From: rpeterso@redhat.com <rpeterso@redhat.com>
To: cluster-devel.redhat.com
Subject: [Cluster-devel] [PATCH 17/66] fsck.gfs2: fsck.gfs2: Don't stop invalidating blocks on invalid
Date: Fri, 20 Jan 2012 09:09:58 -0600 [thread overview]
Message-ID: <1327072247-26275-18-git-send-email-rpeterso@redhat.com> (raw)
In-Reply-To: <1327072247-26275-1-git-send-email-rpeterso@redhat.com>
From: Bob Peterson <rpeterso@redhat.com>
When fsck found a duplicate reference to a block it invalidated the dinode's
metadata. But if it encountered an invalid block, for example, out of range,
the invalidating would stop. If we encounter a block that isn't valid, we
obviously can't invalidate it. However, if we return an error, all future
invalidating will stop for that dinode. That's wrong because we need it to
continue to invalidate the other valid blocks. If we don't do this, block
references that follow the bad one that are also referenced elsewhere
(duplicates) won't be flagged as such. As a result, they'll be freed when
this corrupt dinode is deleted, despite being used by another dinode as a
valid block. This patch makes it return a good return code so the invalidating
continues.
rhbz#675723
---
gfs2/fsck/pass1.c | 11 +++++++++--
1 files changed, 9 insertions(+), 2 deletions(-)
diff --git a/gfs2/fsck/pass1.c b/gfs2/fsck/pass1.c
index 49a8f4e..271075c 100644
--- a/gfs2/fsck/pass1.c
+++ b/gfs2/fsck/pass1.c
@@ -825,8 +825,15 @@ static int mark_block_invalid(struct gfs2_inode *ip, uint64_t block,
{
uint8_t q;
- if (!valid_block(ip->i_sbd, block) != 0)
- return -EFAULT;
+ /* If the block isn't valid, we obviously can't invalidate it.
+ * However, if we return an error, invalidating will stop, and
+ * we want it to continue to invalidate the valid blocks. If we
+ * don't do this, block references that follow that are also
+ * referenced elsewhere (duplicates) won't be flagged as such,
+ * and as a result, they'll be freed when this dinode is deleted,
+ * despite being used by another dinode as a valid block. */
+ if (!valid_block(ip->i_sbd, block))
+ return 0;
q = block_type(block);
if (q != gfs2_block_free) {
--
1.7.7.5
next prev parent reply other threads:[~2012-01-20 15:09 UTC|newest]
Thread overview: 67+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-01-20 15:09 [Cluster-devel] [PATCH 00/66] fsck.gfs2: add ability to fix GFS (gfs1) file systems rpeterso
2012-01-20 15:09 ` [Cluster-devel] [PATCH 01/66] fsck.gfs2: Make functions use sdp rather than sbp rpeterso
2012-01-20 15:09 ` [Cluster-devel] [PATCH 02/66] fsck.gfs2: Change "if(" to "if (" rpeterso
2012-01-20 15:09 ` [Cluster-devel] [PATCH 03/66] libgfs1: Add gfs1 variable to superblock structure rpeterso
2012-01-20 15:09 ` [Cluster-devel] [PATCH 04/66] libgfs2: Make check_sb and read_sb operate on gfs1 rpeterso
2012-01-20 15:09 ` [Cluster-devel] [PATCH 05/66] libgfs2: move gfs1 structures to libgfs2 rpeterso
2012-01-20 15:09 ` [Cluster-devel] [PATCH 06/66] fsck.gfs2: Check for blocks wrongly inside resource groups rpeterso
2012-01-20 15:09 ` [Cluster-devel] [PATCH 07/66] fsck.gfs2: Rename check_leaf to check_ealeaf_block rpeterso
2012-01-20 15:09 ` [Cluster-devel] [PATCH 08/66] fsck.gfs2: fsck.gfs2: Delete vestigial buffer_head in check_leaf rpeterso
2012-01-20 15:09 ` [Cluster-devel] [PATCH 09/66] fsck.gfs2: fsck.gfs2: Rename nlink functions to be intuitive rpeterso
2012-01-20 15:09 ` [Cluster-devel] [PATCH 10/66] fsck.gfs2: fsck.gfs2: Sync di_nlink adding links for lost+found rpeterso
2012-01-20 15:09 ` [Cluster-devel] [PATCH 11/66] fsck.gfs2: fsck.gfs2: Make dir entry count 32 bits rpeterso
2012-01-20 15:09 ` [Cluster-devel] [PATCH 12/66] fsck.gfs2: get rid of triple negative logic rpeterso
2012-01-20 15:09 ` [Cluster-devel] [PATCH 13/66] dirent_repair needs to mark the buffer as modified rpeterso
2012-01-20 15:09 ` [Cluster-devel] [PATCH 14/66] fsck.gfs2: fsck.gfs2: Ask to reclaim unlinked meta per-rgrp only rpeterso
2012-01-20 15:09 ` [Cluster-devel] [PATCH 15/66] fsck.gfs2: fsck.gfs2: Refactor add_dotdot function in lost+found rpeterso
2012-01-20 15:09 ` [Cluster-devel] [PATCH 16/66] libgfs2: libgfs2: Use __FUNCTION__ rather than __FILE__ rpeterso
2012-01-20 15:09 ` rpeterso [this message]
2012-01-20 15:09 ` [Cluster-devel] [PATCH 18/66] fsck.gfs2: fsck.gfs2: Find and clear duplicate leaf blocks refs rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 19/66] fsck.gfs2: fsck.gfs2: Move check_num_ptrs from metawalk to pass1 rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 20/66] fsck.gfs2: fsck.gfs2: Duplicate ref processing for leaf blocks rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 21/66] fsck.gfs2: fsck.gfs2: split check_leaf_blks to be more readable rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 22/66] fsck.gfs2: Shorten output rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 23/66] fsck.gfs2: Make output messages more sensible rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 24/66] fsck.gfs pass2: Refactor function set_dotdot_dir rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 25/66] fsck.gfs2 pass2: Delete extended attributes with inode rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 26/66] fsck.gfs2 pass2: Don't delete invalid inode metadata rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 27/66] fsck.gfs2 pass3: Refactor mark_and_return_parent rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 28/66] fsck.gfs2: misc cosmetic changes rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 29/66] fsck.gfs2: Don't use old_leaf if it was a duplicate rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 30/66] fsck.gfs2: Add find_remove_dup, free_block_if_notdup rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 31/66] fsck.gfs2: don't free prev rgrp list repairing rgrps rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 32/66] libgfs2: eliminate gfs1_readi in favor of gfs2_readi rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 33/66] libgfs2: Mark buffer modified adding a new GFS1 block rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 34/66] libgfs2: Use dinode buffer to map gfs1 dinode blocks rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 35/66] libgfs2: move block_map functions to fsck.gfs2 rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 36/66] libgfs2: eliminate gfs1_rindex_read rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 37/66] libgfs2: combine ri_update and gfs1_ri_update rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 38/66] libgfs2: combine gfs_inode_read and gfs_inode_get rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 39/66] libgfs2: move gfs1 functions from edit to libgfs2 rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 40/66] gfs2_edit savemeta: save_inode_data backward for gfs1 rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 41/66] libgfs2: expand capabilities to operate on gfs1 rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 42/66] fsck.gfs2: Combine block and char device inode types rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 43/66] fsck.gfs2: four-step duplicate elimination process rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 44/66] fsck.gfs2: Add ability to check gfs1 file systems rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 45/66] fsck.gfs2: Remove bad inodes from duplicate tree rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 46/66] fsck.gfs2: Handle duplicate reference to dinode blocks rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 47/66] fsck.gfs2: Bad extended attributes not deleted properly rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 48/66] libgfs2: Make rebuild functions not re-read ip rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 49/66] fsck.gfs2: Shorten debug output rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 50/66] fsck.gfs2: Increment link count reporting wrong dinode rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 51/66] fsck.gfs2: system dinodes take priority over user dinodes rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 52/66] fsck.gfs2: Recognize partially gfs2-converted dinodes rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 53/66] fsck.gfs2: Print step 2 duplicate debug msg first rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 54/66] fsck.gfs2: pass1c counts percentage backward rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 55/66] fsck.gfs2: Speed up rangecheck functions rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 56/66] libgfs2: Make in-core rgrps use rbtree rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 57/66] fsck.gfs2: Fix memory leaks rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 58/66] Change man pages and gfs2_convert messages to include GFS rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 59/66] gfs2_edit: Fix memory leaks rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 60/66] fsck.gfs2: Journals not properly checked rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 61/66] fsck.gfs2: Rearrange block types to group all inode types rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 62/66] fsck.gfs2: Fix initialization error return codes rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 63/66] fsck.gfs2: Don't use strerror for libgfs2 errors rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 64/66] fsck.gfs2: Fix memory leak in initialize.c rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 65/66] fsck.gfs2: Add return code checks and initializations rpeterso
2012-01-20 15:10 ` [Cluster-devel] [PATCH 66/66] libgfs2: Fix null pointer dereference in linked_leaf_search rpeterso
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1327072247-26275-18-git-send-email-rpeterso@redhat.com \
--to=rpeterso@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).