[Skiboot] [RFC PATCH] phb4: Make verbose EEH output less verbose

Oliver oohall at gmail.com
Fri Aug 3 13:33:57 AEST 2018


On Fri, Aug 3, 2018 at 1:15 PM, Stewart Smith <stewart at linux.ibm.com> wrote:
> Russell Currey <ruscur at russell.cc> writes:
>> On Tue, 2018-07-31 at 18:48 +1000, Oliver O'Halloran wrote:
>>> Each PHB4 component has a error status register and a set of
>>> component-specific registers that describe that error in more detail.
>>> If the status register of a component has no bits set there's not
>>> much
>>> point in dumping the error detail registers.
>>>
>>> This patch modifies the verbose EEH dumping code so that it only
>>> dumps the error registers of a component if there are bits set in the
>>> error status register of that component. This limits the output to
>>> the
>>> registers that are actually relevant to the error that occured which
>>> makes interpreting the dumps slightly easier.
>>>
>>> Signed-off-by: Oliver O'Halloran <oohall at gmail.com>
>>> ---
>>> I've been using this in my debug builds for a while, but I'm not sure
>>> if it's a great idea since sometimes other people want to full dump.
>>> It's probably fine...
>>
>> This is the same thing the kernel does, I don't see any reason why we
>> wouldn't do it here.  The kernel also puts multiple registers on the
>> same line to save space, this does require looking at the code to
>> remember what order the regs are in but if you're decoding this stuff
>> that's the least of your problems so *shrug*
>>
>> In any case,
>>
>> Acked-by: Russell Currey <ruscur at russell.cc>
>
> I think this means I can make good on my joke^W promise to merge some of
> Oliver's RFC patches at 4:59pm on a friday before he takes some time
> off? :)

I think tradition requires you to be drunk when merging RFC patches,
so 4:59 might be a bit early.

>
> --
> Stewart Smith
> OPAL Architect, IBM.
>


More information about the Skiboot mailing list