[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Qemu-devel] [PULL 6/6] main-loop: Acquire main_context lock around os_h
From: |
Paolo Bonzini |
Subject: |
[Qemu-devel] [PULL 6/6] main-loop: Acquire main_context lock around os_host_main_loop_wait. |
Date: |
Mon, 3 Apr 2017 21:44:09 +0200 |
From: "Richard W.M. Jones" <address@hidden>
When running virt-rescue the serial console hangs from time to time.
Virt-rescue runs an ordinary Linux kernel "appliance", but there is
only a single idle process running inside, so the qemu main loop is
largely idle. With virt-rescue >= 1.37 you may be able to observe the
hang by doing:
$ virt-rescue -e ^] --scratch
><rescue> while true; do ls -l /usr/bin; done
The hang in virt-rescue can be resolved by pressing a key on the
serial console.
Possibly with the same root cause, we also observed hangs during very
early boot of regular Linux VMs with a serial console. Those hangs
are extremely rare, but you may be able to observe them by running
this command on baremetal for a sufficiently long time:
$ while libguestfs-test-tool -t 60 >& /tmp/log ; do echo -n . ; done
(Check in /tmp/log that the failure was caused by a hang during early
boot, and not some other reason)
During investigation of this bug, Paolo Bonzini wrote:
> glib is expecting QEMU to use g_main_context_acquire around accesses to
> GMainContext. However QEMU is not doing that, instead it is taking its
> own mutex. So we should add g_main_context_acquire and
> g_main_context_release in the two implementations of
> os_host_main_loop_wait; these should undo the effect of Frediano's
> glib patch.
This patch exactly implements Paolo's suggestion in that paragraph.
This fixes the serial console hang in my testing, across 3 different
physical machines (AMD, Intel Core i7 and Intel Xeon), over many hours
of automated testing. I wasn't able to reproduce the early boot hangs
(but as noted above, these are extremely rare in any case).
Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=1435432
Reported-by: Richard W.M. Jones <address@hidden>
Tested-by: Richard W.M. Jones <address@hidden>
Signed-off-by: Richard W.M. Jones <address@hidden>
Message-Id: <address@hidden>
[Paolo: this is actually a glib bug: recent glib versions are also
expecting g_main_context_acquire around g_poll---but that is not
documented and probably not even intended].
Signed-off-by: Paolo Bonzini <address@hidden>
---
util/main-loop.c | 11 +++++++++++
1 file changed, 11 insertions(+)
diff --git a/util/main-loop.c b/util/main-loop.c
index 4534c89..19cad6b 100644
--- a/util/main-loop.c
+++ b/util/main-loop.c
@@ -218,9 +218,12 @@ static void glib_pollfds_poll(void)
static int os_host_main_loop_wait(int64_t timeout)
{
+ GMainContext *context = g_main_context_default();
int ret;
static int spin_counter;
+ g_main_context_acquire(context);
+
glib_pollfds_fill(&timeout);
/* If the I/O thread is very busy or we are incorrectly busy waiting in
@@ -256,6 +259,9 @@ static int os_host_main_loop_wait(int64_t timeout)
}
glib_pollfds_poll();
+
+ g_main_context_release(context);
+
return ret;
}
#else
@@ -412,12 +418,15 @@ static int os_host_main_loop_wait(int64_t timeout)
fd_set rfds, wfds, xfds;
int nfds;
+ g_main_context_acquire(context);
+
/* XXX: need to suppress polling by better using win32 events */
ret = 0;
for (pe = first_polling_entry; pe != NULL; pe = pe->next) {
ret |= pe->func(pe->opaque);
}
if (ret != 0) {
+ g_main_context_release(context);
return ret;
}
@@ -472,6 +481,8 @@ static int os_host_main_loop_wait(int64_t timeout)
g_main_context_dispatch(context);
}
+ g_main_context_release(context);
+
return select_ret || g_poll_ret;
}
#endif
--
1.8.3.1
- [Qemu-devel] [PULL 0/6] Misc patches for QEMU 2.9.0-rc3, Paolo Bonzini, 2017/04/03
- [Qemu-devel] [PULL 1/6] iscsi: drop unused IscsiAIOCB.qiov field, Paolo Bonzini, 2017/04/03
- [Qemu-devel] [PULL 2/6] target-i386: fix "info lapic" segfault on isapc, Paolo Bonzini, 2017/04/03
- [Qemu-devel] [PULL 3/6] ipmi: Fix macro issues, Paolo Bonzini, 2017/04/03
- [Qemu-devel] [PULL 4/6] nbd: fix memory leak on socket_connect failed, Paolo Bonzini, 2017/04/03
- [Qemu-devel] [PULL 5/6] exec: revert MemoryRegionCache, Paolo Bonzini, 2017/04/03
- [Qemu-devel] [PULL 6/6] main-loop: Acquire main_context lock around os_host_main_loop_wait.,
Paolo Bonzini <=
- Re: [Qemu-devel] [PULL 0/6] Misc patches for QEMU 2.9.0-rc3, Peter Maydell, 2017/04/04