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 20C3EC38A2D for ; Tue, 25 Oct 2022 01:17:15 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S231529AbiJYBRN (ORCPT ); Mon, 24 Oct 2022 21:17:13 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:52824 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S231430AbiJYBQr (ORCPT ); Mon, 24 Oct 2022 21:16:47 -0400 Received: from us-smtp-delivery-124.mimecast.com (us-smtp-delivery-124.mimecast.com [170.10.129.124]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 736D9836EA for ; Mon, 24 Oct 2022 17:35:08 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1666658107; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: in-reply-to:in-reply-to:references:references; bh=NGsqal3yawtP1okvtIaH5vVGgtlRu2D64gOPqImXnSU=; b=KK7CuPw0yMJcmB9CK+l/5R+3zRyHPzrWVU/4cHG5/GxKHMbfLMNWhi/wV1Wz4oM69YlFhZ x8WsKKBJoDaZPn61AFjEGipyi35gpsNwsq0K+2JZk3ZCiERzbPigxe0UwTRTs5AozZLnQU YM51D1d36kGwj3HTB6z850VPJsc2UlQ= Received: from mimecast-mx02.redhat.com (mx3-rdu2.redhat.com [66.187.233.73]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id us-mta-277-neb12vkCOBCxQ9aH5www-w-1; Mon, 24 Oct 2022 20:35:04 -0400 X-MC-Unique: neb12vkCOBCxQ9aH5www-w-1 Received: from smtp.corp.redhat.com (int-mx02.intmail.prod.int.rdu2.redhat.com [10.11.54.2]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx02.redhat.com (Postfix) with ESMTPS id 6EED93C0D188; Tue, 25 Oct 2022 00:35:03 +0000 (UTC) Received: from T590 (ovpn-8-20.pek2.redhat.com [10.72.8.20]) by smtp.corp.redhat.com (Postfix) with ESMTPS id 7E94740C6E16; Tue, 25 Oct 2022 00:34:54 +0000 (UTC) Date: Tue, 25 Oct 2022 08:34:47 +0800 From: Ming Lei To: John Garry Cc: axboe@kernel.dk, linux-kernel@vger.kernel.org, linux-block@vger.kernel.org, hch@lst.de, Bart Van Assche Subject: Re: [PATCH] blk-mq: Properly init bios from blk_mq_alloc_request_hctx() Message-ID: References: <1666454846-11749-1-git-send-email-john.garry@huawei.com> <99c6ca81-746d-85f4-04d3-49d7a3de611b@huawei.com> <360c78dc-65ce-362f-389d-075f2259ce5b@huawei.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <360c78dc-65ce-362f-389d-075f2259ce5b@huawei.com> X-Scanned-By: MIMEDefang 3.1 on 10.11.54.2 Precedence: bulk List-ID: X-Mailing-List: linux-block@vger.kernel.org On Mon, Oct 24, 2022 at 05:56:15PM +0100, John Garry wrote: > On 24/10/2022 14:27, Ming Lei wrote: > > > > > - rq->bio = rq->biotail = NULL; > > > > This patch looks not good, why do you switch to initialize the three fields > > > > twice in fast path? > > > Can you please show me how these are initialized twice? > > blk_mq_bio_to_request() is one which setup these fields, then you add > > another one in blk_mq_rq_ctx_init(). > > ok, understood. > > > > > > If there is a real concern with this then we go with my original idea, which > > > was to copy the init method of blk_mq_alloc_request() (in > > > blk_mq_alloc_request_hctx()) > > > > > > > BTW, we know blk_mq_alloc_request_hctx() has big trouble, so please > > > > avoid to extend it to other use cases. > > > Yeah, I know this, > > Did you know the exact issue on nvme-tcp, nvme-rdma or nvme-fc maybe > > with blk_mq_alloc_request_hctx()? > > I thought that the original issue was an OoO bounds issue, fixed in > 14dc7a18. Now there is still some issue in the following link, which is > still unresolved as I understand: > https://lore.kernel.org/linux-block/5bd886f1-a7c6-b765-da29-777be0328bc2@grimberg.me/#t > > But I think that 14dc7a18 may still leave undesirable scenario: > - all cpus in HW queue cpumask may go offline after cpu_online_mask read in > blk_mq_alloc_request_hctx() and before we get the driver tag and set > rq->hctx Yeah. > > > > > > but sometimes we just need to allocate for a specific HW > > > queue... > > > > > > For my usecase of interest, it should not impact if the cpumask of the HW > > > queue goes offline after selecting the cpu in blk_mq_alloc_request_hctx(), > > > so any race is ok ... I think. > > > > > > However it should be still possible to make blk_mq_alloc_request_hctx() more > > > robust. How about using something like work_on_cpu_safe() to allocate and > > > execute the request with blk_mq_alloc_request() on a cpu associated with the > > > HW queue, such that we know the cpu is online and stays online until we > > > execute it? Or also extent to work_on_cpumask_safe() variant, so that we > > > don't need to try all cpus in the mask (to see if online)? > > But all cpus on this hctx->cpumask could become offline. > > If all hctx->cpumask are offline then we should not allocate a request and > this is acceptable. Maybe I am missing your point. As you saw, this API has the above problem too, but any one of CPUs may become online later, maybe just during blk_mq_alloc_request_hctx(), and it is easy to cause inconsistence. You didn't share your use case, but for nvme connection request, if it is 1:1 mapping, if any one of CPU becomes offline, the controller initialization could be failed, that isn't good from user viewpoint at all. Thanks, Ming