mke2fs -G option allows root user to set flex_bg count (power of 2).
However ext4 has bad metadata layout if we specify more than or equal to
2^32 to mke2fs -G, because of the 32bit shift operation
in ext2fs_allocate_group_table().
And the maximum block group count of ext4 is 2^32 -1 (ext4_group_t s_groups_count),
so diallow more than 2^32 flex_bg count.
Steps to reproduce:
# mke2fs -t ext4 -G 4294967296 DEV
# dumpe2fs DEV
<snip>
Flex block group size: 1 <----- flex_bg is 1!
<snip>
Group 0: (Blocks 0-32767)
Checksum 0x4afd, unused inodes 7541
Primary superblock at 0, Group descriptors at 1-1
Reserved GDT blocks at 2-59
Block bitmap at 60 (+60), Inode bitmap at 61 (+61)
Inode table at 62-533 (+62)
32228 free blocks, 7541 free inodes, 2 directories, 7541 unused inodes
Free blocks: 540-32767
Free inodes: 12-7552
Group 1: (Blocks 32768-65535) [INODE_UNINIT]
Checksum 0xc890, unused inodes 7552
Backup superblock at 32768, Group descriptors at 32769-32769
Reserved GDT blocks at 32770-32827
Block bitmap at 32828 (+60), Inode bitmap at 32829 (+61)
Inode table at 32830-33301 (+62)
32234 free blocks, 7552 free inodes, 0 directories, 7552 unused inodes
Free blocks: 33302-65535
Free inodes: 7553-15104
<snip>
Signed-off-by: Akira Fujita <[email protected]>
---
misc/mke2fs.c | 6 ++++++
1 file changed, 6 insertions(+)
diff --git a/misc/mke2fs.c b/misc/mke2fs.c
index b451cc3..b9145d1 100644
--- a/misc/mke2fs.c
+++ b/misc/mke2fs.c
@@ -1605,6 +1605,12 @@ profile_error:
_("flex_bg size must be a power of 2"));
exit(1);
}
+ if (flex_bg_size > MAX_32_NUM) {
+ com_err(program_name, 0,
+ _("flex_bg size (%lu) must be less than"
+ " or equal to 2^31"), flex_bg_size);
+ exit(1);
+ }
break;
case 'i':
inode_ratio = strtoul(optarg, &tmp, 0);
On Wed, Jun 11, 2014 at 08:37:22AM +0000, Akira Fujita wrote:
> mke2fs -G option allows root user to set flex_bg count (power of 2).
> However ext4 has bad metadata layout if we specify more than or equal to
> 2^32 to mke2fs -G, because of the 32bit shift operation
> in ext2fs_allocate_group_table().
> And the maximum block group count of ext4 is 2^32 -1 (ext4_group_t s_groups_count),
> so diallow more than 2^32 flex_bg count.
>
> Steps to reproduce:
>
> # mke2fs -t ext4 -G 4294967296 DEV
>
> # dumpe2fs DEV
> <snip>
> Flex block group size: 1 <----- flex_bg is 1!
> <snip>
> Group 0: (Blocks 0-32767)
> Checksum 0x4afd, unused inodes 7541
> Primary superblock at 0, Group descriptors at 1-1
> Reserved GDT blocks at 2-59
> Block bitmap at 60 (+60), Inode bitmap at 61 (+61)
> Inode table at 62-533 (+62)
> 32228 free blocks, 7541 free inodes, 2 directories, 7541 unused inodes
> Free blocks: 540-32767
> Free inodes: 12-7552
> Group 1: (Blocks 32768-65535) [INODE_UNINIT]
> Checksum 0xc890, unused inodes 7552
> Backup superblock at 32768, Group descriptors at 32769-32769
> Reserved GDT blocks at 32770-32827
> Block bitmap at 32828 (+60), Inode bitmap at 32829 (+61)
> Inode table at 32830-33301 (+62)
> 32234 free blocks, 7552 free inodes, 0 directories, 7552 unused inodes
> Free blocks: 33302-65535
> Free inodes: 7553-15104
> <snip>
>
> Signed-off-by: Akira Fujita <[email protected]>
Looks reasonable to me, you can add my Reviewed-by.
--D
> ---
> misc/mke2fs.c | 6 ++++++
> 1 file changed, 6 insertions(+)
> diff --git a/misc/mke2fs.c b/misc/mke2fs.c
> index b451cc3..b9145d1 100644
> --- a/misc/mke2fs.c
> +++ b/misc/mke2fs.c
> @@ -1605,6 +1605,12 @@ profile_error:
> _("flex_bg size must be a power of 2"));
> exit(1);
> }
> + if (flex_bg_size > MAX_32_NUM) {
> + com_err(program_name, 0,
> + _("flex_bg size (%lu) must be less than"
> + " or equal to 2^31"), flex_bg_size);
> + exit(1);
> + }
> break;
> case 'i':
> inode_ratio = strtoul(optarg, &tmp, 0);
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html