Date: Sun, 01 Sep 2024 17:59:36 -0700
To: mm-commits@vger.kernel.org,stable@vger.kernel.org,konishi.ryusuke@gmail.com,akpm@linux-foundation.org
From: Andrew Morton <akpm@linux-foundation.org>
Subject: [merged mm-hotfixes-stable] nilfs2-fix-state-management-in-error-path-of-log-writing-function.patch removed from -mm tree
Message-Id: <20240902005936.D084EC4CEC3@smtp.kernel.org>
Precedence: bulk


The quilt patch titled
     Subject: nilfs2: fix state management in error path of log writing function
has been removed from the -mm tree.  Its filename was
     nilfs2-fix-state-management-in-error-path-of-log-writing-function.patch

This patch was dropped because it was merged into the mm-hotfixes-stable branch
of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm

------------------------------------------------------
From: Ryusuke Konishi <konishi.ryusuke@gmail.com>
Subject: nilfs2: fix state management in error path of log writing function
Date: Wed, 14 Aug 2024 19:11:19 +0900

After commit a694291a6211 ("nilfs2: separate wait function from
nilfs_segctor_write") was applied, the log writing function
nilfs_segctor_do_construct() was able to issue I/O requests continuously
even if user data blocks were split into multiple logs across segments,
but two potential flaws were introduced in its error handling.

First, if nilfs_segctor_begin_construction() fails while creating the
second or subsequent logs, the log writing function returns without
calling nilfs_segctor_abort_construction(), so the writeback flag set on
pages/folios will remain uncleared.  This causes page cache operations to
hang waiting for the writeback flag.  For example,
truncate_inode_pages_final(), which is called via nilfs_evict_inode() when
an inode is evicted from memory, will hang.

Second, the NILFS_I_COLLECTED flag set on normal inodes remain uncleared. 
As a result, if the next log write involves checkpoint creation, that's
fine, but if a partial log write is performed that does not, inodes with
NILFS_I_COLLECTED set are erroneously removed from the "sc_dirty_files"
list, and their data and b-tree blocks may not be written to the device,
corrupting the block mapping.

Fix these issues by uniformly calling nilfs_segctor_abort_construction()
on failure of each step in the loop in nilfs_segctor_do_construct(),
having it clean up logs and segment usages according to progress, and
correcting the conditions for calling nilfs_redirty_inodes() to ensure
that the NILFS_I_COLLECTED flag is cleared.

Link: https://lkml.kernel.org/r/20240814101119.4070-1-konishi.ryusuke@gmail.com
Fixes: a694291a6211 ("nilfs2: separate wait function from nilfs_segctor_write")
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@gmail.com>
Tested-by: Ryusuke Konishi <konishi.ryusuke@gmail.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---

 fs/nilfs2/segment.c |   10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

--- a/fs/nilfs2/segment.c~nilfs2-fix-state-management-in-error-path-of-log-writing-function
+++ a/fs/nilfs2/segment.c
@@ -1812,6 +1812,9 @@ static void nilfs_segctor_abort_construc
 	nilfs_abort_logs(&logs, ret ? : err);
 
 	list_splice_tail_init(&sci->sc_segbufs, &logs);
+	if (list_empty(&logs))
+		return; /* if the first segment buffer preparation failed */
+
 	nilfs_cancel_segusage(&logs, nilfs->ns_sufile);
 	nilfs_free_incomplete_logs(&logs, nilfs);
 
@@ -2056,7 +2059,7 @@ static int nilfs_segctor_do_construct(st
 
 		err = nilfs_segctor_begin_construction(sci, nilfs);
 		if (unlikely(err))
-			goto out;
+			goto failed;
 
 		/* Update time stamp */
 		sci->sc_seg_ctime = ktime_get_real_seconds();
@@ -2120,10 +2123,9 @@ static int nilfs_segctor_do_construct(st
 	return err;
 
  failed_to_write:
-	if (sci->sc_stage.flags & NILFS_CF_IFILE_STARTED)
-		nilfs_redirty_inodes(&sci->sc_dirty_files);
-
  failed:
+	if (mode == SC_LSEG_SR && nilfs_sc_cstage_get(sci) >= NILFS_ST_IFILE)
+		nilfs_redirty_inodes(&sci->sc_dirty_files);
 	if (nilfs_doing_gc())
 		nilfs_redirty_inodes(&sci->sc_gc_inodes);
 	nilfs_segctor_abort_construction(sci, nilfs, err);
_

Patches currently in -mm which might be from konishi.ryusuke@gmail.com are

nilfs2-add-support-for-fs_ioc_getuuid.patch
nilfs2-add-support-for-fs_ioc_getfssysfspath.patch
nilfs2-add-support-for-fs_ioc_getfslabel.patch
nilfs2-add-support-for-fs_ioc_setfslabel.patch
nilfs2-do-not-output-warnings-when-clearing-dirty-buffers.patch
nilfs2-add-missing-argument-description-for-__nilfs_error.patch
nilfs2-add-missing-argument-descriptions-for-ioctl-related-helpers.patch
nilfs2-improve-kernel-doc-comments-for-b-tree-node-helpers.patch
nilfs2-fix-incorrect-kernel-doc-declaration-of-nilfs_palloc_req-structure.patch
nilfs2-add-missing-description-of-nilfs_btree_path-structure.patch
nilfs2-describe-the-members-of-nilfs_bmap_operations-structure.patch
nilfs2-fix-inconsistencies-in-kernel-doc-comments-in-segmenth.patch
nilfs2-fix-missing-initial-short-descriptions-of-kernel-doc-comments.patch
nilfs2-treat-missing-sufile-header-block-as-metadata-corruption.patch
nilfs2-treat-missing-cpfile-header-block-as-metadata-corruption.patch
nilfs2-do-not-propagate-enoent-error-from-sufile-during-recovery.patch
nilfs2-do-not-propagate-enoent-error-from-sufile-during-gc.patch
nilfs2-do-not-propagate-enoent-error-from-nilfs_sufile_mark_dirty.patch
nilfs2-use-the-bits_per_long-macro.patch
nilfs2-separate-inode-type-information-from-i_state-field.patch
nilfs2-eliminate-the-shared-counter-and-spinlock-for-i_generation.patch
nilfs2-do-not-repair-reserved-inode-bitmap-in-nilfs_new_inode.patch
nilfs2-remove-sc_timer_task.patch
nilfs2-use-kthread_create-and-kthread_stop-for-the-log-writer-thread.patch
nilfs2-refactor-nilfs_segctor_thread.patch