From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:37237) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1br5sz-0004QN-9G for qemu-devel@nongnu.org; Mon, 03 Oct 2016 12:14:38 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1br5su-0005AW-2B for qemu-devel@nongnu.org; Mon, 03 Oct 2016 12:14:37 -0400 Received: from mx1.redhat.com ([209.132.183.28]:57804) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1br5st-00059M-Of for qemu-devel@nongnu.org; Mon, 03 Oct 2016 12:14:32 -0400 Received: from int-mx13.intmail.prod.int.phx2.redhat.com (int-mx13.intmail.prod.int.phx2.redhat.com [10.5.11.26]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id AF24210571 for ; Mon, 3 Oct 2016 16:14:20 +0000 (UTC) From: Paolo Bonzini Date: Mon, 3 Oct 2016 18:14:15 +0200 Message-Id: <1475511256-20051-2-git-send-email-pbonzini@redhat.com> In-Reply-To: <1475511256-20051-1-git-send-email-pbonzini@redhat.com> References: <1475511256-20051-1-git-send-email-pbonzini@redhat.com> Subject: [Qemu-devel] [PATCH 1/2] async: add aio_bh_schedule_oneshot List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: qemu-devel@nongnu.org Cc: kwolf@redhat.com, famz@redhat.com, stefanha@redhat.com qemu_bh_delete is already clearing bh->scheduled at the same time as it's setting bh->deleted. Since it's not using any memory barriers, there is no synchronization going on for bh->deleted, and this makes the bh->deleted checks superfluous in aio_compute_timeout, aio_bh_poll and aio_ctx_check. Just remove them, and put the (bh->scheduled && bh->deleted) combo to work in a new function aio_bh_schedule_oneshot. The new function removes the need to save the QEMUBH pointer between the creation and the execution of the bottom half. Signed-off-by: Paolo Bonzini --- async.c | 27 +++++++++++++++++++++++---- include/block/aio.h | 10 ++++++++++ 2 files changed, 33 insertions(+), 4 deletions(-) diff --git a/async.c b/async.c index 3bca9b0..f30d011 100644 --- a/async.c +++ b/async.c @@ -44,6 +44,25 @@ struct QEMUBH { bool deleted; }; +void aio_bh_schedule_oneshot(AioContext *ctx, QEMUBHFunc *cb, void *opaque) +{ + QEMUBH *bh; + bh = g_new(QEMUBH, 1); + *bh = (QEMUBH){ + .ctx = ctx, + .cb = cb, + .opaque = opaque, + }; + qemu_mutex_lock(&ctx->bh_lock); + bh->next = ctx->first_bh; + bh->scheduled = 1; + bh->deleted = 1; + /* Make sure that the members are ready before putting bh into list */ + smp_wmb(); + ctx->first_bh = bh; + qemu_mutex_unlock(&ctx->bh_lock); +} + QEMUBH *aio_bh_new(AioContext *ctx, QEMUBHFunc *cb, void *opaque) { QEMUBH *bh; @@ -86,7 +105,7 @@ int aio_bh_poll(AioContext *ctx) * thread sees the zero before bh->cb has run, and thus will call * aio_notify again if necessary. */ - if (!bh->deleted && atomic_xchg(&bh->scheduled, 0)) { + if (atomic_xchg(&bh->scheduled, 0)) { /* Idle BHs and the notify BH don't count as progress */ if (!bh->idle && bh != ctx->notify_dummy_bh) { ret = 1; @@ -104,7 +123,7 @@ int aio_bh_poll(AioContext *ctx) bhp = &ctx->first_bh; while (*bhp) { bh = *bhp; - if (bh->deleted) { + if (bh->deleted && !bh->scheduled) { *bhp = bh->next; g_free(bh); } else { @@ -168,7 +187,7 @@ aio_compute_timeout(AioContext *ctx) QEMUBH *bh; for (bh = ctx->first_bh; bh; bh = bh->next) { - if (!bh->deleted && bh->scheduled) { + if (bh->scheduled) { if (bh->idle) { /* idle bottom halves will be polled at least * every 10ms */ @@ -216,7 +235,7 @@ aio_ctx_check(GSource *source) aio_notify_accept(ctx); for (bh = ctx->first_bh; bh; bh = bh->next) { - if (!bh->deleted && bh->scheduled) { + if (bh->scheduled) { return true; } } diff --git a/include/block/aio.h b/include/block/aio.h index 173c1ed..be7dd3b 100644 --- a/include/block/aio.h +++ b/include/block/aio.h @@ -181,6 +181,16 @@ void aio_context_acquire(AioContext *ctx); void aio_context_release(AioContext *ctx); /** + * aio_bh_schedule_oneshot: Allocate a new bottom half structure that will run + * only once and as soon as possible. + * + * Bottom halves are lightweight callbacks whose invocation is guaranteed + * to be wait-free, thread-safe and signal-safe. The #QEMUBH structure + * is opaque and must be allocated prior to its use. + */ +void aio_bh_schedule_oneshot(AioContext *ctx, QEMUBHFunc *cb, void *opaque); + +/** * aio_bh_new: Allocate a new bottom half structure. * * Bottom halves are lightweight callbacks whose invocation is guaranteed -- 2.7.4