Messages in this thread | | | Date | Sun, 23 Aug 2009 19:54:48 +0200 | From | Jarek Poplawski <> | Subject | Re: [PATCH 2.6.30-rc4] r8169: avoid losing MSI interrupts |
| |
On Sun, Aug 23, 2009 at 07:43:52PM +0200, Michal Soltys wrote: > Jarek Poplawski wrote: >> David Dillow wrote, On 08/22/2009 10:43 PM: >> >>> On Sat, 2009-08-22 at 05:07 -0700, Eric W. Biederman wrote: >>>> ebiederm@xmission.com (Eric W. Biederman) writes: >>>> >>>>> David Dillow <dave@thedillows.org> writes: >>>>> >>>>>> Re-looking at the code, I'd guess that some IRQ status line is getting >>>>>> stuck high, but I don't see why -- we should acknowledge all outstanding >>>>>> interrupts each time through the loop, whether we care about them or >>>>>> not. >>>>>> >>>>>> Could reproduce a problem with the following patch applied, and send the >>>>>> full dmesg, please? >>>>> Here is what I get. >>>>> >>>>> r8169 screaming irq status 00000085 mask 0000ffff event 0000803f napi 0000001d >>>> And now that the machine has come out of it, that was followed by: >>>> Looks like the soft lockup did not manage to trigger in this case. >>> >>> I need some more context, please. What is the network load through this >>> NIC when you have the issues? Light, heavy? Can you give me more details >>> about the machine? A full dmesg from boot until this happens would help >>> quite a bit. At a minimum it would help answer which version of the chip >>> we're dealing with and what the machine it is in looks like. >>> >>> Can you reproduce this with pci=nomsi? I'm assuming it the chip running >>> in MSI mode. >>> >>> Also, can you reproduce it when booting UP (or maxcpus=1)? I'm thinking >>> about a race between rtl8169_interrupt() and rtl8169_poll(), but it >>> isn't jumping out at me. >>> >>> Also, I'm having connectivity troubles this weekend, so my response may >>> be spotty. :( >>> >> >> >> BTW, FYI, it seems Michal stopped tracking this problem, but he >> found this commit problematic as well. >> >> From: Michal Soltys <soltys@ziu.info> >> Subject: Re: r8169 (+others ?) and note_interrupt performance hit on 2.6.30.x >> Date: Wed, 05 Aug 2009 20:54:47 +0200 >> http://marc.info/?l=linux-netdev&m=124949848110710&w=2 >> > > Well - not really stopped, but not sure what to look at before that > particular commit (as cpu load for the tests I've done increased rather > significantly as well before that, and after 2.6.29 - but it doesn't > seem to be related to the driver). And I was away for over a week... > > As fot the changes that commit introduced, here's is link to the mail > with the oprofile I did back then: > > http://www.spinics.net/lists/netdev/msg102709.html > > I'm happy to assist any way I can.
Very nice :-) I guess David might be interested in your dmesg etc. (as above) too.
Jarek P.
| |