Re: [PATCH -v2 3/7] x86, NMI, Rename memory parity error to PCI SERR error

[Robert Richter <robert.richter@amd.com>]

On 27.09.10 04:39:20, Huang Ying wrote:
> On Mon, 2010-09-27 at 16:01 +0800, Robert Richter wrote:
> > On 26.09.10 20:57:02, Huang Ying wrote:
> > > memory parity error is only valid for IBM PC-AT, newer machine use 7
> > > bit (0x80) of 0x61 port for PCI SERR. While memory error is usually
> > > reported via MCE. So corresponding function name and kernel log string
> > > is changed.
> > > 
> > > But on some machines, PCI SERR line is still used to report memory
> > > errors. This is used by EDAC, so corresponding EDAC call is reserved.
> > > 
> > > 
> > > v2:
> > > 
> > > - EDAC call in pci_serr_error is reserved.
> > > 
> > > Signed-off-by: Huang Ying <ying.huang@intel.com>
> > > ---
> > >  arch/x86/include/asm/mach_traps.h |    6 +++---
> > >  arch/x86/kernel/traps.c           |   21 ++++++++++-----------
> > >  2 files changed, 13 insertions(+), 14 deletions(-)
> > > 
> > > --- a/arch/x86/include/asm/mach_traps.h
> > > +++ b/arch/x86/include/asm/mach_traps.h
> > > @@ -9,11 +9,11 @@
> > >  
> > >  #define NMI_REASON_PORT		0x61
> > >  
> > > -#define NMI_REASON_MEMPAR	0x80
> > > +#define NMI_REASON_SERR		0x80
> > >  #define NMI_REASON_IOCHK	0x40
> > > -#define NMI_REASON_MASK		(NMI_REASON_MEMPAR | NMI_REASON_IOCHK)
> > > +#define NMI_REASON_MASK		(NMI_REASON_SERR | NMI_REASON_IOCHK)
> > >  
> > > -#define NMI_REASON_CLEAR_MEMPAR	0x04
> > > +#define NMI_REASON_CLEAR_SERR	0x04
> > 
> > I already commented on this, patch #1 and #3 are basically the same in
> > most parts which should be merged. What remains then in this patch is
> > the modified printk() and the comment. Both could be added to #1 too
> > which is then some sort of code cleanup patch.
> 
> Don thinks it is Ok to keep 2 patches.

I don't like reviewing new changes which are thrown away with the next
patch. I review things twice and it is much harder to see what really
changed then. Also we should have a clean history.

And with git it is fairly easy to join patches.

> 
> > >  #define NMI_REASON_CLEAR_IOCHK	0x08
> > >  #define NMI_REASON_CLEAR_MASK	0x0f
> > >  
> > > --- a/arch/x86/kernel/traps.c
> > > +++ b/arch/x86/kernel/traps.c
> > > @@ -301,15 +301,14 @@ gp_in_kernel:
> > >  }
> > >  
> > >  static notrace __kprobes void
> > > -mem_parity_error(unsigned char reason, struct pt_regs *regs)
> > > +pci_serr_error(unsigned char reason, struct pt_regs *regs)
> > >  {
> > > -	printk(KERN_EMERG
> > > -		"Uhhuh. NMI received for unknown reason %02x on CPU %d.\n",
> > > -			reason, smp_processor_id());
> > > -
> > > -	printk(KERN_EMERG
> > > -		"You have some hardware problem, likely on the PCI bus.\n");
> > > +	printk(KERN_EMERG "NMI: PCI system error (SERR).\n");
> > 
> > You should keep reporting the cpu id to identify the affected node and
> > also the reason.
> 
> Ok. I will add CPU ID in message. Because we know the reason, I don't
> think we need the reason in message.

You only know that bit 7 is set, not the rest. As this is an error
message we should provide as much information as possible.

-Robert

-- 
Advanced Micro Devices, Inc.
Operating System Research Center