在 2026/8/26 19:21, Namjae Jeon 写道:
> On Fri, Aug 21, 2026 at 2:32 PM Baolin Liu <liubaolin12138@163.com> wrote:
>>
>> From: Baolin Liu <liubaolin@kylinos.cn>
>>
>> This series fixes a number of problems on error paths in the ntfs
>> driver: errors that are swallowed or collapsed so operations report
>> success or a wrong errno (patches 1-6), a kmap_local leak (patch 7),
>> and free cluster miscounting on failed frees (patches 8-9).
>>
>> Baolin Liu (9):
>> ntfs: return DT_UNKNOWN on inode lookup failure in readdir
>> ntfs: fail EA update when reading $EA_INFORMATION fails
>> ntfs: propagate reparse index insertion failure
>> ntfs: return -ERANGE for undersized xattr buffer
>> ntfs: preserve error code in ntfs_resident_attr_record_add()
>> ntfs: return real error from ntfs_non_resident_attr_record_add()
>> ntfs: fix kmap_local leak in write_mft_record_nolock() error paths
>> ntfs: only count successfully cleared runs when freeing clusters
>> ntfs: skip free cluster decrement when rollback fails
>
> Applied them except 0002 patch to #ntfs-next.
> Could you please recreate 0002 patch based on the #ntfs-next branch?
>
> Thanks.
Hi Namjae,
The 0002 patch has been recreated on top of ntfs-next and sent as
"[PATCH v2] ntfs: leave HasEA flag untouched on setxattr failure".
Note that the subject changed: on ntfs-next, the -EIO return for a
failed $EA_INFORMATION read is already handled by 56cc42a7b387
("ntfs: return errors from ntfs_attr_readall"), so only the HasEA
flag part of the original patch remains.
Thanks,
Baolin