Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751471AbbEZWaX (ORCPT ); Tue, 26 May 2015 18:30:23 -0400 Received: from userp1040.oracle.com ([156.151.31.81]:50428 "EHLO userp1040.oracle.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751300AbbEZWaR (ORCPT ); Tue, 26 May 2015 18:30:17 -0400 Message-ID: <5564F3C9.70707@oracle.com> Date: Tue, 26 May 2015 17:29:29 -0500 From: Dave Kleikamp User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:31.0) Gecko/20100101 Thunderbird/31.7.0 MIME-Version: 1.0 To: Ming Lei , Jens Axboe , linux-kernel@vger.kernel.org CC: "Justin M. Forbes" , Jeff Moyer , Tejun Heo , Christoph Hellwig , "v4.0" Subject: Re: [PATCH 1/2] block: loop: convert to per-device workqueue References: <1430826595-5888-1-git-send-email-ming.lei@canonical.com> <1430826595-5888-2-git-send-email-ming.lei@canonical.com> In-Reply-To: <1430826595-5888-2-git-send-email-ming.lei@canonical.com> Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: 7bit X-Source-IP: aserv0021.oracle.com [141.146.126.233] Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 4351 Lines: 138 On 05/05/2015 06:49 AM, Ming Lei wrote: > Documentation/workqueue.txt: > If there is dependency among multiple work items used > during memory reclaim, they should be queued to separate > wq each with WQ_MEM_RECLAIM. > > Loop devices can be stacked, so we have to convert to per-device > workqueue. One example is Fedora live CD. I'm just now looking at this, but I found a minor nit. I'm not sure if it's worth replacing the patch or just fixing it with a trivial cleanup patch. > > Fixes: b5dd2f6047ca108001328aac0e8588edd15f1778 > Cc: stable@vger.kernel.org (v4.0) > Cc: Justin M. Forbes > Signed-off-by: Ming Lei > --- > drivers/block/loop.c | 30 ++++++++++++++---------------- > drivers/block/loop.h | 1 + > 2 files changed, 15 insertions(+), 16 deletions(-) > > diff --git a/drivers/block/loop.c b/drivers/block/loop.c > index ae3fcb4..3dc1598 100644 > --- a/drivers/block/loop.c > +++ b/drivers/block/loop.c > @@ -86,8 +86,6 @@ static DEFINE_MUTEX(loop_index_mutex); > static int max_part; > static int part_shift; > > -static struct workqueue_struct *loop_wq; > - > static int transfer_xor(struct loop_device *lo, int cmd, > struct page *raw_page, unsigned raw_off, > struct page *loop_page, unsigned loop_off, > @@ -725,6 +723,12 @@ static int loop_set_fd(struct loop_device *lo, fmode_t mode, > size = get_loop_size(lo, file); > if ((loff_t)(sector_t)size != size) > goto out_putf; > + error = -ENOMEM; > + lo->wq = alloc_workqueue("kloopd%d", > + WQ_MEM_RECLAIM | WQ_HIGHPRI | WQ_UNBOUND, 0, > + lo->lo_number); > + if (!lo->wq) > + goto out_putf; > > error = 0; > > @@ -872,6 +876,8 @@ static int loop_clr_fd(struct loop_device *lo) > lo->lo_flags = 0; > if (!part_shift) > lo->lo_disk->flags |= GENHD_FL_NO_PART_SCAN; > + destroy_workqueue(lo->wq); > + lo->wq = NULL; > mutex_unlock(&lo->lo_ctl_mutex); > /* > * Need not hold lo_ctl_mutex to fput backing file. > @@ -1425,9 +1431,13 @@ static int loop_queue_rq(struct blk_mq_hw_ctx *hctx, > const struct blk_mq_queue_data *bd) > { > struct loop_cmd *cmd = blk_mq_rq_to_pdu(bd->rq); > + struct loop_device *lo = cmd->rq->q->queuedata; > > blk_mq_start_request(bd->rq); > > + if (lo->lo_state != Lo_bound) > + return -EIO; > + > if (cmd->rq->cmd_flags & REQ_WRITE) { > struct loop_device *lo = cmd->rq->q->queuedata; The above definition of lo becomes redundant now. > bool need_sched = true; > @@ -1441,9 +1451,9 @@ static int loop_queue_rq(struct blk_mq_hw_ctx *hctx, > spin_unlock_irq(&lo->lo_lock); > > if (need_sched) > - queue_work(loop_wq, &lo->write_work); > + queue_work(lo->wq, &lo->write_work); > } else { > - queue_work(loop_wq, &cmd->read_work); > + queue_work(lo->wq, &cmd->read_work); > } > > return BLK_MQ_RQ_QUEUE_OK; > @@ -1455,9 +1465,6 @@ static void loop_handle_cmd(struct loop_cmd *cmd) > struct loop_device *lo = cmd->rq->q->queuedata; > int ret = -EIO; > > - if (lo->lo_state != Lo_bound) > - goto failed; > - > if (write && (lo->lo_flags & LO_FLAGS_READ_ONLY)) > goto failed; > > @@ -1806,13 +1813,6 @@ static int __init loop_init(void) > goto misc_out; > } > > - loop_wq = alloc_workqueue("kloopd", > - WQ_MEM_RECLAIM | WQ_HIGHPRI | WQ_UNBOUND, 0); > - if (!loop_wq) { > - err = -ENOMEM; > - goto misc_out; > - } > - > blk_register_region(MKDEV(LOOP_MAJOR, 0), range, > THIS_MODULE, loop_probe, NULL, NULL); > > @@ -1850,8 +1850,6 @@ static void __exit loop_exit(void) > blk_unregister_region(MKDEV(LOOP_MAJOR, 0), range); > unregister_blkdev(LOOP_MAJOR, "loop"); > > - destroy_workqueue(loop_wq); > - > misc_deregister(&loop_misc); > } > > diff --git a/drivers/block/loop.h b/drivers/block/loop.h > index 301c27f..49564ed 100644 > --- a/drivers/block/loop.h > +++ b/drivers/block/loop.h > @@ -54,6 +54,7 @@ struct loop_device { > gfp_t old_gfp_mask; > > spinlock_t lo_lock; > + struct workqueue_struct *wq; > struct list_head write_cmd_head; > struct work_struct write_work; > bool write_started; > -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/