add a builtin function for every llvm C library intrinsic and bit manipulation instrinsics #767

andrewrk · 2018-02-09T21:48:02Z

http://llvm.org/docs/LangRef.html#standard-c-library-intrinsics
http://llvm.org/docs/LangRef.html#bit-manipulation-intrinsics

LLVM has these because

they can sometimes map to hardware instructions (such as sqrt) (see Add hw sqrt for x86_64 #681)
LLVM codegen assumes that it can call these functions, which is why we have to provide compiler_rt.o
additional optimization awareness and integration with fast-math

This issue is to

implement all the intrinsics in std/special/builtin.zig
add all the intrinsics as @ builtin functions in zig
most std.math.* functions can just call out to the builtin functions

related: #514

The text was updated successfully, but these errors were encountered:

jfo · 2018-02-09T21:56:49Z

I like the idea of std library functions calling directly into intrinsics when possible, but that makes me wonder why the intrinsic would be available to user space at all, or at least why it would look the same as the other built-ins.

The number of potential @ functions is quite large, but there is no native name spacing scheme right now for keeping them grouped logically. Would that be something worth considering?

andrewrk · 2018-02-09T22:03:25Z

I like the idea of std library functions calling directly into intrinsics when possible, but that makes me wonder why the intrinsic would be available to user space at all, or at least why it would look the same as the other built-ins.

The standard library is user space. The other built-ins are very similar. These are no different than @ctz, @clz, @memset, @memcpy, @fence, @addWithOverflow, etc.

The number of potential @ functions is quite large, but there is no native name spacing scheme right now for keeping them grouped logically. Would that be something worth considering?

It's not that large. We're talking about adding 25 functions, and there are no name collisions, and we already use @ to prevent collisions with user space functions.

jfo · 2018-02-09T22:05:48Z

Fair enough, but at the least it would help to group them in the documentation at some point, I think.

See #767

* add `@bswap` builtin function. See #767 * comptime evaluation facilities are improved to be able to handle a `@ptrCast` with a backing array. * `@truncate` allows "truncating" a u0 value to any integer type, and the result is always comptime known to be `0`. * when specifying pointer alignment in a type expression, the alignment value of pointers which do not have addresses at runtime is ignored, and always has the default/ABI alignment * threw in a fix to freebsd/x86_64.zig to update syntax from language changes * some improvements are pending #863 closes #638 closes #1733 std lib API changes * io.InStream().readIntNe renamed to readIntNative * io.InStream().readIntLe renamed to readIntLittle * io.InStream().readIntBe renamed to readIntBig * introduced io.InStream().readIntForeign * io.InStream().readInt has parameter order changed * io.InStream().readVarInt has parameter order changed * io.InStream().writeIntNe renamed to writeIntNative * introduced io.InStream().writeIntForeign * io.InStream().writeIntLe renamed to writeIntLittle * io.InStream().writeIntBe renamed to writeIntBig * io.InStream().writeInt has parameter order changed * mem.readInt has different parameters and semantics * introduced mem.readIntNative * introduced mem.readIntForeign * mem.readIntBE renamed to mem.readIntBig and different API * mem.readIntLE renamed to mem.readIntLittle and different API * introduced mem.readIntSliceNative * introduced mem.readIntSliceForeign * introduced mem.readIntSliceLittle * introduced mem.readIntSliceBig * introduced mem.readIntSlice * mem.writeInt has different parameters and semantics * introduced mem.writeIntNative * introduced mem.writeIntForeign * mem.writeIntBE renamed to mem.readIntBig and different semantics * mem.writeIntLE renamed to mem.readIntLittle and different semantics * introduced mem.writeIntSliceForeign * introduced mem.writeIntSliceNative * introduced mem.writeIntSliceBig * introduced mem.writeIntSliceLittle * introduced mem.writeIntSlice * removed mem.endianSwapIfLe * removed mem.endianSwapIfBe * removed mem.endianSwapIf * added mem.littleToNative * added mem.bigToNative * added mem.toNative * added mem.nativeTo * added mem.nativeToLittle * added mem.nativeToBig

@bitreverse

* bitreverse - give bswap behavior * bitreverse, comptime_ints, negative values still not working? * bitreverse working for negative comptime ints * Finished bitreverse test cases * Undo exporting a bigint function. @bitreverse test name includes ampersand * added docs entry for @bitreverse

emekoi · 2019-05-27T18:16:03Z

shawnl · 2019-06-19T16:09:37Z

This requires adding a libm implementation to zig,(which i started here #2709) as llvm generates libcalls to libm. Also, once this is complete, I want to move implementations into c.zig and libm.zig and then have the std library versions just call the intrinsics, so that llvm (or other frontends) can optimize these as is appropriate.

Tetralux · 2019-09-02T19:23:41Z

Does LLVM also have a pause intrinsic?
https://www.felixcloutier.com/x86/pause

Would this be considered to be added for this too, or should that be a seperate issue?

shawnl · 2019-09-02T23:42:04Z

That one is a little tricky to do (if llvm already exports it which it probably does) so I think you should open another issue. El lun., 2 sept. 2019 14:23, Tetralux <[email protected]> escribió:

…

Does LLVM also have a pause intrinsic? https://www.felixcloutier.com/x86/pause Would this be considered to be added for this too, or should that be a seperate issue? — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#767?email_source=notifications&email_token=AAD4W4SNQSQ7PWCGCGZWPP3QHVR4NA5CNFSM4EQCJ7RKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD5WN5BA#issuecomment-527228548>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAD4W4VDTT3ODCLK2DYF76LQHVR4NANCNFSM4EQCJ7RA> .

matu3ba · 2023-02-09T23:44:45Z

semi-related to #1290, ie for powi

expikr · 2024-03-15T03:54:37Z

semi-related:

sincos: #18527

rsqrt: #19302

fpatan: #19330

achan1989 · 2024-09-23T10:19:02Z

std/special/builtin.zig no longer exists. Can I get a summary of where the intrinsics are implemented, and how they are exposed as @ builtins?

nektro · 2024-09-23T19:51:32Z

mainly here
https://github.com/ziglang/zig/blob/master/lib/std/zig/BuiltinFn.zig
https://github.com/ziglang/zig/blob/master/lib/std/zig/Zir.zig#L133 (std.zig.Zir.Inst.Tag)
https://github.com/ziglang/zig/blob/master/src/Sema.zig#L1009 fn analyzeBodyInner
compile errors will guide you to everything else

andrewrk added the enhancement Solving this issue will likely involve adding new logic or components to the codebase. label Feb 9, 2018

andrewrk added this to the 0.3.0 milestone Feb 9, 2018

andrewrk modified the milestones: 0.3.0, 0.4.0 Feb 28, 2018

andrewrk mentioned this issue Mar 30, 2018

Add ability to determine if we are currently executing at compile time #868

Closed

andrewrk added a commit that referenced this issue Apr 15, 2018

add @sqrt built-in function

b5459eb

See #767

andrewrk mentioned this issue Jun 8, 2018

Add arbitrary-precision integer to std #1081

Merged

justinbalexander mentioned this issue Jan 2, 2019

@bitreverse intrinsic, part of #767 #1865

Merged

andrewrk mentioned this issue Jan 31, 2019

Add SIMD Support #903

Open

andrewrk modified the milestones: 0.4.0, 0.5.0 Mar 22, 2019

andrewrk mentioned this issue Apr 2, 2019

make all bit manipulation builtin functions take an explicit type parameter #2119

Closed

shawnl mentioned this issue Apr 5, 2019

breaking: polish @clz() @ctz() @popcount() @bswap() @bitReverse() #2134

Closed

andrewrk modified the milestones: 0.5.0, 0.6.0 Sep 2, 2019

data-man mentioned this issue Nov 5, 2019

Proposal: builtin cycle counter #3597

Open

This was referenced Nov 5, 2019

Proposal: @prefetch builtin #3600

Closed

Proposal: @cpuid builtin #3622

Closed

andrewrk modified the milestones: 0.6.0, 0.7.0 Dec 9, 2019

andrewrk modified the milestones: 0.7.0, 0.8.0 Oct 9, 2020

timotheecour mentioned this issue Jan 19, 2021

misc intrinsics, builtins timotheecour/Nim#537

Open

andrewrk modified the milestones: 0.8.0, 0.9.0 May 19, 2021

andrewrk modified the milestones: 0.9.0, 0.10.0 Nov 20, 2021

andrewrk modified the milestones: 0.10.0, 0.11.0 Apr 16, 2022

andrewrk mentioned this issue Jul 7, 2022

powf not inlining or optimizing in builtin implementation (cf. Rust, clang) #12032

Open

andrewrk modified the milestones: 0.11.0, 0.12.0 Apr 9, 2023

andrewrk modified the milestones: 0.13.0, 0.12.0 Jun 29, 2023

castholm mentioned this issue Mar 14, 2024

rsqrt as builtin #19302

Closed

dweiller mentioned this issue Jan 25, 2025

add @memmove builtin #22605

Open

andrewrk modified the milestones: 0.14.0, 0.15.0 Feb 10, 2025

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

add a builtin function for every llvm C library intrinsic and bit manipulation instrinsics #767

add a builtin function for every llvm C library intrinsic and bit manipulation instrinsics #767

andrewrk commented Feb 9, 2018 •

edited

Loading

jfo commented Feb 9, 2018

andrewrk commented Feb 9, 2018

jfo commented Feb 9, 2018

emekoi commented May 27, 2019 •

edited

Loading

shawnl commented Jun 19, 2019

Tetralux commented Sep 2, 2019

shawnl commented Sep 2, 2019 via email

matu3ba commented Feb 9, 2023

expikr commented Mar 15, 2024 •

edited

Loading

achan1989 commented Sep 23, 2024

nektro commented Sep 23, 2024

add a builtin function for every llvm C library intrinsic and bit manipulation instrinsics #767

add a builtin function for every llvm C library intrinsic and bit manipulation instrinsics #767

Comments

andrewrk commented Feb 9, 2018 • edited Loading

jfo commented Feb 9, 2018

andrewrk commented Feb 9, 2018

jfo commented Feb 9, 2018

emekoi commented May 27, 2019 • edited Loading

Standard C Library Intrinsics

Bit Manipulation Intrinsics

shawnl commented Jun 19, 2019

Tetralux commented Sep 2, 2019

shawnl commented Sep 2, 2019 via email

matu3ba commented Feb 9, 2023

expikr commented Mar 15, 2024 • edited Loading

achan1989 commented Sep 23, 2024

nektro commented Sep 23, 2024

andrewrk commented Feb 9, 2018 •

edited

Loading

emekoi commented May 27, 2019 •

edited

Loading

expikr commented Mar 15, 2024 •

edited

Loading