Re: gpio panic

[Patrick Welche <prlw1%cam.ac.uk@localhost>]

On Sun, Apr 11, 2010 at 02:16:23PM +0100, Patrick Welche wrote:
> On Sat, Apr 10, 2010 at 02:53:40PM +0100, Patrick Welche wrote:
> > On Sat, Apr 10, 2010 at 02:36:55PM +0100, Patrick Welche wrote:
> > > With a current/i386 kernel built from source just now, I can't boot with 
> > > gpio.
> > > Switching it (gpio*) off with userconf allows me to boot - not many clues
> > > here I'm afraid:
> > > 
> > > ichlpcib0 at pci0 dev 31 function 0: Intel 82801DB LPC Interface Bridge 
> > > (rev. 0x02)
> > > ichlpcib0: 24-bit timer
> > > gpio0 at ichlpcib0: 64 pins
> > > uvm_fault(0xc07af8c0, 0, 1) -> 0xe
> > > fatal page fault in supervisor mode
> > > trap type 6 code 0 eip c0692b9f cs 8 eflags 10206 cr2 67 ilevel 8
> > > kernel: supervisor trap page fault, code=0
> > > Stopped in pid 0.1 (system) at  netbsd:strcmp+0xf:      movb    
> > > 0(%ebx),%al
> > > 
> > > db{0}> bt
> > > strcmp(cc7de800,c074b078,c089c668,c068c8b1,cc057f4c,0,c089c668,c068c326,cc058fc4
> > > ,c0752c20) at netbsd:strcmp+0xf
> > > uvm_fault(0xc07af8c0, 0x6e206000, 1) -> 0xe
> > > fatal page fault in supervisor mode
> > > trap type 6 code 0 eip c01b79d0 cs 8 eflags 10246 cr2 6e206680 ilevel 8
> > > kernel: supervisor trap page fault, code=0
> > > Faulted in DDB; continuing...
> > > 
> > > 0    >   1 7   0       200           c0757020            swapper
> > 
> > Same on another (identical) box. No problem with kernels from
> 
> > 5.99.24 Feb 20 2010 <- actually that one had gpio disabled with userconf
> > 5.99.22 Dec 18 2009
> 
> Just for more entertainment, all is well with a DEBUG DIAGNOSTIC LOCKDEBUG
> kernel. (remove options and it breaks again - on 2 computers with the
> 82801DB and one with a 82801FBM)
> 
> Now for some kgdb...

Finishing the binary chop of commits lead nowhere.
Written up as port-i386/43284.
Looks like some sort of timing problem, but I don't see what to try next...

Cheers,

Patrick