Skip to content

add specialization for 24MHz QueryPerformanceFrequency#3832

Merged
StephanTLavavej merged 9 commits intomicrosoft:mainfrom
fsb4000:fix3828
Jul 14, 2023
Merged

add specialization for 24MHz QueryPerformanceFrequency#3832
StephanTLavavej merged 9 commits intomicrosoft:mainfrom
fsb4000:fix3828

Conversation

@fsb4000
Copy link
Contributor

@fsb4000 fsb4000 commented Jun 25, 2023 •
edited
Loading

Fixes #3828

Fixes partially #3834

If I do

#define _LIKELY_ARM [[likely]]
#define _LIKELY_X86 [[unlikely]]
...
if (_Freq == _TenMHz) _LIKELY_X86 {
    static_assert(period::den % _TenMHz == 0, "It should never fail.");
    constexpr long long _Multiplier = period::den / _TenMHz;
    return time_point(duration(_Ctr * _Multiplier));
}
...

then clang-format format the code this way:

#define _LIKELY_ARM [[likely]]
#define _LIKELY_X86 [[unlikely]]
...

if (_Freq == _TenMHz) {
    _LIKELY_X86 {
        static_assert(period::den % _TenMHz == 0, "It should never fail.");
        constexpr long long _Multiplier = period::den / _TenMHz;
        return time_point(duration(_Ctr * _Multiplier));
    }
}

I'm not sure how well the Microsoft Visual C++ compiler's optimizer handles this, but clang-cl does a great job with inlining and simplfying the integer divisions into shifts and multiplies.

A repro that the MSVC compiler simplfying the integer divisions into shifts and multiplies: https://godbolt.org/z/zb5fob6xc

@fsb4000 @tycho
add specialization for 24MHz QueryPerformanceFrequency
f7479ab 
Co-authored-by: Steven Noonan <steven@uplinklabs.net>
@fsb4000 fsb4000 requested a review from a team as a code owner June 25, 2023 05:49
@fsb4000

This comment was marked as resolved.

Sign in to view
@achabense

This comment was marked as resolved.

Sign in to view
@frederick-vs-ja

This comment was marked as resolved.

Sign in to view
@achabense achabense mentioned this pull request Jun 26, 2023
Closed
AlexGuteniev
AlexGuteniev suggested changes Jun 27, 2023
View reviewed changes
Alcaro
Alcaro reviewed Jun 27, 2023
View reviewed changes
AlexGuteniev
AlexGuteniev reviewed Jun 29, 2023
View reviewed changes
@StephanTLavavej StephanTLavavej added the performance Must go faster label Jun 30, 2023
@StephanTLavavej StephanTLavavej self-assigned this Jul 4, 2023
StephanTLavavej
StephanTLavavej reviewed Jul 6, 2023
View reviewed changes
@StephanTLavavej StephanTLavavej removed their assignment Jul 6, 2023
strega-nil-ms
strega-nil-ms reviewed Jul 7, 2023
View reviewed changes
stl/inc/__msvc_chrono.hpp
Copy link
Contributor

@strega-nil-ms strega-nil-ms Jul 7, 2023

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Do we need to do the push_macro/undef/pop_macro magic incantation here for likely and unlikely?

Copy link
Member

@StephanTLavavej StephanTLavavej Jul 7, 2023

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

likely and unlikely are commonly defined as function-like macros, but not object-like macros. <xkeycheck.h> avoids rejecting them for that reason.

Unlike msvc etc., users are technically not supposed to macroize likely and unlikely, so we technically don't need to defend against them. We could but I don't think it's necessary at the moment.

@StephanTLavavej StephanTLavavej self-assigned this Jul 13, 2023
@StephanTLavavej
Copy link
Member

StephanTLavavej commented Jul 13, 2023

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

@StephanTLavavej StephanTLavavej merged commit ab57910 into microsoft:main Jul 14, 2023
@StephanTLavavej
Copy link
Member

StephanTLavavej commented Jul 14, 2023

Thanks for implementing this performance improvement! 🚀 🐇 🐆

@fsb4000 fsb4000 deleted the fix3828 branch July 14, 2023 02:26
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@StephanTLavavej StephanTLavavej StephanTLavavej approved these changes

+3 more reviewers

@AlexGuteniev AlexGuteniev AlexGuteniev requested changes

@Alcaro Alcaro Alcaro left review comments

@strega-nil-ms strega-nil-ms strega-nil-ms approved these changes

Reviewers whose approvals may not affect merge requirements

Assignees

@StephanTLavavej StephanTLavavej

Labels

performance Must go faster

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

steady_clock: specialization for 24MHz QueryPerformanceFrequency?

7 participants

@fsb4000 @achabense @frederick-vs-ja @StephanTLavavej @AlexGuteniev @Alcaro @strega-nil-ms

Comments