You missed out the diffs!
I just hit a corruption problem myself, on the i386, this may be
something to do with moving up to test5 - and may or may not be
the same problem you hit. If I get that out of the way then we
can establish if you are seeing the same issue or not. Did you
select KIOBUF based I/O in the configuration options - this only
affects SCSI drives right now, but at the moment I am suspecting
it being related to the corruption I saw.
On a topic related to porting XFS to other architectures, I now
have a system which does not need the 64 bit divide and modulus
support functions. All the places XFS does this turn out to be
dividing by a 32 bit quantity and there is a function called
do_div provided for that on all architectures. This will show
up after we get the corruption sorted out.
An ia64 port has been attempted, not sure what the current status
is, and XFS does come from a 64 bit background in the first place,
all modern Irix platforms are 64 bit. However, this is a good check
that we did not lose information in the port to Linux.
-----
other things: there are some tests in cmd/xfs/stress - getting these
to build may or may not be a problem. Also check out this web site, you
can download yet more tests for xfs from here:
http://oss.sgi.com/projects/ltp/ - again porting to other platforms may
be some work.
I will take a look at packaging the changes you did for the pagebuf
debug mechanism next week.
Thanks for all your work by the way!
Steve
>
> ok - i played around a bit more with xfs on the alpha now - have
> made the test machine a bit more comfortable and looked at the
> compiler warnings - at the end of this mail is a list of changes
> i did to make the compiler happy - but please have a look at them
> if they are basically ok (you know the code - i so far only see
> the warnings :-) ... with this code it so far runs quite fine
> (no panic, no fs crashes) but it silently corrupts data and has
> other problems:
>
> * the file which was ther from the beginning (XFS containig "i am
> here") changed its contents to the ones from another file (i copied
> the /etc tree some times to the xfs filesystem and it looks like
> the XFS file now contains the contents of the later removed file
> from etc/hostname - cyan)
>
> root@cyan:/mnt/xfs# cat XFS
> i am here
> root@cyan:/mnt/xfs#
> ...
> some disk operation (cp to or on the xfs filesystem, bonnie, deleting
> files nothing acting on the file XFS)
> ...
> root@cyan:/mnt/xfs# cat XFS
> cyan
> root@cyan:/mnt/xfs#
>
> * once i had an unremovable directory (something i know from
> testing xfs about 8 weeks ago)
>
> * see the following:
>
> root@cyan:/mnt/xfs# tar xvzf /root/Bonnie.tar.Z
> Bonnie/
> Bonnie/Bonnie.c
> Bonnie/Commentary
> Bonnie/Makefile
> root@cyan:/mnt/xfs# cd Bonnie/
> root@cyan:/mnt/xfs/Bonnie# make
> cc -O2 Bonnie.c -o Bonnie
> collect2: ld terminated with signal 11 [Segmentation fault], core dumped
> make: *** [Bonnie] Error 1
> root@cyan:/mnt/xfs/Bonnie# cd /tmp
> root@cyan:/tmp# tar xvzf /root/Bonnie.tar.Z
> Bonnie/
> Bonnie/Bonnie.c
> Bonnie/Commentary
> Bonnie/Makefile
> root@cyan:/tmp# cd Bonnie/
> root@cyan:/tmp/Bonnie# make
> cc -O2 Bonnie.c -o Bonnie
> root@cyan:/tmp/Bonnie#
>
> ok - so far i got ... if anyone has any comments to the things above
> or other things which come to mind on a 64bit machine please let me
> know - i'll try my best to try out everything to get this moving
> forward ... btw. do you have xfs running on ia64 (so a 64bit
> arch of the linux port of it) ?
>
> t
>
> p.s.: btw. alpha is little endian - just as info
>
> --
> thomas.graichen@xxxxxxxxxxxxx
> Technical Director innominate AG
> Clustering & Security networking people
> tel: +49.30.308806-13 fax: -77 http://innominate.de
|