[Jaunty] Proposing some ext4 patches

[Theodore Tso]

On Thu, Jul 02, 2009 at 10:18:57PM +0200, Stefan Bader wrote:
>
> I did not get to the karmic kernel, but have now done the current Jaunty  
> against a version including the picked set of patches. Both had two tests 
> failing (213, 214). Strangely those seem to be tests that look like only 
> useful for xfs. Also in theory they seem to be requiring the fallocate 
> command which is not included in the tools. So I guess that is a rather a 
> problem with the test than one with the fs.
> So at least it seems to assure, that applying the patches does not make 
> things worse (aka causes regressions).

Hmm, I don't get a failure; I just get:

213	[not_run] xfs_io fallocate support is missing
214	[not_run] xfs_io fallocate support is missing

>From the git log in question:

    fallocate + read/write tests, ext4 regression tests
    
    New test to test basic mixed fallocate + read & write,
    includes a couple regression tests for bugs that ext4
    hit.  Uses xfs_io to generate fallocate calls, so requires
    git xfsprogs and very recent glibc at this point.
    
    Ext4 folks, this is hopefully a reasonable example of
    how to add a new test.   :)
    
    Signed-off-by: Eric Sandeen <sandeen at sandeen.net>
    Reviewed-by: Christoph Hellwig <hch at lst.de>

Note the "requires git xfsprogs and very recent glibc".  (The
fallocate() support missed latest glibc release, although it's in
glibc's CVS tree, so it requires a backport patch.)  So basically you
need the latest bleeding-edge development support for tests 213 and
214 to run, but it *should* have failed with a clean "not_run" status.

If you got something else, it probably is a bug in the tests, and we
should report it to Eric Sandeen, who coded up the test.

       	      	    	 	      	    - Ted