Re: 2.6.11 oops in skb_drop_fraglist

Andrew Morton wrote:

Chuck Lever [off-list ref] wrote:

quoted

testing NFS client workloads on a dual Pentium-III system running 2.6.11 
with some NFS patches.  i hit this oops while doing simple-minded ftps 
and tars.

the system locks up once or twice a day under this workload.  this is 
the first time i had the console and captured the oops output.

Chuck, I didn't see any followup to this.  Is it still happening in current
kernels?

i have not been able to reproduce it with the aforementioned NFS patches 
removed.  i'm now convinced it was a bug in one of the NFS patches i had 
applied, even though none of them come near the fraglist stuff, but i 
haven't had a chance to nail it down.

i had implemented a patch to cause the RPC client to reuse the port 
number when reconnecting to the server after the server drops the 
connection... this is a standard practice for other RPC implementations. 
  i suspect it was that patch that was causing the trouble.

thanks for the follow-up!