On Thu, 28 Sep 2000, Marko Kreen wrote:
> root@snoopy:~# debugfs
> debugfs 1.19, 13-Jul-2000 for EXT2 FS 0.5b, 95/08/09
> debugfs: open /dev/sda6
> debugfs: stat lib/postgres/data/base
> Inode: 14479 Type: directory Mode: 0700 Flags: 0x0 Generation: 883245
> User: 31 Group: 32 Size: 1024
Ugh. With such size it should have 1 chunk, so there should be no way it
could get ext2_check_dir_entry() called at offset 1024. Weird.
Could you replace
if (!ext2_check_dir_entry("ext2_add_entry", dir,de, page))
goto out_page;
with
if (!....) {
printk(KERN_ERROR "ext2_add_link: %d %d %d %d %d %d\n",
chunk_offs,de-(char*)page_address(page),(int)n,chunk_size,
chunk-(char*)page_address(page), (int)dir->i_size);
goto out_page;
}
and see what will it do?
Look: ->i_size == 1024 => dir_chunks() == (1024+1023)>>10, i.e. 1. That
is, loop in ext2_add_link is
for (n=0; n<1; n++) {
...
}
There is only one place in ext2_add_link() that calls ext2_check_dir_entry()
You managed to get offs == 1024. It means de == page_address(page)+1024.
IOW, either de >= chunk+chunk_size and we shouldn't get into that point
in the while() {}, or chunk > page_address(page) or chunk_size > 1024.
ext2_dir_chunk(...,0,p) _must_ return page_address(*p), no matter what.
And chunk_size is a ->i_sb->s_blocksize - it would better be equal to
1024, or many things would break badly. Weird...
I have a next version of the patch, but I would really like to understand
what happens here...
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Sat Sep 30 2000 - 21:00:22 EST