Re: Re:if_sk patch to get more info from people with problems

On Mon, 2004-11-01 at 05:54, User Valera wrote:
> #dmesg | grep sk
> skc0: <Marvell Gigabit Ethernet> port 0xd800-0xd8ff mem 
> 0xfeafc000-0xfeafffff irq 22 at device 5.0 on pci2
> skc0: Yukon Gigabit Ethernet 10/100/1000Base-T Adapter
> skc0: PN: Yukon 88E8001
> skc0: EC: Rev. 1.3
> skc0: MN: Marvell
> skc0: SN: AbCdEfG334454
> sk0: <Marvell Semiconductor, Inc. Yukon> on skc0
> sk0: Ethernet address: 00:11:2f:54:07:95
> miibus0: <MII bus> on sk0
> 
> I can 100% reproduce the my problem by follows
> ftp throw sk0 on another host
> get file
> 
> transfer speed starting at ~10MB/s monotonly decresing down to ~60KB/s 
> (in 1-2 minutes)
> netstat -I sk0 -w 1 -- show no errors but only decrising speed
> 
> Syptoms (from my point if view) very similar to SACK issue which (as i 
> think)
> i expirienced during last week but diferences in two moments:
> 1) before - transfer speed have decrising down to --stalled--
> 2) i have another interface fxp0 which now (after declared SACK issue 
> resolved)
>     working without problems
> 
> my local cure was  - debug.mpsafenet="0"
> it have worked last week and worked now (transfer speed is ~10MB/s)
> 
> You can ask me about any information/action which can help resolve the 
> problem - i will help as i can.

debug.mpsafenet=0 has been my cure here as well for basically the same
symptoms.  I'll apply this patch and get info on what hardware I have
soon.

-- 
Eric Anholt                                eta_at_lclark.edu          
http://people.freebsd.org/~anholt/         anholt_at_FreeBSD.org