From: Christoph Hellwig <hch@infradead.org> To: Mike Christie <michael.christie@oracle.com> Cc: Chaitanya.Kulkarni@wdc.com, loberman@redhat.com, martin.petersen@oracle.com, linux-scsi@vger.kernel.org, target-devel@vger.kernel.org, mst@redhat.com, stefanha@redhat.com, virtualization@lists.linux-foundation.org Subject: Re: [PATCH 11/13] target: replace work per cmd in completion path Date: Wed, 10 Feb 2021 08:42:41 +0000 [thread overview] Message-ID: <20210210084241.GE2045726@infradead.org> (raw) In-Reply-To: <20210209123845.4856-12-michael.christie@oracle.com> On Tue, Feb 09, 2021 at 06:38:43AM -0600, Mike Christie wrote: > Doing a work per cmd can lead to lots of threads being created. > This patch just replaces the completion work per cmd with a per cpu > list. Combined with the first patches this allows tcm loop on top of > initiators like iser to go from around 700K IOPs to 1000K and reduces > the number of threads that get created when the system is under heavy > load and hitting the initiator drivers tagging limits. OTOH it does increase completion latency, which might be the preference for some workloads. Do we need a tunable here? > +static void target_queue_cmd_work(struct se_cmd_queue *q, struct se_cmd *se_cmd, > + int cpu, struct workqueue_struct *wq) > { > - struct se_cmd *cmd = container_of(work, struct se_cmd, work); > + llist_add(&se_cmd->se_cmd_list, &q->cmd_list); > + queue_work_on(cpu, wq, &q->work); > +} Do we need this helper at all? Having it open coded in the two callers would seem easier to follow to me.
WARNING: multiple messages have this Message-ID (diff)
From: Christoph Hellwig <hch@infradead.org> To: Mike Christie <michael.christie@oracle.com> Cc: loberman@redhat.com, Chaitanya.Kulkarni@wdc.com, martin.petersen@oracle.com, mst@redhat.com, virtualization@lists.linux-foundation.org, target-devel@vger.kernel.org, stefanha@redhat.com, linux-scsi@vger.kernel.org Subject: Re: [PATCH 11/13] target: replace work per cmd in completion path Date: Wed, 10 Feb 2021 08:42:41 +0000 [thread overview] Message-ID: <20210210084241.GE2045726@infradead.org> (raw) In-Reply-To: <20210209123845.4856-12-michael.christie@oracle.com> On Tue, Feb 09, 2021 at 06:38:43AM -0600, Mike Christie wrote: > Doing a work per cmd can lead to lots of threads being created. > This patch just replaces the completion work per cmd with a per cpu > list. Combined with the first patches this allows tcm loop on top of > initiators like iser to go from around 700K IOPs to 1000K and reduces > the number of threads that get created when the system is under heavy > load and hitting the initiator drivers tagging limits. OTOH it does increase completion latency, which might be the preference for some workloads. Do we need a tunable here? > +static void target_queue_cmd_work(struct se_cmd_queue *q, struct se_cmd *se_cmd, > + int cpu, struct workqueue_struct *wq) > { > - struct se_cmd *cmd = container_of(work, struct se_cmd, work); > + llist_add(&se_cmd->se_cmd_list, &q->cmd_list); > + queue_work_on(cpu, wq, &q->work); > +} Do we need this helper at all? Having it open coded in the two callers would seem easier to follow to me. _______________________________________________ Virtualization mailing list Virtualization@lists.linux-foundation.org https://lists.linuxfoundation.org/mailman/listinfo/virtualization
next prev parent reply other threads:[~2021-02-10 8:49 UTC|newest] Thread overview: 50+ messages / expand[flat|nested] mbox.gz Atom feed top 2021-02-09 12:38 [PATCH 00/12 V2] target: fix cmd plugging and completion Mike Christie 2021-02-09 12:38 ` [PATCH 01/13] target: move t_task_cdb initialization Mike Christie 2021-02-10 8:35 ` Christoph Hellwig 2021-02-10 8:35 ` Christoph Hellwig 2021-02-09 12:38 ` [PATCH 02/13] target: split target_submit_cmd_map_sgls Mike Christie 2021-02-09 16:15 ` kernel test robot 2021-02-09 16:15 ` kernel test robot 2021-02-09 16:15 ` kernel test robot 2021-02-09 18:40 ` Mike Christie 2021-02-09 18:40 ` Mike Christie 2021-02-10 8:36 ` Christoph Hellwig 2021-02-10 8:36 ` Christoph Hellwig 2021-02-09 12:38 ` [PATCH 03/13] target: add workqueue based cmd submission Mike Christie 2021-02-09 15:48 ` Bodo Stroesser 2021-02-09 18:43 ` Mike Christie 2021-02-09 19:10 ` Mike Christie 2021-02-09 12:38 ` [PATCH 04/13] vhost scsi: use lio wq cmd submission helper Mike Christie 2021-02-09 12:38 ` [PATCH 05/13] tcm loop: use blk cmd allocator for se_cmds Mike Christie 2021-02-10 8:37 ` Christoph Hellwig 2021-02-10 8:37 ` Christoph Hellwig 2021-02-09 12:38 ` [PATCH 06/13] tcm loop: use lio wq cmd submission helper Mike Christie 2021-02-09 15:59 ` Bodo Stroesser 2021-02-09 18:44 ` Mike Christie 2021-02-09 17:39 ` kernel test robot 2021-02-09 17:39 ` kernel test robot 2021-02-09 12:38 ` [PATCH 07/13] target: cleanup cmd flag bits Mike Christie 2021-02-10 8:38 ` Christoph Hellwig 2021-02-10 8:38 ` Christoph Hellwig 2021-02-09 12:38 ` [PATCH 08/13] target: fix backend plugging Mike Christie 2021-02-09 12:38 ` [PATCH 09/13] target iblock: add backend plug/unplug callouts Mike Christie 2021-02-09 12:38 ` [PATCH 10/13] target_core_user: " Mike Christie 2021-02-09 16:32 ` Bodo Stroesser 2021-02-09 18:59 ` Mike Christie 2021-02-09 12:38 ` [PATCH 11/13] target: replace work per cmd in completion path Mike Christie 2021-02-09 17:01 ` Bodo Stroesser 2021-02-09 18:50 ` Mike Christie 2021-02-10 8:42 ` Christoph Hellwig [this message] 2021-02-10 8:42 ` Christoph Hellwig 2021-02-10 18:33 ` Mike Christie 2021-02-09 12:38 ` [PATCH 12/13] target, vhost-scsi: don't switch cpus on completion Mike Christie 2021-02-10 8:44 ` Christoph Hellwig 2021-02-10 8:44 ` Christoph Hellwig 2021-02-10 18:43 ` Mike Christie 2021-02-10 18:45 ` Mike Christie 2021-02-09 12:38 ` [PATCH 13/13] target: flush submission work during TMR processing Mike Christie 2021-02-09 14:31 ` Laurence Oberman 2021-02-10 14:25 ` Laurence Oberman 2021-02-09 17:05 ` Bodo Stroesser 2021-02-09 18:49 ` Mike Christie 2021-02-10 4:55 [PATCH 00/13 V3] target: fix cmd plugging and completion Mike Christie 2021-02-10 4:55 ` [PATCH 11/13] target: replace work per cmd in completion path Mike Christie
Reply instructions: You may reply publicly to this message via plain-text email using any one of the following methods: * Save the following mbox file, import it into your mail client, and reply-to-all from there: mbox Avoid top-posting and favor interleaved quoting: https://en.wikipedia.org/wiki/Posting_style#Interleaved_style * Reply using the --to, --cc, and --in-reply-to switches of git-send-email(1): git send-email \ --in-reply-to=20210210084241.GE2045726@infradead.org \ --to=hch@infradead.org \ --cc=Chaitanya.Kulkarni@wdc.com \ --cc=linux-scsi@vger.kernel.org \ --cc=loberman@redhat.com \ --cc=martin.petersen@oracle.com \ --cc=michael.christie@oracle.com \ --cc=mst@redhat.com \ --cc=stefanha@redhat.com \ --cc=target-devel@vger.kernel.org \ --cc=virtualization@lists.linux-foundation.org \ /path/to/YOUR_REPLY https://kernel.org/pub/software/scm/git/docs/git-send-email.html * If your mail client supports setting the In-Reply-To header via mailto: links, try the mailto: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.