[dpdk-dev] [PATCH v2 01/17] eal: add max SIMD bitwidth

Bruce Richardson bruce.richardson at intel.com
Thu Sep 17 18:43:32 CEST 2020


On Thu, Sep 17, 2020 at 05:31:52PM +0100, Kinsella, Ray wrote:
> 
> 
> On 09/09/2020 20:30, Honnappa Nagarahalli wrote:
> > <snip>
> > 
> >>>
> >>  > diff --git a/lib/librte_eal/include/rte_eal.h b/lib/librte_eal/include/rte_eal.h
> >>>> index ddcf6a2e7a..8148f650f2 100644
> >>>> --- a/lib/librte_eal/include/rte_eal.h
> >>>> +++ b/lib/librte_eal/include/rte_eal.h
> >>>> @@ -43,6 +43,13 @@ enum rte_proc_type_t {
> >>>>  	RTE_PROC_INVALID
> >>>>  };
> >>>>
> >>>> +enum rte_max_simd_t {
> >>> We could add a RTE_MAX_SIMD = 0. Arm platforms can use this to choose
> >> SVE.
> >>>
> >>
> >> Is zero the best value for this? Would setting it to MAX_INT or some other big
> >> number be better, in terms of comparisons operations, or does that just not
> >> apply at all with SVE?
> > I suggested zero as the bitwidth can be specified from the command line. It would be much easier to input zero vs other number.
> 
> Right, but it doesn't end up being that intuitive as interface 
> 0 is enabled, 64 is not, 128 is enabled etc .... 
> 
> Suggest we use a max 16bit integer as 0xFFFF?
> 
I can actually see 0 on command-line as being "unlimited", but for the APIs
and internally, I think that it should be converted to a MAX_INT value so
that the comparisons don't need to special-case zero. I agree with
Honnappa, that a -1 or maxint value is awkward on commandline, but
internally it's just an enum, so we can set it to whatever the most
practical value is.


More information about the dev mailing list