Re: [Qemu-ppc] [Qemu-devel] [PATCH v3 2/8] target/ppc: rework vmrg{l, h}{b, h, w} instructions to use Vsr* macros

[Mark Cave-Ayland]

On 27/01/2019 17:26, Richard Henderson wrote:

> On 1/27/19 7:19 AM, Mark Cave-Ayland wrote:
>> Could this make the loop slower? I certainly haven't noticed any obvious
>> performance difference during testing (OS X uses merge quite a bit for
>> display rendering), and I'd hope that with a good compiler and modern branch
>> prediction then any effect here would be negligible.
>
> I would expect the i < n/2 loop to be faster, because the assignments are
> unconditional.  FWIW.

Do you have any idea as to how much faster? Is it something that would show up 
as
significant within the context of QEMU?

As well as eliminating the HI_IDX/LO_IDX constants I do find the updated 
version much
easier to read, so I would prefer to keep it if possible. What about unrolling 
the
loop into 2 separate ones e.g.

    for (i = 0; i < ARRAY_SIZE(r->element); i+=2) {
        result.access(i) = a->access(i >> 1);
    }

    for (i = 1; i < ARRAY_SIZE(r->element); i+=2) {
        result.access(i) = b->access(i >> 1);
    }

Would you expect this to perform better than the version proposed in the 
patchset?


ATB,

Mark.