cache_purge > cache_zap segmentation fault

[Ali Bahar]

Hi all,

this post may be of interest to people familiar with the filesystem code. 

It concerns a recurring seg fault which occurs in the kernel (a
FreeBSD 4.7 kernel, with a new networking driver module). Similar
areas seem to have been noted before in this list, and it may still be a
valid bug.

Two threads are prevalent (mainly the first, but a mix of the two has
been seen as well):

  syscall2 > open > vn_open > namei > lookup > ufs_vnoperate > 
    vfs_cache_lookup > ufs_vnoperate > ufs_lookup > ffs_vget > getnewvnode >
      cache_purge > cache_zap

  syscall2 > stat > namei > lookup > ufs_vnoperate > 
    vfs_cache_lookup > ufs_vnoperate > ufs_lookup >
      cache_purge > cache_zap

I am not at all familiar with these. But I am doing remote debugging
of the kernel, so I could be of help to anyone wishing to debug this. I
could dump data structs etc in gdb.

Context: any kind of file access may lead to this. No specific
activity is needed. Eg left idle over the weekend, I've found it
crashed on Monday.

thanks and regards,
ali

P.S. Gnats does not appear to show any such past bug. But I have seen
posts by Matt Dillon and phk discussing similar territory.

P.P.S. It's been occuring intermittently, and increasingly,
recently. (Due to its increased prevalence, I even suspected that the
frequency of kernel crashes, might have corrupted the filesystem in a
way ignorable/imperceptible by fsck/me!)


-- 
             Jesus was an Arab.