From patchwork Wed Oct 25 06:19:51 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Xiubo Li X-Patchwork-Id: 738361 Received: from lindbergh.monkeyblade.net (lindbergh.monkeyblade.net [23.128.96.19]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id D6A2D15AD6 for ; Wed, 25 Oct 2023 06:22:28 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b="KkZOwXp+" 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 70FCD10E for ; Tue, 24 Oct 2023 23:22:27 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1698214946; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=TdXlbgwkfMWVl5lLYEP376cZGsnbcGUbwYjHotg3pFE=; b=KkZOwXp+0280W3ZFNViicp/kV+uB7Hgh3K7bRaCrQNiaJRKcffZUqT3lR9ANpKezd3LiEd bGRAcWp/jPlBV8TqPFM05E3vTixTgfklM3elh8diW5JMaRPez3IaI3c3gyQfZzroFJmNMO QvuGUFgvNR5NH3lJT+s7UoPW/k72gb0= Received: from mimecast-mx02.redhat.com (mimecast-mx02.redhat.com [66.187.233.88]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.3, cipher=TLS_AES_256_GCM_SHA384) id us-mta-581-ENjFjHsQOKuPKZgsnVXlsg-1; Wed, 25 Oct 2023 02:22:18 -0400 X-MC-Unique: ENjFjHsQOKuPKZgsnVXlsg-1 Received: from smtp.corp.redhat.com (int-mx05.intmail.prod.int.rdu2.redhat.com [10.11.54.5]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (2048 bits) server-digest SHA256) (No client certificate requested) by mimecast-mx02.redhat.com (Postfix) with ESMTPS id 2C6F58370E2; Wed, 25 Oct 2023 06:22:18 +0000 (UTC) Received: from li-a71a4dcc-35d1-11b2-a85c-951838863c8d.ibm.com.com (unknown [10.72.112.168]) by smtp.corp.redhat.com (Postfix) with ESMTP id 7FF6C8C0A; Wed, 25 Oct 2023 06:22:15 +0000 (UTC) From: xiubli@redhat.com To: ceph-devel@vger.kernel.org Cc: idryomov@gmail.com, jlayton@kernel.org, vshankar@redhat.com, mchangir@redhat.com, Xiubo Li Subject: [PATCH v2 1/2] ceph: always queue a writeback when revoking the Fb caps Date: Wed, 25 Oct 2023 14:19:51 +0800 Message-ID: <20231025061952.288730-2-xiubli@redhat.com> In-Reply-To: <20231025061952.288730-1-xiubli@redhat.com> References: <20231025061952.288730-1-xiubli@redhat.com> Precedence: bulk X-Mailing-List: ceph-devel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 X-Scanned-By: MIMEDefang 3.4.1 on 10.11.54.5 From: Xiubo Li In case there is 'Fw' dirty caps and 'CHECK_CAPS_FLUSH' is set we will always ignore queue a writeback. Queue a writeback is very important because it will block kclient flushing the snapcaps to MDS and which will block MDS waiting for revoking the 'Fb' caps. URL: https://tracker.ceph.com/issues/50223 Signed-off-by: Xiubo Li --- fs/ceph/caps.c | 48 ++++++++++++++++++++++++------------------------ 1 file changed, 24 insertions(+), 24 deletions(-) diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c index c72fcb5a6420..9b9ec1adc19d 100644 --- a/fs/ceph/caps.c +++ b/fs/ceph/caps.c @@ -2156,6 +2156,30 @@ void ceph_check_caps(struct ceph_inode_info *ci, int flags) ceph_cap_string(cap->implemented), ceph_cap_string(revoking)); + /* completed revocation? going down and there are no caps? */ + if (revoking) { + if ((revoking & cap_used) == 0) { + doutc(cl, "completed revocation of %s\n", + ceph_cap_string(cap->implemented & ~cap->issued)); + goto ack; + } + + /* + * If the "i_wrbuffer_ref" was increased by mmap or generic + * cache write just before the ceph_check_caps() is called, + * the Fb capability revoking will fail this time. Then we + * must wait for the BDI's delayed work to flush the dirty + * pages and to release the "i_wrbuffer_ref", which will cost + * at most 5 seconds. That means the MDS needs to wait at + * most 5 seconds to finished the Fb capability's revocation. + * + * Let's queue a writeback for it. + */ + if (S_ISREG(inode->i_mode) && ci->i_wrbuffer_ref && + (revoking & CEPH_CAP_FILE_BUFFER)) + queue_writeback = true; + } + if (cap == ci->i_auth_cap && (cap->issued & CEPH_CAP_FILE_WR)) { /* request larger max_size from MDS? */ @@ -2183,30 +2207,6 @@ void ceph_check_caps(struct ceph_inode_info *ci, int flags) } } - /* completed revocation? going down and there are no caps? */ - if (revoking) { - if ((revoking & cap_used) == 0) { - doutc(cl, "completed revocation of %s\n", - ceph_cap_string(cap->implemented & ~cap->issued)); - goto ack; - } - - /* - * If the "i_wrbuffer_ref" was increased by mmap or generic - * cache write just before the ceph_check_caps() is called, - * the Fb capability revoking will fail this time. Then we - * must wait for the BDI's delayed work to flush the dirty - * pages and to release the "i_wrbuffer_ref", which will cost - * at most 5 seconds. That means the MDS needs to wait at - * most 5 seconds to finished the Fb capability's revocation. - * - * Let's queue a writeback for it. - */ - if (S_ISREG(inode->i_mode) && ci->i_wrbuffer_ref && - (revoking & CEPH_CAP_FILE_BUFFER)) - queue_writeback = true; - } - /* want more caps from mds? */ if (want & ~cap->mds_wanted) { if (want & ~(cap->mds_wanted | cap->issued)) From patchwork Wed Oct 25 06:19:52 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Xiubo Li X-Patchwork-Id: 737883 Received: from lindbergh.monkeyblade.net (lindbergh.monkeyblade.net [23.128.96.19]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 83D1C156E7 for ; Wed, 25 Oct 2023 06:22:32 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b="ZMttTmm6" 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 18589137 for ; Tue, 24 Oct 2023 23:22:31 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1698214950; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=VnedWDCkdpVsct9cYBxihx7SgvdemxFh3LIWHm4FBBE=; b=ZMttTmm6y2BvIpsvITKw08d8ofzbuOTwiPB6x2DcBpgU1DI48S0JV3nHKUpQbwtzO4Zvh6 VWFo/Vv3LNd/WgLpS5Z4dpdHR84dw5UR6AN71j+OFj7HGHWAJ3t0bKJNjfZIuRls3QEDOV dwiVNOMeqn1rxNELElyCuMspaLDtOvw= Received: from mimecast-mx02.redhat.com (mimecast-mx02.redhat.com [66.187.233.88]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.3, cipher=TLS_AES_256_GCM_SHA384) id us-mta-21-ew1I2qyhM0C3pbKIUZZETQ-1; Wed, 25 Oct 2023 02:22:21 -0400 X-MC-Unique: ew1I2qyhM0C3pbKIUZZETQ-1 Received: from smtp.corp.redhat.com (int-mx05.intmail.prod.int.rdu2.redhat.com [10.11.54.5]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (2048 bits) server-digest SHA256) (No client certificate requested) by mimecast-mx02.redhat.com (Postfix) with ESMTPS id 7B53B857F8C; Wed, 25 Oct 2023 06:22:21 +0000 (UTC) Received: from li-a71a4dcc-35d1-11b2-a85c-951838863c8d.ibm.com.com (unknown [10.72.112.168]) by smtp.corp.redhat.com (Postfix) with ESMTP id CDEBB8C0B; Wed, 25 Oct 2023 06:22:18 +0000 (UTC) From: xiubli@redhat.com To: ceph-devel@vger.kernel.org Cc: idryomov@gmail.com, jlayton@kernel.org, vshankar@redhat.com, mchangir@redhat.com, Xiubo Li Subject: [PATCH v2 2/2] ceph: add ceph_cap_unlink_work to fire check caps immediately Date: Wed, 25 Oct 2023 14:19:52 +0800 Message-ID: <20231025061952.288730-3-xiubli@redhat.com> In-Reply-To: <20231025061952.288730-1-xiubli@redhat.com> References: <20231025061952.288730-1-xiubli@redhat.com> Precedence: bulk X-Mailing-List: ceph-devel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 X-Scanned-By: MIMEDefang 3.4.1 on 10.11.54.5 From: Xiubo Li When unlinking a file the check caps could be delayed for more than 5 seconds, but in MDS side it maybe waiting for the clients to release caps. This will add a dedicated work queue and list to help trigger to fire the check caps and dirty buffer flushing immediately. URL: https://tracker.ceph.com/issues/50223 Signed-off-by: Xiubo Li --- fs/ceph/caps.c | 17 ++++++++++++++++- fs/ceph/mds_client.c | 34 ++++++++++++++++++++++++++++++++++ fs/ceph/mds_client.h | 4 ++++ 3 files changed, 54 insertions(+), 1 deletion(-) diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c index 9b9ec1adc19d..be4f986e082d 100644 --- a/fs/ceph/caps.c +++ b/fs/ceph/caps.c @@ -4790,7 +4790,22 @@ int ceph_drop_caps_for_unlink(struct inode *inode) if (__ceph_caps_dirty(ci)) { struct ceph_mds_client *mdsc = ceph_inode_to_fs_client(inode)->mdsc; - __cap_delay_requeue_front(mdsc, ci); + + doutc(mdsc->fsc->client, "%p %llx.%llx\n", inode, + ceph_vinop(inode)); + spin_lock(&mdsc->cap_unlink_delay_lock); + ci->i_ceph_flags |= CEPH_I_FLUSH; + if (!list_empty(&ci->i_cap_delay_list)) + list_del_init(&ci->i_cap_delay_list); + list_add_tail(&ci->i_cap_delay_list, + &mdsc->cap_unlink_delay_list); + spin_unlock(&mdsc->cap_unlink_delay_lock); + + /* + * Fire the work immediately, because the MDS maybe + * waiting for caps release. + */ + schedule_work(&mdsc->cap_unlink_work); } } spin_unlock(&ci->i_ceph_lock); diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c index 6fa7134beec8..a7bffb030036 100644 --- a/fs/ceph/mds_client.c +++ b/fs/ceph/mds_client.c @@ -2500,6 +2500,37 @@ void ceph_reclaim_caps_nr(struct ceph_mds_client *mdsc, int nr) } } +static void ceph_cap_unlink_work(struct work_struct *work) +{ + struct ceph_mds_client *mdsc = + container_of(work, struct ceph_mds_client, cap_unlink_work); + struct ceph_client *cl = mdsc->fsc->client; + + doutc(cl, "begin\n"); + spin_lock(&mdsc->cap_unlink_delay_lock); + while (!list_empty(&mdsc->cap_unlink_delay_list)) { + struct ceph_inode_info *ci; + struct inode *inode; + + ci = list_first_entry(&mdsc->cap_unlink_delay_list, + struct ceph_inode_info, + i_cap_delay_list); + list_del_init(&ci->i_cap_delay_list); + + inode = igrab(&ci->netfs.inode); + if (inode) { + spin_unlock(&mdsc->cap_unlink_delay_lock); + doutc(cl, "on %p %llx.%llx\n", inode, + ceph_vinop(inode)); + ceph_check_caps(ci, CHECK_CAPS_FLUSH); + iput(inode); + spin_lock(&mdsc->cap_unlink_delay_lock); + } + } + spin_unlock(&mdsc->cap_unlink_delay_lock); + doutc(cl, "done\n"); +} + /* * requests */ @@ -5372,6 +5403,8 @@ int ceph_mdsc_init(struct ceph_fs_client *fsc) INIT_LIST_HEAD(&mdsc->cap_delay_list); INIT_LIST_HEAD(&mdsc->cap_wait_list); spin_lock_init(&mdsc->cap_delay_lock); + INIT_LIST_HEAD(&mdsc->cap_unlink_delay_list); + spin_lock_init(&mdsc->cap_unlink_delay_lock); INIT_LIST_HEAD(&mdsc->snap_flush_list); spin_lock_init(&mdsc->snap_flush_lock); mdsc->last_cap_flush_tid = 1; @@ -5380,6 +5413,7 @@ int ceph_mdsc_init(struct ceph_fs_client *fsc) spin_lock_init(&mdsc->cap_dirty_lock); init_waitqueue_head(&mdsc->cap_flushing_wq); INIT_WORK(&mdsc->cap_reclaim_work, ceph_cap_reclaim_work); + INIT_WORK(&mdsc->cap_unlink_work, ceph_cap_unlink_work); err = ceph_metric_init(&mdsc->metric); if (err) goto err_mdsmap; diff --git a/fs/ceph/mds_client.h b/fs/ceph/mds_client.h index 2e6ddaa13d72..f25117fa910f 100644 --- a/fs/ceph/mds_client.h +++ b/fs/ceph/mds_client.h @@ -462,6 +462,8 @@ struct ceph_mds_client { unsigned long last_renew_caps; /* last time we renewed our caps */ struct list_head cap_delay_list; /* caps with delayed release */ spinlock_t cap_delay_lock; /* protects cap_delay_list */ + struct list_head cap_unlink_delay_list; /* caps with delayed release for unlink */ + spinlock_t cap_unlink_delay_lock; /* protects cap_unlink_delay_list */ struct list_head snap_flush_list; /* cap_snaps ready to flush */ spinlock_t snap_flush_lock; @@ -475,6 +477,8 @@ struct ceph_mds_client { struct work_struct cap_reclaim_work; atomic_t cap_reclaim_pending; + struct work_struct cap_unlink_work; + /* * Cap reservations *