From: Paolo Bonzini <pbonzini@redhat.com>
To: qemu-devel@nongnu.org
Subject: [Qemu-devel] [PATCH 03/10] aio-win32: Factor out duplicate code into aio_dispatch_handlers
Date: Wed, 9 Jul 2014 11:53:03 +0200 [thread overview]
Message-ID: <1404899590-24973-4-git-send-email-pbonzini@redhat.com> (raw)
In-Reply-To: <1404899590-24973-1-git-send-email-pbonzini@redhat.com>
Later, the call to aio_dispatch will move int the GSource wrapper, while the
standalone case will still be call the component functions aio_bh_poll,
aio_dispatch_handlers and timerlistgroup_run_timers.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
aio-win32.c | 89 +++++++++++++++++++++++++++----------------------------------
1 file changed, 39 insertions(+), 50 deletions(-)
diff --git a/aio-win32.c b/aio-win32.c
index 7b28411..5e37b42 100644
--- a/aio-win32.c
+++ b/aio-win32.c
@@ -89,29 +89,12 @@ bool aio_pending(AioContext *ctx)
return false;
}
-bool aio_poll(AioContext *ctx, bool blocking)
+static bool aio_dispatch_handlers(AioContext *ctx, HANDLE event)
{
AioHandler *node;
- HANDLE events[MAXIMUM_WAIT_OBJECTS + 1];
- bool progress;
- int count;
- int timeout;
-
- progress = false;
-
- /*
- * If there are callbacks left that have been queued, we need to call then.
- * Do not call select in this case, because it is possible that the caller
- * does not need a complete flush (as is the case for aio_poll loops).
- */
- if (aio_bh_poll(ctx)) {
- blocking = false;
- progress = true;
- }
+ bool progress = false;
/*
- * Then dispatch any pending callbacks from the GSource.
- *
* We have to walk very carefully in case aio_set_fd_handler is
* called while we're walking.
*/
@@ -121,7 +104,9 @@ bool aio_poll(AioContext *ctx, bool blocking)
ctx->walking_handlers++;
- if (node->pfd.revents && node->io_notify) {
+ if (!node->deleted &&
+ (node->pfd.revents || event_notifier_get_handle(node->e) == event) &&
+ node->io_notify) {
node->pfd.revents = 0;
node->io_notify(node->e);
@@ -142,8 +127,40 @@ bool aio_poll(AioContext *ctx, bool blocking)
}
}
- /* Run timers */
+ return progress;
+}
+
+static bool aio_dispatch(AioContext *ctx)
+{
+ bool progress;
+
+ progress = aio_dispatch_handlers(ctx, INVALID_HANDLE_VALUE);
progress |= timerlistgroup_run_timers(&ctx->tlg);
+ return progress;
+}
+
+bool aio_poll(AioContext *ctx, bool blocking)
+{
+ AioHandler *node;
+ HANDLE events[MAXIMUM_WAIT_OBJECTS + 1];
+ bool progress;
+ int count;
+ int timeout;
+
+ progress = false;
+
+ /*
+ * If there are callbacks left that have been queued, we need to call then.
+ * Do not call select in this case, because it is possible that the caller
+ * does not need a complete flush (as is the case for aio_poll loops).
+ */
+ if (aio_bh_poll(ctx)) {
+ blocking = false;
+ progress = true;
+ }
+
+ /* Dispatch any pending callbacks from the GSource. */
+ progress |= aio_dispatch(ctx);
if (progress && !blocking) {
return true;
@@ -176,35 +193,7 @@ bool aio_poll(AioContext *ctx, bool blocking)
blocking = false;
- /* we have to walk very carefully in case
- * aio_set_fd_handler is called while we're walking */
- node = QLIST_FIRST(&ctx->aio_handlers);
- while (node) {
- AioHandler *tmp;
-
- ctx->walking_handlers++;
-
- if (!node->deleted &&
- event_notifier_get_handle(node->e) == events[ret - WAIT_OBJECT_0] &&
- node->io_notify) {
- node->io_notify(node->e);
-
- /* aio_notify() does not count as progress */
- if (node->e != &ctx->notifier) {
- progress = true;
- }
- }
-
- tmp = node;
- node = QLIST_NEXT(node, node);
-
- ctx->walking_handlers--;
-
- if (!ctx->walking_handlers && tmp->deleted) {
- QLIST_REMOVE(tmp, node);
- g_free(tmp);
- }
- }
+ progress |= aio_dispatch_handlers(ctx, events[ret - WAIT_OBJECT_0]);
/* Try again, but only call each handler once. */
events[ret - WAIT_OBJECT_0] = events[--count];
--
1.9.3
next prev parent reply other threads:[~2014-07-09 9:53 UTC|newest]
Thread overview: 25+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-07-09 9:53 [Qemu-devel] [PATCH for-2.2 00/10] AioContext cleanups and Win32 socket support Paolo Bonzini
2014-07-09 9:53 ` [Qemu-devel] [PATCH 01/10] AioContext: take bottom halves into account when computing aio_poll timeout Paolo Bonzini
2014-08-01 14:34 ` Stefan Hajnoczi
2014-08-01 16:03 ` Paolo Bonzini
2014-07-09 9:53 ` [Qemu-devel] [PATCH 02/10] aio-win32: Evaluate timers after handles Paolo Bonzini
2014-07-09 9:53 ` Paolo Bonzini [this message]
2014-07-09 9:53 ` [Qemu-devel] [PATCH 04/10] AioContext: run bottom halves after polling Paolo Bonzini
2014-07-09 9:53 ` [Qemu-devel] [PATCH 05/10] AioContext: export and use aio_dispatch Paolo Bonzini
2014-07-09 9:53 ` [Qemu-devel] [PATCH 06/10] test-aio: test timers on Windows too Paolo Bonzini
2014-07-09 9:53 ` [Qemu-devel] [PATCH 07/10] aio-win32: add aio_set_dispatching optimization Paolo Bonzini
2014-07-09 9:53 ` [Qemu-devel] [PATCH 08/10] AioContext: introduce aio_prepare Paolo Bonzini
2014-07-09 9:53 ` [Qemu-devel] [PATCH 09/10] qemu-coroutine-io: fix for Win32 Paolo Bonzini
2014-07-09 9:53 ` [Qemu-devel] [PATCH 10/10] aio-win32: add support for sockets Paolo Bonzini
2014-09-12 1:39 ` TeLeMan
2014-09-12 10:05 ` Paolo Bonzini
2014-09-13 2:22 ` TeLeMan
2014-09-13 10:33 ` Paolo Bonzini
2014-09-15 1:18 ` TeLeMan
2014-09-15 15:16 ` Paolo Bonzini
2014-09-12 12:51 ` Stefan Hajnoczi
2014-09-12 12:52 ` Paolo Bonzini
2014-09-12 1:43 ` TeLeMan
2014-08-01 14:52 ` [Qemu-devel] [PATCH for-2.2 00/10] AioContext cleanups and Win32 socket support Stefan Hajnoczi
2014-08-01 15:07 ` Paolo Bonzini
2014-08-28 14:00 ` Stefan Hajnoczi
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=1404899590-24973-4-git-send-email-pbonzini@redhat.com \
--to=pbonzini@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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).