Re: mm snapshot broken-out-2007-03-18-02-44.tar.gz uploaded

[Andi Kleen]

On Tue, Mar 20, 2007 at 08:36:27AM -0800, Andrew Morton wrote:
> > caller is avail_to_resrv_perfctr_nmi_bit+0x2b/0x43
> >  [<c0105256>] show_trace_log_lvl+0x1a/0x2f
> >  [<c010597b>] show_trace+0x12/0x14
> >  [<c0105a3d>] dump_stack+0x16/0x18
> >  [<c0213313>] debug_smp_processor_id+0xb3/0xc8
> >  [<c0116a26>] avail_to_resrv_perfctr_nmi_bit+0x2b/0x43
> >  [<fdc819b9>] nmi_create_files+0x2a/0x10e [oprofile]
> >  [<fdc80f52>] oprofile_create_files+0xe6/0xec [oprofile]
> >  [<fdc81157>] oprofilefs_fill_super+0x78/0x7e [oprofile]
> >  [<c0182d2e>] get_sb_single+0x59/0x9f
> >  [<fdc8108f>] oprofilefs_get_sb+0x1c/0x1e [oprofile]
> >  [<c0182792>] vfs_kern_mount+0x81/0xf1
> >  [<c0182852>] do_kern_mount+0x38/0xde
> >  [<c0196671>] do_mount+0x605/0x693
> >  [<c019677f>] sys_mount+0x80/0xb5
> >  [<c0104270>] syscall_call+0x7/0xb
> >  =======================
> 
> Odd.  It looks like oprofile has been doing this for some time.  Andi,
> there are a few changes in the NMI area - can you think of one whihc would
> have triggered this?

Looks like it was always broken. avail_to_resrv_perfctr_nmi_bit() must 
always do all this for all possible CPUs, not just the current one.
I can cook up a patch.

-Andi


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/