From patchwork Fri Mar 21 07:09:16 2025 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: zoudongjie X-Patchwork-Id: 14024902 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 8435FC36000 for ; Fri, 21 Mar 2025 07:12:00 +0000 (UTC) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1tvWX8-0003Ne-66; Fri, 21 Mar 2025 03:11:10 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1tvWX0-0003MX-6G; Fri, 21 Mar 2025 03:11:02 -0400 Received: from szxga07-in.huawei.com ([45.249.212.35]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1tvWWw-00085R-8B; Fri, 21 Mar 2025 03:11:01 -0400 Received: from mail.maildlp.com (unknown [172.19.88.163]) by szxga07-in.huawei.com (SkyGuard) with ESMTP id 4ZJtlT1V9wz1f1L5; Fri, 21 Mar 2025 15:06:13 +0800 (CST) Received: from kwepemk500007.china.huawei.com (unknown [7.202.194.92]) by mail.maildlp.com (Postfix) with ESMTPS id A00AE1800B1; Fri, 21 Mar 2025 15:10:46 +0800 (CST) Received: from huawei.com (10.246.99.19) by kwepemk500007.china.huawei.com (7.202.194.92) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.2.1544.11; Fri, 21 Mar 2025 15:10:45 +0800 To: CC: , , , , , , , , , , , , , , , , , Subject: [PATCH v2 1/2] io/block: Refactoring the bdrv_drained_begin() function and implement a timeout mechanism. Date: Fri, 21 Mar 2025 15:09:16 +0800 Message-ID: <20250321070917.2889854-2-zoudongjie@huawei.com> X-Mailer: git-send-email 2.33.0 In-Reply-To: <20250321070917.2889854-1-zoudongjie@huawei.com> References: <20250321070917.2889854-1-zoudongjie@huawei.com> MIME-Version: 1.0 X-Originating-IP: [10.246.99.19] X-ClientProxiedBy: dggems706-chm.china.huawei.com (10.3.19.183) To kwepemk500007.china.huawei.com (7.202.194.92) Received-SPF: pass client-ip=45.249.212.35; envelope-from=zoudongjie@huawei.com; helo=szxga07-in.huawei.com X-Spam_score_int: -41 X-Spam_score: -4.2 X-Spam_bar: ---- X-Spam_report: (-4.2 / 5.0 requ) BAYES_00=-1.9, RCVD_IN_DNSWL_MED=-2.3, RCVD_IN_VALIDITY_CERTIFIED_BLOCKED=0.001, RCVD_IN_VALIDITY_RPBL_BLOCKED=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Reply-to: zoudongjie X-Patchwork-Original-From: zoudongjie via From: zoudongjie Errors-To: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Sender: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org From: Zhu Yangyang The bdrv_drained_begin() function is a blocking function. In scenarios where network storage is used and network links fail, it may block for a long time. Therefore, we add a timeout parameter to control the duration of the block. Since bdrv_drained_begin() has been widely adopted, both bdrv_drained_begin() and bdrv_drained_begin_timeout() will be retained. Signed-off-by: Zhu Yangyang --- block/io.c | 58 +++++++++++++++++++++++++++++++++------- include/block/aio-wait.h | 49 +++++++++++++++++++++++++++++++++ include/block/block-io.h | 22 ++++++++++++++- util/aio-wait.c | 5 ++++ 4 files changed, 123 insertions(+), 11 deletions(-) diff --git a/block/io.c b/block/io.c index 1ba8d1aeea..912b76c4a4 100644 --- a/block/io.c +++ b/block/io.c @@ -255,6 +255,8 @@ typedef struct { bool begin; bool poll; BdrvChild *parent; + uint64_t timeout_ns; + int ret; } BdrvCoDrainData; /* Returns true if BDRV_POLL_WHILE() should go into a blocking aio_poll() */ @@ -283,6 +285,10 @@ static bool bdrv_drain_poll_top_level(BlockDriverState *bs, return bdrv_drain_poll(bs, ignore_parent, false); } +static int bdrv_do_drained_begin_timeout(BlockDriverState *bs, + BdrvChild *parent, + bool poll, + uint64_t timeout_ns); static void bdrv_do_drained_begin(BlockDriverState *bs, BdrvChild *parent, bool poll); static void bdrv_do_drained_end(BlockDriverState *bs, BdrvChild *parent); @@ -296,7 +302,9 @@ static void bdrv_co_drain_bh_cb(void *opaque) if (bs) { bdrv_dec_in_flight(bs); if (data->begin) { - bdrv_do_drained_begin(bs, data->parent, data->poll); + data->ret = bdrv_do_drained_begin_timeout(bs, data->parent, + data->poll, + data->timeout_ns); } else { assert(!data->poll); bdrv_do_drained_end(bs, data->parent); @@ -310,10 +318,11 @@ static void bdrv_co_drain_bh_cb(void *opaque) aio_co_wake(co); } -static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs, - bool begin, - BdrvChild *parent, - bool poll) +static int coroutine_fn bdrv_co_yield_to_drain_timeout(BlockDriverState *bs, + bool begin, + BdrvChild *parent, + bool poll, + uint64_t timeout_ns) { BdrvCoDrainData data; Coroutine *self = qemu_coroutine_self(); @@ -329,6 +338,8 @@ static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs, .begin = begin, .parent = parent, .poll = poll, + .timeout_ns = timeout_ns, + .ret = 0 }; if (bs) { @@ -342,16 +353,27 @@ static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs, /* If we are resumed from some other event (such as an aio completion or a * timer callback), it is a bug in the caller that should be fixed. */ assert(data.done); + return data.ret; } -static void bdrv_do_drained_begin(BlockDriverState *bs, BdrvChild *parent, - bool poll) +static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs, + bool begin, + BdrvChild *parent, + bool poll) +{ + bdrv_co_yield_to_drain_timeout(bs, begin, parent, poll, 0); +} + +static int bdrv_do_drained_begin_timeout(BlockDriverState *bs, + BdrvChild *parent, + bool poll, + uint64_t timeout_ns) { IO_OR_GS_CODE(); if (qemu_in_coroutine()) { - bdrv_co_yield_to_drain(bs, true, parent, poll); - return; + return bdrv_co_yield_to_drain_timeout(bs, true, parent, poll, + timeout_ns); } GLOBAL_STATE_CODE(); @@ -375,8 +397,17 @@ static void bdrv_do_drained_begin(BlockDriverState *bs, BdrvChild *parent, * nodes. */ if (poll) { - BDRV_POLL_WHILE(bs, bdrv_drain_poll_top_level(bs, parent)); + return BDRV_POLL_WHILE_TIMEOUT(bs, + bdrv_drain_poll_top_level(bs, parent), + timeout_ns); } + return 0; +} + +static void bdrv_do_drained_begin(BlockDriverState *bs, BdrvChild *parent, + bool poll) +{ + bdrv_do_drained_begin_timeout(bs, parent, poll, 0); } void bdrv_do_drained_begin_quiesce(BlockDriverState *bs, BdrvChild *parent) @@ -391,6 +422,13 @@ bdrv_drained_begin(BlockDriverState *bs) bdrv_do_drained_begin(bs, NULL, true); } +int coroutine_mixed_fn +bdrv_drained_begin_timeout(BlockDriverState *bs, uint64_t timeout_ns) +{ + IO_OR_GS_CODE(); + return bdrv_do_drained_begin_timeout(bs, NULL, true, timeout_ns); +} + /** * This function does not poll, nor must any of its recursively called * functions. diff --git a/include/block/aio-wait.h b/include/block/aio-wait.h index cf5e8bde1c..24c31df620 100644 --- a/include/block/aio-wait.h +++ b/include/block/aio-wait.h @@ -99,6 +99,53 @@ extern AioWait global_aio_wait; qatomic_dec(&wait_->num_waiters); \ waited_; }) +/** + * AIO_WAIT_WHILE_TIMEOUT: + * + * Refer to the implementation of AIO_WAIT_WHILE_INTERNAL, + * the timeout parameter is added. + * @timeout_ns: maximum duration to wait, in nanoseconds, except the value + * is unsigned, 0 means infinite. + */ +#define AIO_WAIT_WHILE_TIMEOUT(ctx, cond, timeout_ns) ({ \ + int ret_ = 0; \ + uint64_t timeout_ = (timeout_ns); \ + AioWait *wait_ = &global_aio_wait; \ + AioContext *ctx_ = (ctx); \ + AioContext *current_ctx_ = NULL; \ + QEMUTimer timer_; \ + /* Increment wait_->num_waiters before evaluating cond. */ \ + qatomic_inc(&wait_->num_waiters); \ + /* Paired with smp_mb in aio_wait_kick(). */ \ + smp_mb__after_rmw(); \ + if (ctx_ && in_aio_context_home_thread(ctx_)) { \ + current_ctx_ = ctx_; \ + } else { \ + assert(qemu_get_current_aio_context() == \ + qemu_get_aio_context()); \ + current_ctx_ = qemu_get_aio_context(); \ + } \ + if (timeout_ > 0) { \ + timer_init_full(&timer_, ¤t_ctx_->tlg, \ + QEMU_CLOCK_REALTIME, \ + SCALE_NS, 0, aio_wait_timer_cb, NULL); \ + timer_mod_ns(&timer_, \ + qemu_clock_get_ns(QEMU_CLOCK_REALTIME) + \ + timeout_); \ + } \ + while ((cond)) { \ + aio_poll(current_ctx_, true); \ + if (timeout_ > 0 && !timer_pending(&timer_)) { \ + ret_ = -ETIMEDOUT; \ + break; \ + } \ + } \ + if (timeout_ > 0) { \ + timer_del(&timer_); \ + } \ + qatomic_dec(&wait_->num_waiters); \ + ret_; }) + #define AIO_WAIT_WHILE(ctx, cond) \ AIO_WAIT_WHILE_INTERNAL(ctx, cond) @@ -149,4 +196,6 @@ static inline bool in_aio_context_home_thread(AioContext *ctx) } } +void aio_wait_timer_cb(void *opaque); + #endif /* QEMU_AIO_WAIT_H */ diff --git a/include/block/block-io.h b/include/block/block-io.h index b49e0537dd..844e9cf350 100644 --- a/include/block/block-io.h +++ b/include/block/block-io.h @@ -354,6 +354,11 @@ bdrv_co_copy_range(BdrvChild *src, int64_t src_offset, AIO_WAIT_WHILE(bdrv_get_aio_context(bs_), \ cond); }) +#define BDRV_POLL_WHILE_TIMEOUT(bs, cond, timeout_ns) ({ \ + BlockDriverState *bs_ = (bs); \ + AIO_WAIT_WHILE_TIMEOUT(bdrv_get_aio_context(bs_), \ + cond, timeout_ns); }) + void bdrv_drain(BlockDriverState *bs); int co_wrapper_mixed_bdrv_rdlock @@ -432,7 +437,22 @@ bdrv_drain_poll(BlockDriverState *bs, BdrvChild *ignore_parent, void bdrv_drained_begin(BlockDriverState *bs); /** - * bdrv_do_drained_begin_quiesce: + * bdrv_drained_begin_timeout: + * + * Added timeout parameter for bdrv_drained_begin() to make a time limited. + * + * @timeout_ns: maximum duration to wait; 0 means infinite, equal to call + * bdrv_drained_begin(). + * + * Returns: 0 if succeeded; -ETIMEDOUT when a timeout occurs. + * + * Note: when the timeout fails, we've already begin aquiesced section, so we + * still need to call bdrv_drained_end() to end the quiescent section. + */ +int bdrv_drained_begin_timeout(BlockDriverState *bs, uint64_t timeout_ns); + +/** + * bdrv_do_drained_badegin_quiesce: * * Quiesces a BDS like bdrv_drained_begin(), but does not wait for already * running requests to complete. diff --git a/util/aio-wait.c b/util/aio-wait.c index b5336cf5fd..64c3714fb8 100644 --- a/util/aio-wait.c +++ b/util/aio-wait.c @@ -84,3 +84,8 @@ void aio_wait_bh_oneshot(AioContext *ctx, QEMUBHFunc *cb, void *opaque) aio_bh_schedule_oneshot(ctx, aio_wait_bh, &data); AIO_WAIT_WHILE_UNLOCKED(NULL, !data.done); } + +void aio_wait_timer_cb(void *opaque) +{ + /* The point is to make AIO_WAIT_WHILE_TIMEOUT()'s aio_poll() return */ +} From patchwork Fri Mar 21 07:09:17 2025 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: zoudongjie X-Patchwork-Id: 14024903 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id D282DC35FF3 for ; Fri, 21 Mar 2025 07:12:28 +0000 (UTC) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1tvWXT-0003QM-S8; Fri, 21 Mar 2025 03:11:32 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1tvWX0-0003Ma-Rx; Fri, 21 Mar 2025 03:11:03 -0400 Received: from szxga01-in.huawei.com ([45.249.212.187]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1tvWWy-000833-8Q; Fri, 21 Mar 2025 03:11:02 -0400 Received: from mail.maildlp.com (unknown [172.19.163.48]) by szxga01-in.huawei.com (SkyGuard) with ESMTP id 4ZJtmF2ngfzvWr0; Fri, 21 Mar 2025 15:06:53 +0800 (CST) Received: from kwepemk500007.china.huawei.com (unknown [7.202.194.92]) by mail.maildlp.com (Postfix) with ESMTPS id DE2F21800B3; Fri, 21 Mar 2025 15:10:47 +0800 (CST) Received: from huawei.com (10.246.99.19) by kwepemk500007.china.huawei.com (7.202.194.92) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.2.1544.11; Fri, 21 Mar 2025 15:10:46 +0800 To: CC: , , , , , , , , , , , , , , , , , Subject: [PATCH v2 2/2] qapi/throttle: add timeout parameter for qmp_block_set_io_throttle() Date: Fri, 21 Mar 2025 15:09:17 +0800 Message-ID: <20250321070917.2889854-3-zoudongjie@huawei.com> X-Mailer: git-send-email 2.33.0 In-Reply-To: <20250321070917.2889854-1-zoudongjie@huawei.com> References: <20250321070917.2889854-1-zoudongjie@huawei.com> MIME-Version: 1.0 X-Originating-IP: [10.246.99.19] X-ClientProxiedBy: dggems706-chm.china.huawei.com (10.3.19.183) To kwepemk500007.china.huawei.com (7.202.194.92) Received-SPF: pass client-ip=45.249.212.187; envelope-from=zoudongjie@huawei.com; helo=szxga01-in.huawei.com X-Spam_score_int: -41 X-Spam_score: -4.2 X-Spam_bar: ---- X-Spam_report: (-4.2 / 5.0 requ) BAYES_00=-1.9, RCVD_IN_DNSWL_MED=-2.3, RCVD_IN_VALIDITY_CERTIFIED_BLOCKED=0.001, RCVD_IN_VALIDITY_RPBL_BLOCKED=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Reply-to: zoudongjie X-Patchwork-Original-From: zoudongjie via From: zoudongjie Errors-To: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Sender: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org From: Zhu Yangyang Calling qmp_block_set_io_throttle() will be blocked for a long time when a network disk is configured and the network failure is just about to occur. Therefore, we add a timeout parameter for qmp_block_set_io_throttle to control its execution duration. The default value of timeout is 0, that is infinite wait, consistent with previous behavior. Signed-off-by: Zhu Yangyang Reviewed-by: Stefan Hajnoczi --- block/block-backend.c | 14 +++++++++++++- block/qapi-system.c | 10 +++++++++- include/system/block-backend-global-state.h | 1 + qapi/block-core.json | 5 ++++- 4 files changed, 27 insertions(+), 3 deletions(-) diff --git a/block/block-backend.c b/block/block-backend.c index a402db13f2..fc1554ea55 100644 --- a/block/block-backend.c +++ b/block/block-backend.c @@ -2679,20 +2679,32 @@ void blk_set_io_limits(BlockBackend *blk, ThrottleConfig *cfg) } void blk_io_limits_disable(BlockBackend *blk) +{ + blk_io_limits_disable_timeout(blk, 0); +} + +int blk_io_limits_disable_timeout(BlockBackend *blk, uint64_t timeout_ns) { BlockDriverState *bs = blk_bs(blk); ThrottleGroupMember *tgm = &blk->public.throttle_group_member; + int ret = 0; + assert(tgm->throttle_state); GLOBAL_STATE_CODE(); if (bs) { bdrv_ref(bs); - bdrv_drained_begin(bs); + ret = bdrv_drained_begin_timeout(bs, timeout_ns); + if (ret < 0) { + goto out; + } } throttle_group_unregister_tgm(tgm); +out: if (bs) { bdrv_drained_end(bs); bdrv_unref(bs); } + return ret; } /* should be called before blk_set_io_limits if a limit is set */ diff --git a/block/qapi-system.c b/block/qapi-system.c index 54b7409b2b..3b7ba0959c 100644 --- a/block/qapi-system.c +++ b/block/qapi-system.c @@ -423,6 +423,7 @@ void qmp_block_set_io_throttle(BlockIOThrottle *arg, Error **errp) ThrottleConfig cfg; BlockDriverState *bs; BlockBackend *blk; + uint64_t timeout_ns; blk = qmp_get_blk(arg->device, arg->id, errp); if (!blk) { @@ -444,6 +445,10 @@ void qmp_block_set_io_throttle(BlockIOThrottle *arg, Error **errp) cfg.buckets[THROTTLE_OPS_READ].avg = arg->iops_rd; cfg.buckets[THROTTLE_OPS_WRITE].avg = arg->iops_wr; + timeout_ns = 0; /* 0 means infinite */ + if (arg->has_timeout) { + timeout_ns = arg->timeout * NANOSECONDS_PER_SECOND; + } if (arg->has_bps_max) { cfg.buckets[THROTTLE_BPS_TOTAL].max = arg->bps_max; } @@ -502,7 +507,10 @@ void qmp_block_set_io_throttle(BlockIOThrottle *arg, Error **errp) blk_set_io_limits(blk, &cfg); } else if (blk_get_public(blk)->throttle_group_member.throttle_state) { /* If all throttling settings are set to 0, disable I/O limits */ - blk_io_limits_disable(blk); + if (blk_io_limits_disable_timeout(blk, timeout_ns) < 0) { + error_setg(errp, "Blk io limits disable timeout"); + return; + } } } diff --git a/include/system/block-backend-global-state.h b/include/system/block-backend-global-state.h index 35b5e8380b..8a2a6cbda4 100644 --- a/include/system/block-backend-global-state.h +++ b/include/system/block-backend-global-state.h @@ -110,6 +110,7 @@ int blk_probe_geometry(BlockBackend *blk, HDGeometry *geo); void blk_set_io_limits(BlockBackend *blk, ThrottleConfig *cfg); void blk_io_limits_disable(BlockBackend *blk); +int blk_io_limits_disable_timeout(BlockBackend *blk, uint64_t timeout_ns); void blk_io_limits_enable(BlockBackend *blk, const char *group); void blk_io_limits_update_group(BlockBackend *blk, const char *group); void blk_set_force_allow_inactivate(BlockBackend *blk); diff --git a/qapi/block-core.json b/qapi/block-core.json index b1937780e1..88ef593efd 100644 --- a/qapi/block-core.json +++ b/qapi/block-core.json @@ -2626,6 +2626,9 @@ # # @group: throttle group name (Since 2.4) # +# @timeout: In seconds. Timeout for block_set_io_throttle, +# 0 means no limit. Defaults to 0. (Since 10.0) +# # Features: # # @deprecated: Member @device is deprecated. Use @id instead. @@ -2642,7 +2645,7 @@ '*bps_max_length': 'int', '*bps_rd_max_length': 'int', '*bps_wr_max_length': 'int', '*iops_max_length': 'int', '*iops_rd_max_length': 'int', '*iops_wr_max_length': 'int', - '*iops_size': 'int', '*group': 'str' } } + '*iops_size': 'int', '*group': 'str', '*timeout': 'uint32'} } ## # @ThrottleLimits: