Re: re0 fix that works with polling

From: John-Mark Gurney <gurney_j_at_resnet.uoregon.edu>
Date: Mon, 18 Oct 2004 22:45:36 -0700
Sean McNeil wrote this message on Mon, Oct 18, 2004 at 22:27 -0700:
> > Have you tried w/o running X?   also, tried the bios setting to not
> > assign an irq to the video card?  I doubt that the extra interrupts
> > would cause a problem...
> 
> I can try this if you think it worthwhile.  Remember, though, that all I
> have to do to get it to work for the interrupt case is to decrease the
> count number from 0x800 to 0x200.  For polling, it doesn't work without
> the timer interrupt because of apparent starvation.  This sort of rules
> out an interrupt sharing issue as polling is interrupt-less doesn't it?

Hmmm...  All of my testing was doing in interrupt mode, I haven't tried
polling...  I'm going to send you a patch that will print a message to
the console if we run out of tx buffers...  do you get a message from
vls about ENOBUFS?  tried running truss/ktrace to see if any of the
send requests return ENOBUFS?

-- 
  John-Mark Gurney				Voice: +1 415 225 5579

     "All that I will do, has been done, All that I have, has not."
Received on Tue Oct 19 2004 - 03:45:37 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:38:18 UTC