[Qemu-devel] is it just me or is ne2k broken in qemu?

[Qemu-devel] is it just me or is ne2k broken in qemu?

[Michael Tokarev]

I performed several tests of various emulated NICs
booting with iPXE, and discovered that ne2k_pci (*)
emulated device does not quite work, for quite some
time already, at least with linux guests.

The NIC works for a while, but after a few packets,
or a few 1000s of packets, it stalls.  In tcpdump
on the host I see many ARP requests coming from the
guest and each has corresponding ARP reply, but
nothing is actually reaching the guest.

This behavour is consistent since qemu-kvm 0.12,
and is demonstrated by guest kernel 2.6.32 and
2.6.38, either 32 or 64bit.

Anyone seen something similar?

Thanks!

/mjt

(*) can we make ne2k-pci to ebe an alias for ne2k_pci,
to be consistent with virtio-net-pci?

Re: [Qemu-devel] is it just me or is ne2k broken in qemu?

[Mulyadi Santosa]

On Thu, Apr 14, 2011 at 09:31, Michael Tokarev <mjt@tls.msk.ru> wrote:
> I performed several tests of various emulated NICs
> booting with iPXE, and discovered that ne2k_pci (*)
> emulated device does not quite work, for quite some
> time already, at least with linux guests.
>
> The NIC works for a while, but after a few packets,
> or a few 1000s of packets, it stalls.  In tcpdump
> on the host I see many ARP requests coming from the
> guest and each has corresponding ARP reply, but
> nothing is actually reaching the guest.

IIRC, several people complained roughly the same thing about
ne2k...and it's about timer implementation inside the code.

For workaround, if possible, use e1000...works for me so far....

-- 
regards,

Mulyadi Santosa
Freelance Linux trainer and consultant

blog: the-hydra.blogspot.com
training: mulyaditraining.blogspot.com

Re: [Qemu-devel] is it just me or is ne2k broken in qemu?

[Alex Williamson]

On Thu, 2011-04-14 at 12:31 +0400, Michael Tokarev wrote:
> I performed several tests of various emulated NICs
> booting with iPXE, and discovered that ne2k_pci (*)
> emulated device does not quite work, for quite some
> time already, at least with linux guests.
> 
> The NIC works for a while, but after a few packets,
> or a few 1000s of packets, it stalls.  In tcpdump
> on the host I see many ARP requests coming from the
> guest and each has corresponding ARP reply, but
> nothing is actually reaching the guest.
> 
> This behavour is consistent since qemu-kvm 0.12,
> and is demonstrated by guest kernel 2.6.32 and
> 2.6.38, either 32 or 64bit.
> 
> Anyone seen something similar?

For testing the iPXE ROMs I booted up each NIC, including ne2k_pci, to a
network loaded kernel (~4M) and installation initrd (~8M).  I stopped
the test at the point where the installer kernel was able to
successfully DHCP with the boot NIC.  ne2k_pci was definitely the
slowest of the cards at loading the images, but I didn't notice any
functionality issues.  Maybe I didn't let it run long enough, but the
boot ROM seems ok with it.

Alex

Re: [Qemu-devel] is it just me or is ne2k broken in qemu?

[Michael Tokarev]

15.04.2011 18:17, Alex Williamson wrote:
> On Thu, 2011-04-14 at 12:31 +0400, Michael Tokarev wrote:

>> The NIC works for a while, but after a few packets,
>> or a few 1000s of packets, it stalls.  In tcpdump
>> on the host I see many ARP requests coming from the
>> guest and each has corresponding ARP reply, but
>> nothing is actually reaching the guest.

> For testing the iPXE ROMs I booted up each NIC, including ne2k_pci, to a
> network loaded kernel (~4M) and installation initrd (~8M).  I stopped
> the test at the point where the installer kernel was able to
> successfully DHCP with the boot NIC.  ne2k_pci was definitely the
> slowest of the cards at loading the images, but I didn't notice any
> functionality issues.  Maybe I didn't let it run long enough, but the
> boot ROM seems ok with it.

I'm doing exactly the same here, -- testing iPXE booting,
so booting linux kernel over network.  I haven't been able
to boot linux on ne2k so far, it fails somewhere down the
road after loading initrd+kernel - either during initrd
run or about after switching to new init (still running
off network ofcourse) after mounting nfs root.

So when I encountered this issue I tried non-network boot
and various versions of (linux) guest and qemu-kvm - and
for me, ne2k always fail (stalls) after some time.

And Mulyadi Santosa mentioned it's apparently a known issue
due to some timer-related problem in the code.

/mjt

Re: [Qemu-devel] is it just me or is ne2k broken in qemu?

[Luiz Capitulino]

On Sun, 17 Apr 2011 00:37:35 +0400
Michael Tokarev <mjt@tls.msk.ru> wrote:

> 15.04.2011 18:17, Alex Williamson wrote:
> > On Thu, 2011-04-14 at 12:31 +0400, Michael Tokarev wrote:
> 
> >> The NIC works for a while, but after a few packets,
> >> or a few 1000s of packets, it stalls.  In tcpdump
> >> on the host I see many ARP requests coming from the
> >> guest and each has corresponding ARP reply, but
> >> nothing is actually reaching the guest.
> 
> > For testing the iPXE ROMs I booted up each NIC, including ne2k_pci, to a
> > network loaded kernel (~4M) and installation initrd (~8M).  I stopped
> > the test at the point where the installer kernel was able to
> > successfully DHCP with the boot NIC.  ne2k_pci was definitely the
> > slowest of the cards at loading the images, but I didn't notice any
> > functionality issues.  Maybe I didn't let it run long enough, but the
> > boot ROM seems ok with it.
> 
> I'm doing exactly the same here, -- testing iPXE booting,
> so booting linux kernel over network.  I haven't been able
> to boot linux on ne2k so far, it fails somewhere down the
> road after loading initrd+kernel - either during initrd
> run or about after switching to new init (still running
> off network ofcourse) after mounting nfs root.
> 
> So when I encountered this issue I tried non-network boot
> and various versions of (linux) guest and qemu-kvm - and
> for me, ne2k always fail (stalls) after some time.
> 
> And Mulyadi Santosa mentioned it's apparently a known issue
> due to some timer-related problem in the code.

I've hit this issue today and I can reproduce it by downloading a
80M file with wget. My network configuration is:

 -net nic,model=ne2k_pci -net tap,ifname=vnet0,script=./qemu-ifup-switch

It also happens with qemu 0.12.5. I've also tried to enable the device's
debug and the last lines I see before it stalls *usually* are:

  E2000: read=0x07 val=0

or

  E2000: asic readl val=0xe283fad3

I'll try to reproduce with an older kernel version, but debugging
tips are welcome.