Difference between revisions of "Download/kernel/2.6.8/022stab034.1/changes"

From OpenVZ Virtuozzo Containers Wiki
Jump to: navigation, search
m (Protected "Download/kernel/2.6.8/022stab034.1/changes": Robot: Protecting a list of files. [edit=autoconfirmed:move=autoconfirmed])
(No difference)

Latest revision as of 18:23, 22 October 2009


  • Mainstream kernel security fixes
  • Fairscheduler debug code and fixes
  • Statistics fixes


  • Kernel headers resides under /lib/modules/`$kernel_ver`/build/include directory now. Package vzkernel-headers is removed.



Patch from Denis, based on idea from Dmitry:

This patch removes permission granting for unnamed devices, registered during VPS creation.


Patch from Pavel:

Relocates fairsched_init_late into proper place.

Bug 50457.


Patch from Pavel, edited by Andrey:
  • fixup of nlink counter and mode for proc inodes;
  • small code cleanup.


Patch from Andrey:

Fixes oops in VPS fini hook in vzdq_file if the VPS doesn't have disk quota.
Bug 50161.


Patch from Alexey, modified by Pavel:
  • dgets dentry instead of incrementing its d_count
  • mntgets vfsmount instead of incrementing its mnt_count (at Andrey's request)


Patch from Andrey, rewritten version of a patch by Pavel:

Comment about barrier properties of op_sem.


Patch from Denis:

This patch fixes race between do_env_cleanup (removes VPS from list, vzctl finishes, VPS start allowed) and subsequent do_env_create->fairsched_mknod, where the fairsched node still exist.


Patch from Andrey:

Adds error checks in fairsched node creation/switch/removal at VPS start/stop. Eliminates one of the reasons of vcpu_attach oops.


Patch from Denis:

This patch fixes oops in free_ve_proc called on cleanup path in do_env_create.
Bug 50322.


Patch from Pavel:
  • no need to set exec environment in vzlmond before calling env_cleanup, since env_cleanup sets it itself;
  • make code look nicer with list_first_entry.


Patch from Pavel:

misprint in migrate task. Wrong macro was used, but it happened to be harmless since both macros expand into the same expression.


Patch from Pavel:

read ve idle time in proc and vecalls.

(Andrey's comments, renamed get_ve_idle_cpu into ve_sched_get_idle_time)


Patch from Pavel:

Calculates ve's per-vcpu iowait time


Patch from Pavel:
  • account ve idle time per-vcpu;
  • removed fairsched idle time accounting.

Makes comparison between cycles like in time_after.


Patch from Pavel:
  • fixup of wall_to_monotonic access in show_stat();
  • optimisation and code cleanup in show_stat().


Patch from Pavel:

Introduces nr_running_ve, nr_iowait_ve and ve_uninterruptible_ve that are used in /proc to show statistics in VPS. Fixups these functions to check returned result for "< 0".


Patch from Pavel:
  • global accounting for running/sleeping/etc tasks;
  • removes cycles from ve_nr_sunning_inc/_dec;
  • ve's nr_running etc accounting is per Vcpu;
  • fixed type of sym in nr_sleeping and nr_stopped;
  • removed inc/dec of ve_nr_running from pull_task.


Patch from Andrey:

This patch fixes/improves printk() locking thus making it possible to call it from NMI/scheduler code using special printk_no_wake variable.


Patch from Andrey:

Improvements of printk from schedule/NMI:

  • per-cpu variables can't be used (they are initialized too late),
  • operations in nmi_watchdog_tick() reordered a bit,
  • check for printk_no_wake added to bust_spinlocks,
  • a timer to wake up processes after missing wakeup from printk implemented.


Patch from Andrey:

Debug print in case of loop in schedule_vcpu. no new issues


Patch from Andrey:

Additional fairscheduler lookup debug.


Patch from Denis:

This patch does the following:

  • incorrect VCPU id in deactivate_task (in __migrate_task)
  • too small counter in schedule_vcpu deadlock detection
  • FSN->pcpu decreased if schedule_vcpu deadlock detection fired
  • incorrect spin_lock_irq in __show_vsched cleared
  • rq_vcpu macro introduced

Bug 50622.
Bug 50635.


Patch from Kirill:

This patches cleanups fairsched code:

  • set_task_cpu replaced by set_task_vcpu in many places
  • additional BUG_ON added to __migrate_task
  • fixed BUG in migration_thread: vcpu_offline should be checked
  • reorganized vsched_del_vcpu to make it more obvious
  • irqsave replaced by irq where save not needed


Patch from mainstream:

[PATCH] binfmt_elf: clearing bss may fail

So we discover that Borland's Kylix application builder emits weird elf files which describe a non-writeable bss segment. So remove the clear_user() check at the place where we zero out the bss. I don't _think_ there are any security implications here (plus we've never checked that clear_user() return value, so whoops if it is a problem).

Signed-off-by: Pavel Machek <pavel@suse.cz>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

OpenVz Bug 34


Patch from Pavel:

This patch fixes vm_rss accounting in vma_merge.
Bug 50328.


Patch from Pavel:

Corrected error path on pbc allocation failure in do_swap_page.
Bug 50329.


Patch from Pavel:

Fix of memory leak in __add_vcpu on error path: not installed vcpu must be kfree-ed


Patch from Pavel:

Fixup of compilation of alloc_ve_tty_drivers in x86_64. When size was int gcc thought it was __builtin_constant and compiled symbol __you_cannot_kmalloc_that_much into module.


Patch from Pavel:

Fix include/linux/vsyscall.h to be includable from user space (for vzctl)


Patch from Andrey Mirkin:

Added VPS info in show_registers on x86_64 arch.


Patch from Vasily:

fixed missed context restoring in xprt_create_socket()


Patch from Denis:

This patch fixes quota.h to make it includable from user space.


Patch from Alexey:

Virtuozzo kernel does not compile without initrd. Reason: previous patch for GRUB + initrd.


Patch from mainstream:

[PATCH] Fix DST leak in icmp_push_reply()

Based upon a bug report and initial patch by Ollie Wild:
>>> If the ip_append_data() call in icmp_push_reply() fails,
>>> ip_flush_pending_frames() needs to be called. Otherwise, ip_rt_put()
>>> is never called on inet_sk(icmp_socket->sk)->cork.rt, which prevents
>>> the route (and net_device) from ever being freed.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: "David S. Miller" <davem@davemloft.net>
Signed-off-by: Chris Wright <chrisw@osdl.org>

X-Git-Tag: v2.6.12.6


Patch from mainstream:

[PATCH] Restrict socket policy loading to CAP_NET_ADMIN - CAN-2005-2555

The interface needs much redesigning if we wish to allow normal users to do this in some way.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: "David S. Miller" <davem@davemloft.net>
Signed-off-by: Chris Wright <chrisw@osdl.org>

X-Git-Tag: v2.6.12.6


Patch from mainstream:

[PATCH] Check input buffer size in zisofs

Add fakey 'deflateBound()' function to the in-kernel zlib routines It's not the real deflateBound() in newer zlib libraries, partly because the upcoming usage of it won't have the "stream" available, so we can't have the same interfaces anyway.

This uses the new deflateBound() thing to sanity-check the input to the zlib decompressor before we even bother to start reading in the blocks.


Problem noted by Tim Yamin <plasmaroo@gentoo.org>
Signed-off-by: Chris Wright <chrisw@osdl.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>

X-Git-Tag: v2.6.12.5


Patch from mainstream:

[PATCH] Update in-kernel zlib routines (CAN-2005-2458, CAN-2005-2459)

Fix outstanding security bugs in the Linux zlib implementations. See:

Signed-off-by: Tim Yamin <plasmaroo@gentoo.org>
Signed-off-by: Tavis Ormandy <taviso@gentoo.org>
Signed-off-by: Chris Wright <chrisw@osdl.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

X-Git-Tag: v2.6.12.5


Patch from mainstream:

[PATCH] autofs4: fix infamous "Busy inodes after umount ..." message

If the automount daemon receives a signal which causes it to sumarily terminate the autofs4 module leaks dentries. The same problem exists with detached mount requests without the warning.

This patch cleans these dentries at umount.

Signed-off-by: Ian Kent <raven@themaw.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

Bug 50326.