_badness_v0 vc_data; ++ ++ vc_data.bias = vxi->vx_badness_bias; ++ ++ if (copy_to_user(data, &vc_data, sizeof(vc_data))) ++ return -EFAULT; ++ return 0; ++} ++ ++int vc_set_badness(struct vx_info *vxi, void __user *data) ++{ ++ struct vcmd_badness_v0 vc_data; ++ ++ if (copy_from_user(&vc_data, data, sizeof(vc_data))) ++ return -EFAULT; ++ ++ vxi->vx_badness_bias = vc_data.bias; ++ return 0; ++} ++ ++#include <linux/module.h> ++ ++EXPORT_SYMBOL_GPL(free_vx_info); ++ +diff -NurpP --minimal linux-3.0.9/kernel/vserver/cvirt.c linux-3.0.9-vs2.3.2.1/kernel/vserver/cvirt.c +--- linux-3.0.9/kernel/vserver/cvirt.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/kernel/vserver/cvirt.c 2011-06-13 14:57:26.000000000 +0200 +@@ -0,0 +1,303 @@ ++/* ++ * linux/kernel/vserver/cvirt.c ++ * ++ * Virtual Server: Context Virtualization ++ * ++ * Copyright (C) 2004-2007 Herbert Pötzl ++ * ++ * V0.01 broken out from limit.c ++ * V0.02 added utsname stuff ++ * V0.03 changed vcmds to vxi arg ++ * ++ */ ++ ++#include <linux/types.h> ++#include <linux/utsname.h> ++#include <linux/vs_cvirt.h> ++#include <linux/vserver/switch.h> ++#include <linux/vserver/cvirt_cmd.h> ++ ++#include <asm/uaccess.h> ++ ++ ++void vx_vsi_uptime(struct timespec *uptime, struct timespec *idle) ++{ ++ struct vx_info *vxi = current_vx_info(); ++ ++ set_normalized_timespec(uptime, ++ uptime->tv_sec - vxi->cvirt.bias_uptime.tv_sec, ++ uptime->tv_nsec - vxi->cvirt.bias_uptime.tv_nsec); ++ if (!idle) ++ return; ++ set_normalized_timespec(idle, ++ idle->tv_sec - vxi->cvirt.bias_idle.tv_sec, ++ idle->tv_nsec - vxi->cvirt.bias_idle.tv_nsec); ++ return; ++} ++ ++uint64_t vx_idle_jiffies(void) ++{ ++ return init_task.utime + init_task.stime; ++} ++ ++ ++ ++static inline uint32_t __update_loadavg(uint32_t load, ++ int wsize, int delta, int n) ++{ ++ unsigned long long calc, prev; ++ ++ /* just set it to n */ ++ if (unlikely(delta >= wsize)) ++ return (n << FSHIFT); ++ ++ calc = delta * n; ++ calc <<= FSHIFT; ++ prev = (wsize - delta); ++ prev *= load; ++ calc += prev; ++ do_div(calc, wsize); ++ return calc; ++} ++ ++ ++void vx_update_load(struct vx_info *vxi) ++{ ++ uint32_t now, last, delta; ++ unsigned int nr_running, nr_uninterruptible; ++ unsigned int total; ++ unsigned long flags; ++ ++ spin_lock_irqsave(&vxi->cvirt.load_lock, flags); ++ ++ now = jiffies; ++ last = vxi->cvirt.load_last; ++ delta = now - last; ++ ++ if (delta < 5*HZ) ++ goto out; ++ ++ nr_running = atomic_read(&vxi->cvirt.nr_running); ++ nr_uninterruptible = atomic_read(&vxi->cvirt.nr_uninterruptible); ++ total = nr_running + nr_uninterruptible; ++ ++ vxi->cvirt.load[0] = __update_loadavg(vxi->cvirt.load[0], ++ 60*HZ, delta, total); ++ vxi->cvirt.load[1] = __update_loadavg(vxi->cvirt.load[1], ++ 5*60*HZ, delta, total); ++ vxi->cvirt.load[2] = __update_loadavg(vxi->cvirt.load[2], ++ 15*60*HZ, delta, total); ++ ++ vxi->cvirt.load_last = now; ++out: ++ atomic_inc(&vxi->cvirt.load_updates); ++ spin_unlock_irqrestore(&vxi->cvirt.load_lock, flags); ++} ++ ++ ++/* ++ * Commands to do_syslog: ++ * ++ * 0 -- Close the log. Currently a NOP. ++ * 1 -- Open the log. Currently a NOP. ++ * 2 -- Read from the log. ++ * 3 -- Read all messages remaining in the ring buffer. ++ * 4 -- Read and clear all messages remaining in the ring buffer ++ * 5 -- Clear ring buffer. ++ * 6 -- Disable printk's to console ++ * 7 -- Enable printk's to console ++ * 8 -- Set level of messages printed to console ++ * 9 -- Return number of unread characters in the log buffer ++ * 10 -- Return size of the log buffer ++ */ ++int vx_do_syslog(int type, char __user *buf, int len) ++{ ++ int error = 0; ++ int do_clear = 0; ++ struct vx_info *vxi = current_vx_info(); ++ struct _vx_syslog *log; ++ ++ if (!vxi) ++ return -EINVAL; ++ log = &vxi->cvirt.syslog; ++ ++ switch (type) { ++ case 0: /* Close log */ ++ case 1: /* Open log */ ++ break; ++ case 2: /* Read from log */ ++ error = wait_event_interruptible(log->log_wait, ++ (log->log_start - log->log_end)); ++ if (error) ++ break; ++ spin_lock_irq(&log->logbuf_lock); ++ spin_unlock_irq(&log->logbuf_lock); ++ break; ++ case 4: /* Read/clear last kernel messages */ ++ do_clear = 1; ++ /* fall through */ ++ case 3: /* Read last kernel messages */ ++ return 0; ++ ++ case 5: /* Clear ring buffer */ ++ return 0; ++ ++ case 6: /* Disable logging to console */ ++ case 7: /* Enable logging to console */ ++ case 8: /* Set level of messages printed to console */ ++ break; ++ ++ case 9: /* Number of chars in the log buffer */ ++ return 0; ++ case 10: /* Size of the log buffer */ ++ return 0; ++ default: ++ error = -EINVAL; ++ break; ++ } ++ return error; ++} ++ ++ ++/* virtual host info names */ ++ ++static char *vx_vhi_name(struct vx_info *vxi, int id) ++{ ++ struct nsproxy *nsproxy; ++ struct uts_namespace *uts; ++ ++ if (id == VHIN_CONTEXT) ++ return vxi->vx_name; ++ ++ nsproxy = vxi->space[0].vx_nsproxy; ++ if (!nsproxy) ++ return NULL; ++ ++ uts = nsproxy->uts_ns; ++ if (!uts) ++ return NULL; ++ ++ switch (id) { ++ case VHIN_SYSNAME: ++ return uts->name.sysname; ++ case VHIN_NODENAME: ++ return uts->name.nodename; ++ case VHIN_RELEASE: ++ return uts->name.release; ++ case VHIN_VERSION: ++ return uts->name.version; ++ case VHIN_MACHINE: ++ return uts->name.machine; ++ case VHIN_DOMAINNAME: ++ return uts->name.domainname; ++ default: ++ return NULL; ++ } ++ return NULL; ++} ++ ++int vc_set_vhi_name(struct vx_info *vxi, void __user *data) ++{ ++ struct vcmd_vhi_name_v0 vc_data; ++ char *name; ++ ++ if (copy_from_user(&vc_data, data, sizeof(vc_data))) ++ return -EFAULT; ++ ++ name = vx_vhi_name(vxi, vc_data.field); ++ if (!name) ++ return -EINVAL; ++ ++ memcpy(name, vc_data.name, 65); ++ return 0; ++} ++ ++int vc_get_vhi_name(struct vx_info *vxi, void __user *data) ++{ ++ struct vcmd_vhi_name_v0 vc_data; ++ char *name; ++ ++ if (copy_from_user(&vc_data, data, sizeof(vc_data))) ++ return -EFAULT; ++ ++ name = vx_vhi_name(vxi, vc_data.field); ++ if (!name) ++ return -EINVAL; ++ ++ memcpy(vc_data.name, name, 65); ++ if (copy_to_user(data, &vc_data, sizeof(vc_data))) ++ return -EFAULT; ++ return 0; ++} ++ ++ ++int vc_virt_stat(struct vx_info *vxi, void __user *data) ++{ ++ struct vcmd_virt_stat_v0 vc_data; ++ struct _vx_cvirt *cvirt = &vxi->cvirt; ++ struct timespec uptime; ++ ++ do_posix_clock_monotonic_gettime(&uptime); ++ set_normalized_timespec(&uptime, ++ uptime.tv_sec - cvirt->bias_uptime.tv_sec, ++ uptime.tv_nsec - cvirt->bias_uptime.tv_nsec); ++ ++ vc_data.offset = timespec_to_ns(&cvirt->bias_ts); ++ vc_data.uptime = timespec_to_ns(&uptime); ++ vc_data.nr_threads = atomic_read(&cvirt->nr_threads); ++ vc_data.nr_running = atomic_read(&cvirt->nr_running); ++ vc_data.nr_uninterruptible = atomic_read(&cvirt->nr_uninterruptible); ++ vc_data.nr_onhold = atomic_read(&cvirt->nr_onhold); ++ vc_data.nr_forks = atomic_read(&cvirt->total_forks); ++ vc_data.load[0] = cvirt->load[0]; ++ vc_data.load[1] = cvirt->load[1]; ++ vc_data.load[2] = cvirt->load[2]; ++ ++ if (copy_to_user(data, &vc_data, sizeof(vc_data))) ++ return -EFAULT; ++ return 0; ++} ++ ++ ++#ifdef CONFIG_VSERVER_VTIME ++ ++/* virtualized time base */ ++ ++void vx_adjust_timespec(struct timespec *ts) ++{ ++ struct vx_info *vxi; ++ ++ if (!vx_flags(VXF_VIRT_TIME, 0)) ++ return; ++ ++ vxi = current_vx_info(); ++ ts->tv_sec += vxi->cvirt.bias_ts.tv_sec; ++ ts->tv_nsec += vxi->cvirt.bias_ts.tv_nsec; ++ ++ if (ts->tv_nsec >= NSEC_PER_SEC) { ++ ts->tv_sec++; ++ ts->tv_nsec -= NSEC_PER_SEC; ++ } else if (ts->tv_nsec < 0) { ++ ts->tv_sec--; ++ ts->tv_nsec += NSEC_PER_SEC; ++ } ++} ++ ++int vx_settimeofday(const struct timespec *ts) ++{ ++ struct timespec ats, delta; ++ struct vx_info *vxi; ++ ++ if (!vx_flags(VXF_VIRT_TIME, 0)) ++ return do_settimeofday(ts); ++ ++ getnstimeofday(&ats); ++ delta = timespec_sub(*ts, ats); ++ ++ vxi = current_vx_info(); ++ vxi->cvirt.bias_ts = timespec_add(vxi->cvirt.bias_ts, delta); ++ return 0; ++} ++ ++#endif ++ +diff -NurpP --minimal linux-3.0.9/kernel/vserver/cvirt_init.h linux-3.0.9-vs2.3.2.1/kernel/vserver/cvirt_init.h +--- linux-3.0.9/kernel/vserver/cvirt_init.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/kernel/vserver/cvirt_init.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,70 @@ ++ ++ ++extern uint64_t vx_idle_jiffies(void); ++ ++static inline void vx_info_init_cvirt(struct _vx_cvirt *cvirt) ++{ ++ uint64_t idle_jiffies = vx_idle_jiffies(); ++ uint64_t nsuptime; ++ ++ do_posix_clock_monotonic_gettime(&cvirt->bias_uptime); ++ nsuptime = (unsigned long long)cvirt->bias_uptime.tv_sec ++ * NSEC_PER_SEC + cvirt->bias_uptime.tv_nsec; ++ cvirt->bias_clock = nsec_to_clock_t(nsuptime); ++ cvirt->bias_ts.tv_sec = 0; ++ cvirt->bias_ts.tv_nsec = 0; ++ ++ jiffies_to_timespec(idle_jiffies, &cvirt->bias_idle); ++ atomic_set(&cvirt->nr_threads, 0); ++ atomic_set(&cvirt->nr_running, 0); ++ atomic_set(&cvirt->nr_uninterruptible, 0); ++ atomic_set(&cvirt->nr_onhold, 0); ++ ++ spin_lock_init(&cvirt->load_lock); ++ cvirt->load_last = jiffies; ++ atomic_set(&cvirt->load_updates, 0); ++ cvirt->load[0] = 0; ++ cvirt->load[1] = 0; ++ cvirt->load[2] = 0; ++ atomic_set(&cvirt->total_forks, 0); ++ ++ spin_lock_init(&cvirt->syslog.logbuf_lock); ++ init_waitqueue_head(&cvirt->syslog.log_wait); ++ cvirt->syslog.log_start = 0; ++ cvirt->syslog.log_end = 0; ++ cvirt->syslog.con_start = 0; ++ cvirt->syslog.logged_chars = 0; ++} ++ ++static inline ++void vx_info_init_cvirt_pc(struct _vx_cvirt_pc *cvirt_pc, int cpu) ++{ ++ // cvirt_pc->cpustat = { 0 }; ++} ++ ++static inline void vx_info_exit_cvirt(struct _vx_cvirt *cvirt) ++{ ++#ifdef CONFIG_VSERVER_WARN ++ int value; ++#endif ++ vxwprintk_xid((value = atomic_read(&cvirt->nr_threads)), ++ "!!! cvirt: %p[nr_threads] = %d on exit.", ++ cvirt, value); ++ vxwprintk_xid((value = atomic_read(&cvirt->nr_running)), ++ "!!! cvirt: %p[nr_running] = %d on exit.", ++ cvirt, value); ++ vxwprintk_xid((value = atomic_read(&cvirt->nr_uninterruptible)), ++ "!!! cvirt: %p[nr_uninterruptible] = %d on exit.", ++ cvirt, value); ++ vxwprintk_xid((value = atomic_read(&cvirt->nr_onhold)), ++ "!!! cvirt: %p[nr_onhold] = %d on exit.", ++ cvirt, value); ++ return; ++} ++ ++static inline ++void vx_info_exit_cvirt_pc(struct _vx_cvirt_pc *cvirt_pc, int cpu) ++{ ++ return; ++} ++ +diff -NurpP --minimal linux-3.0.9/kernel/vserver/cvirt_proc.h linux-3.0.9-vs2.3.2.1/kernel/vserver/cvirt_proc.h +--- linux-3.0.9/kernel/vserver/cvirt_proc.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/kernel/vserver/cvirt_proc.h 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,135 @@ ++#ifndef _VX_CVIRT_PROC_H ++#define _VX_CVIRT_PROC_H ++ ++#include <linux/nsproxy.h> ++#include <linux/mnt_namespace.h> ++#include <linux/ipc_namespace.h> ++#include <linux/utsname.h> ++#include <linux/ipc.h> ++ ++ ++static inline ++int vx_info_proc_nsproxy(struct nsproxy *nsproxy, char *buffer) ++{ ++ struct mnt_namespace *ns; ++ struct uts_namespace *uts; ++ struct ipc_namespace *ipc; ++ struct path path; ++ char *pstr, *root; ++ int length = 0; ++ ++ if (!nsproxy) ++ goto out; ++ ++ length += sprintf(buffer + length, ++ "NSProxy:\t%p [%p,%p,%p]\n", ++ nsproxy, nsproxy->mnt_ns, ++ nsproxy->uts_ns, nsproxy->ipc_ns); ++ ++ ns = nsproxy->mnt_ns; ++ if (!ns) ++ goto skip_ns; ++ ++ pstr = kmalloc(PATH_MAX, GFP_KERNEL); ++ if (!pstr) ++ goto skip_ns; ++ ++ path.mnt = ns->root; ++ path.dentry = ns->root->mnt_root; ++ root = d_path(&path, pstr, PATH_MAX - 2); ++ length += sprintf(buffer + length, ++ "Namespace:\t%p [#%u]\n" ++ "RootPath:\t%s\n", ++ ns, atomic_read(&ns->count), ++ root); ++ kfree(pstr); ++skip_ns: ++ ++ uts = nsproxy->uts_ns; ++ if (!uts) ++ goto skip_uts; ++ ++ length += sprintf(buffer + length, ++ "SysName:\t%.*s\n" ++ "NodeName:\t%.*s\n" ++ "Release:\t%.*s\n" ++ "Version:\t%.*s\n" ++ "Machine:\t%.*s\n" ++ "DomainName:\t%.*s\n", ++ __NEW_UTS_LEN, uts->name.sysname, ++ __NEW_UTS_LEN, uts->name.nodename, ++ __NEW_UTS_LEN, uts->name.release, ++ __NEW_UTS_LEN, uts->name.version, ++ __NEW_UTS_LEN, uts->name.machine, ++ __NEW_UTS_LEN, uts->name.domainname); ++skip_uts: ++ ++ ipc = nsproxy->ipc_ns; ++ if (!ipc) ++ goto skip_ipc; ++ ++ length += sprintf(buffer + length, ++ "SEMS:\t\t%d %d %d %d %d\n" ++ "MSG:\t\t%d %d %d\n" ++ "SHM:\t\t%lu %lu %d %d\n", ++ ipc->sem_ctls[0], ipc->sem_ctls[1], ++ ipc->sem_ctls[2], ipc->sem_ctls[3], ++ ipc->used_sems, ++ ipc->msg_ctlmax, ipc->msg_ctlmnb, ipc->msg_ctlmni, ++ (unsigned long)ipc->shm_ctlmax, ++ (unsigned long)ipc->shm_ctlall, ++ ipc->shm_ctlmni, ipc->shm_tot); ++skip_ipc: ++out: ++ return length; ++} ++ ++ ++#include <linux/sched.h> ++ ++#define LOAD_INT(x) ((x) >> FSHIFT) ++#define LOAD_FRAC(x) LOAD_INT(((x) & (FIXED_1 - 1)) * 100) ++ ++static inline ++int vx_info_proc_cvirt(struct _vx_cvirt *cvirt, char *buffer) ++{ ++ int length = 0; ++ int a, b, c; ++ ++ length += sprintf(buffer + length, ++ "BiasUptime:\t%lu.%02lu\n", ++ (unsigned long)cvirt->bias_uptime.tv_sec, ++ (cvirt->bias_uptime.tv_nsec / (NSEC_PER_SEC / 100))); ++ ++ a = cvirt->load[0] + (FIXED_1 / 200); ++ b = cvirt->load[1] + (FIXED_1 / 200); ++ c = cvirt->load[2] + (FIXED_1 / 200); ++ length += sprintf(buffer + length, ++ "nr_threads:\t%d\n" ++ "nr_running:\t%d\n" ++ "nr_unintr:\t%d\n" ++ "nr_onhold:\t%d\n" ++ "load_updates:\t%d\n" ++ "loadavg:\t%d.%02d %d.%02d %d.%02d\n" ++ "total_forks:\t%d\n", ++ atomic_read(&cvirt->nr_threads), ++ atomic_read(&cvirt->nr_running), ++ atomic_read(&cvirt->nr_uninterruptible), ++ atomic_read(&cvirt->nr_onhold), ++ atomic_read(&cvirt->load_updates), ++ LOAD_INT(a), LOAD_FRAC(a), ++ LOAD_INT(b), LOAD_FRAC(b), ++ LOAD_INT(c), LOAD_FRAC(c), ++ atomic_read(&cvirt->total_forks)); ++ return length; ++} ++ ++static inline ++int vx_info_proc_cvirt_pc(struct _vx_cvirt_pc *cvirt_pc, ++ char *buffer, int cpu) ++{ ++ int length = 0; ++ return length; ++} ++ ++#endif /* _VX_CVIRT_PROC_H */ +diff -NurpP --minimal linux-3.0.9/kernel/vserver/debug.c linux-3.0.9-vs2.3.2.1/kernel/vserver/debug.c +--- linux-3.0.9/kernel/vserver/debug.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/kernel/vserver/debug.c 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,32 @@ ++/* ++ * kernel/vserver/debug.c ++ * ++ * Copyright (C) 2005-2007 Herbert Pötzl ++ * ++ * V0.01 vx_info dump support ++ * ++ */ ++ ++#include <linux/module.h> ++ ++#include <linux/vserver/context.h> ++ ++ ++void dump_vx_info(struct vx_info *vxi, int level) ++{ ++ printk("vx_info %p[#%d, %d.%d, %4x]\n", vxi, vxi->vx_id, ++ atomic_read(&vxi->vx_usecnt), ++ atomic_read(&vxi->vx_tasks), ++ vxi->vx_state); ++ if (level > 0) { ++ __dump_vx_limit(&vxi->limit); ++ __dump_vx_sched(&vxi->sched); ++ __dump_vx_cvirt(&vxi->cvirt); ++ __dump_vx_cacct(&vxi->cacct); ++ } ++ printk("---\n"); ++} ++ ++ ++EXPORT_SYMBOL_GPL(dump_vx_info); ++ +diff -NurpP --minimal linux-3.0.9/kernel/vserver/device.c linux-3.0.9-vs2.3.2.1/kernel/vserver/device.c +--- linux-3.0.9/kernel/vserver/device.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/kernel/vserver/device.c 2011-06-10 23:20:56.000000000 +0200 +@@ -0,0 +1,443 @@ ++/* ++ * linux/kernel/vserver/device.c ++ * ++ * Linux-VServer: Device Support ++ * ++ * Copyright (C) 2006 Herbert Pötzl ++ * Copyright (C) 2007 Daniel Hokka Zakrisson ++ * ++ * V0.01 device mapping basics ++ * V0.02 added defaults ++ * ++ */ ++ ++#include <linux/slab.h> ++#include <linux/rcupdate.h> ++#include <linux/fs.h> ++#include <linux/namei.h> ++#include <linux/hash.h> ++ ++#include <asm/errno.h> ++#include <asm/uaccess.h> ++#include <linux/vserver/base.h> ++#include <linux/vserver/debug.h> ++#include <linux/vserver/context.h> ++#include <linux/vserver/device.h> ++#include <linux/vserver/device_cmd.h> ++ ++ ++#define DMAP_HASH_BITS 4 ++ ++ ++struct vs_mapping { ++ union { ++ struct hlist_node hlist; ++ struct list_head list; ++ } u; ++#define dm_hlist u.hlist ++#define dm_list u.list ++ xid_t xid; ++ dev_t device; ++ struct vx_dmap_target target; ++}; ++ ++ ++static struct hlist_head dmap_main_hash[1 << DMAP_HASH_BITS]; ++ ++static DEFINE_SPINLOCK(dmap_main_hash_lock); ++ ++static struct vx_dmap_target dmap_defaults[2] = { ++ { .flags = DATTR_OPEN }, ++ { .flags = DATTR_OPEN }, ++}; ++ ++ ++struct kmem_cache *dmap_cachep __read_mostly; ++ ++int __init dmap_cache_init(void) ++{ ++ dmap_cachep = kmem_cache_create("dmap_cache", ++ sizeof(struct vs_mapping), 0, ++ SLAB_HWCACHE_ALIGN|SLAB_PANIC, NULL); ++ return 0; ++} ++ ++__initcall(dmap_cache_init); ++ ++ ++static inline unsigned int __hashval(dev_t dev, int bits) ++{ ++ return hash_long((unsigned long)dev, bits); ++} ++ ++ ++/* __hash_mapping() ++ * add the mapping to the hash table ++ */ ++static inline void __hash_mapping(struct vx_info *vxi, struct vs_mapping *vdm) ++{ ++ spinlock_t *hash_lock = &dmap_main_hash_lock; ++ struct hlist_head *head, *hash = dmap_main_hash; ++ int device = vdm->device; ++ ++ spin_lock(hash_lock); ++ vxdprintk(VXD_CBIT(misc, 8), "__hash_mapping: %p[#%d] %08x:%08x", ++ vxi, vxi ? vxi->vx_id : 0, device, vdm->target.target); ++ ++ head = &hash[__hashval(device, DMAP_HASH_BITS)]; ++ hlist_add_head(&vdm->dm_hlist, head); ++ spin_unlock(hash_lock); ++} ++ ++ ++static inline int __mode_to_default(umode_t mode) ++{ ++ switch (mode) { ++ case S_IFBLK: ++ return 0; ++ case S_IFCHR: ++ return 1; ++ default: ++ BUG(); ++ } ++} ++ ++ ++/* __set_default() ++ * set a default ++ */ ++static inline void __set_default(struct vx_info *vxi, umode_t mode, ++ struct vx_dmap_target *vdmt) ++{ ++ spinlock_t *hash_lock = &dmap_main_hash_lock; ++ spin_lock(hash_lock); ++ ++ if (vxi) ++ vxi->dmap.targets[__mode_to_default(mode)] = *vdmt; ++ else ++ dmap_defaults[__mode_to_default(mode)] = *vdmt; ++ ++ ++ spin_unlock(hash_lock); ++ ++ vxdprintk(VXD_CBIT(misc, 8), "__set_default: %p[#%u] %08x %04x", ++ vxi, vxi ? vxi->vx_id : 0, vdmt->target, vdmt->flags); ++} ++ ++ ++/* __remove_default() ++ * remove a default ++ */ ++static inline int __remove_default(struct vx_info *vxi, umode_t mode) ++{ ++ spinlock_t *hash_lock = &dmap_main_hash_lock; ++ spin_lock(hash_lock); ++ ++ if (vxi) ++ vxi->dmap.targets[__mode_to_default(mode)].flags = 0; ++ else /* remove == reset */ ++ dmap_defaults[__mode_to_default(mode)].flags = DATTR_OPEN | mode; ++ ++ spin_unlock(hash_lock); ++ return 0; ++} ++ ++ ++/* __find_mapping() ++ * find a mapping in the hash table ++ * ++ * caller must hold hash_lock ++ */ ++static inline int __find_mapping(xid_t xid, dev_t device, umode_t mode, ++ struct vs_mapping **local, struct vs_mapping **global) ++{ ++ struct hlist_head *hash = dmap_main_hash; ++ struct hlist_head *head = &hash[__hashval(device, DMAP_HASH_BITS)]; ++ struct hlist_node *pos; ++ struct vs_mapping *vdm; ++ ++ *local = NULL; ++ if (global) ++ *global = NULL; ++ ++ hlist_for_each(pos, head) { ++ vdm = hlist_entry(pos, struct vs_mapping, dm_hlist); ++ ++ if ((vdm->device == device) && ++ !((vdm->target.flags ^ mode) & S_IFMT)) { ++ if (vdm->xid == xid) { ++ *local = vdm; ++ return 1; ++ } else if (global && vdm->xid == 0) ++ *global = vdm; ++ } ++ } ++ ++ if (global && *global) ++ return 0; ++ else ++ return -ENOENT; ++} ++ ++ ++/* __lookup_mapping() ++ * find a mapping and store the result in target and flags ++ */ ++static inline int __lookup_mapping(struct vx_info *vxi, ++ dev_t device, dev_t *target, int *flags, umode_t mode) ++{ ++ spinlock_t *hash_lock = &dmap_main_hash_lock; ++ struct vs_mapping *vdm, *global; ++ struct vx_dmap_target *vdmt; ++ int ret = 0; ++ xid_t xid = vxi->vx_id; ++ int index; ++ ++ spin_lock(hash_lock); ++ if (__find_mapping(xid, device, mode, &vdm, &global) > 0) { ++ ret = 1; ++ vdmt = &vdm->target; ++ goto found; ++ } ++ ++ index = __mode_to_default(mode); ++ if (vxi && vxi->dmap.targets[index].flags) { ++ ret = 2; ++ vdmt = &vxi->dmap.targets[index]; ++ } else if (global) { ++ ret = 3; ++ vdmt = &global->target; ++ goto found; ++ } else { ++ ret = 4; ++ vdmt = &dmap_defaults[index]; ++ } ++ ++found: ++ if (target && (vdmt->flags & DATTR_REMAP)) ++ *target = vdmt->target; ++ else if (target) ++ *target = device; ++ if (flags) ++ *flags = vdmt->flags; ++ ++ spin_unlock(hash_lock); ++ ++ return ret; ++} ++ ++ ++/* __remove_mapping() ++ * remove a mapping from the hash table ++ */ ++static inline int __remove_mapping(struct vx_info *vxi, dev_t device, ++ umode_t mode) ++{ ++ spinlock_t *hash_lock = &dmap_main_hash_lock; ++ struct vs_mapping *vdm = NULL; ++ int ret = 0; ++ ++ spin_lock(hash_lock); ++ ++ ret = __find_mapping((vxi ? vxi->vx_id : 0), device, mode, &vdm, ++ NULL); ++ vxdprintk(VXD_CBIT(misc, 8), "__remove_mapping: %p[#%d] %08x %04x", ++ vxi, vxi ? vxi->vx_id : 0, device, mode); ++ if (ret < 0) ++ goto out; ++ hlist_del(&vdm->dm_hlist); ++ ++out: ++ spin_unlock(hash_lock); ++ if (vdm) ++ kmem_cache_free(dmap_cachep, vdm); ++ return ret; ++} ++ ++ ++ ++int vs_map_device(struct vx_info *vxi, ++ dev_t device, dev_t *target, umode_t mode) ++{ ++ int ret, flags = DATTR_MASK; ++ ++ if (!vxi) { ++ if (target) ++ *target = device; ++ goto out; ++ } ++ ret = __lookup_mapping(vxi, device, target, &flags, mode); ++ vxdprintk(VXD_CBIT(misc, 8), "vs_map_device: %08x target: %08x flags: %04x mode: %04x mapped=%d", ++ device, target ? *target : 0, flags, mode, ret); ++out: ++ return (flags & DATTR_MASK); ++} ++ ++ ++ ++static int do_set_mapping(struct vx_info *vxi, ++ dev_t device, dev_t target, int flags, umode_t mode) ++{ ++ if (device) { ++ struct vs_mapping *new; ++ ++ new = kmem_cache_alloc(dmap_cachep, GFP_KERNEL); ++ if (!new) ++ return -ENOMEM; ++ ++ INIT_HLIST_NODE(&new->dm_hlist); ++ new->device = device; ++ new->target.target = target; ++ new->target.flags = flags | mode; ++ new->xid = (vxi ? vxi->vx_id : 0); ++ ++ vxdprintk(VXD_CBIT(misc, 8), "do_set_mapping: %08x target: %08x flags: %04x", device, target, flags); ++ __hash_mapping(vxi, new); ++ } else { ++ struct vx_dmap_target new = { ++ .target = target, ++ .flags = flags | mode, ++ }; ++ __set_default(vxi, mode, &new); ++ } ++ return 0; ++} ++ ++ ++static int do_unset_mapping(struct vx_info *vxi, ++ dev_t device, dev_t target, int flags, umode_t mode) ++{ ++ int ret = -EINVAL; ++ ++ if (device) { ++ ret = __remove_mapping(vxi, device, mode); ++ if (ret < 0) ++ goto out; ++ } else { ++ ret = __remove_default(vxi, mode); ++ if (ret < 0) ++ goto out; ++ } ++ ++out: ++ return ret; ++} ++ ++ ++static inline int __user_device(const char __user *name, dev_t *dev, ++ umode_t *mode) ++{ ++ struct nameidata nd; ++ int ret; ++ ++ if (!name) { ++ *dev = 0; ++ return 0; ++ } ++ ret = user_lpath(name, &nd.path); ++ if (ret) ++ return ret; ++ if (nd.path.dentry->d_inode) { ++ *dev = nd.path.dentry->d_inode->i_rdev; ++ *mode = nd.path.dentry->d_inode->i_mode; ++ } ++ path_put(&nd.path); ++ return 0; ++} ++ ++static inline int __mapping_mode(dev_t device, dev_t target, ++ umode_t device_mode, umode_t target_mode, umode_t *mode) ++{ ++ if (device) ++ *mode = device_mode & S_IFMT; ++ else if (target) ++ *mode = target_mode & S_IFMT; ++ else ++ return -EINVAL; ++ ++ /* if both given, device and target mode have to match */ ++ if (device && target && ++ ((device_mode ^ target_mode) & S_IFMT)) ++ return -EINVAL; ++ return 0; ++} ++ ++ ++static inline int do_mapping(struct vx_info *vxi, const char __user *device_path, ++ const char __user *target_path, int flags, int set) ++{ ++ dev_t device = ~0, target = ~0; ++ umode_t device_mode = 0, target_mode = 0, mode; ++ int ret; ++ ++ ret = __user_device(device_path, &device, &device_mode); ++ if (ret) ++ return ret; ++ ret = __user_device(target_path, &target, &target_mode); ++ if (ret) ++ return ret; ++ ++ ret = __mapping_mode(device, target, ++ device_mode, target_mode, &mode); ++ if (ret) ++ return ret; ++ ++ if (set) ++ return do_set_mapping(vxi, device, target, ++ flags, mode); ++ else ++ return do_unset_mapping(vxi, device, target, ++ flags, mode); ++} ++ ++ ++int vc_set_mapping(struct vx_info *vxi, void __user *data) ++{ ++ struct vcmd_set_mapping_v0 vc_data; ++ ++ if (copy_from_user(&vc_data, data, sizeof(vc_data))) ++ return -EFAULT; ++ ++ return do_mapping(vxi, vc_data.device, vc_data.target, ++ vc_data.flags, 1); ++} ++ ++int vc_unset_mapping(struct vx_info *vxi, void __user *data) ++{ ++ struct vcmd_set_mapping_v0 vc_data; ++ ++ if (copy_from_user(&vc_data, data, sizeof(vc_data))) ++ return -EFAULT; ++ ++ return do_mapping(vxi, vc_data.device, vc_data.target, ++ vc_data.flags, 0); ++} ++ ++ ++#ifdef CONFIG_COMPAT ++ ++int vc_set_mapping_x32(struct vx_info *vxi, void __user *data) ++{ ++ struct vcmd_set_mapping_v0_x32 vc_data; ++ ++ if (copy_from_user(&vc_data, data, sizeof(vc_data))) ++ return -EFAULT; ++ ++ return do_mapping(vxi, compat_ptr(vc_data.device_ptr), ++ compat_ptr(vc_data.target_ptr), vc_data.flags, 1); ++} ++ ++int vc_unset_mapping_x32(struct vx_info *vxi, void __user *data) ++{ ++ struct vcmd_set_mapping_v0_x32 vc_data; ++ ++ if (copy_from_user(&vc_data, data, sizeof(vc_data))) ++ return -EFAULT; ++ ++ return do_mapping(vxi, compat_ptr(vc_data.device_ptr), ++ compat_ptr(vc_data.target_ptr), vc_data.flags, 0); ++} ++ ++#endif /* CONFIG_COMPAT */ ++ ++ +diff -NurpP --minimal linux-3.0.9/kernel/vserver/dlimit.c linux-3.0.9-vs2.3.2.1/kernel/vserver/dlimit.c +--- linux-3.0.9/kernel/vserver/dlimit.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/kernel/vserver/dlimit.c 2011-06-10 23:20:56.000000000 +0200 +@@ -0,0 +1,531 @@ ++/* ++ * linux/kernel/vserver/dlimit.c ++ * ++ * Virtual Server: Context Disk Limits ++ * ++ * Copyright (C) 2004-2009 Herbert Pötzl ++ * ++ * V0.01 initial version ++ * V0.02 compat32 splitup ++ * V0.03 extended interface ++ * ++ */ ++ ++#include <linux/statfs.h> ++#include <linux/sched.h> ++#include <linux/namei.h> ++#include <linux/vs_tag.h> ++#include <linux/vs_dlimit.h> ++#include <linux/vserver/dlimit_cmd.h> ++#include <linux/slab.h> ++// #include <linux/gfp.h> ++ ++#include <asm/uaccess.h> ++ ++/* __alloc_dl_info() ++ ++ * allocate an initialized dl_info struct ++ * doesn't make it visible (hash) */ ++ ++static struct dl_info *__alloc_dl_info(struct super_block *sb, tag_t tag) ++{ ++ struct dl_info *new = NULL; ++ ++ vxdprintk(VXD_CBIT(dlim, 5), ++ "alloc_dl_info(%p,%d)*", sb, tag); ++ ++ /* would this benefit from a slab cache? */ ++ new = kmalloc(sizeof(struct dl_info), GFP_KERNEL); ++ if (!new) ++ return 0; ++ ++ memset(new, 0, sizeof(struct dl_info)); ++ new->dl_tag = tag; ++ new->dl_sb = sb; ++ // INIT_RCU_HEAD(&new->dl_rcu); ++ INIT_HLIST_NODE(&new->dl_hlist); ++ spin_lock_init(&new->dl_lock); ++ atomic_set(&new->dl_refcnt, 0); ++ atomic_set(&new->dl_usecnt, 0); ++ ++ /* rest of init goes here */ ++ ++ vxdprintk(VXD_CBIT(dlim, 4), ++ "alloc_dl_info(%p,%d) = %p", sb, tag, new); ++ return new; ++} ++ ++/* __dealloc_dl_info() ++ ++ * final disposal of dl_info */ ++ ++static void __dealloc_dl_info(struct dl_info *dli) ++{ ++ vxdprintk(VXD_CBIT(dlim, 4), ++ "dealloc_dl_info(%p)", dli); ++ ++ dli->dl_hlist.next = LIST_POISON1; ++ dli->dl_tag = -1; ++ dli->dl_sb = 0; ++ ++ BUG_ON(atomic_read(&dli->dl_usecnt)); ++ BUG_ON(atomic_read(&dli->dl_refcnt)); ++ ++ kfree(dli); ++} ++ ++ ++/* hash table for dl_info hash */ ++ ++#define DL_HASH_SIZE 13 ++ ++struct hlist_head dl_info_hash[DL_HASH_SIZE]; ++ ++static DEFINE_SPINLOCK(dl_info_hash_lock); ++ ++ ++static inline unsigned int __hashval(struct super_block *sb, tag_t tag) ++{ ++ return ((tag ^ (unsigned long)sb) % DL_HASH_SIZE); ++} ++ ++ ++ ++/* __hash_dl_info() ++ ++ * add the dli to the global hash table ++ * requires the hash_lock to be held */ ++ ++static inline void __hash_dl_info(struct dl_info *dli) ++{ ++ struct hlist_head *head; ++ ++ vxdprintk(VXD_CBIT(dlim, 6), ++ "__hash_dl_info: %p[#%d]", dli, dli->dl_tag); ++ get_dl_info(dli); ++ head = &dl_info_hash[__hashval(dli->dl_sb, dli->dl_tag)]; ++ hlist_add_head_rcu(&dli->dl_hlist, head); ++} ++ ++/* __unhash_dl_info() ++ ++ * remove the dli from the global hash table ++ * requires the hash_lock to be held */ ++ ++static inline void __unhash_dl_info(struct dl_info *dli) ++{ ++ vxdprintk(VXD_CBIT(dlim, 6), ++ "__unhash_dl_info: %p[#%d]", dli, dli->dl_tag); ++ hlist_del_rcu(&dli->dl_hlist); ++ put_dl_info(dli); ++} ++ ++ ++/* __lookup_dl_info() ++ ++ * requires the rcu_read_lock() ++ * doesn't increment the dl_refcnt */ ++ ++static inline struct dl_info *__lookup_dl_info(struct super_block *sb, tag_t tag) ++{ ++ struct hlist_head *head = &dl_info_hash[__hashval(sb, tag)]; ++ struct hlist_node *pos; ++ struct dl_info *dli; ++ ++ hlist_for_each_entry_rcu(dli, pos, head, dl_hlist) { ++ ++ if (dli->dl_tag == tag && dli->dl_sb == sb) { ++ return dli; ++ } ++ } ++ return NULL; ++} ++ ++ ++struct dl_info *locate_dl_info(struct super_block *sb, tag_t tag) ++{ ++ struct dl_info *dli; ++ ++ rcu_read_lock(); ++ dli = get_dl_info(__lookup_dl_info(sb, tag)); ++ vxdprintk(VXD_CBIT(dlim, 7), ++ "locate_dl_info(%p,#%d) = %p", sb, tag, dli); ++ rcu_read_unlock(); ++ return dli; ++} ++ ++void rcu_free_dl_info(struct rcu_head *head) ++{ ++ struct dl_info *dli = container_of(head, struct dl_info, dl_rcu); ++ int usecnt, refcnt; ++ ++ BUG_ON(!dli || !head); ++ ++ usecnt = atomic_read(&dli->dl_usecnt); ++ BUG_ON(usecnt < 0); ++ ++ refcnt = atomic_read(&dli->dl_refcnt); ++ BUG_ON(refcnt < 0); ++ ++ vxdprintk(VXD_CBIT(dlim, 3), ++ "rcu_free_dl_info(%p)", dli); ++ if (!usecnt) ++ __dealloc_dl_info(dli); ++ else ++ printk("!!! rcu didn't free\n"); ++} ++ ++ ++ ++ ++static int do_addrem_dlimit(uint32_t id, const char __user *name, ++ uint32_t flags, int add) ++{ ++ struct path path; ++ int ret; ++ ++ ret = user_lpath(name, &path); ++ if (!ret) { ++ struct super_block *sb; ++ struct dl_info *dli; ++ ++ ret = -EINVAL; ++ if (!path.dentry->d_inode) ++ goto out_release; ++ if (!(sb = path.dentry->d_inode->i_sb)) ++ goto out_release; ++ ++ if (add) { ++ dli = __alloc_dl_info(sb, id); ++ spin_lock(&dl_info_hash_lock); ++ ++ ret = -EEXIST; ++ if (__lookup_dl_info(sb, id)) ++ goto out_unlock; ++ __hash_dl_info(dli); ++ dli = NULL; ++ } else { ++ spin_lock(&dl_info_hash_lock); ++ dli = __lookup_dl_info(sb, id); ++ ++ ret = -ESRCH; ++ if (!dli) ++ goto out_unlock; ++ __unhash_dl_info(dli); ++ } ++ ret = 0; ++ out_unlock: ++ spin_unlock(&dl_info_hash_lock); ++ if (add && dli) ++ __dealloc_dl_info(dli); ++ out_release: ++ path_put(&path); ++ } ++ return ret; ++} ++ ++int vc_add_dlimit(uint32_t id, void __user *data) ++{ ++ struct vcmd_ctx_dlimit_base_v0 vc_data; ++ ++ if (copy_from_user(&vc_data, data, sizeof(vc_data))) ++ return -EFAULT; ++ ++ return do_addrem_dlimit(id, vc_data.name, vc_data.flags, 1); ++} ++ ++int vc_rem_dlimit(uint32_t id, void __user *data) ++{ ++ struct vcmd_ctx_dlimit_base_v0 vc_data; ++ ++ if (copy_from_user(&vc_data, data, sizeof(vc_data))) ++ return -EFAULT; ++ ++ return do_addrem_dlimit(id, vc_data.name, vc_data.flags, 0); ++} ++ ++#ifdef CONFIG_COMPAT ++ ++int vc_add_dlimit_x32(uint32_t id, void __user *data) ++{ ++ struct vcmd_ctx_dlimit_base_v0_x32 vc_data; ++ ++ if (copy_from_user(&vc_data, data, sizeof(vc_data))) ++ return -EFAULT; ++ ++ return do_addrem_dlimit(id, ++ compat_ptr(vc_data.name_ptr), vc_data.flags, 1); ++} ++ ++int vc_rem_dlimit_x32(uint32_t id, void __user *data) ++{ ++ struct vcmd_ctx_dlimit_base_v0_x32 vc_data; ++ ++ if (copy_from_user(&vc_data, data, sizeof(vc_data))) ++ return -EFAULT; ++ ++ return do_addrem_dlimit(id, ++ compat_ptr(vc_data.name_ptr), vc_data.flags, 0); ++} ++ ++#endif /* CONFIG_COMPAT */ ++ ++ ++static inline ++int do_set_dlimit(uint32_t id, const char __user *name, ++ uint32_t space_used, uint32_t space_total, ++ uint32_t inodes_used, uint32_t inodes_total, ++ uint32_t reserved, uint32_t flags) ++{ ++ struct path path; ++ int ret; ++ ++ ret = user_lpath(name, &path); ++ if (!ret) { ++ struct super_block *sb; ++ struct dl_info *dli; ++ ++ ret = -EINVAL; ++ if (!path.dentry->d_inode) ++ goto out_release; ++ if (!(sb = path.dentry->d_inode->i_sb)) ++ goto out_release; ++ ++ /* sanity checks */ ++ if ((reserved != CDLIM_KEEP && ++ reserved > 100) || ++ (inodes_used != CDLIM_KEEP && ++ inodes_used > inodes_total) || ++ (space_used != CDLIM_KEEP && ++ space_used > space_total)) ++ goto out_release; ++ ++ ret = -ESRCH; ++ dli = locate_dl_info(sb, id); ++ if (!dli) ++ goto out_release; ++ ++ spin_lock(&dli->dl_lock); ++ ++ if (inodes_used != CDLIM_KEEP) ++ dli->dl_inodes_used = inodes_used; ++ if (inodes_total != CDLIM_KEEP) ++ dli->dl_inodes_total = inodes_total; ++ if (space_used != CDLIM_KEEP) ++ dli->dl_space_used = dlimit_space_32to64( ++ space_used, flags, DLIMS_USED); ++ ++ if (space_total == CDLIM_INFINITY) ++ dli->dl_space_total = DLIM_INFINITY; ++ else if (space_total != CDLIM_KEEP) ++ dli->dl_space_total = dlimit_space_32to64( ++ space_total, flags, DLIMS_TOTAL); ++ ++ if (reserved != CDLIM_KEEP) ++ dli->dl_nrlmult = (1 << 10) * (100 - reserved) / 100; ++ ++ spin_unlock(&dli->dl_lock); ++ ++ put_dl_info(dli); ++ ret = 0; ++ ++ out_release: ++ path_put(&path); ++ } ++ return ret; ++} ++ ++int vc_set_dlimit(uint32_t id, void __user *data) ++{ ++ struct vcmd_ctx_dlimit_v0 vc_data; ++ ++ if (copy_from_user(&vc_data, data, sizeof(vc_data))) ++ return -EFAULT; ++ ++ return do_set_dlimit(id, vc_data.name, ++ vc_data.space_used, vc_data.space_total, ++ vc_data.inodes_used, vc_data.inodes_total, ++ vc_data.reserved, vc_data.flags); ++} ++ ++#ifdef CONFIG_COMPAT ++ ++int vc_set_dlimit_x32(uint32_t id, void __user *data) ++{ ++ struct vcmd_ctx_dlimit_v0_x32 vc_data; ++ ++ if (copy_from_user(&vc_data, data, sizeof(vc_data))) ++ return -EFAULT; ++ ++ return do_set_dlimit(id, compat_ptr(vc_data.name_ptr), ++ vc_data.space_used, vc_data.space_total, ++ vc_data.inodes_used, vc_data.inodes_total, ++ vc_data.reserved, vc_data.flags); ++} ++ ++#endif /* CONFIG_COMPAT */ ++ ++ ++static inline ++int do_get_dlimit(uint32_t id, const char __user *name, ++ uint32_t *space_used, uint32_t *space_total, ++ uint32_t *inodes_used, uint32_t *inodes_total, ++ uint32_t *reserved, uint32_t *flags) ++{ ++ struct path path; ++ int ret; ++ ++ ret = user_lpath(name, &path); ++ if (!ret) { ++ struct super_block *sb; ++ struct dl_info *dli; ++ ++ ret = -EINVAL; ++ if (!path.dentry->d_inode) ++ goto out_release; ++ if (!(sb = path.dentry->d_inode->i_sb)) ++ goto out_release; ++ ++ ret = -ESRCH; ++ dli = locate_dl_info(sb, id); ++ if (!dli) ++ goto out_release; ++ ++ spin_lock(&dli->dl_lock); ++ *inodes_used = dli->dl_inodes_used; ++ *inodes_total = dli->dl_inodes_total; ++ ++ *space_used = dlimit_space_64to32( ++ dli->dl_space_used, flags, DLIMS_USED); ++ ++ if (dli->dl_space_total == DLIM_INFINITY) ++ *space_total = CDLIM_INFINITY; ++ else ++ *space_total = dlimit_space_64to32( ++ dli->dl_space_total, flags, DLIMS_TOTAL); ++ ++ *reserved = 100 - ((dli->dl_nrlmult * 100 + 512) >> 10); ++ spin_unlock(&dli->dl_lock); ++ ++ put_dl_info(dli); ++ ret = -EFAULT; ++ ++ ret = 0; ++ out_release: ++ path_put(&path); ++ } ++ return ret; ++} ++ ++ ++int vc_get_dlimit(uint32_t id, void __user *data) ++{ ++ struct vcmd_ctx_dlimit_v0 vc_data; ++ int ret; ++ ++ if (copy_from_user(&vc_data, data, sizeof(vc_data))) ++ return -EFAULT; ++ ++ ret = do_get_dlimit(id, vc_data.name, ++ &vc_data.space_used, &vc_data.space_total, ++ &vc_data.inodes_used, &vc_data.inodes_total, ++ &vc_data.reserved, &vc_data.flags); ++ if (ret) ++ return ret; ++ ++ if (copy_to_user(data, &vc_data, sizeof(vc_data))) ++ return -EFAULT; ++ return 0; ++} ++ ++#ifdef CONFIG_COMPAT ++ ++int vc_get_dlimit_x32(uint32_t id, void __user *data) ++{ ++ struct vcmd_ctx_dlimit_v0_x32 vc_data; ++ int ret; ++ ++ if (copy_from_user(&vc_data, data, sizeof(vc_data))) ++ return -EFAULT; ++ ++ ret = do_get_dlimit(id, compat_ptr(vc_data.name_ptr), ++ &vc_data.space_used, &vc_data.space_total, ++ &vc_data.inodes_used, &vc_data.inodes_total, ++ &vc_data.reserved, &vc_data.flags); ++ if (ret) ++ return ret; ++ ++ if (copy_to_user(data, &vc_data, sizeof(vc_data))) ++ return -EFAULT; ++ return 0; ++} ++ ++#endif /* CONFIG_COMPAT */ ++ ++ ++void vx_vsi_statfs(struct super_block *sb, struct kstatfs *buf) ++{ ++ struct dl_info *dli; ++ __u64 blimit, bfree, bavail; ++ __u32 ifree; ++ ++ dli = locate_dl_info(sb, dx_current_tag()); ++ if (!dli) ++ return; ++ ++ spin_lock(&dli->dl_lock); ++ if (dli->dl_inodes_total == (unsigned long)DLIM_INFINITY) ++ goto no_ilim; ++ ++ /* reduce max inodes available to limit */ ++ if (buf->f_files > dli->dl_inodes_total) ++ buf->f_files = dli->dl_inodes_total; ++ ++ ifree = dli->dl_inodes_total - dli->dl_inodes_used; ++ /* reduce free inodes to min */ ++ if (ifree < buf->f_ffree) ++ buf->f_ffree = ifree; ++ ++no_ilim: ++ if (dli->dl_space_total == DLIM_INFINITY) ++ goto no_blim; ++ ++ blimit = dli->dl_space_total >> sb->s_blocksize_bits; ++ ++ if (dli->dl_space_total < dli->dl_space_used) ++ bfree = 0; ++ else ++ bfree = (dli->dl_space_total - dli->dl_space_used) ++ >> sb->s_blocksize_bits; ++ ++ bavail = ((dli->dl_space_total >> 10) * dli->dl_nrlmult); ++ if (bavail < dli->dl_space_used) ++ bavail = 0; ++ else ++ bavail = (bavail - dli->dl_space_used) ++ >> sb->s_blocksize_bits; ++ ++ /* reduce max space available to limit */ ++ if (buf->f_blocks > blimit) ++ buf->f_blocks = blimit; ++ ++ /* reduce free space to min */ ++ if (bfree < buf->f_bfree) ++ buf->f_bfree = bfree; ++ ++ /* reduce avail space to min */ ++ if (bavail < buf->f_bavail) ++ buf->f_bavail = bavail; ++ ++no_blim: ++ spin_unlock(&dli->dl_lock); ++ put_dl_info(dli); ++ ++ return; ++} ++ ++#include <linux/module.h> ++ ++EXPORT_SYMBOL_GPL(locate_dl_info); ++EXPORT_SYMBOL_GPL(rcu_free_dl_info); ++ +diff -NurpP --minimal linux-3.0.9/kernel/vserver/helper.c linux-3.0.9-vs2.3.2.1/kernel/vserver/helper.c +--- linux-3.0.9/kernel/vserver/helper.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/kernel/vserver/helper.c 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,223 @@ ++/* ++ * linux/kernel/vserver/helper.c ++ * ++ * Virtual Context Support ++ * ++ * Copyright (C) 2004-2007 Herbert Pötzl ++ * ++ * V0.01 basic helper ++ * ++ */ ++ ++#include <linux/kmod.h> ++#include <linux/reboot.h> ++#include <linux/vs_context.h> ++#include <linux/vs_network.h> ++#include <linux/vserver/signal.h> ++ ++ ++char vshelper_path[255] = "/sbin/vshelper"; ++ ++ ++static int do_vshelper(char *name, char *argv[], char *envp[], int sync) ++{ ++ int ret; ++ ++ if ((ret = call_usermodehelper(name, argv, envp, sync))) { ++ printk( KERN_WARNING ++ "%s: (%s %s) returned %s with %d\n", ++ name, argv[1], argv[2], ++ sync ? "sync" : "async", ret); ++ } ++ vxdprintk(VXD_CBIT(switch, 4), ++ "%s: (%s %s) returned %s with %d", ++ name, argv[1], argv[2], sync ? "sync" : "async", ret); ++ return ret; ++} ++ ++/* ++ * vshelper path is set via /proc/sys ++ * invoked by vserver sys_reboot(), with ++ * the following arguments ++ * ++ * argv [0] = vshelper_path; ++ * argv [1] = action: "restart", "halt", "poweroff", ... ++ * argv [2] = context identifier ++ * ++ * envp [*] = type-specific parameters ++ */ ++ ++long vs_reboot_helper(struct vx_info *vxi, int cmd, void __user *arg) ++{ ++ char id_buf[8], cmd_buf[16]; ++ char uid_buf[16], pid_buf[16]; ++ int ret; ++ ++ char *argv[] = {vshelper_path, NULL, id_buf, 0}; ++ char *envp[] = {"HOME=/", "TERM=linux", ++ "PATH=/sbin:/usr/sbin:/bin:/usr/bin", ++ uid_buf, pid_buf, cmd_buf, 0}; ++ ++ if (vx_info_state(vxi, VXS_HELPER)) ++ return -EAGAIN; ++ vxi->vx_state |= VXS_HELPER; ++ ++ snprintf(id_buf, sizeof(id_buf)-1, "%d", vxi->vx_id); ++ ++ snprintf(cmd_buf, sizeof(cmd_buf)-1, "VS_CMD=%08x", cmd); ++ snprintf(uid_buf, sizeof(uid_buf)-1, "VS_UID=%d", current_uid()); ++ snprintf(pid_buf, sizeof(pid_buf)-1, "VS_PID=%d", current->pid); ++ ++ switch (cmd) { ++ case LINUX_REBOOT_CMD_RESTART: ++ argv[1] = "restart"; ++ break; ++ ++ case LINUX_REBOOT_CMD_HALT: ++ argv[1] = "halt"; ++ break; ++ ++ case LINUX_REBOOT_CMD_POWER_OFF: ++ argv[1] = "poweroff"; ++ break; ++ ++ case LINUX_REBOOT_CMD_SW_SUSPEND: ++ argv[1] = "swsusp"; ++ break; ++ ++ case LINUX_REBOOT_CMD_OOM: ++ argv[1] = "oom"; ++ break; ++ ++ default: ++ vxi->vx_state &= ~VXS_HELPER; ++ return 0; ++ } ++ ++ ret = do_vshelper(vshelper_path, argv, envp, 0); ++ vxi->vx_state &= ~VXS_HELPER; ++ __wakeup_vx_info(vxi); ++ return (ret) ? -EPERM : 0; ++} ++ ++ ++long vs_reboot(unsigned int cmd, void __user *arg) ++{ ++ struct vx_info *vxi = current_vx_info(); ++ long ret = 0; ++ ++ vxdprintk(VXD_CBIT(misc, 5), ++ "vs_reboot(%p[#%d],%u)", ++ vxi, vxi ? vxi->vx_id : 0, cmd); ++ ++ ret = vs_reboot_helper(vxi, cmd, arg); ++ if (ret) ++ return ret; ++ ++ vxi->reboot_cmd = cmd; ++ if (vx_info_flags(vxi, VXF_REBOOT_KILL, 0)) { ++ switch (cmd) { ++ case LINUX_REBOOT_CMD_RESTART: ++ case LINUX_REBOOT_CMD_HALT: ++ case LINUX_REBOOT_CMD_POWER_OFF: ++ vx_info_kill(vxi, 0, SIGKILL); ++ vx_info_kill(vxi, 1, SIGKILL); ++ default: ++ break; ++ } ++ } ++ return 0; ++} ++ ++long vs_oom_action(unsigned int cmd) ++{ ++ struct vx_info *vxi = current_vx_info(); ++ long ret = 0; ++ ++ vxdprintk(VXD_CBIT(misc, 5), ++ "vs_oom_action(%p[#%d],%u)", ++ vxi, vxi ? vxi->vx_id : 0, cmd); ++ ++ ret = vs_reboot_helper(vxi, cmd, NULL); ++ if (ret) ++ return ret; ++ ++ vxi->reboot_cmd = cmd; ++ if (vx_info_flags(vxi, VXF_REBOOT_KILL, 0)) { ++ vx_info_kill(vxi, 0, SIGKILL); ++ vx_info_kill(vxi, 1, SIGKILL); ++ } ++ return 0; ++} ++ ++/* ++ * argv [0] = vshelper_path; ++ * argv [1] = action: "startup", "shutdown" ++ * argv [2] = context identifier ++ * ++ * envp [*] = type-specific parameters ++ */ ++ ++long vs_state_change(struct vx_info *vxi, unsigned int cmd) ++{ ++ char id_buf[8], cmd_buf[16]; ++ char *argv[] = {vshelper_path, NULL, id_buf, 0}; ++ char *envp[] = {"HOME=/", "TERM=linux", ++ "PATH=/sbin:/usr/sbin:/bin:/usr/bin", cmd_buf, 0}; ++ ++ if (!vx_info_flags(vxi, VXF_SC_HELPER, 0)) ++ return 0; ++ ++ snprintf(id_buf, sizeof(id_buf)-1, "%d", vxi->vx_id); ++ snprintf(cmd_buf, sizeof(cmd_buf)-1, "VS_CMD=%08x", cmd); ++ ++ switch (cmd) { ++ case VSC_STARTUP: ++ argv[1] = "startup"; ++ break; ++ case VSC_SHUTDOWN: ++ argv[1] = "shutdown"; ++ break; ++ default: ++ return 0; ++ } ++ ++ return do_vshelper(vshelper_path, argv, envp, 1); ++} ++ ++ ++/* ++ * argv [0] = vshelper_path; ++ * argv [1] = action: "netup", "netdown" ++ * argv [2] = context identifier ++ * ++ * envp [*] = type-specific parameters ++ */ ++ ++long vs_net_change(struct nx_info *nxi, unsigned int cmd) ++{ ++ char id_buf[8], cmd_buf[16]; ++ char *argv[] = {vshelper_path, NULL, id_buf, 0}; ++ char *envp[] = {"HOME=/", "TERM=linux", ++ "PATH=/sbin:/usr/sbin:/bin:/usr/bin", cmd_buf, 0}; ++ ++ if (!nx_info_flags(nxi, NXF_SC_HELPER, 0)) ++ return 0; ++ ++ snprintf(id_buf, sizeof(id_buf)-1, "%d", nxi->nx_id); ++ snprintf(cmd_buf, sizeof(cmd_buf)-1, "VS_CMD=%08x", cmd); ++ ++ switch (cmd) { ++ case VSC_NETUP: ++ argv[1] = "netup"; ++ break; ++ case VSC_NETDOWN: ++ argv[1] = "netdown"; ++ break; ++ default: ++ return 0; ++ } ++ ++ return do_vshelper(vshelper_path, argv, envp, 1); ++} ++ +diff -NurpP --minimal linux-3.0.9/kernel/vserver/history.c linux-3.0.9-vs2.3.2.1/kernel/vserver/history.c +--- linux-3.0.9/kernel/vserver/history.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.0.9-vs2.3.2.1/kernel/vserver/history.c 2011-06-10 22:11:24.000000000 +0200 +@@ -0,0 +1,258 @@ ++/* ++ * kernel/vserver/history.c ++ * ++ * Virtual Context History Backtrace ++ * ++ * Copyright (C) 2004-2007 Herbert Pötzl ++ * ++ * V0.01 basic structure ++ * V0.02 hash/unhash and trace ++ * V0.03 preemption fixes ++ * ++ */ ++ ++#include <linux/module.h> ++#include <asm/uaccess.h> ++ ++#include <linux/vserver/context.h> ++#include <linux/vserver/debug.h> ++#include <linux/vserver/debug_cmd.h> ++#include <linux/vserver/history.h> ++ ++ ++#ifdef CONFIG_VSERVER_HISTORY ++#define VXH_SIZE CONFIG_VSERVER_HISTORY_SIZE ++#else ++#define VXH_SIZE 64 ++#endif ++ ++struct _vx_history { ++ unsigned int counter; ++ ++ struct _vx_hist_entry entry[VXH_SIZE + 1]; ++}; ++ ++ ++DEFINE_PER_CPU(struct _vx_history, vx_history_buffer); ++ ++unsigned volatile int vxh_active = 1; ++ ++static atomic_t sequence = ATOMIC_INIT(0); ++ ++ ++/* vxh_advance() ++ ++ * requires disabled preemption */ ++ ++struct _vx_hist_entry *vxh_advance(void *loc) ++{ ++ unsigned int cpu = smp_processor_id(); ++ struct _vx_history *hist = &per_cpu(vx_history_buffer, cpu); ++ struct _vx_hist_entry *entry; ++ unsigned int index; ++ ++ index = vxh_active ? (hist->counter++ % VXH_SIZE) : VXH_SIZE; ++ entry = &hist->entry[index]; ++ ++ entry->seq = atomic_inc_return(&sequence); ++ entry->loc = loc; ++ return entry; ++} ++ ++EXPORT_SYMBOL_GPL(vxh_advance); ++ ++ ++#define VXH_LOC_FMTS "(#%04x,*%d):%p" ++ ++#define VXH_LOC_ARGS(e) (e)->seq, cpu, (e)->loc ++ ++ ++#define VXH_VXI_FMTS "%p[#%d,%d.%d]" ++ ++#define VXH_VXI_ARGS(e) (e)->vxi.ptr, \ ++ (e)->vxi.ptr ? (e)->vxi.xid : 0, \ ++ (e)->vxi.ptr ? (e)->vxi.usecnt : 0, \ ++ (e)->vxi.ptr ? (e)->vxi.tasks : 0 ++ ++void vxh_dump_entry(struct _vx_hist_entry *e, unsigned cpu) ++{ ++ switch (e->type) { ++ case VXH_THROW_OOPS: ++ printk( VXH_LOC_FMTS " oops \n", VXH_LOC_ARGS(e)); ++ break; ++ ++ case VXH_GET_VX_INFO: ++ case VXH_PUT_VX_INFO: ++ printk( VXH_LOC_FMTS " %s_vx_info " VXH_VXI_FMTS "\n", ++ VXH_LOC_ARGS(e), ++ (e->type == VXH_GET_VX_INFO) ? "get" : "put", ++ VXH_VXI_ARGS(e)); ++ break; ++ ++ case VXH_INIT_VX_INFO: ++ case VXH_SET_VX_INFO: ++ case VXH_CLR_VX_INFO: ++ printk( VXH_LOC_FMTS " %s_vx_info " VXH_VXI_FMTS " @%p\n", ++ VXH_LOC_ARGS(e), ++ (e->type == VXH_INIT_VX_INFO) ? "init" : ++ ((e->type == VXH_SET_VX_INFO) ? "set" : "clr"), ++ VXH_VXI_ARGS(e), e->sc.data); ++ break; ++ ++ case VXH_CLAIM_VX_INFO: ++ case VXH_RELEASE_VX_INFO: ++ printk( VXH_LOC_FMTS " %s_vx_info " VXH_VXI_FMTS " @%p\n", ++ VXH_LOC_ARGS(e), ++ (e->type == VXH_CLAIM_VX_INFO) ? "claim" : "release", ++ VXH_VXI_ARGS(e), e->sc.data); ++ break; ++ ++ case VXH_ALLOC_VX_INFO: ++ case VXH_DEALLOC_VX_INFO: ++ printk( VXH_LOC_FMTS " %s_vx_info " VXH_VXI_FMTS "\n", ++ VXH_LOC_ARGS(e), ++ (e->type == VXH_ALLOC_VX_INFO) ? "alloc" : "dealloc", ++ VXH_VXI_ARGS(e)); ++ break; ++ ++ case VXH_HASH_VX_INFO: ++ case VXH_UNHASH_VX_INFO: ++ printk( VXH_LOC_FMTS " __%s_vx_info " VXH_VXI_FMTS "\n", ++ VXH_LOC_ARGS(e), ++ (e->type == VXH_HASH_VX_INFO) ? "hash" : "unhash", ++ VXH_VXI_ARGS(e)); ++ break; ++ ++ case VXH_LOC_VX_INFO: ++ case VXH_LOOKUP_VX_INFO: ++ case VXH_CREATE_VX_INFO: ++ printk( VXH_LOC_FMTS " __%s_vx_info [#%d] -> " VXH_VXI_FMTS "\n", ++ VXH_LOC_ARGS(e), ++ (e->type == VXH_CREATE_VX_INFO) ? "create" : ++ ((e->type == VXH_LOC_VX_INFO) ? "loc" : "lookup"), ++ e->ll.arg, VXH_VXI_ARGS(e)); ++ break; ++ } ++} ++ ++static void __vxh_dump_history(void) ++{ ++ unsigned int i, cpu; ++ ++ printk("History:\tSEQ: %8x\tNR_CPUS: %d\n", ++ atomic_read(&sequence), NR_CPUS); ++ ++ for (i = 0; i < VXH_SIZE; i++) { ++ for_each_online_cpu(cpu) { ++ struct _vx_history *hist = ++ &per_cpu(vx_history_buffer, cpu); ++ unsigned int index = (hist->counter - i) % VXH_SIZE; ++ struct _vx_hist_entry *entry = &hist->entry[index]; ++ ++ vxh_dump_entry(entry, cpu); ++ } ++ } ++} ++ ++void vxh_dump_history(void) ++{ ++ vxh_active = 0; ++#ifdef CONFIG_SMP ++ local_irq_enable(); ++ smp_send_stop(); ++ local_irq_disable(); ++#endif ++ __vxh_dump_history(); ++} ++ ++ ++/* vserver syscall commands below here */ ++ ++ ++int vc_dump_history(uint32_t id) ++{ ++ vxh_active = 0; ++ __vxh_dump_history(); ++ vxh_active = 1; ++ ++ return 0; ++} ++ ++ ++int do_read_history(struct __user _vx_hist_entry *data, ++ int cpu, uint32_t *index, uint32_t *count) ++{ ++ int pos, ret = 0; ++ struct _vx_history *hist = &per_cpu(vx_history_buffer, cpu); ++ int end = hist->counter; ++ int start = end - VXH_SIZE + 2; ++ int idx = *index; ++ ++ /* special case: get current pos */ ++ if (!*count) { ++ *index = end; ++ return 0; ++ } ++ ++ /* have we lost some data? */ ++ if (idx < start) ++ idx = start; ++ ++ for (pos = 0; (pos < *count) && (idx < end); pos++, idx++) { ++ struct _vx_hist_entry *entry = ++ &hist->entry[idx % VXH_SIZE]; ++ ++ /* send entry to userspace */ ++ ret = copy_to_user(&data[pos], entry, sizeof(*entry)); ++ if (ret) ++ break; ++ } ++ /* save new index and count */ ++ *index = idx; ++ *count = pos; ++ return ret ? ret : (*index < end); ++} ++ ++int vc_read_history(uint32_t id, void __user *data) ++{ ++ struct vcmd_read_history_v0 vc_data; ++ int ret; ++ ++ if (id >= NR_CPUS) ++ return -EINVAL; ++ ++ if (copy_from_user(&vc_data, data, sizeof(vc_data))) ++ return -EFAULT; ++ ++ ret = do_read_history((struct __user _vx_hist_entry *)vc_data.data, ++ id, &vc_data.index, &vc_data.count); ++ ++ if (copy_to_user(data, &vc_data, sizeof(vc_data))) ++ return -EFAULT; ++ return ret; ++} ++ ++#ifdef CONFIG_COMPAT ++ ++int vc_read_history_x32(uint32_t id, void __user *data) ++{ ++ struct vcmd_read_history_v0_x32 vc_data; ++ int ret; ++ ++ if (id >= NR_CPUS) ++ return -EINVAL; ++ ++ if (copy_from_user(&vc_data, data, sizeof(vc_data))) ++ return -EFAULT; ++ ++ ret = do_read_history((struct __user _vx_hist_entry *) ++ compat_ptr(vc_data.data_ptr), ++ i |