Re: PR/52068 CVS commit: src/sys/dev/pci

[Frank Kardel <kardel%netbsd.org@localhost>]

Hi knakahara@n.o !

I checked the new version and no more checksum errors for UDP and TCP.
Thanks for the fix.

I do notice errors on ICMP though:

icmp:
        806 calls to icmp_error
        0 errors not generated because old message was icmp
        Output histogram:
                echoreply: 72
                unreach: 643
                redirect: 124
                timxceed: 3
        0 messages with bad code fields
        0 messages < minimum length
        15 bad checksums
        0 messages with bad length
...
Gateway# ping www.heise.de
PING www.heise.de (193.99.144.85): 56 data bytes
64 bytes from 193.99.144.85: icmp_seq=0 ttl=243 time=19.751926 ms
64 bytes from 193.99.144.85: icmp_seq=1 ttl=243 time=20.527536 ms
64 bytes from 193.99.144.85: icmp_seq=2 ttl=243 time=17.968474 ms
64 bytes from 193.99.144.85: icmp_seq=3 ttl=243 time=19.501118 ms
64 bytes from 193.99.144.85: icmp_seq=4 ttl=243 time=28.901272 ms
64 bytes from 193.99.144.85: icmp_seq=5 ttl=243 time=16.245018 ms
64 bytes from 193.99.144.85: icmp_seq=6 ttl=243 time=29.275564 ms
64 bytes from 193.99.144.85: icmp_seq=7 ttl=243 time=18.914712 ms
64 bytes from 193.99.144.85: icmp_seq=8 ttl=243 time=25.853516 ms
64 bytes from 193.99.144.85: icmp_seq=9 ttl=243 time=17.357554 ms
^C
----www.heise.de PING Statistics----
10 packets transmitted, 10 packets received, 0.0% packet loss
round-trip min/avg/max/stddev = 16.245018/21.429669/29.275564/4.785034 ms

icmp:
        807 calls to icmp_error
        0 errors not generated because old message was icmp
        Output histogram:
                echoreply: 72
                unreach: 644
                redirect: 124
                timxceed: 3
        0 messages with bad code fields
        0 messages < minimum length
        24 bad checksums
        0 messages with bad length

that is 10 ping exchanges (ICMP ECHO/ICMP REPLY) and 9 additional
icmp bad checksums.

That seems a bit surprising and may have been there before the fix - I 
just noticed it now.

Best regards,
  Frank

On 03/21/17 11:50, Kengo NAKAHARA wrote:
> The following reply was made to PR kern/52068; it has been noted by GNATS.
>
> From: Kengo NAKAHARA <k-nakahara%iij.ad.jp@localhost>
> To: gnats-bugs%NetBSD.org@localhost, kern-bug-people%netbsd.org@localhost, gnats-admin%netbsd.org@localhost,
>          netbsd-bugs%netbsd.org@localhost, kardel%netbsd.org@localhost
> Cc:
> Subject: Re: PR/52068 CVS commit: src/sys/dev/pci
> Date: Tue, 21 Mar 2017 19:45:16 +0900
>
>   Hi kardel@n.o,
>   
>   I fix kern/52068 by if_wm.c:r1.498 and it works well in my 82574 environment.
>   
>   Could you try it?
>   
>   
>   Thanks,
>   
>   On 2017/03/21 19:40, Kengo NAKAHARA wrote:
>   > The following reply was made to PR kern/52068; it has been noted by GNATS.
>   >
>   > From: "Kengo NAKAHARA" <knakahara%netbsd.org@localhost>
>   > To: gnats-bugs%gnats.NetBSD.org@localhost
>   > Cc:
>   > Subject: PR/52068 CVS commit: src/sys/dev/pci
>   > Date: Tue, 21 Mar 2017 10:39:52 +0000
>   >
>   >  Module Name:	src
>   >  Committed By:	knakahara
>   >  Date:		Tue Mar 21 10:39:52 UTC 2017
>   >
>   >  Modified Files:
>   >  	src/sys/dev/pci: if_wm.c
>   >
>   >  Log Message:
>   >  fix PR kern/52068 (82574 Tx csum offload bug in not only i386 but also amd64).
>   >
>   >  The packets sent from txq[1] could not be enabled any tx offload.
>   >  This has been enbugged since r1.454.
>   >
>   >
>   >  To generate a diff of this commit:
>   >  cvs rdiff -u -r1.497 -r1.498 src/sys/dev/pci/if_wm.c
>   >
>   >  Please note that diffs are not public domain; they are subject to the
>   >  copyright notices on the relevant files.
>   
>   --
>   //////////////////////////////////////////////////////////////////////
>   Internet Initiative Japan Inc.
>   
>   Device Engineering Section,
>   IoT Platform Development Department,
>   Network Division,
>   Technology Unit
>   
>   Kengo NAKAHARA <k-nakahara%iij.ad.jp@localhost>
>