From patchwork Thu Dec 10 14:46:06 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Brian Foster X-Patchwork-Id: 11965119 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-15.8 required=3.0 tests=BAYES_00,DKIMWL_WL_HIGH, DKIM_SIGNED,DKIM_VALID,DKIM_VALID_AU,HEADER_FROM_DIFFERENT_DOMAINS, INCLUDES_CR_TRAILER,INCLUDES_PATCH,MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id D95D5C433FE for ; Thu, 10 Dec 2020 15:10:25 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id 9C48123BE5 for ; Thu, 10 Dec 2020 15:10:25 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S2388251AbgLJPKR (ORCPT ); Thu, 10 Dec 2020 10:10:17 -0500 Received: from us-smtp-delivery-124.mimecast.com ([63.128.21.124]:46462 "EHLO us-smtp-delivery-124.mimecast.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S2391659AbgLJOrm (ORCPT ); Thu, 10 Dec 2020 09:47:42 -0500 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1607611574; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=fKNomnAFTTOKrWab3r7X/KDNNbCd6qD9rsgGVqV2YyM=; b=Ru3g/DV+oDGSLEZCNtt6vKZf3IQFXEi7wy1HWP2UnOVheX42RbnVLdDTOF6ifTe6y+/JLp 3ZkYKtlW8tN0jLtp+wgM8IaTRlwWjHzrL+u8IYwCZg85GasKPtlZRrV453jqkFgwCgYkdO HgXYPPqu8iXzxcISI+HyEpSnaU8/XFg= Received: from mimecast-mx01.redhat.com (mimecast-mx01.redhat.com [209.132.183.4]) (Using TLS) by relay.mimecast.com with ESMTP id us-mta-276-smY8caPMPyG6Pr2cE1JfWA-1; Thu, 10 Dec 2020 09:46:12 -0500 X-MC-Unique: smY8caPMPyG6Pr2cE1JfWA-1 Received: from smtp.corp.redhat.com (int-mx08.intmail.prod.int.phx2.redhat.com [10.5.11.23]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx01.redhat.com (Postfix) with ESMTPS id 69DE1802EA6 for ; Thu, 10 Dec 2020 14:46:11 +0000 (UTC) Received: from bfoster.redhat.com (ovpn-112-184.rdu2.redhat.com [10.10.112.184]) by smtp.corp.redhat.com (Postfix) with ESMTP id 0E4B719714 for ; Thu, 10 Dec 2020 14:46:10 +0000 (UTC) From: Brian Foster To: linux-xfs@vger.kernel.org Subject: [PATCH 1/2] xfs: rename xfs_wait_buftarg() to xfs_buftarg_drain() Date: Thu, 10 Dec 2020 09:46:06 -0500 Message-Id: <20201210144607.1922026-2-bfoster@redhat.com> In-Reply-To: <20201210144607.1922026-1-bfoster@redhat.com> References: <20201210144607.1922026-1-bfoster@redhat.com> MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.84 on 10.5.11.23 Precedence: bulk List-ID: X-Mailing-List: linux-xfs@vger.kernel.org xfs_wait_buftarg() is vaguely named and somewhat overloaded. Its primary purpose is to reclaim all buffers from the provided buffer target LRU. In preparation to refactor xfs_wait_buftarg() into serialization and LRU draining components, rename the function and associated helpers to something more descriptive. This patch has no functional changes with the minor exception of renaming a tracepoint. Signed-off-by: Brian Foster Reviewed-by: Chandan Babu R Reviewed-by: Darrick J. Wong --- fs/xfs/xfs_buf.c | 12 ++++++------ fs/xfs/xfs_buf.h | 10 +++++----- fs/xfs/xfs_log.c | 6 +++--- fs/xfs/xfs_mount.c | 4 ++-- fs/xfs/xfs_trace.h | 2 +- 5 files changed, 17 insertions(+), 17 deletions(-) diff --git a/fs/xfs/xfs_buf.c b/fs/xfs/xfs_buf.c index 4e4cf91f4f9f..db918ed20c40 100644 --- a/fs/xfs/xfs_buf.c +++ b/fs/xfs/xfs_buf.c @@ -43,7 +43,7 @@ static kmem_zone_t *xfs_buf_zone; * pag_buf_lock * lru_lock * - * xfs_buftarg_wait_rele + * xfs_buftarg_drain_rele * lru_lock * b_lock (trylock due to inversion) * @@ -88,7 +88,7 @@ xfs_buf_vmap_len( * because the corresponding decrement is deferred to buffer release. Buffers * can undergo I/O multiple times in a hold-release cycle and per buffer I/O * tracking adds unnecessary overhead. This is used for sychronization purposes - * with unmount (see xfs_wait_buftarg()), so all we really need is a count of + * with unmount (see xfs_buftarg_drain()), so all we really need is a count of * in-flight buffers. * * Buffers that are never released (e.g., superblock, iclog buffers) must set @@ -1786,7 +1786,7 @@ __xfs_buf_mark_corrupt( * while freeing all the buffers only held by the LRU. */ static enum lru_status -xfs_buftarg_wait_rele( +xfs_buftarg_drain_rele( struct list_head *item, struct list_lru_one *lru, spinlock_t *lru_lock, @@ -1798,7 +1798,7 @@ xfs_buftarg_wait_rele( if (atomic_read(&bp->b_hold) > 1) { /* need to wait, so skip it this pass */ - trace_xfs_buf_wait_buftarg(bp, _RET_IP_); + trace_xfs_buf_drain_buftarg(bp, _RET_IP_); return LRU_SKIP; } if (!spin_trylock(&bp->b_lock)) @@ -1816,7 +1816,7 @@ xfs_buftarg_wait_rele( } void -xfs_wait_buftarg( +xfs_buftarg_drain( struct xfs_buftarg *btp) { LIST_HEAD(dispose); @@ -1841,7 +1841,7 @@ xfs_wait_buftarg( /* loop until there is nothing left on the lru list. */ while (list_lru_count(&btp->bt_lru)) { - list_lru_walk(&btp->bt_lru, xfs_buftarg_wait_rele, + list_lru_walk(&btp->bt_lru, xfs_buftarg_drain_rele, &dispose, LONG_MAX); while (!list_empty(&dispose)) { diff --git a/fs/xfs/xfs_buf.h b/fs/xfs/xfs_buf.h index bfd2907e7bc4..ea32369f8f77 100644 --- a/fs/xfs/xfs_buf.h +++ b/fs/xfs/xfs_buf.h @@ -152,7 +152,7 @@ typedef struct xfs_buf { struct list_head b_list; struct xfs_perag *b_pag; /* contains rbtree root */ struct xfs_mount *b_mount; - xfs_buftarg_t *b_target; /* buffer target (device) */ + struct xfs_buftarg *b_target; /* buffer target (device) */ void *b_addr; /* virtual address of buffer */ struct work_struct b_ioend_work; struct completion b_iowait; /* queue for I/O waiters */ @@ -344,11 +344,11 @@ xfs_buf_update_cksum(struct xfs_buf *bp, unsigned long cksum_offset) /* * Handling of buftargs. */ -extern xfs_buftarg_t *xfs_alloc_buftarg(struct xfs_mount *, - struct block_device *, struct dax_device *); +extern struct xfs_buftarg *xfs_alloc_buftarg(struct xfs_mount *, + struct block_device *, struct dax_device *); extern void xfs_free_buftarg(struct xfs_buftarg *); -extern void xfs_wait_buftarg(xfs_buftarg_t *); -extern int xfs_setsize_buftarg(xfs_buftarg_t *, unsigned int); +extern void xfs_buftarg_drain(struct xfs_buftarg *); +extern int xfs_setsize_buftarg(struct xfs_buftarg *, unsigned int); #define xfs_getsize_buftarg(buftarg) block_size((buftarg)->bt_bdev) #define xfs_readonly_buftarg(buftarg) bdev_read_only((buftarg)->bt_bdev) diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c index fa2d05e65ff1..5ad4d5e78019 100644 --- a/fs/xfs/xfs_log.c +++ b/fs/xfs/xfs_log.c @@ -741,7 +741,7 @@ xfs_log_mount_finish( xfs_log_force(mp, XFS_LOG_SYNC); xfs_ail_push_all_sync(mp->m_ail); } - xfs_wait_buftarg(mp->m_ddev_targp); + xfs_buftarg_drain(mp->m_ddev_targp); if (readonly) mp->m_flags |= XFS_MOUNT_RDONLY; @@ -936,13 +936,13 @@ xfs_log_quiesce( /* * The superblock buffer is uncached and while xfs_ail_push_all_sync() - * will push it, xfs_wait_buftarg() will not wait for it. Further, + * will push it, xfs_buftarg_drain() will not wait for it. Further, * xfs_buf_iowait() cannot be used because it was pushed with the * XBF_ASYNC flag set, so we need to use a lock/unlock pair to wait for * the IO to complete. */ xfs_ail_push_all_sync(mp->m_ail); - xfs_wait_buftarg(mp->m_ddev_targp); + xfs_buftarg_drain(mp->m_ddev_targp); xfs_buf_lock(mp->m_sb_bp); xfs_buf_unlock(mp->m_sb_bp); diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c index 7110507a2b6b..29a553f0877d 100644 --- a/fs/xfs/xfs_mount.c +++ b/fs/xfs/xfs_mount.c @@ -1023,8 +1023,8 @@ xfs_mountfs( xfs_log_mount_cancel(mp); out_fail_wait: if (mp->m_logdev_targp && mp->m_logdev_targp != mp->m_ddev_targp) - xfs_wait_buftarg(mp->m_logdev_targp); - xfs_wait_buftarg(mp->m_ddev_targp); + xfs_buftarg_drain(mp->m_logdev_targp); + xfs_buftarg_drain(mp->m_ddev_targp); out_free_perag: xfs_free_perag(mp); out_free_dir: diff --git a/fs/xfs/xfs_trace.h b/fs/xfs/xfs_trace.h index 86951652d3ed..7b4d8a5f2a49 100644 --- a/fs/xfs/xfs_trace.h +++ b/fs/xfs/xfs_trace.h @@ -340,7 +340,7 @@ DEFINE_BUF_EVENT(xfs_buf_get_uncached); DEFINE_BUF_EVENT(xfs_buf_item_relse); DEFINE_BUF_EVENT(xfs_buf_iodone_async); DEFINE_BUF_EVENT(xfs_buf_error_relse); -DEFINE_BUF_EVENT(xfs_buf_wait_buftarg); +DEFINE_BUF_EVENT(xfs_buf_drain_buftarg); DEFINE_BUF_EVENT(xfs_trans_read_buf_shut); /* not really buffer traces, but the buf provides useful information */ From patchwork Thu Dec 10 14:46:07 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Brian Foster X-Patchwork-Id: 11965059 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-15.8 required=3.0 tests=BAYES_00,DKIMWL_WL_HIGH, DKIM_SIGNED,DKIM_VALID,DKIM_VALID_AU,HEADER_FROM_DIFFERENT_DOMAINS, INCLUDES_CR_TRAILER,INCLUDES_PATCH,MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 5B631C4167B for ; Thu, 10 Dec 2020 14:48:10 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id 1A02322EBE for ; Thu, 10 Dec 2020 14:48:10 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S2389571AbgLJOru (ORCPT ); Thu, 10 Dec 2020 09:47:50 -0500 Received: from us-smtp-delivery-124.mimecast.com ([63.128.21.124]:47754 "EHLO us-smtp-delivery-124.mimecast.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S2391661AbgLJOrn (ORCPT ); Thu, 10 Dec 2020 09:47:43 -0500 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1607611576; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=HEbJBrzCLJ0gwC8ya5yChC9F2Jtk/4QoE40fjbAsGM0=; b=Jmi0z2hV22+qLW52mc25xTrVQ7UsXL6C7xG82Bu7xy5MiqnEcta5uLSDMmfElKZlUwow0F luOx+Y9CxetW+hRGDn7uPOykOjDUKJM8Y5KoHOM5ltI+oVKY1cZYpmFx2Z2tvXwSHvhmUQ 5Ek0HsrDpYiLdyYaaQLgR00gGx+fTCg= Received: from mimecast-mx01.redhat.com (mimecast-mx01.redhat.com [209.132.183.4]) (Using TLS) by relay.mimecast.com with ESMTP id us-mta-320-NWKWKKShOxOap-MxbHFTNg-1; Thu, 10 Dec 2020 09:46:13 -0500 X-MC-Unique: NWKWKKShOxOap-MxbHFTNg-1 Received: from smtp.corp.redhat.com (int-mx08.intmail.prod.int.phx2.redhat.com [10.5.11.23]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx01.redhat.com (Postfix) with ESMTPS id D6E61800D55 for ; Thu, 10 Dec 2020 14:46:11 +0000 (UTC) Received: from bfoster.redhat.com (ovpn-112-184.rdu2.redhat.com [10.10.112.184]) by smtp.corp.redhat.com (Postfix) with ESMTP id 917F819713 for ; Thu, 10 Dec 2020 14:46:11 +0000 (UTC) From: Brian Foster To: linux-xfs@vger.kernel.org Subject: [PATCH 2/2] xfs: don't drain buffer lru on freeze and read-only remount Date: Thu, 10 Dec 2020 09:46:07 -0500 Message-Id: <20201210144607.1922026-3-bfoster@redhat.com> In-Reply-To: <20201210144607.1922026-1-bfoster@redhat.com> References: <20201210144607.1922026-1-bfoster@redhat.com> MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.84 on 10.5.11.23 Precedence: bulk List-ID: X-Mailing-List: linux-xfs@vger.kernel.org xfs_buftarg_drain() is called from xfs_log_quiesce() to ensure the buffer cache is reclaimed during unmount. xfs_log_quiesce() is also called from xfs_quiesce_attr(), however, which means that cache state is completely drained for filesystem freeze and read-only remount. While technically harmless, this is unnecessarily heavyweight. Both freeze and read-only mounts allow reads and thus allow population of the buffer cache. Therefore, the transitional sequence in either case really only needs to quiesce outstanding writes to return the filesystem in a generally read-only state. Additionally, some users have reported that attempts to freeze a filesystem concurrent with a read-heavy workload causes the freeze process to stall for a significant amount of time. This occurs because, as mentioned above, the read workload repopulates the buffer LRU while the freeze task attempts to drain it. To improve this situation, replace the drain in xfs_log_quiesce() with a buffer I/O quiesce and lift the drain into the unmount path. This removes buffer LRU reclaim from freeze and read-only [re]mount, but ensures the LRU is still drained before the filesystem unmounts. Signed-off-by: Brian Foster Reviewed-by: Chandan Babu R Reviewed-by: Darrick J. Wong --- fs/xfs/xfs_buf.c | 20 +++++++++++++++----- fs/xfs/xfs_buf.h | 1 + fs/xfs/xfs_log.c | 6 ++++-- 3 files changed, 20 insertions(+), 7 deletions(-) diff --git a/fs/xfs/xfs_buf.c b/fs/xfs/xfs_buf.c index db918ed20c40..d3fce3129f6e 100644 --- a/fs/xfs/xfs_buf.c +++ b/fs/xfs/xfs_buf.c @@ -1815,14 +1815,13 @@ xfs_buftarg_drain_rele( return LRU_REMOVED; } +/* + * Wait for outstanding I/O on the buftarg to complete. + */ void -xfs_buftarg_drain( +xfs_buftarg_wait( struct xfs_buftarg *btp) { - LIST_HEAD(dispose); - int loop = 0; - bool write_fail = false; - /* * First wait on the buftarg I/O count for all in-flight buffers to be * released. This is critical as new buffers do not make the LRU until @@ -1838,6 +1837,17 @@ xfs_buftarg_drain( while (percpu_counter_sum(&btp->bt_io_count)) delay(100); flush_workqueue(btp->bt_mount->m_buf_workqueue); +} + +void +xfs_buftarg_drain( + struct xfs_buftarg *btp) +{ + LIST_HEAD(dispose); + int loop = 0; + bool write_fail = false; + + xfs_buftarg_wait(btp); /* loop until there is nothing left on the lru list. */ while (list_lru_count(&btp->bt_lru)) { diff --git a/fs/xfs/xfs_buf.h b/fs/xfs/xfs_buf.h index ea32369f8f77..96c6b478e26e 100644 --- a/fs/xfs/xfs_buf.h +++ b/fs/xfs/xfs_buf.h @@ -347,6 +347,7 @@ xfs_buf_update_cksum(struct xfs_buf *bp, unsigned long cksum_offset) extern struct xfs_buftarg *xfs_alloc_buftarg(struct xfs_mount *, struct block_device *, struct dax_device *); extern void xfs_free_buftarg(struct xfs_buftarg *); +extern void xfs_buftarg_wait(struct xfs_buftarg *); extern void xfs_buftarg_drain(struct xfs_buftarg *); extern int xfs_setsize_buftarg(struct xfs_buftarg *, unsigned int); diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c index 5ad4d5e78019..46ea4017fcec 100644 --- a/fs/xfs/xfs_log.c +++ b/fs/xfs/xfs_log.c @@ -936,13 +936,13 @@ xfs_log_quiesce( /* * The superblock buffer is uncached and while xfs_ail_push_all_sync() - * will push it, xfs_buftarg_drain() will not wait for it. Further, + * will push it, xfs_buftarg_wait() will not wait for it. Further, * xfs_buf_iowait() cannot be used because it was pushed with the * XBF_ASYNC flag set, so we need to use a lock/unlock pair to wait for * the IO to complete. */ xfs_ail_push_all_sync(mp->m_ail); - xfs_buftarg_drain(mp->m_ddev_targp); + xfs_buftarg_wait(mp->m_ddev_targp); xfs_buf_lock(mp->m_sb_bp); xfs_buf_unlock(mp->m_sb_bp); @@ -962,6 +962,8 @@ xfs_log_unmount( { xfs_log_quiesce(mp); + xfs_buftarg_drain(mp->m_ddev_targp); + xfs_trans_ail_destroy(mp); xfs_sysfs_del(&mp->m_log->l_kobj);