On 14 Aug 2004, Mikko Huhtala <[EMAIL PROTECTED]> wrote:
> 
> Thanks for the link to the patch. I'll try it if I can find the time.
> 
> Martin Pool writes:
>  > On 10 Aug 2004, Mikko Huhtala <[EMAIL PROTECTED]> wrote:
>  > > 
>  > > I tried distcc 2.17 with the environment variables DISTCC_MMAP=0
>  > > DISTCC_SENDFILE=0 DISTCC_TCP_CORK=0 set. I still kernel panics on the
>  > > latest Fedora Core 1 kernel. This time the ethernet driver int the
>  > > panicked system was 3c59x.
>  > 
>  > That's no good; that means it's probably not the same bug as before.
>  > Do you get an oops or some other backtrace from the kerne?  You should
>  > probably report it to the kernel list.
>  > 
> 
> I'm no kernel programmer, but as far as I can see the problem (and the
> patch) is in page allocation during zero-copy operations, so the bug
> could affect any driver that does zero-copy, right? Certainly I see
> exactly the same behaviour with both e1000 and 3c59x.

Right.  With both mmap and sendfile turned off I don't think there are
any zerocopy operations taking place though, so that's why I am afraid
it might be a different bug.

--
Martin

Attachment: signature.asc
Description: Digital signature

__ 
distcc mailing list            http://distcc.samba.org/
To unsubscribe or change options: 
http://lists.samba.org/mailman/listinfo/distcc

Reply via email to