Skip to content

Q
qemu

openeuler / qemu

通知 10
Star 0
Fork 0
Q
qemu

体验新版 GitCode，发现更多精彩内容 >>

切换分支/标签
  1. 29 10月, 2018 1 次提交
  3. 26 9月, 2018 3 次提交
  5. 15 8月, 2018 3 次提交
    • F
      aio-posix: Improve comment around marking node deleted · 37a81812
      由 Fam Zheng 提交于 
      The counter is for qemu_lockcnt_inc/dec sections (read side),
qemu_lockcnt_lock/unlock is for the write side.
Suggested-by: NPaolo Bonzini <pbonzini@redhat.com>
Signed-off-by: NFam Zheng <famz@redhat.com>
Message-Id: <20180803063917.30292-1-famz@redhat.com>
Signed-off-by: NFam Zheng <famz@redhat.com>
      37a81812
    • F
      aio: Do aio_notify_accept only during blocking aio_poll · b37548fc
      由 Fam Zheng 提交于 
      An aio_notify() pairs with an aio_notify_accept(). The former should
happen in the main thread or a vCPU thread, and the latter should be
done in the IOThread.

There is one rare case that the main thread or vCPU thread may "steal"
the aio_notify() event just raised by itself, in bdrv_set_aio_context()
[1]. The sequence is like this:

    main thread                     IO Thread
    ===============================================================
    bdrv_drained_begin()
      aio_disable_external(ctx)
                                    aio_poll(ctx, true)
                                      ctx->notify_me += 2
    ...
    bdrv_drained_end()
      ...
        aio_notify()
    ...
    bdrv_set_aio_context()
      aio_poll(ctx, false)
[1]     aio_notify_accept(ctx)
                                      ppoll() /* Hang! */

[1] is problematic. It will clear the ctx->notifier event so that
the blocked ppoll() will not return.

(For the curious, this bug was noticed when booting a number of VMs
simultaneously in RHV.  One or two of the VMs will hit this race
condition, making the VIRTIO device unresponsive to I/O commands. When
it hangs, Seabios is busy waiting for a read request to complete (read
MBR), right after initializing the virtio-blk-pci device, using 100%
guest CPU. See also https://bugzilla.redhat.com/show_bug.cgi?id=1562750
for the original bug analysis.)

aio_notify() only injects an event when ctx->notify_me is set,
correspondingly aio_notify_accept() is only useful when ctx->notify_me
_was_ set. Move the call to it into the "blocking" branch. This will
effectively skip [1] and fix the hang.

Furthermore, blocking aio_poll is only allowed on home thread
(in_aio_context_home_thread), because otherwise two blocking
aio_poll()'s can steal each other's ctx->notifier event and cause
hanging just like described above.

Cc: qemu-stable@nongnu.org
Suggested-by: NPaolo Bonzini <pbonzini@redhat.com>
Signed-off-by: NFam Zheng <famz@redhat.com>
Message-Id: <20180809132259.18402-3-famz@redhat.com>
Signed-off-by: NFam Zheng <famz@redhat.com>
      b37548fc
    • F
      aio-posix: Don't count ctx->notifier as progress when polling · 70232b52
      由 Fam Zheng 提交于 
      The same logic exists in fd polling. This change is especially important
to avoid busy loop once we limit aio_notify_accept() to blocking
aio_poll().

Cc: qemu-stable@nongnu.org
Signed-off-by: NFam Zheng <famz@redhat.com>
Message-Id: <20180809132259.18402-2-famz@redhat.com>
Signed-off-by: NFam Zheng <famz@redhat.com>
      70232b52
  7. 18 5月, 2018 1 次提交
  9. 10 2月, 2018 1 次提交
  11. 06 11月, 2017 1 次提交
  13. 04 10月, 2017 1 次提交
    • S
      aio: fix assert when remove poll during destroy · f708a5e7
      由 Stefan Hajnoczi 提交于 
      After iothread is enabled internally inside QEMU with GMainContext, we
may encounter this warning when destroying the iothread:

(qemu-system-x86_64:19925): GLib-CRITICAL **: g_source_remove_poll:
 assertion '!SOURCE_DESTROYED (source)' failed

The problem is that g_source_remove_poll() does not allow to remove one
source from array if the source is detached from its owner
context. (peterx: which IMHO does not make much sense)

Fix it on QEMU side by avoid calling g_source_remove_poll() if we know
the object is during destruction, and we won't leak anything after all
since the array will be gone soon cleanly even with that fd.
Signed-off-by: NStefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: NFam Zheng <famz@redhat.com>
Signed-off-by: NPeter Xu <peterx@redhat.com>
Message-id: 20170928025958.1420-6-peterx@redhat.com
[peterx: write the commit message]
Signed-off-by: NPeter Xu <peterx@redhat.com>
Signed-off-by: NStefan Hajnoczi <stefanha@redhat.com>
      f708a5e7
  15. 21 2月, 2017 6 次提交
  17. 01 2月, 2017 1 次提交
  19. 26 1月, 2017 1 次提交
    • S
      aio-posix: honor is_external in AioContext polling · 59c9f437
      由 Stefan Hajnoczi 提交于 
      AioHandlers marked ->is_external must be skipped when aio_node_check()
fails.  bdrv_drained_begin() needs this to prevent dataplane from
submitting new I/O requests while another thread accesses the device and
relies on it being quiesced.

This patch fixes the following segfault:

  Program terminated with signal SIGSEGV, Segmentation fault.
  #0  0x00005577f6127dad in bdrv_io_plug (bs=0x5577f7ae52f0) at qemu/block/io.c:2650
  2650            bdrv_io_plug(child->bs);
  [Current thread is 1 (Thread 0x7ff5c4bd1c80 (LWP 10917))]
  (gdb) bt
  #0  0x00005577f6127dad in bdrv_io_plug (bs=0x5577f7ae52f0) at qemu/block/io.c:2650
  #1  0x00005577f6114363 in blk_io_plug (blk=0x5577f7b8ba20) at qemu/block/block-backend.c:1561
  #2  0x00005577f5d4091d in virtio_blk_handle_vq (s=0x5577f9ada030, vq=0x5577f9b3d2a0) at qemu/hw/block/virtio-blk.c:589
  #3  0x00005577f5d4240d in virtio_blk_data_plane_handle_output (vdev=0x5577f9ada030, vq=0x5577f9b3d2a0) at qemu/hw/block/dataplane/virtio-blk.c:158
  #4  0x00005577f5d88acd in virtio_queue_notify_aio_vq (vq=0x5577f9b3d2a0) at qemu/hw/virtio/virtio.c:1304
  #5  0x00005577f5d8aaaf in virtio_queue_host_notifier_aio_poll (opaque=0x5577f9b3d308) at qemu/hw/virtio/virtio.c:2134
  #6  0x00005577f60ca077 in run_poll_handlers_once (ctx=0x5577f79ddbb0) at qemu/aio-posix.c:493
  #7  0x00005577f60ca268 in try_poll_mode (ctx=0x5577f79ddbb0, blocking=true) at qemu/aio-posix.c:569
  #8  0x00005577f60ca331 in aio_poll (ctx=0x5577f79ddbb0, blocking=true) at qemu/aio-posix.c:601
  #9  0x00005577f612722a in bdrv_flush (bs=0x5577f7c20970) at qemu/block/io.c:2403
  #10 0x00005577f60c1b2d in bdrv_close (bs=0x5577f7c20970) at qemu/block.c:2322
  #11 0x00005577f60c20e7 in bdrv_delete (bs=0x5577f7c20970) at qemu/block.c:2465
  #12 0x00005577f60c3ecf in bdrv_unref (bs=0x5577f7c20970) at qemu/block.c:3425
  #13 0x00005577f60bf951 in bdrv_root_unref_child (child=0x5577f7a2de70) at qemu/block.c:1361
  #14 0x00005577f6112162 in blk_remove_bs (blk=0x5577f7b8ba20) at qemu/block/block-backend.c:491
  #15 0x00005577f6111b1b in blk_remove_all_bs () at qemu/block/block-backend.c:245
  #16 0x00005577f60c1db6 in bdrv_close_all () at qemu/block.c:2382
  #17 0x00005577f5e60cca in main (argc=20, argv=0x7ffea6eb8398, envp=0x7ffea6eb8440) at qemu/vl.c:4684

The key thing is that bdrv_close() uses bdrv_drained_begin() and
virtio_queue_host_notifier_aio_poll() must not be called.

Thanks to Fam Zheng <famz@redhat.com> for identifying the root cause of
this crash.
Reported-by: NAlberto Garcia <berto@igalia.com>
Signed-off-by: NStefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: NFam Zheng <famz@redhat.com>
Tested-by: NAlberto Garcia <berto@igalia.com>
Message-id: 20170124095350.16679-1-stefanha@redhat.com
Signed-off-by: NStefan Hajnoczi <stefanha@redhat.com>
      59c9f437
  21. 16 1月, 2017 3 次提交
  23. 04 1月, 2017 5 次提交
    • S
      aio: self-tune polling time · 82a41186
      由 Stefan Hajnoczi 提交于 
      This patch is based on the algorithm for the kvm.ko halt_poll_ns
parameter in Linux.  The initial polling time is zero.

If the event loop is woken up within the maximum polling time it means
polling could be effective, so grow polling time.

If the event loop is woken up beyond the maximum polling time it means
polling is not effective, so shrink polling time.

If the event loop makes progress within the current polling time then
the sweet spot has been reached.

This algorithm adjusts the polling time so it can adapt to variations in
workloads.  The goal is to reach the sweet spot while also recognizing
when polling would hurt more than help.

Two new trace events, poll_grow and poll_shrink, are added for observing
polling time adjustment.
Signed-off-by: NStefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: NPaolo Bonzini <pbonzini@redhat.com>
Message-id: 20161201192652.9509-13-stefanha@redhat.com
Signed-off-by: NStefan Hajnoczi <stefanha@redhat.com>
      82a41186
    • S
      aio: add .io_poll_begin/end() callbacks · 684e508c
      由 Stefan Hajnoczi 提交于 
      The begin and end callbacks can be used to prepare for the polling loop
and clean up when polling stops.  Note that they may only be called once
for multiple aio_poll() calls if polling continues to succeed.  Once
polling fails the end callback is invoked before aio_poll() resumes file
descriptor monitoring.
Signed-off-by: NStefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: NPaolo Bonzini <pbonzini@redhat.com>
Message-id: 20161201192652.9509-11-stefanha@redhat.com
Signed-off-by: NStefan Hajnoczi <stefanha@redhat.com>
      684e508c
    • S
      aio: add polling mode to AioContext · 4a1cba38
      由 Stefan Hajnoczi 提交于 
      The AioContext event loop uses ppoll(2) or epoll_wait(2) to monitor file
descriptors or until a timer expires.  In cases like virtqueues, Linux
AIO, and ThreadPool it is technically possible to wait for events via
polling (i.e. continuously checking for events without blocking).

Polling can be faster than blocking syscalls because file descriptors,
the process scheduler, and system calls are bypassed.

The main disadvantage to polling is that it increases CPU utilization.
In classic polling configuration a full host CPU thread might run at
100% to respond to events as quickly as possible.  This patch implements
a timeout so we fall back to blocking syscalls if polling detects no
activity.  After the timeout no CPU cycles are wasted on polling until
the next event loop iteration.

The run_poll_handlers_begin() and run_poll_handlers_end() trace events
are added to aid performance analysis and troubleshooting.  If you need
to know whether polling mode is being used, trace these events to find
out.

Note that the AioContext is now re-acquired before disabling notify_me
in the non-polling case.  This makes the code cleaner since notify_me
was enabled outside the non-polling AioContext release region.  This
change is correct since it's safe to keep notify_me enabled longer
(disabling is an optimization) but potentially causes unnecessary
event_notifer_set() calls.  I think the chance of performance regression
is small here.
Signed-off-by: NStefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: NPaolo Bonzini <pbonzini@redhat.com>
Message-id: 20161201192652.9509-4-stefanha@redhat.com
Signed-off-by: NStefan Hajnoczi <stefanha@redhat.com>
      4a1cba38
    • S
      aio: add AioPollFn and io_poll() interface · f6a51c84
      由 Stefan Hajnoczi 提交于 
      The new AioPollFn io_poll() argument to aio_set_fd_handler() and
aio_set_event_handler() is used in the next patch.

Keep this code change separate due to the number of files it touches.
Signed-off-by: NStefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: NPaolo Bonzini <pbonzini@redhat.com>
Message-id: 20161201192652.9509-3-stefanha@redhat.com
Signed-off-by: NStefan Hajnoczi <stefanha@redhat.com>
      f6a51c84
    • S
      aio: add flag to skip fds to aio_dispatch() · 721671ad
      由 Stefan Hajnoczi 提交于 
      Polling mode will not call ppoll(2)/epoll_wait(2).  Therefore we know
there are no fds ready and should avoid looping over fd handlers in
aio_dispatch().
Signed-off-by: NStefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: NPaolo Bonzini <pbonzini@redhat.com>
Message-id: 20161201192652.9509-2-stefanha@redhat.com
Signed-off-by: NStefan Hajnoczi <stefanha@redhat.com>
      721671ad
  25. 09 11月, 2016 2 次提交
  27. 28 9月, 2016 1 次提交
  29. 18 7月, 2016 1 次提交
  31. 22 4月, 2016 1 次提交
  33. 17 3月, 2016 1 次提交
  35. 05 2月, 2016 1 次提交
    • P
      all: Clean up includes · d38ea87a
      由 Peter Maydell 提交于 
      Clean up includes so that osdep.h is included first and headers
which it implies are not included manually.

This commit was created with scripts/clean-includes.
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>
Message-id: 1454089805-5470-16-git-send-email-peter.maydell@linaro.org
      d38ea87a
  37. 17 11月, 2015 1 次提交
  39. 09 11月, 2015 2 次提交
    • F
      aio: Introduce aio-epoll.c · fbe3fc5c
      由 Fam Zheng 提交于 
      To minimize code duplication, epoll is hooked into aio-posix's
aio_poll() instead of rolling its own. This approach also has both
compile-time and run-time switchability.

1) When QEMU starts with a small number of fds in the event loop, ppoll
is used.

2) When QEMU starts with a big number of fds, or when more devices are
hot plugged, epoll kicks in when the number of fds hits the threshold.

3) Some fds may not support epoll, such as tty based stdio. In this
case, it falls back to ppoll.

A rough benchmark with scsi-disk on virtio-scsi dataplane (epoll gets
enabled from 64 onward). Numbers are in MB/s.

===============================================
             |     master     |     epoll
             |                |
scsi disks # | read    randrw | read    randrw
-------------|----------------|----------------
1            | 86      36     | 92      45
8            | 87      43     | 86      41
64           | 71      32     | 70      38
128          | 48      24     | 58      31
256          | 37      19     | 57      28
===============================================

To comply with aio_{disable,enable}_external, we always use ppoll when
aio_external_disabled() is true.

[Removed #ifdef CONFIG_EPOLL around AioContext epollfd field declaration
since the field is also referenced outside CONFIG_EPOLL code.
--Stefan]
Signed-off-by: NFam Zheng <famz@redhat.com>
Message-id: 1446177989-6702-4-git-send-email-famz@redhat.com
Signed-off-by: NStefan Hajnoczi <stefanha@redhat.com>
      fbe3fc5c
    • F
      aio: Introduce aio_context_setup · 37fcee5d
      由 Fam Zheng 提交于 
      This is the place to initialize platform specific bits of AioContext.
Signed-off-by: NFam Zheng <famz@redhat.com>
Message-id: 1446177989-6702-3-git-send-email-famz@redhat.com
Signed-off-by: NStefan Hajnoczi <stefanha@redhat.com>
      37fcee5d
  41. 24 10月, 2015 2 次提交
  43. 22 7月, 2015 1 次提交