[OpenAFS-devel] [FOR TESTING] OpenAFS 1.3.87 RPMs for RHEL4 (i386, x86_64)

[chas williams - CONTRACTOR]

In message <ubzy876fnlg.fsf@metacircular-evaluator.mit.edu>,Alex Rolfe writes:
>Using the rhel4 2.6.9-11EL.smp kernel on x86_64, I could cause an oops
>and kernel panic fairly reliably by reading a series of large files.  I

how big and how many?

>can cause an immediate oops and panic trying to use 
>"-memcache -stat 3600 -dcache 3600 -daemons 5 -volumes 196 -chunksize 20
>-dynroot -fakestat -afsdb -nosettime -verbose"

you asked for 2^20 * 3600 bytes of memory for the memcache.  about
3.6G.  do you have more than 3.6G of memory in the machine?

>Traceback:
>
>cache_alloc_refill+609
>__kmalloc+123
>__vmalloc+110
>:libafs:osi_linux_alloc+104
>:libafs:afs_INitMemCache+168
>:libafs:afs_syscall_call+3056
>avc_has_perm+70
>:libafs:afs_syscall+541
>se_linux_file_ioctl+711
>:libafs:afs_ioctl+158
>strncpy_from_user+74
>sys_ioctl+811
>system_call+126

this traceback looks unusual.   is this what an oops looks like
on x86_64?