[ipxe-devel] eepro100 vs qemu

[Alex Williamson]

On Mon, 2011-03-14 at 13:18 -0600, Alex Williamson wrote:
> On Mon, 2011-03-14 at 12:22 -0600, Alex Williamson wrote:
> > On Sat, 2011-03-12 at 12:57 +0000, Michael Brown wrote:
> > > On Saturday 12 Mar 2011 12:05:53 Michael Brown wrote:
> > > > On Friday 11 Mar 2011 18:27:28 Alex Williamson wrote:
> > > > > I'm looking to refresh qemu's PXE ROMs with iPXE, but it looks like the
> > > > > eepro100 ROM has been broken for qemu for quite some time.  I bisected
> > > > > the start of the problem back to this commit:
> > > > 
> > > > I can reproduce this locally.  The "ifstat" command (see
> > > > http://ipxe.org/cmd/ifstat) shows multiple RX errors of "Invalid argument
> > > > (http://ipxe.org/1c384002)".  Turning on DEBUG=eepro100 shows the message
> > > > 
> > > >   Corrupted packet received. Status = 0x0fffa080
> > > 
> > > Should now be fixed:
> > > 
> > >   http://git.ipxe.org/ipxe.git/commitdiff/c315ae0
> > 
> > Awesome, thanks!
> > 
> > Unfortunately I just found another regression in the pcnet32 driver.  It
> > works fine with bare qemu, but adding in --enable-kvm, it hangs right
> > after printing the first bit of the PMM header.  I bisected this one
> > back to f4faa27, which is of course another large driver rewrite.  If
> > you have any insights, please let me know.  Thanks,
> 
> Hmm, well ignore the above.  I'm not sure how I came up with that
> bisect.  At the moment, I'm having trouble with all the roms using
> --enable-kvm, and it seems to point at commit 132c3917.  But I would
> have sworn they were working for me on Friday, so I don't have much
> faith in my build environment today.  I'll save any more speculation
> until I can figure out how things were working before.  Thanks,

Ok, sorry for the distraction, I think I can explain it now.  Last week
I was testing large (8G) guests.  Today I was trying to figure out how
the pcnet32 image ever worked because I couldn't make the old gpxe or
the new ipxe image work.  Eventually I discovered that it worked if I
used a smaller guest memory size (512M).  I then used that to test all
the images under qemu, followed by kvm.  Unfortunately nothing boots
under kvm with a 512M guest.  In fact, I can only boot these ROMs if the
guest has 3.6G or more (note that qemu/kvm currently has a 512MB I/O
hole, so this means there's memory bounding the hole on both sides).

This was introduced with the commit above, 132c3917.  So we've still got
a blocker for refreshing the current ROMs.  Thanks,

Alex