All of lore.kernel.org
 help / color / mirror / Atom feed
From: Paolo Bonzini <pbonzini@redhat.com>
To: qemu-devel@nongnu.org
Cc: "Marc-André Lureau" <marcandre.lureau@redhat.com>,
	"James Le Cuirot" <chewi@gentoo.org>
Subject: [PULL 28/30] win32: work around main-loop busy loop on socket/fd event
Date: Wed,  2 Oct 2019 18:51:51 +0200	[thread overview]
Message-ID: <1570035113-56848-29-git-send-email-pbonzini@redhat.com> (raw)
In-Reply-To: <1570035113-56848-1-git-send-email-pbonzini@redhat.com>

From: Marc-André Lureau <marcandre.lureau@redhat.com>

Commit 05e514b1d4d5bd4209e2c8bbc76ff05c85a235f3 introduced an AIO
context optimization to avoid calling event_notifier_test_and_clear() on
ctx->notifier. On Windows, the same notifier is being used to wakeup the
wait on socket events (see commit
d3385eb448e38f828c78f8f68ec5d79c66a58b5d).

The ctx->notifier event is added to the gpoll sources in
aio_set_event_notifier(), aio_ctx_check() should clear the event
regardless of ctx->notified, since Windows sets the event by itself,
bypassing the aio->notified. This fixes qemu not clearing the event
resulting in a busy loop.

Paolo suggested to me on irc to call event_notifier_test_and_clear()
after select() >0 from aio-win32.c's aio_prepare. Unfortunately, not all
fds associated with ctx->notifiers are in AIO fd handlers set.
(qemu_set_nonblock() in util/oslib-win32.c calls qemu_fd_register()).

This is essentially a v2 of a patch that was sent earlier:
https://lists.gnu.org/archive/html/qemu-devel/2017-01/msg00420.html

that resurfaced when James investigated Spice performance issues on Windows:
https://gitlab.freedesktop.org/spice/spice/issues/36

In order to test that patch, I simply tried running test-char on
win32, and it hangs. Applying that patch solves it. QIO idle sources
are not dispatched. I haven't investigated much further, I suspect
source priorities and busy looping still come into play.

This version keeps the "notified" field, so event_notifier_poll()
should still work as expected.

Cc: James Le Cuirot <chewi@gentoo.org>
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 util/async.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/util/async.c b/util/async.c
index 4e4c7af..ca83e32 100644
--- a/util/async.c
+++ b/util/async.c
@@ -354,7 +354,11 @@ void aio_notify(AioContext *ctx)
 
 void aio_notify_accept(AioContext *ctx)
 {
-    if (atomic_xchg(&ctx->notified, false)) {
+    if (atomic_xchg(&ctx->notified, false)
+#ifdef WIN32
+        || true
+#endif
+    ) {
         event_notifier_test_and_clear(&ctx->notifier);
     }
 }
-- 
1.8.3.1




  parent reply	other threads:[~2019-10-02 17:22 UTC|newest]

Thread overview: 37+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-10-02 16:51 [PULL 00/30] Misc patches for 2010-10-02 Paolo Bonzini
2019-10-02 16:51 ` [PULL 01/30] tests/migration: Add a test for auto converge Paolo Bonzini
2019-10-02 16:51 ` [PULL 02/30] target/i386: handle filtered_features in a new function mark_unavailable_features Paolo Bonzini
2019-10-02 16:51 ` [PULL 03/30] target/i386: introduce generic feature dependency mechanism Paolo Bonzini
2019-10-02 16:51 ` [PULL 04/30] target/i386: expand feature words to 64 bits Paolo Bonzini
2019-10-02 16:51 ` [PULL 05/30] target/i386: add VMX definitions Paolo Bonzini
2019-10-02 16:51 ` [PULL 06/30] vmxcap: correct the name of the variables Paolo Bonzini
2019-10-02 16:51 ` [PULL 07/30] target/i386: add VMX features Paolo Bonzini
2019-10-02 16:51 ` [PULL 08/30] target/i386: work around KVM_GET_MSRS bug for secondary execution controls Paolo Bonzini
2019-10-02 16:51 ` [PULL 09/30] target/i386/kvm: Silence warning from Valgrind about uninitialized bytes Paolo Bonzini
2019-10-02 16:51 ` [PULL 10/30] qemu-pr-helper: fix crash in mpath_reconstruct_sense Paolo Bonzini
2019-10-02 16:51 ` [PULL 11/30] replay: don't synchronize memory operations in replay mode Paolo Bonzini
2019-10-02 16:51 ` [PULL 12/30] Makefile: Remove generated files when doing 'distclean' Paolo Bonzini
2019-10-04 12:20   ` Peter Maydell
2019-10-04 16:48     ` Paolo Bonzini
2019-10-07  6:28       ` Thomas Huth
2019-10-07  7:13         ` Aleksandar Markovic
2019-10-02 16:51 ` [PULL 13/30] hw/isa: Introduce a CONFIG_ISA_SUPERIO switch for isa-superio.c Paolo Bonzini
2019-10-02 16:51 ` [PULL 14/30] ide: fix leak from qemu_allocate_irqs Paolo Bonzini
2019-10-02 16:51 ` [PULL 15/30] microblaze: fix leak of fdevice tree blob Paolo Bonzini
2019-10-02 16:51 ` [PULL 16/30] mcf5208: fix leak from qemu_allocate_irqs Paolo Bonzini
2019-10-02 16:51 ` [PULL 17/30] hppa: fix leak from g_strdup_printf Paolo Bonzini
2019-10-02 16:51 ` [PULL 18/30] mips: fix memory leaks in board initialization Paolo Bonzini
2019-10-02 16:51 ` [PULL 19/30] cris: do not leak struct cris_disasm_data Paolo Bonzini
2019-10-02 16:51 ` [PULL 20/30] lm32: do not leak memory on object_new/object_unref Paolo Bonzini
2019-10-02 16:51 ` [PULL 21/30] docker: test-debug: disable LeakSanitizer Paolo Bonzini
2019-10-02 16:51 ` [PULL 22/30] i386: Add CPUID bit for CLZERO and XSAVEERPTR Paolo Bonzini
2019-10-02 16:51 ` [PULL 23/30] vfio: Turn the container error into an Error handle Paolo Bonzini
2019-10-02 16:51 ` [PULL 24/30] memory: allow memory_region_register_iommu_notifier() to fail Paolo Bonzini
2019-10-02 16:51 ` [PULL 25/30] Fix wrong behavior of cpu_memory_rw_debug() function in SMM Paolo Bonzini
2019-10-07  8:29   ` Laszlo Ersek
2019-10-02 16:51 ` [PULL 26/30] util: WSAEWOULDBLOCK on connect should map to EINPROGRESS Paolo Bonzini
2019-10-02 16:51 ` [PULL 27/30] tests: skip serial test on windows Paolo Bonzini
2019-10-02 16:51 ` Paolo Bonzini [this message]
2019-10-02 16:51 ` [PULL 29/30] tests/docker: only enable ubsan for test-clang Paolo Bonzini
2019-10-02 16:51 ` [PULL 30/30] accel/kvm: ensure ret always set Paolo Bonzini
2019-10-02 18:29 ` [PULL 00/30] Misc patches for 2010-10-02 no-reply

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1570035113-56848-29-git-send-email-pbonzini@redhat.com \
    --to=pbonzini@redhat.com \
    --cc=chewi@gentoo.org \
    --cc=marcandre.lureau@redhat.com \
    --cc=qemu-devel@nongnu.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.