{"id":2225810,"url":"http://patchwork.ozlabs.org/api/1.2/patches/2225810/?format=json","web_url":"http://patchwork.ozlabs.org/project/qemu-devel/patch/20260421161132.99878-3-kwolf@redhat.com/","project":{"id":14,"url":"http://patchwork.ozlabs.org/api/1.2/projects/14/?format=json","name":"QEMU Development","link_name":"qemu-devel","list_id":"qemu-devel.nongnu.org","list_email":"qemu-devel@nongnu.org","web_url":"","scm_url":"","webscm_url":"","list_archive_url":"","list_archive_url_format":"","commit_url_format":""},"msgid":"<20260421161132.99878-3-kwolf@redhat.com>","list_archive_url":null,"date":"2026-04-21T16:11:27","name":"[2/7] block: Add blk_co_start/end_request() and BDRV_REQ_NO_QUEUE","commit_ref":null,"pull_url":null,"state":"new","archived":false,"hash":"f2697cf7167c0c8a5dedf06f2e2a55c783bf1676","submitter":{"id":2714,"url":"http://patchwork.ozlabs.org/api/1.2/people/2714/?format=json","name":"Kevin Wolf","email":"kwolf@redhat.com"},"delegate":null,"mbox":"http://patchwork.ozlabs.org/project/qemu-devel/patch/20260421161132.99878-3-kwolf@redhat.com/mbox/","series":[{"id":500847,"url":"http://patchwork.ozlabs.org/api/1.2/series/500847/?format=json","web_url":"http://patchwork.ozlabs.org/project/qemu-devel/list/?series=500847","date":"2026-04-21T16:11:27","name":"ide: Fix deadlock between TRIM and drain","version":1,"mbox":"http://patchwork.ozlabs.org/series/500847/mbox/"}],"comments":"http://patchwork.ozlabs.org/api/patches/2225810/comments/","check":"pending","checks":"http://patchwork.ozlabs.org/api/patches/2225810/checks/","tags":{},"related":[],"headers":{"Return-Path":"","X-Original-To":"incoming@patchwork.ozlabs.org","Delivered-To":"patchwork-incoming@legolas.ozlabs.org","Authentication-Results":["legolas.ozlabs.org;\n\tdkim=pass (1024-bit key;\n unprotected) header.d=redhat.com header.i=@redhat.com header.a=rsa-sha256\n header.s=mimecast20190719 header.b=LPzFuJe4;\n\tdkim-atps=neutral","legolas.ozlabs.org;\n spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org\n (client-ip=209.51.188.17; helo=lists1p.gnu.org;\n envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org;\n receiver=patchwork.ozlabs.org)"],"Received":["from lists1p.gnu.org (lists1p.gnu.org [209.51.188.17])\n\t(using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits))\n\t(No client certificate requested)\n\tby legolas.ozlabs.org (Postfix) with ESMTPS id 4g0S836Tv6z1yCv\n\tfor ; Wed, 22 Apr 2026 02:13:23 +1000 (AEST)","from localhost ([::1] helo=lists1p.gnu.org)\n\tby lists1p.gnu.org with esmtp (Exim 4.90_1)\n\t(envelope-from )\n\tid 1wFDi3-0001P4-IK; Tue, 21 Apr 2026 12:12:23 -0400","from eggs.gnu.org ([2001:470:142:3::10])\n by lists1p.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)\n (Exim 4.90_1) (envelope-from ) id 1wFDi0-0001LL-3a\n for qemu-devel@nongnu.org; Tue, 21 Apr 2026 12:12:20 -0400","from us-smtp-delivery-124.mimecast.com ([170.10.133.124])\n by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)\n (Exim 4.90_1) (envelope-from ) id 1wFDhu-0000FF-Va\n for qemu-devel@nongnu.org; Tue, 21 Apr 2026 12:12:19 -0400","from mx-prod-mc-05.mail-002.prod.us-west-2.aws.redhat.com\n (ec2-54-186-198-63.us-west-2.compute.amazonaws.com [54.186.198.63]) by\n relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.3,\n cipher=TLS_AES_256_GCM_SHA384) id us-mta-126-pWNiQMyoNfeU1BPgLa0LSA-1; Tue,\n 21 Apr 2026 12:12:07 -0400","from mx-prod-int-08.mail-002.prod.us-west-2.aws.redhat.com\n (mx-prod-int-08.mail-002.prod.us-west-2.aws.redhat.com [10.30.177.111])\n (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits)\n key-exchange X25519 server-signature RSA-PSS (2048 bits) server-digest\n SHA256)\n (No client certificate requested)\n by mx-prod-mc-05.mail-002.prod.us-west-2.aws.redhat.com (Postfix) with ESMTPS\n id 2DF961955F0B; Tue, 21 Apr 2026 16:12:06 +0000 (UTC)","from merkur.fritz.box (unknown [10.44.49.173])\n by mx-prod-int-08.mail-002.prod.us-west-2.aws.redhat.com (Postfix) with ESMTP\n id 59AAC180045E; Tue, 21 Apr 2026 16:12:04 +0000 (UTC)"],"DKIM-Signature":"v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com;\n s=mimecast20190719; t=1776787933;\n h=from:from:reply-to:subject:subject:date:date:message-id:message-id:\n to:to:cc:cc:mime-version:mime-version:\n content-transfer-encoding:content-transfer-encoding:\n in-reply-to:in-reply-to:references:references;\n bh=HaurCmmcFcdQqODCgdgVOtbs7i0XHZRM1ttP+Dd/wVI=;\n b=LPzFuJe4NvC0sYrkIkICyd48gBASD4agI+aYSpbtfcMDtJRnZ7HLn9CbFRbJjh3oSRkxV6\n H/8vzVWLy6gQBC8X5AMvh/pZr3gGT+b/NwM4te3FhrqN4mAF2n4kAncxApQiLhwDpros9p\n d5yfSQVGearvBt6uyj6Gn6YWRpCPZR0=","X-MC-Unique":"pWNiQMyoNfeU1BPgLa0LSA-1","X-Mimecast-MFC-AGG-ID":"pWNiQMyoNfeU1BPgLa0LSA_1776787926","From":"Kevin Wolf ","To":"qemu-block@nongnu.org","Cc":"kwolf@redhat.com, hreitz@redhat.com, jsnow@redhat.com,\n qemu-devel@nongnu.org, qemu-stable@nongnu.org","Subject":"[PATCH 2/7] block: Add blk_co_start/end_request() and\n BDRV_REQ_NO_QUEUE","Date":"Tue, 21 Apr 2026 18:11:27 +0200","Message-ID":"<20260421161132.99878-3-kwolf@redhat.com>","In-Reply-To":"<20260421161132.99878-1-kwolf@redhat.com>","References":"<20260421161132.99878-1-kwolf@redhat.com>","MIME-Version":"1.0","Content-Transfer-Encoding":"8bit","X-Scanned-By":"MIMEDefang 3.4.1 on 10.30.177.111","Received-SPF":"pass client-ip=170.10.133.124; envelope-from=kwolf@redhat.com;\n helo=us-smtp-delivery-124.mimecast.com","X-Spam_score_int":"-20","X-Spam_score":"-2.1","X-Spam_bar":"--","X-Spam_report":"(-2.1 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.001,\n DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1,\n RCVD_IN_DNSWL_NONE=-0.0001, SPF_HELO_PASS=-0.001,\n SPF_PASS=-0.001 autolearn=unavailable autolearn_force=no","X-Spam_action":"no action","X-BeenThere":"qemu-devel@nongnu.org","X-Mailman-Version":"2.1.29","Precedence":"list","List-Id":"qemu development ","List-Unsubscribe":",\n ","List-Archive":"","List-Post":"","List-Help":"","List-Subscribe":",\n ","Errors-To":"qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org","Sender":"qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org"},"content":"If a device uses blk_inc/dec_in_flight() in order to build macro\noperations that involve multiple requests for the block layer and that\nneed to be completed as a unit before the BlockBackend can be considered\ndrained, it sets the stage for a deadlock: When a drain is requested,\nthe inner request at the BlockBackend level will be queued in\nblk_wait_while_drained() and wait until the drained section ends, but at\nthe same time, drain_begin can only return if the whole macro operation\nat the device level has completed.\n\nIntroduce a new interface to allow implementing the logic correctly:\nInstead of queueing individual requests, blk_co_start_request() calls\nblk_wait_while_drained() once at the beginning. The individual requests\nmust then set BDRV_REQ_NO_QUEUE to avoid being queued and running into\nthe deadlock; being wrapped in blk_co_start/end_request() makes sure\nthat drain_begin waits for them and they don't sneak in when the\nBlockBackend is supposed to already be quiescent.\n\nSigned-off-by: Kevin Wolf \n---\n include/block/block-common.h | 11 ++++++++-\n include/system/block-backend-io.h | 2 ++\n block/block-backend.c | 38 +++++++++++++++++++++++--------\n 3 files changed, 41 insertions(+), 10 deletions(-)","diff":"diff --git a/include/block/block-common.h b/include/block/block-common.h\nindex c8c626daeaa..895ea175413 100644\n--- a/include/block/block-common.h\n+++ b/include/block/block-common.h\n@@ -215,8 +215,17 @@ typedef enum {\n */\n BDRV_REQ_NO_WAIT = 0x400,\n \n+ /*\n+ * Used between blk_co_start_request() and blk_end_request() to avoid\n+ * that the request waits in a drained BlockBackend until the drained\n+ * section ends. Waiting would cause a deadlock because drain waits for\n+ * blk_end_request() to be called, but the request never completes\n+ * because it waits for the drain to end.\n+ */\n+ BDRV_REQ_NO_QUEUE = 0x800,\n+\n /* Mask of valid flags */\n- BDRV_REQ_MASK = 0x7ff,\n+ BDRV_REQ_MASK = 0xfff,\n } BdrvRequestFlags;\n \n #define BDRV_O_NO_SHARE 0x0001 /* don't share permissions */\ndiff --git a/include/system/block-backend-io.h b/include/system/block-backend-io.h\nindex 6d5ac476fc0..0248c1c36e2 100644\n--- a/include/system/block-backend-io.h\n+++ b/include/system/block-backend-io.h\n@@ -71,6 +71,8 @@ BlockAIOCB *blk_aio_ioctl(BlockBackend *blk, unsigned long int req, void *buf,\n \n void blk_inc_in_flight(BlockBackend *blk);\n void blk_dec_in_flight(BlockBackend *blk);\n+void coroutine_fn blk_co_start_request(BlockBackend *blk);\n+void blk_end_request(BlockBackend *blk);\n \n bool coroutine_fn GRAPH_RDLOCK blk_co_is_inserted(BlockBackend *blk);\n bool co_wrapper_mixed_bdrv_rdlock blk_is_inserted(BlockBackend *blk);\ndiff --git a/block/block-backend.c b/block/block-backend.c\nindex 99446571201..ee00440e28d 100644\n--- a/block/block-backend.c\n+++ b/block/block-backend.c\n@@ -82,6 +82,7 @@ struct BlockBackend {\n QemuMutex queued_requests_lock; /* protects queued_requests */\n CoQueue queued_requests;\n bool disable_request_queuing; /* atomic */\n+ int start_request_count; /* atomic */\n \n VMChangeStateEntry *vmsh;\n bool force_allow_inactivate;\n@@ -1306,10 +1307,16 @@ bool blk_in_drain(BlockBackend *blk)\n }\n \n /* To be called between exactly one pair of blk_inc/dec_in_flight() */\n-static void coroutine_fn blk_wait_while_drained(BlockBackend *blk)\n+static void coroutine_fn blk_wait_while_drained(BlockBackend *blk,\n+ BdrvRequestFlags flags)\n {\n assert(blk->in_flight > 0);\n \n+ if (flags & BDRV_REQ_NO_QUEUE) {\n+ assert(qatomic_read(&blk->start_request_count));\n+ return;\n+ }\n+\n if (qatomic_read(&blk->quiesce_counter) &&\n !qatomic_read(&blk->disable_request_queuing)) {\n /*\n@@ -1335,7 +1342,7 @@ blk_co_do_preadv_part(BlockBackend *blk, int64_t offset, int64_t bytes,\n BlockDriverState *bs;\n IO_CODE();\n \n- blk_wait_while_drained(blk);\n+ blk_wait_while_drained(blk, flags);\n GRAPH_RDLOCK_GUARD();\n \n /* Call blk_bs() only after waiting, the graph may have changed */\n@@ -1410,7 +1417,7 @@ blk_co_do_pwritev_part(BlockBackend *blk, int64_t offset, int64_t bytes,\n BlockDriverState *bs;\n IO_CODE();\n \n- blk_wait_while_drained(blk);\n+ blk_wait_while_drained(blk, flags);\n GRAPH_RDLOCK_GUARD();\n \n /* Call blk_bs() only after waiting, the graph may have changed */\n@@ -1523,6 +1530,19 @@ void blk_dec_in_flight(BlockBackend *blk)\n aio_wait_kick();\n }\n \n+void coroutine_fn blk_co_start_request(BlockBackend *blk)\n+{\n+ blk_inc_in_flight(blk);\n+ blk_wait_while_drained(blk, 0);\n+ qatomic_inc(&blk->start_request_count);\n+}\n+\n+void blk_end_request(BlockBackend *blk)\n+{\n+ qatomic_dec(&blk->start_request_count);\n+ blk_dec_in_flight(blk);\n+}\n+\n static void error_callback_bh(void *opaque)\n {\n struct BlockBackendAIOCB *acb = opaque;\n@@ -1741,7 +1761,7 @@ blk_co_do_ioctl(BlockBackend *blk, unsigned long int req, void *buf)\n {\n IO_CODE();\n \n- blk_wait_while_drained(blk);\n+ blk_wait_while_drained(blk, 0);\n GRAPH_RDLOCK_GUARD();\n \n if (!blk_co_is_available(blk)) {\n@@ -1788,7 +1808,7 @@ blk_co_do_pdiscard(BlockBackend *blk, int64_t offset, int64_t bytes)\n int ret;\n IO_CODE();\n \n- blk_wait_while_drained(blk);\n+ blk_wait_while_drained(blk, 0);\n GRAPH_RDLOCK_GUARD();\n \n ret = blk_check_byte_request(blk, offset, bytes);\n@@ -1834,7 +1854,7 @@ int coroutine_fn blk_co_pdiscard(BlockBackend *blk, int64_t offset,\n static int coroutine_fn blk_co_do_flush(BlockBackend *blk)\n {\n IO_CODE();\n- blk_wait_while_drained(blk);\n+ blk_wait_while_drained(blk, 0);\n GRAPH_RDLOCK_GUARD();\n \n if (!blk_co_is_available(blk)) {\n@@ -2009,7 +2029,7 @@ int coroutine_fn blk_co_zone_report(BlockBackend *blk, int64_t offset,\n IO_CODE();\n \n blk_inc_in_flight(blk); /* increase before waiting */\n- blk_wait_while_drained(blk);\n+ blk_wait_while_drained(blk, 0);\n GRAPH_RDLOCK_GUARD();\n if (!blk_is_available(blk)) {\n blk_dec_in_flight(blk);\n@@ -2034,7 +2054,7 @@ int coroutine_fn blk_co_zone_mgmt(BlockBackend *blk, BlockZoneOp op,\n IO_CODE();\n \n blk_inc_in_flight(blk);\n- blk_wait_while_drained(blk);\n+ blk_wait_while_drained(blk, 0);\n GRAPH_RDLOCK_GUARD();\n \n ret = blk_check_byte_request(blk, offset, len);\n@@ -2058,7 +2078,7 @@ int coroutine_fn blk_co_zone_append(BlockBackend *blk, int64_t *offset,\n IO_CODE();\n \n blk_inc_in_flight(blk);\n- blk_wait_while_drained(blk);\n+ blk_wait_while_drained(blk, flags);\n GRAPH_RDLOCK_GUARD();\n if (!blk_is_available(blk)) {\n blk_dec_in_flight(blk);\n","prefixes":["2/7"]}