3.13-rc1: eth0 hw csum failure on powerpc (bisected)

[Christian Kujau]

On Sat, 30 Nov 2013 at 23:05, Andreas Schwab wrote:
> >> > [...]
> >> >  Freeing unused kernel memory: 204K (c06ea000 - c071d000)
> >> >  eth0: hw csum failure
> >> >  CPU: 0 PID: 0 Comm: swapper Not tainted 3.13.0-rc2 #1
> >> >  Call Trace:
> >> 
> >> Btw, full dmesg & .config:  http://nerdbynature.de/bits/3.13-rc1/
> >
> > Oh, this has already been reported earlier this week:
> >
> > https://lists.ozlabs.org/pipermail/linuxppc-dev/2013-November/113584.html
> 
> No solution yet, though, 3.13-rc2 is still broken.

As 3.13-rc3 was still broken, I tried a git-bisect, resulting in:


018c5bba052b3a383d83cf0c756da0e7bc748397 is the first bad commit
commit 018c5bba052b3a383d83cf0c756da0e7bc748397
Author: David S. Miller <davem at davemloft.net>
Date:   Fri Nov 15 21:11:16 2013 -0500

    net: Handle CHECKSUM_COMPLETE more adequately in pskb_trim_rcsum().
    
    Currently pskb_trim_rcsum() just balks on CHECKSUM_COMPLETE packets
    and remarks them as CHECKSUM_NONE, forcing a software checksum
    validation later.
    
    We have all of the mechanics available to fixup the skb->csum value,
    even for complicated fragmented packets, via the helpers
    skb_checksum() and csum_sub().
    
    So just use them.
    
    Based upon a suggestion by Herbert Xu.
    
    Signed-off-by: David S. Miller <davem at davemloft.net>

:040000 040000 14e295c30c2d2a1164f4eb70aaec821fbe288c01 456c666ebbb355292b683c71c85585fc6d07e852 M      include


And sure enough, today's 3.13-rc3 with only that commit reverted boots 
just fine on this PowerBook G4 system.

Thanks,
Christian
-- 
BOFH excuse #152:

My pony-tail hit the on/off switch on the power strip.