64-bit e2fsprogs tree for testing

64-bit e2fsprogs tree for testing

[Valerie Aurora Henson]

Hi folks,

The e2fsprogs 64-bit branch is available for public testing.  This
branch adds support for creating, checking, etc. ext4 file systems
with more than 2^32 blocks - i.e., more than 16TB on most systems.
This is for TESTING ONLY - don't put any data you care about on it!

The public git repository is available here:

git://git.kernel.org/pub/scm/fs/ext2/val/e2fsprogs.git

The branch is "64bit".

Known bugs:

* compilation of programs using certain libe2p routines may fail 
  (chattr, lsattr)
* fsck will fail if directories contain blocks above the 32-bit
  boundary

In general, this is very lightly tested.

Please only test against the latest ext4 kernel bits from:

git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4.git

Please report bugs to this mailing list and cc'd to me if possible.

I am personally testing this using a sparse backing file on XFS, e.g.:

 # dd if=/dev/zero of=/xfs/testfile bs=4096 seek=5G count=1
 # ./misc/mke2fs /xfs/testfile
 # mount -o loop /xfs/testfile /mnt
 # dmesg | tail # to make sure it actually mounted as ext4, etc.

Unfortunately, you can't use ext4 itself to host the backing file,
since it has a limit of 2^32 blocks per file.

Thank you in advance!

-VAL

Re: 64-bit e2fsprogs tree for testing

[Valerie Aurora Henson]

On Thu, Jan 22, 2009 at 07:28:15PM -0500, Valerie Aurora Henson wrote:
> Hi folks,
> 
> The e2fsprogs 64-bit branch is available for public testing.  This
> branch adds support for creating, checking, etc. ext4 file systems
> with more than 2^32 blocks - i.e., more than 16TB on most systems.
> This is for TESTING ONLY - don't put any data you care about on it!
> 
> The public git repository is available here:
> 
> git://git.kernel.org/pub/scm/fs/ext2/val/e2fsprogs.git
> 
> The branch is "64bit".
> 
> Known bugs:
> 
> * compilation of programs using certain libe2p routines may fail 
>   (chattr, lsattr)
> * fsck will fail if directories contain blocks above the 32-bit
>   boundary

I just finished and pushed the fix for the second item (64-bit
dblists).

Note that I'll be unavailable for the next week while I recover from
shoulder surgery.

-VAL