Re: Packet corruption in re0

Pyun YongHyeon escribió:

[....]
>  > >Download if_re.c and if_rlreg.h in the following URL and rebuild the 
>  > >driver.
>  > >http://people.freebsd.org/~yongari/re/7.0R/if_re.c
>  > >http://people.freebsd.org/~yongari/re/7.0R/if_rlreg.h
>  > 
>  > I've tested it against my 'netcat | tar' and it improves as it takes 
>  > four iterations (instead of one) to fall down (nearly unusable). I've 
>  > not run it so many times in a row on CURRENT. Must I try it again?. 
>  > That's what pciconf says about my re0:
>  > 
>  > re0_at_pci0:2:15:0:        class=0x020000 card=0xe0001458 chip=0x816710ec 
>  > rev=0x10 hdr=0x00
>  > 
> 
> How about disabling checksum offload on re0 side?

After 6 iterations everything looked good so I started to do other 
things while the netcat was running (build ports). Shortly after 
(seconds), the problems appear again. Now, I'm not sure how reliable the 
testing procedure is and there are other issues with the chipset (AMD 
690G), usb support problems, the sata subsystem is not well recognized 
... I've seen also problems on this box while using em0 so too complex 
environment to isolate the possible re0 problem. While I'll be glad to 
test whatever you ask me for, too loose ends on this box nowadays.

Regards,
Raul