[PATCH 0/6] PowerPc 8xx TLB/MMU fixes

[Joakim Tjernlund]

Rex Feany <RFeany at mrv.com> wrote on 06/10/2009 00:42:18:
>
> Thus spake Joakim Tjernlund (joakim.tjernlund at transmode.se):
>
> > > I got this oops:
> > >
> > > Unable to handle kernel paging request for data at address 0x00000000
> > > Faulting instruction address: 0xc000e110
> > > Oops: Kernel access of bad area, sig: 11 [#1]
> > > MRV NM2
> > > NIP: c000e110 LR: c000d520 CTR: 1006bf40
> > > REGS: c2203dd0 TRAP: 0300   Not tainted  (2.6.32-rc3-00014-gcea49b0-dirty)
> > > MSR: 00009032 <EE,ME,IR,DR>  CR: 44022422  XER: 20000000
> > > DAR: 00000000, DSISR: c0000000
> > > TASK = c21c18c0[69] 'rc.sysinit' THREAD: c2202000
> > > GPR00: 00000401 c2203e80 c21c18c0 c2203f50 00000000 4000d032 00000000 00000000
> > > GPR08: 10093eb4 00000000 00009032 c000d514 021c1ad0 10095c50 03ffb000 10006500
> > > GPR16: 10006508 00000000 00000000 00000000 100a2fc8 021c4238 00000000 00000400
> > > GPR24: c21a2700 00000002 40000000 00000000 00000000 c2203f50 00000000 00000011
> > > NIP [c000e110] do_page_fault+0x44/0x5fc
> > > LR [c000d520] handle_page_fault+0xc/0x80
> > > Call Trace:
> > > [c2203e80] [c000e58c] do_page_fault+0x4c0/0x5fc (unreliable)
> > > [c2203f40] [c000d520] handle_page_fault+0xc/0x80
> > > Instruction dump:
> > > 800300a0 7cba2b78 54170036 2f970400 7c9c2378 830200e0 54b6018c 40be000c
> > > 74ba4820 3ac00000 813d0080 3bc00000 <80a90000> 54a036be 2f80001f 40be0100
> > > ---[ end trace 99a4d88f7e2f1b60 ]---
> > >
> > > this happens in do_page_fault becaose regs->nip is null, so
> >
> > regs or regs->nip is NULL? Either one does not make sense
> > In any case it might be a secondary problem as DAR is NULL already when you
> > enter the page fault.
>
> I assumed it was NIP because ... I'm not sure why. TRAP() above
> dereferences regs, and that didn't fail, but I didn't see that until
> now.

Yes, it is a bit strange though that the kernel isn't allowed to read from
NULL. Is that expected?

 Jocke