From mboxrd@z Thu Jan 1 00:00:00 1970 From: Grant Grundler Subject: Re: HPPA hang with PCI SERR or PARITY enabling Date: Mon, 10 Aug 2009 17:20:36 -0600 Message-ID: <20090810232036.GA20454@lackof.org> References: <20090805174507.GB11458@lackof.org> <20090806141542.bf48febf.berndg@fsim-ev.de> <20090807162242.GA18373@lackof.org> <20090808113100.37e0f4ab.berndg@fsim-ev.de> <20090809052714.GA22362@lackof.org> <4A7F7652.6050209@tausq.org> <7d01f9f00908100101w3d73e91fmce80eb79e85bdc61@mail.gmail.com> <7d01f9f00908100109y6265c3efs31a6a2c46679ae90@mail.gmail.com> <20090810130827.98dfd895.berndg@fsim-ev.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Thibaut VARENE , Randolph Chung , Grant Grundler , linux-parisc@vger.kernel.org To: Bernd Gietzelt Return-path: In-Reply-To: <20090810130827.98dfd895.berndg@fsim-ev.de> List-ID: List-Id: linux-parisc.vger.kernel.org On Mon, Aug 10, 2009 at 01:08:27PM +0200, Bernd Gietzelt wrote: > On Mon, 10 Aug 2009 10:09:36 +0200 > Thibaut VARENE wrote: > > > Bummer, I've been reading the logs a bit too quickly. It seems we have > > two different problems here: > > > > 2.6.30-1-64smp doesn't boot (apparently doesn't load init, hangs after > > the badness in smp.c message) > > 2.6.30-1-64up crashes while loading PCI drivers > > > > A log of the up kernel with pdcchassis=0 would thus be helpful. > > http://pastebin.com/m2a494b9b > > I couldn't see any difference to the log without pdcchassis=0. Yeah, wierd....I must be misremembering then. In any case, the HPMC occurs on the first none-zero numbered Sym2 device: sym53c8xx 0000:00:01.0: enabling device (0000 -> 0003) is OK sym53c8xx 0000:00:01.1: enabling device (0146 -> 0147) is OK sym53c8xx 0000:00:02.0: enabling device (0000 -> 0003) is OK sym53c8xx 0000:00:02.1: enabling device (0000 -> 0003) is OK sym53c8xx 0000:30:00.0: enabling device (0000 -> 0003) seems to be OK sym53c8xx 0000:38:00.0: enabling device (0000 -> 0003) HPMCs? The second cards seems to be the culprit though there are two add-on cards in the machine (I was expecting both to fail in similar ways). I saw you included the PIM dump with a later console capture but I won't have time this week to look at it. :( This PIM dump should confirm if we had problems accessing MMIO space and which device (should) owns that space. thanks, grant > > > > -- > Bernd Gietzelt