From patchwork Wed Aug 25 13:45:43 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Xiubo Li X-Patchwork-Id: 502355 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=-16.5 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 EEB80C4320A for ; Wed, 25 Aug 2021 13:46:14 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id D9AE1610C9 for ; Wed, 25 Aug 2021 13:46:14 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S240148AbhHYNq7 (ORCPT ); Wed, 25 Aug 2021 09:46:59 -0400 Received: from us-smtp-delivery-124.mimecast.com ([170.10.133.124]:43327 "EHLO us-smtp-delivery-124.mimecast.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S240149AbhHYNq5 (ORCPT ); Wed, 25 Aug 2021 09:46:57 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1629899171; 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=47RWAST7KQh92ze1S+hs/1cBTdTSPX7fQAtGKU0lVSU=; b=ZpF6MJAFrCIINhsm+xybd6f9O/FBELfINWAhDTH4DPoxzO8zeAqGid7bwDEXvKxyTDrzhR EhOBZ1KKSs9zTrnfaoHpxGpva1y8AaMAJNKVuW1fKF39eTiw57cHOvkS9b5mp2hxsPWvdi fcYaWYtH0EMJm66xNlx2ZU9Qw8iVGAw= 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-90-b-9rDevwPTaYhUIx_tXcbQ-1; Wed, 25 Aug 2021 09:46:10 -0400 X-MC-Unique: b-9rDevwPTaYhUIx_tXcbQ-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 0F6B9CC647; Wed, 25 Aug 2021 13:46:09 +0000 (UTC) Received: from lxbceph1.gsslab.pek2.redhat.com (unknown [10.72.47.117]) by smtp.corp.redhat.com (Postfix) with ESMTP id DFC3026DED; Wed, 25 Aug 2021 13:46:06 +0000 (UTC) From: xiubli@redhat.com To: jlayton@kernel.org Cc: idryomov@gmail.com, ukernel@gmail.com, pdonnell@redhat.com, ceph-devel@vger.kernel.org, Xiubo Li Subject: [PATCH v3 1/3] ceph: remove the capsnaps when removing the caps Date: Wed, 25 Aug 2021 21:45:43 +0800 Message-Id: <20210825134545.117521-2-xiubli@redhat.com> In-Reply-To: <20210825134545.117521-1-xiubli@redhat.com> References: <20210825134545.117521-1-xiubli@redhat.com> MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.84 on 10.5.11.23 Precedence: bulk List-ID: X-Mailing-List: ceph-devel@vger.kernel.org From: Xiubo Li The capsnaps will ihold the inodes when queuing to flush, so when force umounting it will close the sessions first and if the MDSes respond very fast and the session connections are closed just before killing the superblock, which will flush the msgr queue, then the flush capsnap callback won't ever be called, which will lead the memory leak bug for the ceph_inode_info. URL: https://tracker.ceph.com/issues/52295 Signed-off-by: Xiubo Li --- fs/ceph/caps.c | 67 +++++++++++++++++++++++++++++++++----------- fs/ceph/mds_client.c | 31 +++++++++++++++++++- fs/ceph/super.h | 6 ++++ 3 files changed, 86 insertions(+), 18 deletions(-) diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c index 1e6261a16fb5..61326b490b2b 100644 --- a/fs/ceph/caps.c +++ b/fs/ceph/caps.c @@ -3162,7 +3162,15 @@ void ceph_put_wrbuffer_cap_refs(struct ceph_inode_info *ci, int nr, break; } } - BUG_ON(!found); + + /* + * The capsnap should already be removed when + * removing auth cap in case likes force unmount. + */ + BUG_ON(!found && ci->i_auth_cap); + if (!found) + goto unlock; + capsnap->dirty_pages -= nr; if (capsnap->dirty_pages == 0) { complete_capsnap = true; @@ -3184,6 +3192,7 @@ void ceph_put_wrbuffer_cap_refs(struct ceph_inode_info *ci, int nr, complete_capsnap ? " (complete capsnap)" : ""); } +unlock: spin_unlock(&ci->i_ceph_lock); if (last) { @@ -3658,6 +3667,43 @@ static void handle_cap_flush_ack(struct inode *inode, u64 flush_tid, iput(inode); } +void __ceph_remove_capsnap(struct inode *inode, struct ceph_cap_snap *capsnap, + bool *wake_ci, bool *wake_mdsc) +{ + struct ceph_inode_info *ci = ceph_inode(inode); + struct ceph_mds_client *mdsc = ceph_sb_to_client(inode->i_sb)->mdsc; + bool ret; + + lockdep_assert_held(&ci->i_ceph_lock); + + dout("removing capsnap %p, inode %p ci %p\n", capsnap, inode, ci); + + list_del_init(&capsnap->ci_item); + ret = __detach_cap_flush_from_ci(ci, &capsnap->cap_flush); + if (wake_ci) + *wake_ci = ret; + + spin_lock(&mdsc->cap_dirty_lock); + if (list_empty(&ci->i_cap_flush_list)) + list_del_init(&ci->i_flushing_item); + + ret = __detach_cap_flush_from_mdsc(mdsc, &capsnap->cap_flush); + if (wake_mdsc) + *wake_mdsc = ret; + spin_unlock(&mdsc->cap_dirty_lock); +} + +void ceph_remove_capsnap(struct inode *inode, struct ceph_cap_snap *capsnap, + bool *wake_ci, bool *wake_mdsc) +{ + struct ceph_inode_info *ci = ceph_inode(inode); + + lockdep_assert_held(&ci->i_ceph_lock); + + WARN_ON_ONCE(capsnap->dirty_pages || capsnap->writing); + __ceph_remove_capsnap(inode, capsnap, wake_ci, wake_mdsc); +} + /* * Handle FLUSHSNAP_ACK. MDS has flushed snap data to disk and we can * throw away our cap_snap. @@ -3695,23 +3741,10 @@ static void handle_cap_flushsnap_ack(struct inode *inode, u64 flush_tid, capsnap, capsnap->follows); } } - if (flushed) { - WARN_ON(capsnap->dirty_pages || capsnap->writing); - dout(" removing %p cap_snap %p follows %lld\n", - inode, capsnap, follows); - list_del(&capsnap->ci_item); - wake_ci |= __detach_cap_flush_from_ci(ci, &capsnap->cap_flush); - - spin_lock(&mdsc->cap_dirty_lock); - - if (list_empty(&ci->i_cap_flush_list)) - list_del_init(&ci->i_flushing_item); - - wake_mdsc |= __detach_cap_flush_from_mdsc(mdsc, - &capsnap->cap_flush); - spin_unlock(&mdsc->cap_dirty_lock); - } + if (flushed) + ceph_remove_capsnap(inode, capsnap, &wake_ci, &wake_mdsc); spin_unlock(&ci->i_ceph_lock); + if (flushed) { ceph_put_snap_context(capsnap->context); ceph_put_cap_snap(capsnap); diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c index df3a735f7837..36ad0ebb2295 100644 --- a/fs/ceph/mds_client.c +++ b/fs/ceph/mds_client.c @@ -1604,14 +1604,39 @@ int ceph_iterate_session_caps(struct ceph_mds_session *session, return ret; } +static int remove_capsnaps(struct ceph_mds_client *mdsc, struct inode *inode) +{ + struct ceph_inode_info *ci = ceph_inode(inode); + struct ceph_cap_snap *capsnap; + int capsnap_release = 0; + + lockdep_assert_held(&ci->i_ceph_lock); + + dout("removing capsnaps, ci is %p, inode is %p\n", ci, inode); + + while (!list_empty(&ci->i_cap_snaps)) { + capsnap = list_first_entry(&ci->i_cap_snaps, + struct ceph_cap_snap, ci_item); + __ceph_remove_capsnap(inode, capsnap, NULL, NULL); + ceph_put_snap_context(capsnap->context); + ceph_put_cap_snap(capsnap); + capsnap_release++; + } + wake_up_all(&ci->i_cap_wq); + wake_up_all(&mdsc->cap_flushing_wq); + return capsnap_release; +} + static int remove_session_caps_cb(struct inode *inode, struct ceph_cap *cap, void *arg) { struct ceph_fs_client *fsc = (struct ceph_fs_client *)arg; + struct ceph_mds_client *mdsc = fsc->mdsc; struct ceph_inode_info *ci = ceph_inode(inode); LIST_HEAD(to_remove); bool dirty_dropped = false; bool invalidate = false; + int capsnap_release = 0; dout("removing cap %p, ci is %p, inode is %p\n", cap, ci, &ci->vfs_inode); @@ -1619,7 +1644,6 @@ static int remove_session_caps_cb(struct inode *inode, struct ceph_cap *cap, __ceph_remove_cap(cap, false); if (!ci->i_auth_cap) { struct ceph_cap_flush *cf; - struct ceph_mds_client *mdsc = fsc->mdsc; if (READ_ONCE(fsc->mount_state) >= CEPH_MOUNT_SHUTDOWN) { if (inode->i_data.nrpages > 0) @@ -1683,6 +1707,9 @@ static int remove_session_caps_cb(struct inode *inode, struct ceph_cap *cap, list_add(&ci->i_prealloc_cap_flush->i_list, &to_remove); ci->i_prealloc_cap_flush = NULL; } + + if (!list_empty(&ci->i_cap_snaps)) + capsnap_release = remove_capsnaps(mdsc, inode); } spin_unlock(&ci->i_ceph_lock); while (!list_empty(&to_remove)) { @@ -1699,6 +1726,8 @@ static int remove_session_caps_cb(struct inode *inode, struct ceph_cap *cap, ceph_queue_invalidate(inode); if (dirty_dropped) iput(inode); + while (capsnap_release--) + iput(inode); return 0; } diff --git a/fs/ceph/super.h b/fs/ceph/super.h index 8f4f2747be65..445d13d760d1 100644 --- a/fs/ceph/super.h +++ b/fs/ceph/super.h @@ -1169,6 +1169,12 @@ extern void ceph_put_cap_refs_no_check_caps(struct ceph_inode_info *ci, int had); extern void ceph_put_wrbuffer_cap_refs(struct ceph_inode_info *ci, int nr, struct ceph_snap_context *snapc); +extern void __ceph_remove_capsnap(struct inode *inode, + struct ceph_cap_snap *capsnap, + bool *wake_ci, bool *wake_mdsc); +extern void ceph_remove_capsnap(struct inode *inode, + struct ceph_cap_snap *capsnap, + bool *wake_ci, bool *wake_mdsc); extern void ceph_flush_snaps(struct ceph_inode_info *ci, struct ceph_mds_session **psession); extern bool __ceph_should_report_size(struct ceph_inode_info *ci); From patchwork Wed Aug 25 13:45:45 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Xiubo Li X-Patchwork-Id: 502354 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=-16.5 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 BECE1C4338F for ; Wed, 25 Aug 2021 13:46:20 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id A22B96109F for ; Wed, 25 Aug 2021 13:46:20 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S240231AbhHYNrF (ORCPT ); Wed, 25 Aug 2021 09:47:05 -0400 Received: from us-smtp-delivery-124.mimecast.com ([170.10.133.124]:58159 "EHLO us-smtp-delivery-124.mimecast.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S240410AbhHYNrE (ORCPT ); Wed, 25 Aug 2021 09:47:04 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1629899179; 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=BRarUKPqEQ+HHtxXvW6U/xyjjGu0QEO6gYdU6kLBKW4=; b=PT+29ujDSNn+xp5424YuYV/ZZw5iJR/1x/fUp68P3LSM/zUNBCG+VrV4PaFDCdSYnZ6R7k 1TvXmGg8n55B6O2oQTNVh1S9+94GkCScVnWIsfsNQDGZRMms9k7aSg2Pq2oskoHxua6VHE TaOMGkLR3A3PSfm5hTZssYvUj7uM8YA= 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-368-pHmBKQWfPWSGnF0KPzC3HQ-1; Wed, 25 Aug 2021 09:46:15 -0400 X-MC-Unique: pHmBKQWfPWSGnF0KPzC3HQ-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 64086195D560; Wed, 25 Aug 2021 13:46:14 +0000 (UTC) Received: from lxbceph1.gsslab.pek2.redhat.com (unknown [10.72.47.117]) by smtp.corp.redhat.com (Postfix) with ESMTP id 4087926DED; Wed, 25 Aug 2021 13:46:11 +0000 (UTC) From: xiubli@redhat.com To: jlayton@kernel.org Cc: idryomov@gmail.com, ukernel@gmail.com, pdonnell@redhat.com, ceph-devel@vger.kernel.org, Xiubo Li Subject: [PATCH v3 3/3] ceph: don't WARN if we're iterate removing the session caps Date: Wed, 25 Aug 2021 21:45:45 +0800 Message-Id: <20210825134545.117521-4-xiubli@redhat.com> In-Reply-To: <20210825134545.117521-1-xiubli@redhat.com> References: <20210825134545.117521-1-xiubli@redhat.com> MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.84 on 10.5.11.23 Precedence: bulk List-ID: X-Mailing-List: ceph-devel@vger.kernel.org From: Xiubo Li For example in case force umounting it will remove all the session caps one by one even it's dirty cap. URL: https://tracker.ceph.com/issues/52295 Signed-off-by: Xiubo Li --- fs/ceph/caps.c | 39 ++++++++++++++++++++++++++++++--------- fs/ceph/mds_client.c | 2 +- fs/ceph/super.h | 1 + 3 files changed, 32 insertions(+), 10 deletions(-) diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c index 61326b490b2b..6f20e18023c8 100644 --- a/fs/ceph/caps.c +++ b/fs/ceph/caps.c @@ -1114,17 +1114,16 @@ void __ceph_remove_cap(struct ceph_cap *cap, bool queue_release) return; } + lockdep_assert_held(&ci->i_ceph_lock); + dout("__ceph_remove_cap %p from %p\n", cap, &ci->vfs_inode); mdsc = ceph_inode_to_client(&ci->vfs_inode)->mdsc; /* remove from inode's cap rbtree, and clear auth cap */ rb_erase(&cap->ci_node, &ci->i_caps); - if (ci->i_auth_cap == cap) { - WARN_ON_ONCE(!list_empty(&ci->i_dirty_item) && - !mdsc->fsc->blocklisted); + if (ci->i_auth_cap == cap) ci->i_auth_cap = NULL; - } /* remove from session list */ spin_lock(&session->s_cap_lock); @@ -1176,6 +1175,28 @@ void __ceph_remove_cap(struct ceph_cap *cap, bool queue_release) } } +void ceph_remove_cap(struct ceph_cap *cap, bool queue_release) +{ + struct ceph_inode_info *ci = cap->ci; + struct ceph_fs_client *fsc; + + /* 'ci' being NULL means the remove have already occurred */ + if (!ci) { + dout("%s: cap inode is NULL\n", __func__); + return; + } + + lockdep_assert_held(&ci->i_ceph_lock); + + fsc = ceph_sb_to_client(ci->vfs_inode.i_sb); + WARN_ON_ONCE(ci->i_auth_cap == cap && + !list_empty(&ci->i_dirty_item) && + !fsc->blocklisted && + READ_ONCE(fsc->mount_state) != CEPH_MOUNT_SHUTDOWN); + + __ceph_remove_cap(cap, queue_release); +} + struct cap_msg_args { struct ceph_mds_session *session; u64 ino, cid, follows; @@ -1304,7 +1325,7 @@ void __ceph_remove_caps(struct ceph_inode_info *ci) while (p) { struct ceph_cap *cap = rb_entry(p, struct ceph_cap, ci_node); p = rb_next(p); - __ceph_remove_cap(cap, true); + ceph_remove_cap(cap, true); } spin_unlock(&ci->i_ceph_lock); } @@ -3828,7 +3849,7 @@ static void handle_cap_export(struct inode *inode, struct ceph_mds_caps *ex, goto out_unlock; if (target < 0) { - __ceph_remove_cap(cap, false); + ceph_remove_cap(cap, false); goto out_unlock; } @@ -3863,7 +3884,7 @@ static void handle_cap_export(struct inode *inode, struct ceph_mds_caps *ex, change_auth_cap_ses(ci, tcap->session); } } - __ceph_remove_cap(cap, false); + ceph_remove_cap(cap, false); goto out_unlock; } else if (tsession) { /* add placeholder for the export tagert */ @@ -3880,7 +3901,7 @@ static void handle_cap_export(struct inode *inode, struct ceph_mds_caps *ex, spin_unlock(&mdsc->cap_dirty_lock); } - __ceph_remove_cap(cap, false); + ceph_remove_cap(cap, false); goto out_unlock; } @@ -3991,7 +4012,7 @@ static void handle_cap_import(struct ceph_mds_client *mdsc, ocap->mseq, mds, le32_to_cpu(ph->seq), le32_to_cpu(ph->mseq)); } - __ceph_remove_cap(ocap, (ph->flags & CEPH_CAP_FLAG_RELEASE)); + ceph_remove_cap(ocap, (ph->flags & CEPH_CAP_FLAG_RELEASE)); } *old_issued = issued; diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c index 8758a27e691c..3cfd4654d17c 100644 --- a/fs/ceph/mds_client.c +++ b/fs/ceph/mds_client.c @@ -2016,7 +2016,7 @@ static int trim_caps_cb(struct inode *inode, struct ceph_cap *cap, void *arg) if (oissued) { /* we aren't the only cap.. just remove us */ - __ceph_remove_cap(cap, true); + ceph_remove_cap(cap, true); (*remaining)--; } else { struct dentry *dentry; diff --git a/fs/ceph/super.h b/fs/ceph/super.h index 445d13d760d1..c1add4ed59d2 100644 --- a/fs/ceph/super.h +++ b/fs/ceph/super.h @@ -1144,6 +1144,7 @@ extern void ceph_add_cap(struct inode *inode, unsigned cap, unsigned seq, u64 realmino, int flags, struct ceph_cap **new_cap); extern void __ceph_remove_cap(struct ceph_cap *cap, bool queue_release); +extern void ceph_remove_cap(struct ceph_cap *cap, bool queue_release); extern void __ceph_remove_caps(struct ceph_inode_info *ci); extern void ceph_put_cap(struct ceph_mds_client *mdsc, struct ceph_cap *cap);