watchdog network card

[Pyun YongHyeon]

On Wed, Apr 11, 2007 at 11:53:24AM +0300, Stefan Lambrev wrote:
 > Hi all,
 > 
 > JoaoBR wrote:
 > >On Tuesday 10 April 2007 01:24, Andrew Reilly wrote:
 > >  
 > >>Wow.  Somehow we've slipped through a one-year timewarp:
 > >>
 > >>On Tue, Mar 28, 2006 at 07:47:54AM -0300, JoaoBR wrote:
 > >>    
 > >>>On Tuesday 28 March 2006 07:40, Andrew Reilly wrote:
 > >>>      
 > >>>>After the last rebuild on my amd64-x2 box, both the nve ethernet
 > >>>>on the motherboard and the dc ethernet that I had been using to
 > >>>>work around other problems in the nve driver stopped working in
 > >>>>this way.  DEVICE_POLLING and ifconfig...polling has got me
 > >>>>going again.  I thoroughly recommend it.
 > >>>>        
 > >>>nve does not run polling mode but dc does
 > >>>
 > >>>I guess you have an IRQ conflict, nve and dc on the same hw interrupt, 
 > >>>and that setting dc in polling mode worked around this problem then
 > >>>
 > >>>you could check vmstat -i with and without polling enabled to see it
 > >>>      
 > >>Thanks for the tip.  I haven't been running dc or nve for about
 > >>a year, now :-)  Nfe has been working beautifully for me,
 > >>without polling.  I guess that I should have a look to see if
 > >>nve has improved in the interim, but it's difficult to make
 > >>oneself mess with something that isn't broken...
 > >>
 > >>    
 > >
 > >nfe appears to work much better (also with polling) and flawless. I tried 
 > >one and another time nve but nfe is what works, at least on amd64 and 
 > >newer hardware so probably you don't need to waste your time ;)
 > >
 > >
 > >  
 > I noticed before few months that something changed in freebsd and now 
 > nfe is not very stable :(
 > I said in freebsd because In the beginning I updated only freebsd and 
 > not the drivert itself.
 > I thought getting latest nfe source will help but it doesn't.
 > 
 > Anyway nfe is still lot better then nve, nve for me was useless
 > best uptime for my network card with nve was few minutes.
 > 
 > nfe0: watchdog timeout (missed Tx interrupts) -- recovering
 > 

According to the above message, it seems that you use new nfe(4). :-)

 > interrupt                          total       rate
 > irq1: atkbd0                        6660          0
 > irq12: psm0                       153054          0
 > irq15: ata1                       778797          1
 > irq16: pcm0                      2735727          7
 > irq17: skc0                    115669786        296
 > irq18: nvidia0                  24802500         63
 > irq21: ohci0+                    2418887          6
 > irq22: nfe0 ehci0               92319117        236
          ^^^^^^^^^^
You are using shared interrut so it's possible to get occasional
watchdog timeouts. polling(4) should fix your issue here.

 > cpu0: timer                    780827522       2000
 > Total                         1019712050       2612
 > 
 > FreeBSD 6.2-STABLE i386
 > 
 > nfe0: <NVIDIA nForce4 CK804 MCP9 Networking Adapter> port 0xb000-0xb007 
 > mem 0xd5000000-0xd5000fff irq 22 at device 10.0 on pci0
 > 

nfe(4) should be teached to use MSI/MSI-X for PCI-Express/PCI-X based
adapters but it's not done yet.

-- 
Regards,
Pyun YongHyeon