On 30.03.2023 16:27, Roger Pau Monné wrote:
> On Thu, Mar 30, 2023 at 01:40:43PM +0200, Jan Beulich wrote:
>> On 30.03.2023 11:54, Roger Pau Monné wrote:
>>> On Thu, Mar 30, 2023 at 09:53:23AM +0200, Jan Beulich wrote:
>>>> On 29.03.2023 16:17, Roger Pau Monné wrote:
>>> Patch 8 I'm unsure, I guess it should be up to the user to decide
>>> whether to use -Os or some other optimization?
>>>
>>> I expect introspection users likely care way more about the speed
>>> rather than the size of the generated x86 emulator code?
>>
>> Perhaps. But do we want to introduce another Kconfig control just
>> for that? And wouldn't there likely be other performance concerns,
>> if performance really mattered in the introspection case?
> 
> I don't really have a strong opinion on the usage of -Os or not.  It's
> likely fine as long as we also stay consistent with the flag we use
> when building the test harness and the fuzzer, just in case.

While they might be built with different compilers (and hence things can't
be compared directly anyway), I can certainly make those two use -Os as
well.

Jan

Reply via email to