Re: [PATCH] Revert "net: pskb_trim_rcsum() and CHECKSUM_COMPLETE are friends"

On 06/19/2018 03:32 PM, Andreas Schwab wrote: > On Jun 19 2018, Eric Dumazet wrote: > >> diff --git a/drivers/net/ethernet/sun/sungem.c >> b/drivers/net/ethernet/sun/sungem.c >> index >> 7a16d40a72d13cf1d522e8a3a396c826fe76f9b9..672d6748ab44f0890e92d5ca55d6ff6834c20dc9 >> 100644 >> --- a/dr

Re: [PATCH] Revert "net: pskb_trim_rcsum() and CHECKSUM_COMPLETE are friends"

On Jun 19 2018, Eric Dumazet wrote: > diff --git a/drivers/net/ethernet/sun/sungem.c > b/drivers/net/ethernet/sun/sungem.c > index > 7a16d40a72d13cf1d522e8a3a396c826fe76f9b9..672d6748ab44f0890e92d5ca55d6ff6834c20dc9 > 100644 > --- a/drivers/net/ethernet/sun/sungem.c > +++ b/drivers/net/etherne

Re: [PATCH] Revert "net: pskb_trim_rcsum() and CHECKSUM_COMPLETE are friends"

On 06/19/2018 01:10 PM, Eric Dumazet wrote: > > > On 06/19/2018 12:10 PM, Andreas Schwab wrote: >> On Jun 18 2018, Eric Dumazet wrote: >> >>> DUMP_PREFIX_ADDRESS might give us more information (say alignment problem, >>> or crossing page boundaries) >> >> DUMP_PREFIX_ADDRESS is useless for t

Re: [PATCH] Revert "net: pskb_trim_rcsum() and CHECKSUM_COMPLETE are friends"

On 06/19/2018 12:10 PM, Andreas Schwab wrote: > On Jun 18 2018, Eric Dumazet wrote: > >> DUMP_PREFIX_ADDRESS might give us more information (say alignment problem, >> or crossing page boundaries) > > DUMP_PREFIX_ADDRESS is useless for that purpose. > > Here are some samples of broken csums:

Re: [PATCH] Revert "net: pskb_trim_rcsum() and CHECKSUM_COMPLETE are friends"

On Jun 18 2018, Eric Dumazet wrote: > DUMP_PREFIX_ADDRESS might give us more information (say alignment problem, or > crossing page boundaries) DUMP_PREFIX_ADDRESS is useless for that purpose. Here are some samples of broken csums: [ 853.849225] sungem: sungem wrong csum : 9886/07be, len 94

Re: [PATCH] Revert "net: pskb_trim_rcsum() and CHECKSUM_COMPLETE are friends"

On 06/18/2018 04:29 PM, Eric Dumazet wrote: > > > On 06/18/2018 11:45 AM, Mathieu Malaterre wrote: > >> >> Here is what I get on my side >> >> [ 53.628847] sungem: sungem wrong csum : 4e04/f97, len 64 bytes >> [ 53.667063] sungem: sungem wrong csum : eea8/6eec, len 149 bytes >> [ 58.648

Re: [PATCH] Revert "net: pskb_trim_rcsum() and CHECKSUM_COMPLETE are friends"

On 06/18/2018 11:45 AM, Mathieu Malaterre wrote: > > Here is what I get on my side > > [ 53.628847] sungem: sungem wrong csum : 4e04/f97, len 64 bytes > [ 53.667063] sungem: sungem wrong csum : eea8/6eec, len 149 bytes > [ 58.648952] sungem: sungem wrong csum : 2095/3d06, len 64 bytes >

Re: [PATCH] Revert "net: pskb_trim_rcsum() and CHECKSUM_COMPLETE are friends"

On Mon, Jun 18, 2018 at 8:18 PM Eric Dumazet wrote: > > > > On 06/18/2018 10:54 AM, Andreas Schwab wrote: > > On Jun 17 2018, Eric Dumazet wrote: > > > >> Oh this is silly, please try : > >> > >> diff --git a/net/core/skbuff.c b/net/core/skbuff.c > >> index > >> c642304f178ce0a4e1358d59e45032a39

Re: [PATCH] Revert "net: pskb_trim_rcsum() and CHECKSUM_COMPLETE are friends"

On Mon, Jun 18, 2018 at 7:54 PM Andreas Schwab wrote: > > On Jun 17 2018, Eric Dumazet wrote: > > > Oh this is silly, please try : > > > > diff --git a/net/core/skbuff.c b/net/core/skbuff.c > > index > > c642304f178ce0a4e1358d59e45032a39f76fb3f..54dd9c18ecad817812898d6f335e1794a07dabbe > > 1006

Re: [PATCH] Revert "net: pskb_trim_rcsum() and CHECKSUM_COMPLETE are friends"

On 06/18/2018 10:54 AM, Andreas Schwab wrote: > On Jun 17 2018, Eric Dumazet wrote: > >> Oh this is silly, please try : >> >> diff --git a/net/core/skbuff.c b/net/core/skbuff.c >> index >> c642304f178ce0a4e1358d59e45032a39f76fb3f..54dd9c18ecad817812898d6f335e1794a07dabbe >> 100644 >> --- a/n

Re: [PATCH] Revert "net: pskb_trim_rcsum() and CHECKSUM_COMPLETE are friends"

On Jun 17 2018, Eric Dumazet wrote: > Oh this is silly, please try : > > diff --git a/net/core/skbuff.c b/net/core/skbuff.c > index > c642304f178ce0a4e1358d59e45032a39f76fb3f..54dd9c18ecad817812898d6f335e1794a07dabbe > 100644 > --- a/net/core/skbuff.c > +++ b/net/core/skbuff.c > @@ -1845,10 +18

Re: [PATCH] Revert "net: pskb_trim_rcsum() and CHECKSUM_COMPLETE are friends"

On 06/17/2018 03:27 AM, Andreas Schwab wrote: > > That doesn't change anything. OK, thanks ! Oh this is silly, please try : diff --git a/net/core/skbuff.c b/net/core/skbuff.c index c642304f178ce0a4e1358d59e45032a39f76fb3f..54dd9c18ecad817812898d6f335e1794a07dabbe 100644 --- a/net/core/skb

Re: [PATCH] Revert "net: pskb_trim_rcsum() and CHECKSUM_COMPLETE are friends"

On Jun 16 2018, Eric Dumazet wrote: > I would try something like : > > Basically do not bother using CHECKSUM_COMPLETE for small frames that might > have been padded. > > Since we need to bring one cache line in eth_type_trans() and further header > processing, > computing the checksum in softw

Re: [PATCH] Revert "net: pskb_trim_rcsum() and CHECKSUM_COMPLETE are friends"

On Jun 16 2018, Mathieu Malaterre wrote: > That's odd since it seems to only affect g4+sungem user. None of the > ppc64 seems to be having it. I'm also seeing it on a PowerMac G5. Andreas. -- Andreas Schwab, sch...@linux-m68k.org GPG Key fingerprint = 7578 EB47 D4E5 4D69 2510 2552 DF73 E780

Re: [PATCH] Revert "net: pskb_trim_rcsum() and CHECKSUM_COMPLETE are friends"

On 06/16/2018 12:14 AM, Mathieu Malaterre wrote: > Hi Eric, > > On Fri, Jun 15, 2018 at 9:14 PM Eric Dumazet wrote: >> >> >> >> On 06/15/2018 11:56 AM, Mathieu Malaterre wrote: >>> This reverts commit 88078d98d1bb085d72af8437707279e203524fa5. >>> >>> It causes regressions for people using chip

Re: [PATCH] Revert "net: pskb_trim_rcsum() and CHECKSUM_COMPLETE are friends"

Hi Eric, On Fri, Jun 15, 2018 at 9:14 PM Eric Dumazet wrote: > > > > On 06/15/2018 11:56 AM, Mathieu Malaterre wrote: > > This reverts commit 88078d98d1bb085d72af8437707279e203524fa5. > > > > It causes regressions for people using chips driven by the sungem > > driver. Suspicion is that the skb->