Re: [PATCH] ext4: add error handling when discarding is fail inFITRIM ioctl

[Namjae Jeon]

2012/1/24 Jan Kara <jack@xxxxxxx>:
> On Sat 21-01-12 02:56:12, Namjae Jeon wrote:
>> Although free extents is proper not trimmed(mmc driver return error code while sending trim command), currently FITRIM ioctl return success.
>> I tried to add exception routine to inform user error code.
>>
>> #> ./fitrim_test
>> end_request: I/O error, dev mmcblk0, sector 27232
>> EXT4-fs warning (device mmcblk0): ext4_trim_all_free:4857: Discard command returned error -5
>> #>
>>
>> Signed-off-by: Namjae Jeon <linkinjeon@xxxxxxxxx>
>> Signed-off-by: Amit Sahrawat <amit.sahrawat83@xxxxxxxxx>
>> ---
>> Âfs/ext4/mballoc.c | Â 22 +++++++++++++++-------
>> Â1 files changed, 15 insertions(+), 7 deletions(-)
>>
>> diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
>> index cb990b2..09cc09a 100644
>> --- a/fs/ext4/mballoc.c
>> +++ b/fs/ext4/mballoc.c
>> @@ -4901,10 +4901,11 @@ error_return:
>> Â * one will allocate those blocks, mark it as used in buddy bitmap. This must
>> Â * be called with under the group lock.
>> Â */
>> -static void ext4_trim_extent(struct super_block *sb, int start, int count,
>> +static int ext4_trim_extent(struct super_block *sb, int start, int count,
>> Â Â Â Â Â Â Â Â Â Â Â Â Â Âext4_group_t group, struct ext4_buddy *e4b)
>> Â{
>> Â Â Â struct ext4_free_extent ex;
>> + Â Â int err;
>>
>> Â Â Â trace_ext4_trim_extent(sb, group, start, count);
>>
>> @@ -4920,9 +4921,10 @@ static void ext4_trim_extent(struct super_block *sb, int start, int count,
>> Â Â Â Â*/
>> Â Â Â mb_mark_used(e4b, &ex);
>> Â Â Â ext4_unlock_group(sb, group);
>> - Â Â ext4_issue_discard(sb, group, start, count);
>> + Â Â err = ext4_issue_discard(sb, group, start, count);
>> Â Â Â ext4_lock_group(sb, group);
>> Â Â Â mb_free_blocks(NULL, e4b, start, ex.fe_len);
>> + Â Â return err;
>> Â}
>>
>> Â/**
>> @@ -4978,15 +4980,21 @@ ext4_trim_all_free(struct super_block *sb, ext4_group_t group,
>> Â Â Â Â Â Â Â next = mb_find_next_bit(bitmap, max, start);
>>
>> Â Â Â Â Â Â Â if ((next - start) >= minblocks) {
>> - Â Â Â Â Â Â Â Â Â Â ext4_trim_extent(sb, start,
>> - Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Ânext - start, group, &e4b);
>> - Â Â Â Â Â Â Â Â Â Â count += next - start;
>> + Â Â Â Â Â Â Â Â Â Â ret = ext4_trim_extent(sb, start,
>> + Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â next - start, group, &e4b);
>> + Â Â Â Â Â Â Â Â Â Â if (ret < 0) {
>> + Â Â Â Â Â Â Â Â Â Â Â Â Â Â if (ret != -EOPNOTSUPP)
>> + Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â ext4_warning(sb, "Discard command "
>> + Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â"returned error %d\n", ret);
>> + Â Â Â Â Â Â Â Â Â Â Â Â Â Â break;
>> + Â Â Â Â Â Â Â Â Â Â } else
>> + Â Â Â Â Â Â Â Â Â Â Â Â Â Â count += next - start;
>> Â Â Â Â Â Â Â }
>> Â Â Â Â Â Â Â free_count += next - start;
>> Â Â Â Â Â Â Â start = next + 1;
>>
>> Â Â Â Â Â Â Â if (fatal_signal_pending(current)) {
>> - Â Â Â Â Â Â Â Â Â Â count = -ERESTARTSYS;
>> + Â Â Â Â Â Â Â Â Â Â ret = -ERESTARTSYS;
>> Â Â Â Â Â Â Â Â Â Â Â break;
>> Â Â Â Â Â Â Â }
>>
>> @@ -5009,7 +5017,7 @@ out:
>> Â Â Â ext4_debug("trimmed %d blocks in the group %d\n",
>> Â Â Â Â Â Â Â count, group);
>>
>> - Â Â return count;
>> + Â Â return (ret < 0) ? ret : count;
> ÂI think you need to initialize ret to 0 at the beginning of the function.
> Otherwise you could use it uninitialized. Otherwise the patch looks good to
> me and makes things consistent with ext3 so after fixing that bug you can add:
> ÂReviewed-by: Jan Kara <jack@xxxxxxx>
Hi. Jan.
Thanks for your review.~
>
> Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â ÂHonza
> --
> Jan Kara <jack@xxxxxxx>
> SUSE Labs, CR
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/