Skip to content

cloud-kernel
cloud-kernel

openanolis / cloud-kernel
大约 1 年 前同步成功

通知 158
Star 36
Fork 7
cloud-kernel
cloud-kernel

体验新版 GitCode，发现更多精彩内容 >>

切换分支/标签
  1. 04 1月, 2012 26 次提交
  3. 26 12月, 2011 1 次提交
    • J
      KVM: Don't automatically expose the TSC deadline timer in cpuid · 4d25a066
      由 Jan Kiszka 提交于 
      Unlike all of the other cpuid bits, the TSC deadline timer bit is set
unconditionally, regardless of what userspace wants.

This is broken in several ways:
 - if userspace doesn't use KVM_CREATE_IRQCHIP, and doesn't emulate the TSC
   deadline timer feature, a guest that uses the feature will break
 - live migration to older host kernels that don't support the TSC deadline
   timer will cause the feature to be pulled from under the guest's feet;
   breaking it
 - guests that are broken wrt the feature will fail.

Fix by not enabling the feature automatically; instead report it to userspace.
Because the feature depends on KVM_CREATE_IRQCHIP, which we cannot guarantee
will be called, we expose it via a KVM_CAP_TSC_DEADLINE_TIMER and not
KVM_GET_SUPPORTED_CPUID.

Fixes the Illumos guest kernel, which uses the TSC deadline timer feature.

[avi: add the KVM_CAP + documentation]
Reported-by: NAlexey Zaytsev <alexey.zaytsev@gmail.com>
Tested-by: NAlexey Zaytsev <alexey.zaytsev@gmail.com>
Signed-off-by: NJan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: NAvi Kivity <avi@redhat.com>
      4d25a066
  5. 22 12月, 2011 1 次提交
    • S
      VFS: Fix race between CPU hotplug and lglocks · e30e2fdf
      由 Srivatsa S. Bhat 提交于 
      Currently, the *_global_[un]lock_online() routines are not at all synchronized
with CPU hotplug. Soft-lockups detected as a consequence of this race was
reported earlier at https://lkml.org/lkml/2011/8/24/185. (Thanks to Cong Meng
for finding out that the root-cause of this issue is the race condition
between br_write_[un]lock() and CPU hotplug, which results in the lock states
getting messed up).

Fixing this race by just adding {get,put}_online_cpus() at appropriate places
in *_global_[un]lock_online() is not a good option, because, then suddenly
br_write_[un]lock() would become blocking, whereas they have been kept as
non-blocking all this time, and we would want to keep them that way.

So, overall, we want to ensure 3 things:
1. br_write_lock() and br_write_unlock() must remain as non-blocking.
2. The corresponding lock and unlock of the per-cpu spinlocks must not happen
   for different sets of CPUs.
3. Either prevent any new CPU online operation in between this lock-unlock, or
   ensure that the newly onlined CPU does not proceed with its corresponding
   per-cpu spinlock unlocked.

To achieve all this:
(a) We introduce a new spinlock that is taken by the *_global_lock_online()
    routine and released by the *_global_unlock_online() routine.
(b) We register a callback for CPU hotplug notifications, and this callback
    takes the same spinlock as above.
(c) We maintain a bitmap which is close to the cpu_online_mask, and once it is
    initialized in the lock_init() code, all future updates to it are done in
    the callback, under the above spinlock.
(d) The above bitmap is used (instead of cpu_online_mask) while locking and
    unlocking the per-cpu locks.

The callback takes the spinlock upon the CPU_UP_PREPARE event. So, if the
br_write_lock-unlock sequence is in progress, the callback keeps spinning,
thus preventing the CPU online operation till the lock-unlock sequence is
complete. This takes care of requirement (3).

The bitmap that we maintain remains unmodified throughout the lock-unlock
sequence, since all updates to it are managed by the callback, which takes
the same spinlock as the one taken by the lock code and released only by the
unlock routine. Combining this with (d) above, satisfies requirement (2).

Overall, since we use a spinlock (mentioned in (a)) to prevent CPU hotplug
operations from racing with br_write_lock-unlock, requirement (1) is also
taken care of.

By the way, it is to be noted that a CPU offline operation can actually run
in parallel with our lock-unlock sequence, because our callback doesn't react
to notifications earlier than CPU_DEAD (in order to maintain our bitmap
properly). And this means, since we use our own bitmap (which is stale, on
purpose) during the lock-unlock sequence, we could end up unlocking the
per-cpu lock of an offline CPU (because we had locked it earlier, when the
CPU was online), in order to satisfy requirement (2). But this is harmless,
though it looks a bit awkward.
Debugged-by: NCong Meng <mc@linux.vnet.ibm.com>
Signed-off-by: NSrivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>
Cc: stable@vger.kernel.org
      e30e2fdf
  7. 19 12月, 2011 1 次提交
  9. 17 12月, 2011 1 次提交
    • E
      iommu: Export intel_iommu_enabled to signal when iommu is in use · 8bc1f85c
      由 Eugeni Dodonov 提交于 
      In i915 driver, we do not enable either rc6 or semaphores on SNB when dmar
is enabled. The new 'intel_iommu_enabled' variable signals when the
iommu code is in operation.

Cc: Ted Phelps <phelps@gnusto.com>
Cc: Peter <pab1612@gmail.com>
Cc: Lukas Hejtmanek <xhejtman@fi.muni.cz>
Cc: Andrew Lutomirski <luto@mit.edu>
CC: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Eugeni Dodonov <eugeni.dodonov@intel.com>
Signed-off-by: NKeith Packard <keithp@keithp.com>
      8bc1f85c
  11. 13 12月, 2011 1 次提交
    • L
      linux/log2.h: Fix rounddown_pow_of_two(1) · 13c07b02
      由 Linus Torvalds 提交于 
      Exactly like roundup_pow_of_two(1), the rounddown version was buggy for
the case of a compile-time constant '1' argument.  Probably because it
originated from the same code, sharing history with the roundup version
from before the bugfix (for that one, see commit 1a06a52e: "Fix
roundup_pow_of_two(1)").

However, unlike the roundup version, the fix for rounddown is to just
remove the broken special case entirely.  It's simply not needed - the
generic code

    1UL << ilog2(n)

does the right thing for the constant '1' argment too.  The only reason
roundup needed that special case was because rounding up does so by
subtracting one from the argument (and then adding one to the result)
causing the obvious problems with "ilog2(0)".

But rounddown doesn't do any of that, since ilog2() naturally truncates
(ie "rounds down") to the right rounded down value.  And without the
ilog2(0) case, there's no reason for the special case that had the wrong
value.

tl;dr: rounddown_pow_of_two(1) should be 1, not 0.
Acked-by: NDmitry Torokhov <dtor@vmware.com>
Cc: stable@kernel.org
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      13c07b02
  13. 11 12月, 2011 1 次提交
  15. 09 12月, 2011 1 次提交
  17. 07 12月, 2011 1 次提交
    • A
      fix apparmor dereferencing potentially freed dentry, sanitize __d_path() API · 02125a82
      由 Al Viro 提交于 
      __d_path() API is asking for trouble and in case of apparmor d_namespace_path()
getting just that.  The root cause is that when __d_path() misses the root
it had been told to look for, it stores the location of the most remote ancestor
in *root.  Without grabbing references.  Sure, at the moment of call it had
been pinned down by what we have in *path.  And if we raced with umount -l, we
could have very well stopped at vfsmount/dentry that got freed as soon as
prepend_path() dropped vfsmount_lock.

It is safe to compare these pointers with pre-existing (and known to be still
alive) vfsmount and dentry, as long as all we are asking is "is it the same
address?".  Dereferencing is not safe and apparmor ended up stepping into
that.  d_namespace_path() really wants to examine the place where we stopped,
even if it's not connected to our namespace.  As the result, it looked
at ->d_sb->s_magic of a dentry that might've been already freed by that point.
All other callers had been careful enough to avoid that, but it's really
a bad interface - it invites that kind of trouble.

The fix is fairly straightforward, even though it's bigger than I'd like:
	* prepend_path() root argument becomes const.
	* __d_path() is never called with NULL/NULL root.  It was a kludge
to start with.  Instead, we have an explicit function - d_absolute_root().
Same as __d_path(), except that it doesn't get root passed and stops where
it stops.  apparmor and tomoyo are using it.
	* __d_path() returns NULL on path outside of root.  The main
caller is show_mountinfo() and that's precisely what we pass root for - to
skip those outside chroot jail.  Those who don't want that can (and do)
use d_path().
	* __d_path() root argument becomes const.  Everyone agrees, I hope.
	* apparmor does *NOT* try to use __d_path() or any of its variants
when it sees that path->mnt is an internal vfsmount.  In that case it's
definitely not mounted anywhere and dentry_path() is exactly what we want
there.  Handling of sysctl()-triggered weirdness is moved to that place.
	* if apparmor is asked to do pathname relative to chroot jail
and __d_path() tells it we it's not in that jail, the sucker just calls
d_absolute_path() instead.  That's the other remaining caller of __d_path(),
BTW.
        * seq_path_root() does _NOT_ return -ENAMETOOLONG (it's stupid anyway -
the normal seq_file logics will take care of growing the buffer and redoing
the call of ->show() just fine).  However, if it gets path not reachable
from root, it returns SEQ_SKIP.  The only caller adjusted (i.e. stopped
ignoring the return value as it used to do).
Reviewed-by: NJohn Johansen <john.johansen@canonical.com>
ACKed-by: NJohn Johansen <john.johansen@canonical.com>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>
Cc: stable@vger.kernel.org
      02125a82
  19. 06 12月, 2011 2 次提交
  21. 05 12月, 2011 1 次提交
    • P
      perf: Fix loss of notification with multi-event · 10c6db11
      由 Peter Zijlstra 提交于 
      When you do:
        $ perf record -e cycles,cycles,cycles noploop 10

You expect about 10,000 samples for each event, i.e., 10s at
1000samples/sec. However, this is not what's happening. You
get much fewer samples, maybe 3700 samples/event:

$ perf report -D | tail -15
Aggregated stats:
           TOTAL events:      10998
            MMAP events:         66
            COMM events:          2
          SAMPLE events:      10930
cycles stats:
           TOTAL events:       3644
          SAMPLE events:       3644
cycles stats:
           TOTAL events:       3642
          SAMPLE events:       3642
cycles stats:
           TOTAL events:       3644
          SAMPLE events:       3644

On a Intel Nehalem or even AMD64, there are 4 counters capable
of measuring cycles, so there is plenty of space to measure those
events without multiplexing (even with the NMI watchdog active).
And even with multiplexing, we'd expect roughly the same number
of samples per event.

The root of the problem was that when the event that caused the buffer
to become full was not the first event passed on the cmdline, the user
notification would get lost. The notification was sent to the file
descriptor of the overflowed event but the perf tool was not polling
on it.  The perf tool aggregates all samples into a single buffer,
i.e., the buffer of the first event. Consequently, it assumes
notifications for any event will come via that descriptor.

The seemingly straight forward solution of moving the waitq into the
ringbuffer object doesn't work because of life-time issues. One could
perf_event_set_output() on a fd that you're also blocking on and cause
the old rb object to be freed while its waitq would still be
referenced by the blocked thread -> FAIL.

Therefore link all events to the ringbuffer and broadcast the wakeup
from the ringbuffer object to all possible events that could be waited
upon. This is rather ugly, and we're open to better solutions but it
works for now.
Reported-by: NStephane Eranian <eranian@google.com>
Finished-by: NStephane Eranian <eranian@google.com>
Reviewed-by: NStephane Eranian <eranian@google.com>
Signed-off-by: NPeter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20111126014731.GA7030@quadSigned-off-by: NIngo Molnar <mingo@elte.hu>
      10c6db11
  23. 04 12月, 2011 1 次提交
  25. 29 11月, 2011 2 次提交