Re: bug with large_dir in 5.12.17
From: Благодаренко Артём <hidden>
Date: 2021-08-04 21:15:50
Hello Teodore, Your one-line fix looks good. I have tested it. 1560000 names created successfully. But the patch with refactoring doesn’t work. I got this messages 1480000 names created 1520000 names created ln: failed to access 'n0000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000001520519': Bad message ln: failed to access 'n0000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000001520520': Bad message [ 7699.212018] EXT4-fs error (device loop0): dx_probe:887: inode #2: block 144843: comm ln: Directory index failed checksum [ 7699.216001] EXT4-fs error (device loop0): dx_probe:887: inode #2: block 144843: comm ln: Directory index failed checksum I have no objections to send your one-line fix, but we need to double check refactoring. Best regards, Artem Blagodarenko
quoted hunk ↗ jump to hunk
On 4 Aug 2021, at 22:25, Theodore Ts'o [off-list ref] wrote: On Thu, Jul 29, 2021 at 10:23:35PM +0300, Благодаренко Артём wrote:quoted
Hello, It looks like the fix b5776e7524afbd4569978ff790864755c438bba7 "ext4: fix potential htree index checksum corruption” introduced this regression. I reverted it and my test from previous message passed the dangerous level of 1570000 names count. Now test is still in progress. 2520000 names are already created. I am searching the way to fix this. Best regards, Artem Blagodarenko.Hi Artem, did you have a chance to take a look at some of the possible fixes which I floated on this thread? Do you have any objections if I take this and send it to Linus? Thanks, - Ted From fa8db30806b4e83981c65f18f98de33f804012d9 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o <tytso@mit.edu> Date: Wed, 4 Aug 2021 14:23:55 -0400 Subject: [PATCH] ext4: fix potential htree correuption when growing large_dir directories MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Commit b5776e7524af ("ext4: fix potential htree index checksum corruption) removed a required restart when multiple levels of index nodes need to be split. Fix this to avoid directory htree corruptions when using the large_dir feature. Cc: stable@kernel.org # v5.11 Cc: Благодаренко Артём <redacted> Fixes: b5776e7524af ("ext4: fix potential htree index checksum corruption) Reported-by: Denis <redacted> Signed-off-by: Theodore Ts'o <tytso@mit.edu> --- fs/ext4/namei.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index 5fd56f616cf0..f3bbcd4efb56 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c@@ -2517,7 +2517,7 @@ static int ext4_dx_add_entry(handle_t *handle, struct ext4_filename *fname,goto journal_error; err = ext4_handle_dirty_dx_node(handle, dir, frame->bh); - if (err) + if (restart || err) goto journal_error; } else { struct dx_root *dxroot; -- 2.31.0