From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:54050) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1gTrYE-0007do-Vj for qemu-devel@nongnu.org; Mon, 03 Dec 2018 11:58:32 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1gTrYD-0008K0-MI for qemu-devel@nongnu.org; Mon, 03 Dec 2018 11:58:30 -0500 From: Kevin Wolf Date: Mon, 3 Dec 2018 17:58:09 +0100 Message-Id: <20181203165810.14509-2-kwolf@redhat.com> In-Reply-To: <20181203165810.14509-1-kwolf@redhat.com> References: <20181203165810.14509-1-kwolf@redhat.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Subject: [Qemu-devel] [PULL 1/2] mirror: fix dead-lock List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: qemu-block@nongnu.org Cc: kwolf@redhat.com, peter.maydell@linaro.org, qemu-devel@nongnu.org From: Vladimir Sementsov-Ogievskiy Let start from the beginning: Commit b9e413dd375 (in 2.9) "block: explicitly acquire aiocontext in aio callbacks that need it" added pairs of aio_context_acquire/release to mirror_write_complete and mirror_read_complete, when they were aio callbacks for blk_aio_* calls. Then, commit 2e1990b26e5 (in 3.0) "block/mirror: Convert to coroutines" dropped these blk_aio_* calls, than mirror_write_complete and mirror_read_complete are not callbacks more, and don't need additional aiocontext acquiring. Furthermore, mirror_read_complete calls blk_co_pwritev inside these pair of aio_context_acquire/release, which leads to the following dead-lock with mirror: (gdb) info thr Id Target Id Frame 3 Thread (LWP 145412) "qemu-system-x86" syscall () 2 Thread (LWP 145416) "qemu-system-x86" __lll_lock_wait () * 1 Thread (LWP 145411) "qemu-system-x86" __lll_lock_wait () (gdb) bt #0 __lll_lock_wait () #1 _L_lock_812 () #2 __GI___pthread_mutex_lock #3 qemu_mutex_lock_impl (mutex=3D0x561032dce420 , file=3D0x5610327d8654 "util/main-loop.c", line=3D236) at util/qemu-thread-posix.c:66 #4 qemu_mutex_lock_iothread_impl #5 os_host_main_loop_wait (timeout=3D480116000) at util/main-loop.c:236 #6 main_loop_wait (nonblocking=3D0) at util/main-loop.c:497 #7 main_loop () at vl.c:1892 #8 main Printing contents of qemu_global_mutex, I see that "__owner =3D 145416", so, thr1 is main loop, and now it wants BQL, which is owned by thr2. (gdb) thr 2 (gdb) bt #0 __lll_lock_wait () #1 _L_lock_870 () #2 __GI___pthread_mutex_lock #3 qemu_mutex_lock_impl (mutex=3D0x561034d25dc0, ... #4 aio_context_acquire (ctx=3D0x561034d25d60) #5 dma_blk_cb #6 dma_blk_io #7 dma_blk_read #8 ide_dma_cb #9 bmdma_cmd_writeb #10 bmdma_write #11 memory_region_write_accessor #12 access_with_adjusted_size #15 flatview_write #16 address_space_write #17 address_space_rw #18 kvm_handle_io #19 kvm_cpu_exec #20 qemu_kvm_cpu_thread_fn #21 qemu_thread_start #22 start_thread #23 clone () Printing mutex in fr 2, I see "__owner =3D 145411", so thr2 wants aio context mutex, which is owned by thr1. Classic dead-lock. Then, let's check that aio context is hold by mirror coroutine: just print coroutine stack of first tracked request in mirror job target: (gdb) [...] (gdb) qemu coroutine 0x561035dd0860 #0 qemu_coroutine_switch #1 qemu_coroutine_yield #2 qemu_co_mutex_lock_slowpath #3 qemu_co_mutex_lock #4 qcow2_co_pwritev #5 bdrv_driver_pwritev #6 bdrv_aligned_pwritev #7 bdrv_co_pwritev #8 blk_co_pwritev #9 mirror_read_complete () at block/mirror.c:232 #10 mirror_co_read () at block/mirror.c:370 #11 coroutine_trampoline #12 __start_context Yes it is mirror_read_complete calling blk_co_pwritev after acquiring aio context. Signed-off-by: Vladimir Sementsov-Ogievskiy Reviewed-by: Max Reitz Signed-off-by: Kevin Wolf --- block/mirror.c | 13 +++++-------- 1 file changed, 5 insertions(+), 8 deletions(-) diff --git a/block/mirror.c b/block/mirror.c index 56d9ef7474..8f52c6215d 100644 --- a/block/mirror.c +++ b/block/mirror.c @@ -199,7 +199,6 @@ static void coroutine_fn mirror_write_complete(Mirror= Op *op, int ret) { MirrorBlockJob *s =3D op->s; =20 - aio_context_acquire(blk_get_aio_context(s->common.blk)); if (ret < 0) { BlockErrorAction action; =20 @@ -209,15 +208,14 @@ static void coroutine_fn mirror_write_complete(Mirr= orOp *op, int ret) s->ret =3D ret; } } + mirror_iteration_done(op, ret); - aio_context_release(blk_get_aio_context(s->common.blk)); } =20 static void coroutine_fn mirror_read_complete(MirrorOp *op, int ret) { MirrorBlockJob *s =3D op->s; =20 - aio_context_acquire(blk_get_aio_context(s->common.blk)); if (ret < 0) { BlockErrorAction action; =20 @@ -228,12 +226,11 @@ static void coroutine_fn mirror_read_complete(Mirro= rOp *op, int ret) } =20 mirror_iteration_done(op, ret); - } else { - ret =3D blk_co_pwritev(s->target, op->offset, - op->qiov.size, &op->qiov, 0); - mirror_write_complete(op, ret); + return; } - aio_context_release(blk_get_aio_context(s->common.blk)); + + ret =3D blk_co_pwritev(s->target, op->offset, op->qiov.size, &op->qi= ov, 0); + mirror_write_complete(op, ret); } =20 /* Clip bytes relative to offset to not exceed end-of-file */ --=20 2.19.2