ext4: don't report EOPNOTSUPP errors from discard
authorJan Kara <jack@suse.cz>
Tue, 13 Feb 2024 10:16:01 +0000 (11:16 +0100)
committerTheodore Ts'o <tytso@mit.edu>
Thu, 7 Mar 2024 18:32:54 +0000 (13:32 -0500)
When ext4 is mounted without journal, with discard mount option, and on
a device not supporting trim, we print error for each and every freed
extent. This is not only useless but actively harmful. Instead ignore
the EOPNOTSUPP error. Trim is only advisory anyway and when the
filesystem has journal we silently ignore trim error as well.

Signed-off-by: Jan Kara <jack@suse.cz>
Reviewed-by: Zhang Yi <yi.zhang@huawei.com>
Link: https://lore.kernel.org/r/20240213101601.17463-1-jack@suse.cz
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
fs/ext4/mballoc.c

index bd26f2a9075129cf1a5a25ddc1c2cb23b3d8730a..85a91a61b761c91594143e0b752348fa450c85ae 100644 (file)
@@ -6496,7 +6496,13 @@ do_more:
                if (test_opt(sb, DISCARD)) {
                        err = ext4_issue_discard(sb, block_group, bit,
                                                 count_clusters, NULL);
-                       if (err && err != -EOPNOTSUPP)
+                       /*
+                        * Ignore EOPNOTSUPP error. This is consistent with
+                        * what happens when using journal.
+                        */
+                       if (err == -EOPNOTSUPP)
+                               err = 0;
+                       if (err)
                                ext4_msg(sb, KERN_WARNING, "discard request in"
                                         " group:%u block:%d count:%lu failed"
                                         " with %d", block_group, bit, count,