I can confirm that this patch fixes the issue, thanks!
Den ons 18 okt. 2023 kl 13:15 skrev Jeff Layton <jlayton@xxxxxxxxxx>:
>
> Klara reported seeing mangled mtimes when dealing with FAT. Fix the
> braino in the FAT conversion to the new timestamp accessors.
>
> Fixes: e57260ae3226 (fat: convert to new timestamp accessors)
> Reported-by: Klara Modin <klarasmodin@xxxxxxxxx>
> Signed-off-by: Jeff Layton <jlayton@xxxxxxxxxx>
> ---
> This patch fixes the bug that Klara reported late yesterday. The issue
> is a bad by-hand conversion of __fat_write_inode to the new timestamp
> accessor functions.
>
> Christian, this patch should probably be squashed into e57260ae3226.
>
> Thanks!
> Jeff
> ---
> fs/fat/inode.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/fs/fat/inode.c b/fs/fat/inode.c
> index aa87f323fd44..1fac3dabf130 100644
> --- a/fs/fat/inode.c
> +++ b/fs/fat/inode.c
> @@ -888,9 +888,9 @@ static int __fat_write_inode(struct inode *inode, int wait)
> raw_entry->size = cpu_to_le32(inode->i_size);
> raw_entry->attr = fat_make_attrs(inode);
> fat_set_start(raw_entry, MSDOS_I(inode)->i_logstart);
> + mtime = inode_get_mtime(inode);
> fat_time_unix2fat(sbi, &mtime, &raw_entry->time,
> &raw_entry->date, NULL);
> - inode_set_mtime_to_ts(inode, mtime);
> if (sbi->options.isvfat) {
> struct timespec64 ts = inode_get_atime(inode);
> __le16 atime;
>
> ---
> base-commit: fea0e8fc7829dc85f82c8a1a8249630f6fb85553
> change-id: 20231018-amtime-24d2effcc9a9
>
> Best regards,
> --
> Jeff Layton <jlayton@xxxxxxxxxx>
>
