On Tue, Apr 02, 2024 at 11:30:39PM +0300, Ants Aasma wrote:
> On Tue, 2 Apr 2024 at 00:31, Nathan Bossart <nathandbossart@gmail.com> wrote:
>> On Tue, Apr 02, 2024 at 12:11:59AM +0300, Ants Aasma wrote:
>> > What about using the masking capabilities of AVX-512 to handle the
>> > tail in the same code path? Masked out portions of a load instruction
>> > will not generate an exception. To allow byte level granularity
>> > masking, -mavx512bw is needed. Based on wikipedia this will only
>> > disable this fast path on Knights Mill (Xeon Phi), in all other cases
>> > VPOPCNTQ implies availability of BW.
>>
>> Sounds promising. IMHO we should really be sure that these kinds of loads
>> won't generate segfaults and the like due to the masked-out portions. I
>> searched around a little bit but haven't found anything that seemed
>> definitive.
>
> After sleeping on the problem, I think we can avoid this question
> altogether while making the code faster by using aligned accesses.
> Loads that straddle cache line boundaries run internally as 2 load
> operations. Gut feel says that there are enough out-of-order resources
> available to make it not matter in most cases. But even so, not doing
> the extra work is surely better. Attached is another approach that
> does aligned accesses, and thereby avoids going outside bounds.
>
> Would be interesting to see how well that fares in the small use case.
> Anything that fits into one aligned cache line should be constant
> speed, and there is only one branch, but the mask setup and folding
> the separate popcounts together should add up to about 20-ish cycles
> of overhead.
I tested your patch in comparison to v25 and saw the following:
bytes v25 v25+ants
2 1108.205 1033.132
4 1311.227 1289.373
8 1927.954 2360.113
16 2281.091 2365.408
32 3856.992 2390.688
64 3648.72 3242.498
128 4108.549 3607.148
256 4910.076 4496.852
For 2 bytes and 4 bytes, the inlining should take effect, so any difference
there is likely just noise. At 8 bytes, we are calling the function
pointer, and there is a small regression with the masking approach.
However, by 16 bytes, the masking approach is on par with v25, and it wins
for all larger buffers, although the gains seem to taper off a bit.
If we can verify this approach won't cause segfaults and can stomach the
regression between 8 and 16 bytes, I'd happily pivot to this approach so
that we can avoid the function call dance that I have in v25.
Thoughts?
--
Nathan Bossart
Amazon Web Services: https://aws.amazon.com
