Re: [RFC PATCH for 2.11 12/23] target/arm/translate-a64.c: add FP16 FAGCT to AdvSIMD 3 Same

["Alex Bennée" <alex.bennee@linaro.org>]

Richard Henderson <rth@twiddle.net> writes:

> On 07/20/2017 05:04 AM, Alex Bennée wrote:
>> +static softfloat_flags softfloat_mapping_table[] = {
>> +    { float_flag_inexact  , softfloat_flag_inexact },
>> +    { float_flag_underflow, softfloat_flag_underflow },
>> +    { float_flag_overflow , softfloat_flag_overflow },
>> +    { float_flag_invalid  , softfloat_flag_invalid },
>> +};
>> +/* FIXME: 2a has no infinite flag */
>> +
>> +static uint8_t sync_softfloat_flags_from_2a(float_status *flags2a)
>> +{
>> +    uint8_t flags = flags2a->float_exception_flags;
>> +    int i;
>> +    if (flags) {
>> +        for (i = 0; i < ARRAY_SIZE(softfloat_mapping_table); i++) {
>> +            struct softfloat_flags *map = &softfloat_mapping_table[i];
>> +            if (flags & map->float2a_flag) {
>> +                softfloat_raiseFlags(map->float3c_flag);
>> +            }
>> +        }
>> +    } else {
>> +        softfloat_exceptionFlags = 0;
>> +    }
>> +
>> +    return softfloat_exceptionFlags;
>> +}
>> +
>
> For conversions like this IMO it's better to make the compiler do the
> work. C.f. target/alpha/fpu_helper.c and CONVERT_BIT.

Cool, I'll look at that.

> BTW, I think these TLS variables that softfloat3a are using are going
> to be a real problem.  It's one thing to do it temporarily like you
> are here, coordinating between 2a and 3c, but later, when the front
> end is fully converted?  That's just nonsense.

Wouldn't the other option to be to drop float_status out of the guests
CPUEnv and grab it from the TLS variable instead? Of course all guests
would need to be MTTCG enabled for that to work.

> Is it going to be worthwhile to significantly hack up the incoming sources?
>
> If so, then we might do something like this:  Given a 3c function foo,
>
>   T foo_st(T, T, float3a_status *) { ... }
>   T foo(T x, T y) { return foo_st(x, y, &tls_status); }
>
> And of course pack all of the relevant state into a struct then
>
>   #define softfloat_exceptionFlags tls_status.exceptionflags
>
> etc, instead of having individual tls variables.  This feels like
> something that we could push back upstream, assuming that upstream is
> active.

Peter has found what might be an upstream source repository so we can
certainly look at that.

--
Alex Bennée