* Thu Jun 18 2009 Eric Sandeen <sandeen@redhat.com> 1.41.6-2

- Update journal backup blocks in sb after resize (#505339)
- Fix memory leak in extent handling functions
- Fix bug in inode writing in extent code, clobbered i_extra_isize etc
This commit is contained in:
Eric Sandeen 2009-06-18 14:13:30 +00:00
parent 2355e455db
commit eec0db86fa
5 changed files with 187 additions and 2 deletions

View File

@ -0,0 +1,28 @@
Each time an extent handle is opened and closed, if the inode has an
extent tree which does not fit in the inode's i_block structure, a
filesystem block buffer was not getting released. Since e2fsck opens
an extent handle for every inode using extents, this can translate to
a very large amount of memory getting lost.
Thanks to Henrik 'Mauritz' Johnson for discovering and pointing out
this leak, which he ran into while running the "rdump" command in
debugfs.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
lib/ext2fs/extent.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/lib/ext2fs/extent.c b/lib/ext2fs/extent.c
index b7eb617..2b88739 100644
--- a/lib/ext2fs/extent.c
+++ b/lib/ext2fs/extent.c
@@ -168,7 +168,7 @@ extern void ext2fs_extent_free(ext2_extent_handle_t handle)
if (handle->inode)
ext2fs_free_mem(&handle->inode);
if (handle->path) {
- for (i=1; i < handle->max_depth; i++) {
+ for (i=1; i <= handle->max_depth; i++) {
if (handle->path[i].buf)
ext2fs_free_mem(&handle->path[i].buf);
}

View File

@ -0,0 +1,55 @@
commit 125a36780626cdb0fc4d62fd529486baa8bce54c
Author: Eric Sandeen <sandeen@redhat.com>
Date: Wed Jun 17 18:49:01 2009 -0400
libext2fs: write only core inode in update_path()
The ext2_extent_handle only has a struct ext2_inode allocated on
it, and the same amount copied into it in that same function,
but in update_path() we're possibly writing out more than that -
for example 256 bytes, from that address. This causes uninitialized
memory to get written to disk, overwriting the parts of the
inode past the osd2 member (the end of the smaller structure).
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
diff --git a/lib/ext2fs/extent.c b/lib/ext2fs/extent.c
index 2b88739..35b080e 100644
--- a/lib/ext2fs/extent.c
+++ b/lib/ext2fs/extent.c
@@ -546,8 +546,8 @@ static errcode_t update_path(ext2_extent_handle_t handle)
struct ext3_extent_idx *ix;
if (handle->level == 0) {
- retval = ext2fs_write_inode_full(handle->fs, handle->ino,
- handle->inode, EXT2_INODE_SIZE(handle->fs->super));
+ retval = ext2fs_write_inode(handle->fs, handle->ino,
+ handle->inode);
} else {
ix = handle->path[handle->level - 1].curr;
blk = ext2fs_le32_to_cpu(ix->ei_leaf) +
@@ -1011,8 +1011,8 @@ static errcode_t extent_node_split(ext2_extent_handle_t handle)
/* new node hooked in, so update inode block count (do this here?) */
handle->inode->i_blocks += handle->fs->blocksize / 512;
- retval = ext2fs_write_inode_full(handle->fs, handle->ino,
- handle->inode, EXT2_INODE_SIZE(handle->fs->super));
+ retval = ext2fs_write_inode(handle->fs, handle->ino,
+ handle->inode);
if (retval)
goto done;
@@ -1370,9 +1370,8 @@ errcode_t ext2fs_extent_delete(ext2_extent_handle_t handle, int flags)
retval = ext2fs_extent_delete(handle, flags);
handle->inode->i_blocks -= handle->fs->blocksize / 512;
- retval = ext2fs_write_inode_full(handle->fs,
- handle->ino, handle->inode,
- EXT2_INODE_SIZE(handle->fs->super));
+ retval = ext2fs_write_inode(handle->fs, handle->ino,
+ handle->inode);
ext2fs_block_alloc_stats(handle->fs, extent.e_pblk, -1);
}
} else {

View File

@ -0,0 +1,89 @@
This was reported in Fedora, since the livecd creator does
a lot of resizing.
If we've moved the journal blocks during resize (more likely now,
due to the journal being in the middle) the backup blocks in the
superblock don't get updated, and a subsequent e2fsck will find
issues:
e2fsck 1.41.6 (30-May-2009)
Backing up journal inode block information.
Pass 1: Checking inodes, blocks, and sizes
Pass 2: Checking directory structure
Pass 3: Checking directory connectivity
Pass 4: Checking reference counts
Pass 5: Checking group summary information
/mnt/test/img: ***** FILE SYSTEM WAS MODIFIED *****
/mnt/test/img: 11/16000 files (0.0% non-contiguous), 17789/38400 blocks
This can be shown in a simple test:
# dd if=/dev/zero of=img bs=1 count=0 seek=3000M
# mke2fs -t ext4 -F img
# resize2fs img 150M
# e2fsck -f img
(thanks to the Fedora reporter Mads Kiilerich for the testcase!
https://bugzilla.redhat.com/show_bug.cgi?id=506105#c2)
So, update the backup journal in the superblock before resize2fs exits.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Resolves-RH-Bugzilla: 505339
---
diff --git a/resize/resize2fs.c b/resize/resize2fs.c
index 0d5dc81..63c469f 100644
--- a/resize/resize2fs.c
+++ b/resize/resize2fs.c
@@ -49,6 +49,7 @@ static errcode_t inode_ref_fix(ext2_resize_t rfs);
static errcode_t move_itables(ext2_resize_t rfs);
static errcode_t fix_resize_inode(ext2_filsys fs);
static errcode_t ext2fs_calculate_summary_stats(ext2_filsys fs);
+static errcode_t fix_sb_journal_backup(ext2_filsys fs);
/*
* Some helper CPP macros
@@ -148,6 +149,10 @@ errcode_t resize_fs(ext2_filsys fs, blk_t *new_size, int flags,
if (retval)
goto errout;
+ retval = fix_sb_journal_backup(rfs->new_fs);
+ if (retval)
+ goto errout;
+
rfs->new_fs->flags &= ~EXT2_FLAG_MASTER_SB_ONLY;
retval = ext2fs_close(rfs->new_fs);
if (retval)
@@ -1857,6 +1862,28 @@ static errcode_t ext2fs_calculate_summary_stats(ext2_filsys fs)
}
/*
+ * Journal may have been relocated; update the backup journal blocks
+ * in the superblock.
+ */
+static errcode_t fix_sb_journal_backup(ext2_filsys fs)
+{
+ errcode_t retval;
+ struct ext2_inode inode;
+
+ if (!(fs->super->s_feature_compat & EXT3_FEATURE_COMPAT_HAS_JOURNAL))
+ return 0;
+
+ retval = ext2fs_read_inode(fs, fs->super->s_journal_inum, &inode);
+ if (retval)
+ return retval;
+ memcpy(fs->super->s_jnl_blocks, inode.i_block, EXT2_N_BLOCKS*4);
+ fs->super->s_jnl_blocks[16] = inode.i_size;
+ fs->super->s_jnl_backup_type = EXT3_JNL_BACKUP_BLOCKS;
+ ext2fs_mark_super_dirty(fs);
+ return 0;
+}
+
+/*
* calcluate the minimum number of blocks the given fs can be resized to
*/
blk_t calculate_minimum_resize_size(ext2_filsys fs)

View File

@ -4,7 +4,7 @@
Summary: Utilities for managing ext2, ext3, and ext4 filesystems
Name: e2fsprogs
Version: 1.41.6
Release: 4%{?dist}
Release: 5%{?dist}
# License based on upstream-modified COPYING file,
# which clearly states "V2" intent.
License: GPLv2
@ -14,6 +14,11 @@ Source1: ext2_types-wrapper.h
Source3: uuidd.init
Patch2: e2fsprogs-1.40.4-sb_feature_check_ignore.patch
# Upstream or soon to be:
Patch3: e2fsprogs-1.41.4-update-sb-journal-backup.patch
Patch4: e2fsprogs-1.41.4-extent-open-leak.patch
Patch5: e2fsprogs-1.41.4-no-full-inode-write-in-extentcode.patch
Url: http://e2fsprogs.sourceforge.net/
BuildRoot: %{_tmppath}/%{name}-%{version}-%{release}-root-%(%{__id_u} -n)
Requires: e2fsprogs-libs = %{version}-%{release}, device-mapper
@ -89,6 +94,9 @@ SMP systems.
# mildly unsafe but 'til I get something better, avoid full fsck
# after an selinux install...
%patch2 -p1 -b .featurecheck
%patch3 -p1
%patch4 -p1
%patch5 -p1
%build
%configure --enable-elf-shlibs --enable-nls \
@ -270,6 +278,11 @@ fi
%dir %attr(2775, uuidd, uuidd) /var/lib/libuuid
%changelog
* Thu Jun 18 2009 Eric Sandeen <sandeen@redhat.com> 1.41.6-5
- Update journal backup blocks in sb after resize (#505339)
- Fix memory leak in extent handling functions
- Fix bug in inode writing in extent code, clobbered i_extra_isize etc
* Mon Jun 8 2009 Karel Zak <kzak@redhat.com> 1.41.6-4
- set BuildRequires: libblkid-devel (from util-linux-ng)

View File

@ -65,7 +65,7 @@ case "$1" in
;;
status)
status -p /var/lib/libuuid/uuidd.pid uuidd uuidd
REVAL=$?
RETVAL=$?
;;
*)
echo $"Usage: $0 {start|stop|status|restart|reload|force-reload|condrestart}"