[tulip] Linksys NC100 v2.0 Transmit time out

Mon, 28 Aug 2000 17:52:38 +0000 (CST)

1st, thanks for the quick reply!

I suspected an IRQ conflict from lspci, but a cat of /proc/interrupts shows no problems.  Ran tulip-diag which looks like its waiting for an interrupt... (Interrupt sources are pending!  CSR5 is fc674054.  here's some output:

[root@mors /root]# cat /proc/interrupts
           CPU0
  0:     303496          XT-PIC  timer
  1:       2961          XT-PIC  keyboard
  2:          0          XT-PIC  cascade
  6:        162          XT-PIC  floppy
  8:          1          XT-PIC  rtc
 10:          0          XT-PIC  eth0
 12:       1665          XT-PIC  PS/2 Mouse
 13:          1          XT-PIC  fpu
 14:      89751          XT-PIC  ide0
 15:          7          XT-PIC  ide1
NMI:          0

[root@mors /root]# /usr/src/redhat/SOURCES/tulip-diag -a
tulip-diag.c:v2.03 7/31/2000 Donald Becker (becker@scyld.com)
http://www.scyld.com/diag/index.html

Index #1: Found a ADMtek AL985 Centaur-P adapter at 0xd800.
 * A potential Tulip chip has been found, but it appears to be active.
 * Either shutdown the network, or use the '-f' flag to see all values.
 Port selection is 100mbps-SYM/PCS 100baseTx scrambler, half-duplex.
 Transmit started, Receive started, half-duplex.
  The Rx process state is 'Waiting for packets'.
  The Tx process state is 'Idle'.
  The transmit threshold is 128.
 Interrupt sources are pending!  CSR5 is fc674054.
   Tx out of buffers indication.
   Link passed indication.
   Rx Done indication.
   Early Rx indication.
 The Comet MAC registers are 03782000 ffff1809 filter 8000000000000000.

Messing around in BIOS changed eth0 to an IRQ of 10, but still the same problem.  If this is an IRQ problem anybody been working with the Tyan i810 (got this one from the linuxstore.com)?  Or else any other suggestions, tips etc?  The nic is on a riser in a 2U rack mount case.  I wonder if the nic is competing for an IRQ with the riser...?

Thanks in advance,
-Jonas

On Mon, 28 Aug 2000 Donald Becker  wrote:
> On Mon, 28 Aug 2000, Jonas P. Valkiunas wrote:
> 
> > 
> > I've perused through the last few months of the scyld and other lists
> > regarding the new Linksys NC100 v2.0 problem, but cannot find my specific
> > problem.  Both srpm and individual tulip.c files compile without a
> > problem, but as soon as I try to ping any ip (other than my own), I get
> > the error: 
> > 
> > "eth0: Transmit timed out, status fc67c057, CSR120000000 resetting..."
> 
> IRQ conflict -- the chip is trying to raise an interrupt, and nothing is
> happening.
> 
> Run 'tulip-diag' to verify.
> 
> > > lspci show the already list mentioned Bridgecom ID, while logs show the ADMtek Comet,
> > rev 17, using IRQ11.
> > 
> > other interesting notes from lspci:
> > "
> > 00:1f.3 SMBus Intel Corporation 82801AA 82810 Chipset SMBus (rev02)
> > Subsystem: Intel Corporation: Unknown device 2413
> > Flags: medium devsel, IRQ 11
> > I/O ports at efa0
> 
> If the BIOS is using the SMBus for something unknown to the OS, that might
> prevent interrupts from being delivered.
> 
> Another possible conflict is the USB controller, if the BIOS is doing PS/2
> keyboard emulation.
> 
> > This is a Red Hat 6.2 box running on a celeron 533 in a Tyan S2054 Tomcat
> > i810 motherboard.  Has anybody seen this time out problem (happening on 2
> > identical boxes)?  Sorry for the naivete...PnP disabled in BIOS and
> > kernel, PCI enabled in kernel.  At first thought of an IRQ conflict (at
> > 11), but could not change the nic's IRQ through BIOS (PCI).  Any help
> > tips, etc. MUCH appreciated! 
> 
> Try the LILO option "noapic".
> 
> Donald Becker				becker@scyld.com
> Scyld Computing Corporation		http://www.scyld.com
> 410 Severn Ave. Suite 210		Beowulf-II Cluster Distribution
> Annapolis MD 21403
> 