Re: msk watchdog timeout

On Friday 13 July 2007 10:29:55 am Li-Lun "Leland" Wang wrote:
> On 7/13/07, John Baldwin <jhb_at_freebsd.org> wrote:
> > On Monday 21 May 2007 12:45:39 am Li-Lun "Leland" Wang wrote:
> > > On 5/20/07, Pyun YongHyeon <pyunyh_at_gmail.com> wrote:
> > > > On Sun, May 20, 2007 at 09:39:54PM -0500, Li-Lun Leland Wang wrote:
> > > >  > On 5/20/07, Pyun YongHyeon <pyunyh_at_gmail.com> wrote:
> > > >  > >On Mon, May 21, 2007 at 01:41:24AM +0800, Li-Lun Wang (Leland 
Wang)
> > wrote:
> > > >  > > > I just installed 7.0-current as of May 3 on my new computer 
that
> > comes
> > > >  > > > with an on-board Marvell Yukon Gigabit Ethernet.  Every now and
> > then
> > > >  > > > if the network throughput comes near several hundred kbytes, I 
get
> > the
> > > >  > > > msk0 watchdog timeout messages:
> > > >  > > >
> > > >  > > >      kernel: msk0: watchdog timeout
> > > >  > > >      msk0: watchdog timeout (missed Tx interrupts) -- 
recovering
> > > >  > > >
> > > >  > > > Although it says recovering, the interface never comes back 
alive.
> > > >  > >
> > > >  > >The above message indicates the driver sent all pending 
transmission
> > > >  > >requests but the driver didn't receive corresponding Tx completion
> > > >  > >interrupts. Not recovering from the watchdog timeout means there 
are
> > > >  > >another issues on the driver. However as disabling MSI fixed the
> > > >  > >issue, I guess it's not fault of msk(4) and it comes from 
bad/broken
> > > >  > >MSI implementation of your system. I guess it's time to add your
> > > >  > >chipset to a PCI quirk table in order to blacklist it.
> > > >  >
> > > >  > I do reckon that MSI doesn't work on earlier Intel chipsets.  Mine 
is
> > > >  > P965 (on a gigabyte GA-965P-DS3 rev 1.3), which I suppose is recent
> > > >  > enough to support MSI, isn't it?  Or could there be other problems
> > > >
> > > > Using latest chipsets does not necessarily guarantee working MSI.
> > >
> > > I see.  I think we should maybe add P965 to the PCI quirk list for
> > > broken MSI, then?
> >
> > Possibly.
> >
> > > >  > possible?
> > > >  >
> > > >
> > > > Yes. But I couldn't find possible issue on msk(4) yet.
> > >
> > > Maybe I was not clear enough.  Could there be something else that
> > > causes MSI to not working correctly other than the chipset?  I was
> > > just wondering why I didn't see too many broken MSI reports if most
> > > Intel chipsets are broken.
> >
> > If it's not the driver it would be the chipset.  We already don't use MSI 
on
> > systems that don't support either PCI-X or PCI-express, so that implicitly
> > blacklists most older Intel chipsets.  Do you have any other devices in 
your
> > system that support MSI?  pciconf -lc output would be useful to look at.
> 
> Here is the output of pciconf -lc:
> 
> pcm0_at_pci0:27:0:	class=0x040300 card=0xa0021458 chip=0x284b8086 rev=0x02 
hdr=0x00
>     cap 01[50] = powerspec 2  supports D0 D3  current D0
>     cap 05[60] = MSI supports 1 message, 64 bit
>     cap 10[70] = PCI-Express 1 type 0

Perhaps try the experimental pcm MSI support to see if it works.  If it does 
then you know MSI works ok.  Oh, and if MSI were broken, you wouldn't get any 
interrupts on the adapter at all, if you are only getting timeouts under load 
but the adapter works fine with lower amounts of traffic, then it is probably 
an issue in the msk(4) driver and not the MSI support in the chipset.

-- 
John Baldwin