[illumos-Discuss] recent sunos kernels has big problem with qemu's e1000 driver

Garrett D'Amore garrett at nexenta.com
Wed Sep 15 08:31:12 PDT 2010


On Wed, 2010-09-15 at 17:10 +0200, dpecka wrote:
> well, i'm not expert in any case for such a thing, but i'd like to
> suppose, that the emulated driver attempts to be identical as a hardware
> however it doesn't in real ..

Actually, usually the emulation doesn't emulate the full hardware, just
enough to make the typical OS drivers work with it.  In some cases, the
behaviors are very fragile, depending on specific initialization
sequences, such that when the driver does them differently (perhaps on a
different OS), the emulation fails to work.

	-- Garrett

> anyway, symptoms are suspicious: e1000 driver runs in every other OS
> which i've tested and also with sunos kernel it's just *only driver
> which seems like is taken the best.
> 
> dladm correctly recognizes link state and other
> infos, /dev/{tcp,ip,e1000g$n} seems like supporting all expected
> options. Just only receiving is not ok. in logs is nothing and i was
> been also unable to dig out *anything from mdb
> 
> if i assign address on there and ping some nonexistant ip addr i can see
> on *parent system expectable results:
> 
> # tcpdump -i oi0 broadcast
> tcpdump: verbose output suppressed, use -v or -vv for full protocol
> decode
> listening on oi0, link-type EN10MB (Ethernet), capture size 65535 bytes
> 12:34:07.916352 ARP, Request who-has 192.168.99.100 (Broadcast) tell
> kokina-s10u8, length 46
> 12:34:08.415916 ARP, Request who-has 192.168.99.100 (Broadcast) tell
> kokina-s10u8, length 46
> 12:34:09.415760 ARP, Request who-has 192.168.99.100 (Broadcast) tell
> kokina-s10u8, length 46
> 
> ^^ vice versa not .. i can't see even *any packet by using snoop.
> 
> it could be also very helpful, if someone would point me how to hunt
> this down by using mdb/dtrace hence this issue and its symptoms are just
> exciting ..
> 
> regards, daniel ..
> 
> ps. as i've written in bugreport i experienced this also with snadard
> sunos kernel from solaris 10u7 and 10u8 as well as in opensolaris
> 
> On Wed, 2010-09-15 at 07:56 -0700, Garrett D'Amore wrote:
> > On Wed, 2010-09-15 at 16:01 +0200, dpecka wrote:
> > > hi there,
> > > 
> > > https://bugs.launchpad.net/qemu/+bug/638955
> > > 
> > > i've filled this bugreport ^^ .. all is in there
> > > 
> > > i'd like to know only, if this could be considered as a problem of qemu
> > > or as a problem in sunos kernel ..
> > > 
> > > regards and thanks guys for all your effort and care with illumos, i
> > > believe now, that we'll be happy at the end that oraclists executed
> > > osol ..
> > > 
> > 
> > 
> > Its probably a qemu bug, but I'm not sure.  Its frequently the case that
> > drivers are coded to hardware, and the emulation of the hardware isn't
> > quite identical, so drivers written on different operating systems might
> > not function.
> > 
> > 	- Garrett
> > 
> > 




More information about the Discuss mailing list