Re: [PATCH 2/3] rfi-flush: Make it possible to call setup_rfi_flush() again

[Mauricio Faria de Oliveira <mauricfo@linux.vnet.ibm.com>]

Hi Michael,

On 03/13/2018 08:39 AM, Michael Ellerman wrote:
>> I agree; the 'Using fallback displacement flush' message is misleading
>> (is the system slower/fallback or not? Ô_o)

> That message is actually just wrong.
> 
> It still prints that even if enable=false.
> 
> So we should change all those messages, perhaps:
> 
> 	pr_info("rfi-flush: fallback displacement flush available\n");
> 	pr_info("rfi-flush: ori type flush available\n");
> 	pr_info("rfi-flush: mttrig type flush available\n");

Indeed.

>> So I wrote something with a new function parameter to force the init of
>> the fallback flush area (true in pseries, false in powernv).  Not that
>> contained, but it seemed to convey the intent here in a clear way.
>>
>> That's v2, just sent.

> OK thanks. I don't really like it :D - sorry!

No worries :) fair enough. Well, I didn't like it much, either, TBH.

> It's a lot of plumbing of that bool just to avoid the message, whereas I
> think we could just change the message like above.

Yup.

And what you think about a more descriptive confirmation of what flush
instructions/methods are _actually_ being used?

Currently and w/ your suggestion aobve, all that is known is what is
_available_, not what has gone in (or out, in the disable case) the
nop slots.

cheers,
mauricio