[PATCH] Exit from OOMed process ASAP rather then dig in try_to_free_pages(). it can spend really lots of time in try_to_free_pages() (up to minutes), while process selected for OOM should die ASAP.Bug #79344
[PATCH] grsecurity: TPE feature This patch is a part of big grsecure-2.1.10 patch.This patch doesn't contain virtualization support,
and source code looks like original grsecure source as much as possible.
V2 - added TPE check in mmap()/mprotect()
V3 - removed exec_file; added grsec_lock feature
[PATCH] fix bust_spinlocks() race when doing Alt-SysRq-P via NMI IPI Fix bust_spinlocks() race when doing Alt-SysRq-P via NMI IPI smp_show_regs() function calls bust_spinlocks() which is not protected by any lock. Just call bust_spinlocks() under smp_show_regs lock.Bug #92669
[PATCH] hide global /proc/asound entry from VEcreate_proc_entry() in sound/core/info.c is called with
gobal parent == &proc_root, thus /proc/asound is global.
Make it VE0 local.
Bug #92723
[PATCH] mangle devname in /proc/mounts only for OVZ specific simfs We used to hide devname in /proc/mounts to prevent leaking HN-node information to the VE. Actually only VZ specifics should be hidden, i.e. simfs root. All other information can be shown as is.And the fix is actually required for NFS. its umount scripts
goes crazy when can't read real NFS server mount information.
Bug #87637
[CPT] changes to core required to checkpoint sysv msg queues1. "Standard" set, like SHM and SEM: sysvipc_walk_msg, sysvipc_setup_msg,
additional arqument to newque() to pass ipc id.
2. Two new exports: sysv_msg_store, sysv_msg_load.
The structure is enough weird to deal with it inside CPT.
[PATCH] virtualization of sysv msg queues was incomplete Two variables visible from userspace were global, let's make them per-namespace.Signed-Off-By: Alexey Kuznetsov <alexey@openvz.org>
Signed-Off-By: Kirill Korotaev <dev@openvz.org>
[PATCH] don't mmap zero pages for /dev/zero in advance In current implementation mmap_zero() immedeately mmaps zero page to the whole /dev/zero vma and thus allocates/pollutes all pte's. This is actually not required and can be dangerous if vma is very big.For CPT we do not need such "optimization",
as we can set vm_ops on this vma and return zero page on demand.
Bug #91769
[PATCH] fix dump_trace() when called from NMI For x86_64 architecture NMI interrupt handler (do_NMI) uses exception stack. When show_regs() is called from NMI handler then pt_regs is in the bottom of exception stack (unlike to do_IRQ). show_regs() passes pointer of the traced stack to the show_trace() function which points to the bottom of the exception stack. Therefore problem occurs when try...This patch adjusts pointer of the traced stack so as it doesn't point to the
bottom of the exception stack.
Bugs #91201
[PATCH] CPT: fix restore of process start timeDue to stupid misprint processes start_time were restored incorrectly.
When we restore start_time of process we should sum up value from image with
VE start time. During checkpoint we save to the image relative process
start_time (process start_time - VE start_time).
[PATCH] BC: minor compiler warning fix in case of CONFIG_UBC_DEBUG_KMEM=nfix for:
kernel/ub/ub_mem.c: In function 'ub_page_charge':
kernel/ub/ub_mem.c:385: warning: statement with no effect
[PATCH] another cond_resched() fix cond_resched() should check that it is not nexted via preempt_count() & PREEMPT_ACTIVE flag. Drop the whole super-logic from Den checking preempt count at al.Bug #92140
[PATCH] fix for cond_resched() fix: remove wrong WARN_ON(1) WARN_ON(1) is illegal, since when we return from cond_resched()->schedule() we have preempt_count = PREEMPT_ACTIVE, and if we have still current->need_resched flag set we can get to cond_resched() again from schedule()->reacquire_kernel_lock()->cond_resched() and thus this WARN_ON(1) gets triggered.Bug #92140
[PATCH] Fix needless SysRq help message Every time one does echo p >/proc/sysrq-trigger
newline sneaks into kernel buffer, sysrq code doesn't find it in
handlers table, and spits help banner.
[PATCH] x86_64: Zero extend all registers after ptrace in 32bit entry path.Strictly it's only needed for eax.
It actually does a little more than strictly needed -- the other registers
are already zero extended.
Also remove the now unnecessary and non functional compat task check
in ptrace.
This is CVE-2007-4573
Found by Wojciech Purczynski
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
X-Git-Url: http://git....
[FS]: disable O_DIRECT by default inside VEWe still have to disable O_DIRECT by default inside VE due to:
Compatibility with old broken software (e.g. rpm)
Bug #91550
[PATCH] fairsched: fix warning on preempt kernels rq->curr should be initialized to something to avoid its dereference e.g. in try_to_wakeup() on first process wakeup.Actually it doesn't matter to what it is initialized.
Let's use init_task for initial rq->curr.
[PATCH] cond_resched_lock() doesn't work in 2.6.18 When CONFIG_PREEMPT=n, cond_resched_lock() and cond_resched_softirq() don't work, since they check for preempt_count to be sane, but this counter is not tracked w/o preemption and is always 0.So the fix is:
- ignore preempt count when CONFIG_PREEMPT=n
- plus if we want to check preempt_count in CONFIG_PREEMPT=y case
(just to be on the safe side), we need to account lock_kernel()
effect on preempt_count correctly.
Bug #91012
[PATCH] BC: account fib rules to kmemsizeThis patch adds UB_KMEMSIZE accounting for:
- routing (fib) rules
- routing entries
- inet & inet6 addresses
All these objects are created by userspace and leaves in kernel memory.
[PATCH] VZDQ: report busy dentries on vzquota onIn case of vzquota ON fail, find and pass to the userspace
information about busy dentries.
vzquota must pass user buffer, and kernel fill it with found filenames.
[PATCH] VZDQ: report busy dentries on quota off In case of vzquota off fail, find and pass to the userspace information about busy dentries. vzquota must pass PAGE_SIZE buffer, and kernel fill it with found filenames.V2:
- use free_page() instead of kfree()
- remove unnecessary \n after the last file name
V3:
- use generic __d_path()
- don't call copy_to_user() if both ubuf and buf == NULL
V4:
- remove VZ_DQ_OFF_FORCED declaration from header, it relates to another
patch
Bug #86944
[PATCH] CPT: fix LDT pages leak with 4GB split In case 4GB split kernel is used, CPT can leak some of LDT pages - it allocates pages first, but doesn't set context->size, thus destroy_context() won't try to free these additional LDT pages.Relevant for -ent kernel flavors only.
[NET]: Do not dereference iov if length is zeroWhen msg_iovlen is zero we shouldn't try to dereference
msg_iov. Right now the only thing that tries to do so
is skb_copy_and_csum_datagram_iovec. Since the total
length should also be zero if msg_iovlen is zero, it's
sufficient to check the total length there and simply
return if it's zero.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemlo...
[PATCH] BC: set correct ub context in netlink processing - rtnl netlink socket is asynchronous and can be processed during rtnl_unlock in the other context. - rtnl netlink socket is used to create kernel objects - these objects are planned to be accounted at least to UB_KMEMSIZESo, let's set correct UB context for packets processing.
[PATCH] net: allow SIOCSIFFLAGS in dev_ioctl() This patch allows ioctl SIOCSIFFLAGS from VE on PF_INET6 socket.In old VEs (with redhat-6.2) ifconfig calls ioctl() on IPv6 socket,
while tries to make venet UP inside VE. And fails.
Since this ioctl is prohibited in dev_ioctl(). Newer ifconfig's
call this ioctl() on IPv4 socket and thus end up with inet_ioctl().
Bug #91248
[PATCH][REVERT] proc: don't hash task dentries in VE0 Patch diff-ve-proc-hash-pid-dentries-20070516 is obsoleted by previous patch commit: diff-ubc-pid-uncharge-early-20070717Revert it.
[PATCH] uncharge pids from kmemsize a bit earlier When charged with generic slab charger pids hold the beancounter and may pin it to memory after VE is stopped. The proposal is to charge/uncharge pids manually letting them live longer than the beancounter. This is OK, since the only reason (in rhel5 kernel) they can do so is that proc inodes hold them. However, this is not a big deal, as inodes...Potential fix for bug #83891
BTW, patch named diff-ve-proc-hash-pid-dentries-20070516 may be
dropped from the list.
[IPV6]: /proc/net/anycast6 unbalanced inet6_dev refcntReading /proc/net/anycast6 when there is no anycast address
on an interface results in an ever-increasing inet6_dev reference
count, as well as a reference to the netdevice you can't get rid of.
Signed-off-by: David S. Miller <davem@davemloft.net>
Bug #75822
X-Git-Tag: v2.6.21-rc3~1492~5
X-Git-Url:
http://git.kernel.org/?p=linux%2Fkernel%2Fgit%2Ftorvalds%2Flinux-2.6.git;a=commitdiff_plain;h=a...
[PATCH] improve shrink_dcache_sb() This patch makes shrink_dcache_sb consistent with dentry pruning policy.On the first pass we iterate over dentry unused list and prepare some
dentries for removal.
However, since the existing code moves evicted dentries
to the beginning of the LRU it can happen that fresh dentries from
other superblocks will be inserted *before* our dentries.
This can result in significant slowdown of shrink_dcache_sb().
Moreover, for virtual filesystems like unionfs which can cal...
[PATCH] CPT: lock fd close correct error handlingLock fd close correct error handling:
on undump local variable 'err' was used to store an error,
that is why this error was ignored and undump continues
in spite of the error.
[ATM] he: Fix __init/__devinit conflicthe_init_one() is declared __devinit, but calls lots of init functions
that are marked __init. However, if CONFIG_HOTPLUG is enabled,
__devinit functions go into normal .text, which leads to
WARNING: drivers/atm/he.o - Section mismatch: reference to .init.text: from .text between 'he_start' (at offset 0x2130) and 'he_service_tbrq'
Fix this by changing the __init functions to __devinit.
S...
[PATCH] CPT: core changes to allow checkpointing inotify interface1. Cosmetic changes to export nessecary structs and variables
(time to start to think about moving those cpt bits to core to avoid
pollution. cpt interface could be ugly, but it is _one_ interface)
2. There is one essential change: each watchpoint refers to dentry/mnt.
The drawback: it means fs umount is blocked while some directory
on it is watched. With current uses of inotify it...
[PATCH] vzevent: fix netlink channel numberFix vzevent module. It is incompatible with kobject uevents in reallity.
Current code does *nothing*, since kobj is not fully configured.
Instead let's send messages via a separate netlink channel.
OpenVZ team
