From nobody Wed Dec 17 21:47:32 2025 Delivered-To: importer@patchew.org Received-SPF: pass (zoho.com: domain of gnu.org designates 208.118.235.17 as permitted sender) client-ip=208.118.235.17; envelope-from=qemu-devel-bounces+importer=patchew.org@nongnu.org; helo=lists.gnu.org; Authentication-Results: mx.zoho.com; spf=pass (zoho.com: domain of gnu.org designates 208.118.235.17 as permitted sender) smtp.mailfrom=qemu-devel-bounces+importer=patchew.org@nongnu.org; Return-Path: Received: from lists.gnu.org (lists.gnu.org [208.118.235.17]) by mx.zohomail.com with SMTPS id 1498238366992581.2602393354042; Fri, 23 Jun 2017 10:19:26 -0700 (PDT) Received: from localhost ([::1]:36547 helo=lists.gnu.org) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1dOSEv-0005mO-4p for importer@patchew.org; Fri, 23 Jun 2017 13:19:25 -0400 Received: from eggs.gnu.org ([2001:4830:134:3::10]:60608) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1dORMa-0001tt-N2 for qemu-devel@nongnu.org; Fri, 23 Jun 2017 12:23:18 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1dORMW-00087X-Ht for qemu-devel@nongnu.org; Fri, 23 Jun 2017 12:23:16 -0400 Received: from mx1.redhat.com ([209.132.183.28]:32420) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1dORMT-00082I-Fd; Fri, 23 Jun 2017 12:23:09 -0400 Received: from smtp.corp.redhat.com (int-mx02.intmail.prod.int.phx2.redhat.com [10.5.11.12]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id 7BC51B8153; Fri, 23 Jun 2017 16:23:08 +0000 (UTC) Received: from noname.redhat.com (ovpn-117-196.ams2.redhat.com [10.36.117.196]) by smtp.corp.redhat.com (Postfix) with ESMTP id 952B86EC9E; Fri, 23 Jun 2017 16:23:07 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mx1.redhat.com 7BC51B8153 Authentication-Results: ext-mx02.extmail.prod.ext.phx2.redhat.com; dmarc=none (p=none dis=none) header.from=redhat.com Authentication-Results: ext-mx02.extmail.prod.ext.phx2.redhat.com; spf=pass smtp.mailfrom=kwolf@redhat.com DKIM-Filter: OpenDKIM Filter v2.11.0 mx1.redhat.com 7BC51B8153 From: Kevin Wolf To: qemu-block@nongnu.org Date: Fri, 23 Jun 2017 18:21:48 +0200 Message-Id: <1498234919-27316-51-git-send-email-kwolf@redhat.com> In-Reply-To: <1498234919-27316-1-git-send-email-kwolf@redhat.com> References: <1498234919-27316-1-git-send-email-kwolf@redhat.com> X-Scanned-By: MIMEDefang 2.79 on 10.5.11.12 X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.5.16 (mx1.redhat.com [10.5.110.26]); Fri, 23 Jun 2017 16:23:08 +0000 (UTC) X-detected-operating-system: by eggs.gnu.org: GNU/Linux 2.2.x-3.x [generic] [fuzzy] X-Received-From: 209.132.183.28 Subject: [Qemu-devel] [PULL 50/61] qed: Use CoQueue for serialising allocations X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.21 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: kwolf@redhat.com, qemu-devel@nongnu.org Errors-To: qemu-devel-bounces+importer=patchew.org@nongnu.org Sender: "Qemu-devel" X-ZohoMail: RSF_0 Z_629925259 SPT_0 Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Now that we're running in coroutine context, the ad-hoc serialisation code (which drops a request that has to wait out of coroutine context) can be replaced by a CoQueue. This means that when we resume a serialised request, it is running in coroutine context again and its I/O isn't blocking any more. Signed-off-by: Kevin Wolf Reviewed-by: Stefan Hajnoczi --- block/qed.c | 49 +++++++++++++++++-------------------------------- block/qed.h | 3 ++- 2 files changed, 19 insertions(+), 33 deletions(-) diff --git a/block/qed.c b/block/qed.c index a5111fd..cd3ef55 100644 --- a/block/qed.c +++ b/block/qed.c @@ -269,16 +269,10 @@ static void qed_plug_allocating_write_reqs(BDRVQEDSta= te *s) =20 static void qed_unplug_allocating_write_reqs(BDRVQEDState *s) { - QEDAIOCB *acb; - assert(s->allocating_write_reqs_plugged); =20 s->allocating_write_reqs_plugged =3D false; - - acb =3D QSIMPLEQ_FIRST(&s->allocating_write_reqs); - if (acb) { - qed_aio_start_io(acb); - } + qemu_co_enter_next(&s->allocating_write_reqs); } =20 static void qed_clear_need_check(void *opaque, int ret) @@ -305,7 +299,7 @@ static void qed_need_check_timer_cb(void *opaque) BDRVQEDState *s =3D opaque; =20 /* The timer should only fire when allocating writes have drained */ - assert(!QSIMPLEQ_FIRST(&s->allocating_write_reqs)); + assert(!s->allocating_acb); =20 trace_qed_need_check_timer_cb(s); =20 @@ -388,7 +382,7 @@ static int bdrv_qed_do_open(BlockDriverState *bs, QDict= *options, int flags, int ret; =20 s->bs =3D bs; - QSIMPLEQ_INIT(&s->allocating_write_reqs); + qemu_co_queue_init(&s->allocating_write_reqs); =20 ret =3D bdrv_pread(bs->file, 0, &le_header, sizeof(le_header)); if (ret < 0) { @@ -910,11 +904,6 @@ static void qed_aio_complete_bh(void *opaque) qed_release(s); } =20 -static void qed_resume_alloc_bh(void *opaque) -{ - qed_aio_start_io(opaque); -} - static void qed_aio_complete(QEDAIOCB *acb, int ret) { BDRVQEDState *s =3D acb_to_s(acb); @@ -942,13 +931,10 @@ static void qed_aio_complete(QEDAIOCB *acb, int ret) * next request in the queue. This ensures that we don't cycle through * requests multiple times but rather finish one at a time completely. */ - if (acb =3D=3D QSIMPLEQ_FIRST(&s->allocating_write_reqs)) { - QEDAIOCB *next_acb; - QSIMPLEQ_REMOVE_HEAD(&s->allocating_write_reqs, next); - next_acb =3D QSIMPLEQ_FIRST(&s->allocating_write_reqs); - if (next_acb) { - aio_bh_schedule_oneshot(bdrv_get_aio_context(acb->common.bs), - qed_resume_alloc_bh, next_acb); + if (acb =3D=3D s->allocating_acb) { + s->allocating_acb =3D NULL; + if (!qemu_co_queue_empty(&s->allocating_write_reqs)) { + qemu_co_enter_next(&s->allocating_write_reqs); } else if (s->header.features & QED_F_NEED_CHECK) { qed_start_need_check_timer(s); } @@ -1124,17 +1110,18 @@ static int qed_aio_write_alloc(QEDAIOCB *acb, size_= t len) int ret; =20 /* Cancel timer when the first allocating request comes in */ - if (QSIMPLEQ_EMPTY(&s->allocating_write_reqs)) { + if (s->allocating_acb =3D=3D NULL) { qed_cancel_need_check_timer(s); } =20 /* Freeze this request if another allocating write is in progress */ - if (acb !=3D QSIMPLEQ_FIRST(&s->allocating_write_reqs)) { - QSIMPLEQ_INSERT_TAIL(&s->allocating_write_reqs, acb, next); - } - if (acb !=3D QSIMPLEQ_FIRST(&s->allocating_write_reqs) || - s->allocating_write_reqs_plugged) { - return -EINPROGRESS; /* wait for existing request to finish */ + if (s->allocating_acb !=3D acb || s->allocating_write_reqs_plugged) { + if (s->allocating_acb !=3D NULL) { + qemu_co_queue_wait(&s->allocating_write_reqs, NULL); + assert(s->allocating_acb =3D=3D NULL); + } + s->allocating_acb =3D acb; + return -EAGAIN; /* start over with looking up table entries */ } =20 acb->cur_nclusters =3D qed_bytes_to_clusters(s, @@ -1313,10 +1300,8 @@ static void qed_aio_next_io(QEDAIOCB *acb) ret =3D qed_aio_read_data(acb, ret, offset, len); } =20 - if (ret < 0) { - if (ret !=3D -EINPROGRESS) { - qed_aio_complete(acb, ret); - } + if (ret < 0 && ret !=3D -EAGAIN) { + qed_aio_complete(acb, ret); return; } } diff --git a/block/qed.h b/block/qed.h index 8644fed..37558e4 100644 --- a/block/qed.h +++ b/block/qed.h @@ -163,7 +163,8 @@ typedef struct { uint32_t l2_mask; =20 /* Allocating write request queue */ - QSIMPLEQ_HEAD(, QEDAIOCB) allocating_write_reqs; + QEDAIOCB *allocating_acb; + CoQueue allocating_write_reqs; bool allocating_write_reqs_plugged; =20 /* Periodic flush and clear need check flag */ --=20 1.8.3.1