I just happened to see your post on Linux kernel - you should probably
ask on linux-xfs@xxxxxxxxxxx which is the mailing list for XFS. The
volume on the kernel list is a bit much to always catch mailings
Anyway, if you downloaded a patch file you probably have out of date
code, the best way to get uptodate code is to use the cvs repository.
If you were attempting to use the patch and using XFS as a root filesystem
you will have problems - including a panic during boot which looked similar
to what you reported, these are fixed in the cvs version.
for instructions on how to download the repository.
As for kernel size, we do have people who have built xfs into their
kernel and run that way sucessfully, it maybe that you have a lot of
other code compiled into your kernel. I personally always use modules,
although I just did build and boot a kernel with XFS build in.
> ls -l vmlinux
-rwxr-xr-x 1 lord network 5279892 Apr 25 10:37 vmlinux*
> size vmlinux
text data bss dec hex filename
2083734 640497 568064 3292295 323c87 vmlinux
>> Yesterday, I downloaded the latest version of xfs (a patch to
>> 2.3.99-pre2) and compiled it with xfs statically linked in. After a
>> reboot, I got as long as "Checking if this processor honours the WP
>> bit even in supervisor mode..." and ... panic Found I had been a
>> little too quick when I checked the size of /usr/src/linux/vmlinux -
>> 3.8MB... A little wc told me the xfs filesystem consisted of no less
>> than 110.000 lines of code.
>> Anyone got an idea of how to get this stuff compiled in? It's probably
>> possible doing it in a modular fashion, but how if I want to use xfs
>> on / ?
>> Please CC: to me as I'm not on the list