Re: [PATCH 0/9] ACPI, APEI patches for 2.6.40

[huang ying <huang.ying.caritas@gmail.com>]

On Sun, May 29, 2011 at 2:55 PM, Len Brown <lenb@kernel.org> wrote:
>> [PATCH 1/9] Add Kconfig option ARCH_HAVE_NMI_SAFE_CMPXCHG
>> [PATCH 2/9] lib, Add lock-less NULL terminated single list
>> [PATCH 3/9] lib, Make gen_pool memory allocator lockless
>> [PATCH 4/9] ACPI, APEI, GHES, printk support for recoverable error via NMI
>
> I think the 4 above should go upstream to enable the
> printk for recoverable errors.  If queuing messages
> from NMI context gets re-implemented in the future,
> then so be it.

The GHES records are queued just for printk.  It has no contradiction
with perf based logging.

> I'll skip the rest, however, except #9:-)
>
> thanks,
> -Len
>
>> [PATCH 5/9] HWPoison: add memory_failure_queue()
>> [PATCH 6/9] ACPI, APEI, GHES: Add hardware memory error recovery support
>> [PATCH 7/9] PCIe, AER, add aer_recover_queue
>> [PATCH 8/9] ACPI, APEI, GHES: Add PCIe AER recovery support
>> [PATCH 9/9] ACPI, APEI, ERST, Prevent erst_dbg from loading if ERST is disabled

I think #5 and #7 just some general code that can be used by perf
based recovering too.  Because we are still waiting for perf based
hardware error logging/recovering infrastructure.  I think we can take
#6 and #8 for now and maybe replace them with perf based solution when
that is ready.

BTW: Can you take a look at another patch with following title?

[PATCH] ACPI, APEI, Add APEI _OSC support

https://lkml.org/lkml/2011/5/25/28

This is needed to enable GHES reporting in firmware on some machines.

Best Regards,
Huang Ying