I've got it forward-ported to the latest errata kernel as well, I'll
update that on the ftp site soon (BTW:
ftp://oss.sgi.com/projects/xfs/testing/RHEL for those just joining
I'll have to sort out which of the xfsqa failures are unique to RHEL,
and which are having trouble in general.
Also a note, please read the README in the above dir for details on
what's different in this kernel - of note, devfs is turned on, and this
version of xfs is a bit different from what's currently in cvs, that may
change in the future but this is the result of an internal build, tossed
out onto ftp for the impatient... :)
On Fri, 2004-01-30 at 10:16, Chris Green wrote:
> I downloaded kernel-2.4.21-4.0.1.EL.sgi1.src.rpm and accompanying
> xfs-modules src.rpm and built them on a Fermi Linux LTS3.0 system (based
> on RHEL3.0 update1), and then downloaded, built and ran the xfstests
> suite. Being ambitious, I had started to try to forward-port the kernel to
> the latest RHEL kernel, 2.4.21-9.EL, but reverted to the packaged
> 2.4.21-4.0.1.EL when I encountered xfstest failures.
> The good news is that the failures are the same betwen the "official"
> 2.4.21-4.0.1.EL and my attempt to port to 2.4.21-9.EL. The bad news is
> that there were failures. I tested this on a Polywell server (single-CPU,
> Hyperthreading) with SATA RAID (2x160GB on a 3Ware 8006-2, 11x160GB on
> each of two 8506-12s) with the test partitions, initially ~70GB each and
> one final test with 640MB each, both on the RAID1 system disk.
> The test results are attached above. Apart from the failure of test 005
> which was expected, and the failure of any test that used xfsdump or
> xfsrestore since I don't have access to a tape device, I believe the
> worrying failures are 073, 079 and 080. The failures are identical (within
> expected differences like $$-expansion) between the three attached log
> At the request of Dan Yocum, our local informal contact with the SGI team,
> I'm posting these experiences on the linux-xfs list. If I have mis-run the
> tests in any way (./check -g auto) or misinterpreted the results, I would
> be happy to hear about it.
> Comments appreciated,
Eric Sandeen [C]XFS for Linux http://oss.sgi.com/projects/xfs
sandeen@xxxxxxx SGI, Inc. 651-683-3102