Randy Dunlap <randy.dunlap@xxxxxxxxxx> writes:
> On Wed, 29 Aug 2007 17:27:07 +0800 常青 wrote:
>
> [adding maintainer cc:]
>
>> I've been encountered with a FAT panic, saying,
>> =====================================
>> Filesystem panic (dev mmcblk0p1)
>> fat_get_cluster: invalid cluster chain (i_pos 16021)
>> File system has been set read-only
>> =====================================
>>
>> After tracking into kernel code, I find this is printed out by
>> fat_get_cluster@fs/fat/cache.c ,
>> =====================================
>> nr = fat_access(sb, *dclus, -1);
>> if (nr < 0)
>> return nr;
>> else if (nr == FAT_ENT_FREE) {
>> fat_fs_panic(sb, "%s: invalid cluster chain"
>> " (i_pos %lld)", __FUNCTION__,
>> MSDOS_I(inode)->i_pos);
>> return -EIO;
>> } else if (nr == FAT_ENT_EOF) {
>> fat_cache_add(inode, &cid);
>> return FAT_ENT_EOF;
>> }
>> =====================================
>>
>> Could you please help to explain what does FAT_ENT_FREE mean? and what
>> condition will lead to (nr == FAT_ENT_FREE) to be true?
>
> #define FAT_ENT_FREE (0)
>
> It means that the FAT filesystem cluster chain came upon an entry
> with a value of 0, which is invalid. I.e., it means that this
> cluster (of blocks) is free, not part of a file, but the code was
> trying to follow a chain of clusters for a file, in which case it
> should never see a 0, just valid cluster links until EOF (value)
> is reached.
>
> Sounds like the FAT is corrupted.
Yes.
> Hirofumi, is there second copy of the FAT table? If so, is there
> a way to use it? or just dd copy#2 to copy#1?
If this is FAT12, it doesn't have a copy of FAT. IIRC, "dosfsck -v -n -r"
will check and display whether FATs have difference, and if those have
the difference, it will ask which FAT you use.
Thanks.
--
OGAWA Hirofumi <hirofumi@xxxxxxxxxxxxxxxxxx>
-
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
