Can't you just throw in a __restrict? Works just fine with Clang/libc++: https://godbolt.org/z/Ez1hM11jG

Can't you just throw in a __restrict? Works just fine with Clang/libc++:

No, unfortunately. Pragma loop ivdep also doesn't help

I want to note that ranges::swap does not have any self-swap check, unlike std::swap.

Thanks! 🐈 I pushed fairly minor changes.

Results on my 5950X:

Results on my 5950X:

A big difference in "After" columnbetween std_swap<9000, uint8_t> and std_swap_ranges<uint8_t>/9000 made me curious.
It is big for me too.

Looks like the main reason is vector over-alignment. adding alignas(64) to stack array makes significant improvement there.

I don't think the benchmark should be modified to have that though.

Results on my 5950X:

A big difference in "After" columnbetween std_swap<9000, uint8_t> and std_swap_ranges<uint8_t>/9000 made me curious. It is big for me too.

Looks like the main reason is vector over-alignment. adding alignas(64) to stack array makes significant improvement there.

I don't think the benchmark should be modified to have that though.

If we know that alignment has a significant effect on the results, shouldn't we control it instead of letting it vary and skew the benchmark? Note that "control" doesn't mean "pick an unrealistic value that skews the results" but either pick a value that is representative or average over a set of values that are representative.

but either pick a value that is representative or average over a set of values that are representative

x64 stack is 16-aligned on function entry, but for multiple variables on the stack the location of each is limited by its own alignment and alignment of the variable next to it.

malloc has also 16 bytes alignment.

So 8 is a good skew maybe to imitate being next to a pointer.
Or 16 to imitate top or the only variable on stack or malloc allocation.

I'm mirroring this to the MSVC-internal repo - please notify me if any further changes are pushed.

🔀 🚀 😻