From mboxrd@z Thu Jan 1 00:00:00 1970 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 vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id E71CEC433F5 for ; Fri, 15 Apr 2022 05:18:52 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S236511AbiDOFVQ (ORCPT ); Fri, 15 Apr 2022 01:21:16 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:43928 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S233425AbiDOFVP (ORCPT ); Fri, 15 Apr 2022 01:21:15 -0400 Received: from verein.lst.de (verein.lst.de [213.95.11.211]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id A2F351AD8C; Thu, 14 Apr 2022 22:18:48 -0700 (PDT) Received: by verein.lst.de (Postfix, from userid 2407) id 1808D68B05; Fri, 15 Apr 2022 07:18:45 +0200 (CEST) Date: Fri, 15 Apr 2022 07:18:44 +0200 From: Christoph Hellwig To: Ming Lei Cc: Jens Axboe , linux-block@vger.kernel.org, Christoph Hellwig , linux-mm@kvack.org, linux-xfs@vger.kernel.org, Changhui Zhong Subject: Re: [PATCH V2] block: avoid io timeout in case of sync polled dio Message-ID: <20220415051844.GA22762@lst.de> References: <20220415034703.2081695-1-ming.lei@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline Content-Transfer-Encoding: 8bit In-Reply-To: <20220415034703.2081695-1-ming.lei@redhat.com> User-Agent: Mutt/1.5.17 (2007-11-01) Precedence: bulk List-ID: X-Mailing-List: linux-xfs@vger.kernel.org On Fri, Apr 15, 2022 at 11:47:03AM +0800, Ming Lei wrote: > + /* make sure the bio is issued before polling */ > + if (bio.bi_opf & REQ_POLLED) > + blk_flush_plug(current->plug, false); I still think the core code should handle this. Without that we'd need to export the blk_flush_plug for anything that would want to poll bios from modules, in addition to it generally being a mess. See a proposed patch for that below. I'd also split the flush aspect from the poll aspect into two patches. > + if (bio.bi_opf & REQ_POLLED) > + bio_poll(&bio, NULL, 0); > + else > blk_io_schedule(); Instead of this duplicate logic everywhere I'd just make bio_boll call blk_io_schedule for the !REQ_POLLED case and simplify all the callers. > + if (dio->submit.poll_bio && > + (dio->submit.poll_bio->bi_opf & > + REQ_POLLED)) This indentation looks awfull,normal would be: if (dio->submit.poll_bio && (dio->submit.poll_bio->bi_opf & REQ_POLLED)) --- >From 08ff61b0142eb708fc384cf867c72175561d974a Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Fri, 15 Apr 2022 07:15:42 +0200 Subject: blk-mq: don't plug for synchronously polled requests For synchronous polling to work, the bio must be issued to the driver from the submit_bio call, otherwise ->bi_cookie won't be set. Based on a patch from Ming Lei. Reported-by: Changhui Zhong Signed-off-by: Christoph Hellwig --- block/blk-mq.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/block/blk-mq.c b/block/blk-mq.c index ed3ed86f7dd24..bcc7e3d11296c 100644 --- a/block/blk-mq.c +++ b/block/blk-mq.c @@ -2851,7 +2851,13 @@ void blk_mq_submit_bio(struct bio *bio) return; } - if (plug) + /* + * We can't plug for synchronously polled submissions, otherwise + * bio->bi_cookie won't be set directly after submission, which is the + * indicator used by the submitter to check if a bio needs polling. + */ + if (plug && + (rq->bio->bi_opf & (REQ_POLLED | REQ_NOWAIT)) != REQ_POLLED) blk_add_rq_to_plug(plug, rq); else if ((rq->rq_flags & RQF_ELV) || (rq->mq_hctx->dispatch_busy && -- 2.30.2