* Hang when testing with AMD64 with Tg3
@ 2004-10-05 0:11 Stephen Hemminger
2004-10-05 0:15 ` Andi Kleen
0 siblings, 1 reply; 9+ messages in thread
From: Stephen Hemminger @ 2004-10-05 0:11 UTC (permalink / raw)
To: Andi Kleen, davem; +Cc: netdev
Doing simple iperf test on new opteron with tg3 against existing Xeon
system. I am seeing something wierd, the connection hangs right away.
Is this a TSO bug.
Sender: BK latest (2.6.9-rc3) Tg3 (XX.YY.250.3)
Receiver: BK + netdev(jeffm) + dave's latest e100 (XX.YY.1.73)
Both machines are directly connected with a netgear 100mbit switch.
16:59:30.726551 IP XX.YY.250.3.32948 > XX.YY.1.73.5001: S 4113740578:4113740578(0) win 5840 <mss 1460,sackOK,timestamp 4233878 0,nop,wscale 2>
16:59:30.726574 IP XX.YY.1.73.5001 > XX.YY.250.3.32948: S 4117129705:4117129705(0) ack 4113740579 win 5792 <mss 1460,sackOK,timestamp 25903995 4233878,nop,wscale 2>
16:59:30.726578 IP XX.YY.250.3.32948 > XX.YY.1.73.5001: . ack 1 win 1460 <nop,nop,timestamp 4233879 25903995>
16:59:30.726604 IP XX.YY.250.3.32948 > XX.YY.1.73.5001: P 1:25(24) ack 1 win 1460 <nop,nop,timestamp 4233879 25903995>
16:59:30.726641 IP XX.YY.1.73.5001 > XX.YY.250.3.32948: . ack 25 win 1448 <nop,nop,timestamp 25903995 4233879>
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: Hang when testing with AMD64 with Tg3
2004-10-05 0:11 Hang when testing with AMD64 with Tg3 Stephen Hemminger
@ 2004-10-05 0:15 ` Andi Kleen
2004-10-05 0:29 ` David S. Miller
` (2 more replies)
0 siblings, 3 replies; 9+ messages in thread
From: Andi Kleen @ 2004-10-05 0:15 UTC (permalink / raw)
To: Stephen Hemminger; +Cc: Andi Kleen, davem, netdev
On Mon, Oct 04, 2004 at 05:11:11PM -0700, Stephen Hemminger wrote:
> Doing simple iperf test on new opteron with tg3 against existing Xeon
> system. I am seeing something wierd, the connection hangs right away.
> Is this a TSO bug.
>
> Sender: BK latest (2.6.9-rc3) Tg3 (XX.YY.250.3)
> Receiver: BK + netdev(jeffm) + dave's latest e100 (XX.YY.1.73)
>
> Both machines are directly connected with a netgear 100mbit switch.
I don't see things like this on my boxes with rc3, also with tg3s.
Of what nature is the hang? Does the machine crash?
-Andi
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: Hang when testing with AMD64 with Tg3
2004-10-05 0:15 ` Andi Kleen
@ 2004-10-05 0:29 ` David S. Miller
2004-10-05 0:41 ` Andi Kleen
2004-10-05 0:47 ` Stephen Hemminger
2004-10-05 1:00 ` Stephen Hemminger
2004-10-05 1:05 ` Stephen Hemminger
2 siblings, 2 replies; 9+ messages in thread
From: David S. Miller @ 2004-10-05 0:29 UTC (permalink / raw)
To: Andi Kleen; +Cc: shemminger, ak, davem, netdev
On Tue, 5 Oct 2004 02:15:27 +0200
Andi Kleen <ak@suse.de> wrote:
> I don't see things like this on my boxes with rc3, also with tg3s.
>
> Of what nature is the hang? Does the machine crash?
And what tg3 revisions do you have?
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: Hang when testing with AMD64 with Tg3
2004-10-05 0:29 ` David S. Miller
@ 2004-10-05 0:41 ` Andi Kleen
2004-10-05 0:47 ` Stephen Hemminger
1 sibling, 0 replies; 9+ messages in thread
From: Andi Kleen @ 2004-10-05 0:41 UTC (permalink / raw)
To: David S. Miller; +Cc: Andi Kleen, shemminger, davem, netdev
On Mon, Oct 04, 2004 at 05:29:48PM -0700, David S. Miller wrote:
> On Tue, 5 Oct 2004 02:15:27 +0200
> Andi Kleen <ak@suse.de> wrote:
>
> > I don't see things like this on my boxes with rc3, also with tg3s.
> >
> > Of what nature is the hang? Does the machine crash?
>
> And what tg3 revisions do you have?
Mostly 5701s.
-Andi
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: Hang when testing with AMD64 with Tg3
2004-10-05 0:29 ` David S. Miller
2004-10-05 0:41 ` Andi Kleen
@ 2004-10-05 0:47 ` Stephen Hemminger
1 sibling, 0 replies; 9+ messages in thread
From: Stephen Hemminger @ 2004-10-05 0:47 UTC (permalink / raw)
To: David S. Miller; +Cc: Andi Kleen, davem, netdev
On Mon, 2004-10-04 at 17:29 -0700, David S. Miller wrote:
> On Tue, 5 Oct 2004 02:15:27 +0200
> Andi Kleen <ak@suse.de> wrote:
>
> > I don't see things like this on my boxes with rc3, also with tg3s.
> >
> > Of what nature is the hang? Does the machine crash?
> And what tg3 revisions do you have?
# ethtool -i eth0
driver: tg3
version: 3.10
firmware-version:
bus-info: 0000:03:02.0
# ethtool -k eth0
Offload parameters for eth0:
rx-checksumming: on
tx-checksumming: on
scatter-gather: on
tcp segmentation offload: off
# lspci -v
0000:03:02.0 Ethernet controller: Broadcom Corporation NetXtreme BCM5703 Gigabit Ethernet (rev 02)
Subsystem: IBM: Unknown device 026f
Flags: bus master, 66Mhz, medium devsel, latency 64, IRQ 28
Memory at c0200000 (64-bit, non-prefetchable)
Capabilities: [40] PCI-X non-bridge device.
Capabilities: [48] Power Management version 2
Capabilities: [50] Vital Product Data
Capabilities: [58] Message Signalled Interrupts: 64bit+ Queue=0/3 Enable-
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: Hang when testing with AMD64 with Tg3
2004-10-05 0:15 ` Andi Kleen
2004-10-05 0:29 ` David S. Miller
@ 2004-10-05 1:00 ` Stephen Hemminger
2004-10-05 1:05 ` Stephen Hemminger
2 siblings, 0 replies; 9+ messages in thread
From: Stephen Hemminger @ 2004-10-05 1:00 UTC (permalink / raw)
To: Andi Kleen; +Cc: davem, netdev
On Tue, 2004-10-05 at 02:15 +0200, Andi Kleen wrote:
> On Mon, Oct 04, 2004 at 05:11:11PM -0700, Stephen Hemminger wrote:
> > Doing simple iperf test on new opteron with tg3 against existing Xeon
> > system. I am seeing something wierd, the connection hangs right away.
> > Is this a TSO bug.
> >
> > Sender: BK latest (2.6.9-rc3) Tg3 (XX.YY.250.3)
> > Receiver: BK + netdev(jeffm) + dave's latest e100 (XX.YY.1.73)
> >
> > Both machines are directly connected with a netgear 100mbit switch.
>
>
> I don't see things like this on my boxes with rc3, also with tg3s.
>
> Of what nature is the hang? Does the machine crash?
The sender just doesn't send any more data??? I can reproduce it every
time and it even happens with stock Suse 9.1 kernel (2.6.4-XX)
Will try a different ether card tommorrow to see if it is there.
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: Hang when testing with AMD64 with Tg3
2004-10-05 0:15 ` Andi Kleen
2004-10-05 0:29 ` David S. Miller
2004-10-05 1:00 ` Stephen Hemminger
@ 2004-10-05 1:05 ` Stephen Hemminger
2004-10-05 14:58 ` Nivedita Singhvi
2 siblings, 1 reply; 9+ messages in thread
From: Stephen Hemminger @ 2004-10-05 1:05 UTC (permalink / raw)
To: Andi Kleen; +Cc: davem, netdev
On Tue, 2004-10-05 at 02:15 +0200, Andi Kleen wrote:
> On Mon, Oct 04, 2004 at 05:11:11PM -0700, Stephen Hemminger wrote:
> > Doing simple iperf test on new opteron with tg3 against existing Xeon
> > system. I am seeing something wierd, the connection hangs right away.
> > Is this a TSO bug.
> >
> > Sender: BK latest (2.6.9-rc3) Tg3 (XX.YY.250.3)
> > Receiver: BK + netdev(jeffm) + dave's latest e100 (XX.YY.1.73)
> >
> > Both machines are directly connected with a netgear 100mbit switch.
>
IT ISN'T A NETWORK PROBLEM. The problem is that iperf uses posix
pthread mutex's to sychronize and it looks like a futex bug.
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: Hang when testing with AMD64 with Tg3
2004-10-05 1:05 ` Stephen Hemminger
@ 2004-10-05 14:58 ` Nivedita Singhvi
2004-10-05 17:06 ` Stephen Hemminger
0 siblings, 1 reply; 9+ messages in thread
From: Nivedita Singhvi @ 2004-10-05 14:58 UTC (permalink / raw)
To: Stephen Hemminger; +Cc: Andi Kleen, davem, netdev
Stephen Hemminger wrote:
> On Tue, 2004-10-05 at 02:15 +0200, Andi Kleen wrote:
>
>>On Mon, Oct 04, 2004 at 05:11:11PM -0700, Stephen Hemminger wrote:
>>
>>>Doing simple iperf test on new opteron with tg3 against existing Xeon
>>>system. I am seeing something wierd, the connection hangs right away.
>>>Is this a TSO bug.
>>>
>>>Sender: BK latest (2.6.9-rc3) Tg3 (XX.YY.250.3)
>>>Receiver: BK + netdev(jeffm) + dave's latest e100 (XX.YY.1.73)
>>>
>>>Both machines are directly connected with a netgear 100mbit switch.
>>
>
> IT ISN'T A NETWORK PROBLEM. The problem is that iperf uses posix
> pthread mutex's to sychronize and it looks like a futex bug.
might not be a futex bug - check to see if iperf does
any writes, fflushes, etc, in a signal handler - i.e.
that its signal handling is thread-safe. glibc now
makes it a fatal bug..
thanks,
Nivedita
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: Hang when testing with AMD64 with Tg3
2004-10-05 14:58 ` Nivedita Singhvi
@ 2004-10-05 17:06 ` Stephen Hemminger
0 siblings, 0 replies; 9+ messages in thread
From: Stephen Hemminger @ 2004-10-05 17:06 UTC (permalink / raw)
To: Nivedita Singhvi; +Cc: Andi Kleen, davem, netdev
On Tue, 2004-10-05 at 07:58 -0700, Nivedita Singhvi wrote:
> Stephen Hemminger wrote:
> > On Tue, 2004-10-05 at 02:15 +0200, Andi Kleen wrote:
> >
> >>On Mon, Oct 04, 2004 at 05:11:11PM -0700, Stephen Hemminger wrote:
> >>
> >>>Doing simple iperf test on new opteron with tg3 against existing Xeon
> >>>system. I am seeing something wierd, the connection hangs right away.
> >>>Is this a TSO bug.
> >>>
> >>>Sender: BK latest (2.6.9-rc3) Tg3 (XX.YY.250.3)
> >>>Receiver: BK + netdev(jeffm) + dave's latest e100 (XX.YY.1.73)
> >>>
> >>>Both machines are directly connected with a netgear 100mbit switch.
> >>
> >
> > IT ISN'T A NETWORK PROBLEM. The problem is that iperf uses posix
> > pthread mutex's to sychronize and it looks like a futex bug.
>
> might not be a futex bug - check to see if iperf does
> any writes, fflushes, etc, in a signal handler - i.e.
> that its signal handling is thread-safe. glibc now
> makes it a fatal bug..
It doesn't do I/O from signal handler, but the problem goes away when
not compiling with -O2.
^ permalink raw reply [flat|nested] 9+ messages in thread
end of thread, other threads:[~2004-10-05 17:06 UTC | newest]
Thread overview: 9+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2004-10-05 0:11 Hang when testing with AMD64 with Tg3 Stephen Hemminger
2004-10-05 0:15 ` Andi Kleen
2004-10-05 0:29 ` David S. Miller
2004-10-05 0:41 ` Andi Kleen
2004-10-05 0:47 ` Stephen Hemminger
2004-10-05 1:00 ` Stephen Hemminger
2004-10-05 1:05 ` Stephen Hemminger
2004-10-05 14:58 ` Nivedita Singhvi
2004-10-05 17:06 ` Stephen Hemminger
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).