ziglang/zig
266
6.0k
Fork
You've already forked zig
786

compiler_rt: optimize memset #35754

Open
rpkak wants to merge 3 commits from rpkak/zig:memset into master
pull from: rpkak/zig:memset
merge into: ziglang:master
ziglang:master
ziglang:elfv2
ziglang:spork8
ziglang:restricted
ziglang:0.16.x
ziglang:panic-rewrite
ziglang:parking-futex-lockfree
ziglang:windows-Io-cleanup
ziglang:Io-watch
ziglang:ParseCommandLineOptions
ziglang:windows-async-files
ziglang:poll-ring
ziglang:debug-file-leaks-differently
ziglang:debug-file-leaks
ziglang:ProcessPrng
ziglang:elfv2-dyn
ziglang:jobserver
ziglang:threadtheft
ziglang:io-threaded-no-queue
ziglang:0.15.x
ziglang:Io.net
ziglang:comptime-allocator
ziglang:restricted-function-pointers
ziglang:cli
ziglang:wasm-linker-writer
ziglang:wrangle-writer-buffering
ziglang:sha1-stream
ziglang:async-await-demo
ziglang:fixes
ziglang:0.14.x
ziglang:ast-node-methods
ziglang:macos-debug-info
ziglang:make-vs-configure
ziglang:fuzz-macos
ziglang:sans-aro
ziglang:ArrayList-reserve
ziglang:incr-bug
ziglang:llvm-ir-nosanitize-metadata
ziglang:ci-tarballs
ziglang:ci-scripts
ziglang:threadpool
ziglang:0.12.x
ziglang:new-pkg-hash
ziglang:json-diagnostics
ziglang:more-doctests
ziglang:rework-comptime-mutation
ziglang:0.11.x
ziglang:ci-perf-comment
ziglang:stage2-async
ziglang:0.10.x
ziglang:autofix
ziglang:0.9.x
ziglang:aro
ziglang:hcs
ziglang:0.8.x
ziglang:0.7.x
Contributor
Copy link

Benchmark code: https://codeberg.org/rpkak/zig-memset-bench

Use these benchmark results from a AMD Ryzen 5 7640U or your own benchmark results with the plot.py to see the benchmark graph.

Here are also images, but use the plot.py with the python packages matplotlib and e.g. PyQt6 installed, to better zoom into the graph to be less distracted by e.g. glibc being slow at memseting 0 bytes.

short_unaligned_uncached:
short_unaligned_uncached

short_unaligned_cached:
short_unaligned_cached

short_aligned_uncached:
short_aligned_uncached

short_aligned_cached:
short_aligned_cached

long_unaligned_uncached:
long_unaligned_uncached

long_unaligned_cached:
long_unaligned_cached

long_aligned_uncached:
long_aligned_uncached

long_aligned_cached:
long_aligned_cached

alternative to #35723
closes #32091

Edit: This now also removes __memset and fixes the function arg type of memset and __memset_chk.

Benchmark code: https://codeberg.org/rpkak/zig-memset-bench Use [these](/attachments/1f6bad17-28f4-4ac1-b613-7403070dcd25) benchmark results from a AMD Ryzen 5 7640U or your own benchmark results with the [`plot.py`](https://codeberg.org/rpkak/zig-memset-bench/src/branch/main/plot.py) to see the benchmark graph. <details> <summary> Here are also images, but use the [`plot.py`](https://codeberg.org/rpkak/zig-memset-bench/src/branch/main/plot.py) with the python packages `matplotlib` and e.g. `PyQt6` installed, to better zoom into the graph to be less distracted by e.g. glibc being slow at `memset`ing `0` bytes. </summary> short_unaligned_uncached: ![short_unaligned_uncached](/attachments/c3463690-bced-4cb4-b2c6-803477a7b421) short_unaligned_cached: ![short_unaligned_cached](/attachments/564011e4-75f4-491f-ac7e-bbc6105c8634) short_aligned_uncached: ![short_aligned_uncached](/attachments/8dde4e44-ea6b-4268-baa7-4f2a2b903df2) short_aligned_cached: ![short_aligned_cached](/attachments/942008d1-bd11-40f8-8f31-e1079394d4e6) long_unaligned_uncached: ![long_unaligned_uncached](/attachments/fa88f192-024a-49af-85c0-435fc86c4996) long_unaligned_cached: ![long_unaligned_cached](/attachments/43187131-cd01-4184-a548-6d9102193756) long_aligned_uncached: ![long_aligned_uncached](/attachments/d1cd6a52-f7cf-41cb-839d-65626ab8ea43) long_aligned_cached: ![long_aligned_cached](/attachments/7edb71d9-2f05-46c9-b8e0-f492aab37941) </details> alternative to #35723 closes #32091 **Edit:** This now also removes `__memset` and fixes the function arg type of `memset` and `__memset_chk`.
Author
Contributor
Copy link

@andrewrk, you seem to oppose the use of @ptrCast to a pointer to a vector here and here. I tried to not use such @ptrCasts in this PR, but this resulted in worse machine code.

@andrewrk, you seem to oppose the use of `@ptrCast` to a pointer to a vector [here](https://ziggit.dev/t/memset-beaten/15844/8?u=rpkak) and [here](https://codeberg.org/ziglang/zig/pulls/35723#issuecomment-17325590). I tried to not use such `@ptrCast`s in this PR, but this resulted in worse machine code.
Owner
Copy link

you seem to oppose the use of @ptrCast to a pointer to a vector

The code you have written is incorrect. The in-memory representation of vectors is entirely implementation-defined: Zig implementations are permitted to insert arbitrary padding bits/bytes, or even to shuffle the bits around into crazy non-contiguous arrangements if they wish. To implement bulk memory operations, the values you store/load to/from memory must be arrays, never vectors.

> you seem to oppose the use of `@ptrCast` to a pointer to a vector The code you have written is incorrect. The in-memory representation of vectors is entirely implementation-defined: Zig implementations are permitted to insert arbitrary padding bits/bytes, or even to shuffle the bits around into crazy non-contiguous arrangements if they wish. To implement bulk memory operations, the values you store/load to/from memory must be *arrays*, never vectors.
rpkak force-pushed memset from f1ecbe6f8f
Some checks failed
ci / aarch64-macos-release (pull_request) Failing after 15m59s
ci / aarch64-macos-debug (pull_request) Failing after 16m52s
ci / x86_64-linux-debug (pull_request) Failing after 56m6s
ci / x86_64-netbsd-release (pull_request) Failing after 30m43s
ci / powerpc64le-linux-release (pull_request) Failing after 1h10m39s
ci / x86_64-freebsd-release (pull_request) Failing after 32m55s
ci / powerpc64le-linux-debug (pull_request) Failing after 1h16m49s
ci / x86_64-netbsd-debug (pull_request) Failing after 41m7s
ci / x86_64-freebsd-debug (pull_request) Failing after 39m10s
ci / s390x-linux-release (pull_request) Failing after 1h24m19s
ci / aarch64-linux-release (pull_request) Failing after 1h29m28s
ci / x86_64-openbsd-debug (pull_request) Failing after 48m36s
ci / x86_64-openbsd-release (pull_request) Failing after 42m39s
ci / x86_64-windows-release (pull_request) Failing after 49m29s
ci / x86_64-windows-debug (pull_request) Failing after 1h2m40s
ci / aarch64-linux-debug (pull_request) Failing after 2h5m45s
ci / s390x-linux-debug (pull_request) Failing after 2h16m20s
ci / x86_64-linux-release (pull_request) Failing after 2h33m50s
ci / x86_64-linux-debug-llvm (pull_request) Failing after 3h20m37s
ci / aarch64-freebsd-debug (pull_request) Has been cancelled
ci / aarch64-freebsd-release (pull_request) Has been cancelled
ci / aarch64-netbsd-debug (pull_request) Has been cancelled
ci / aarch64-netbsd-release (pull_request) Has been cancelled
ci / loongarch64-linux-debug (pull_request) Has been cancelled
ci / loongarch64-linux-release (pull_request) Has been cancelled
ci / riscv64-linux-debug (pull_request) Has been cancelled
ci / riscv64-linux-release (pull_request) Has been cancelled
to 9d69ab5195
Some checks failed
ci / aarch64-macos-release (pull_request) Failing after 15m54s
ci / aarch64-macos-debug (pull_request) Failing after 16m44s
ci / x86_64-netbsd-release (pull_request) Failing after 29m10s
ci / x86_64-freebsd-release (pull_request) Failing after 28m12s
ci / x86_64-netbsd-debug (pull_request) Failing after 39m5s
ci / x86_64-freebsd-debug (pull_request) Failing after 38m25s
ci / x86_64-openbsd-release (pull_request) Failing after 41m51s
ci / x86_64-windows-release (pull_request) Failing after 46m37s
ci / x86_64-openbsd-debug (pull_request) Failing after 48m35s
ci / x86_64-linux-debug (pull_request) Failing after 1h0m9s
ci / x86_64-windows-debug (pull_request) Failing after 1h0m6s
ci / powerpc64le-linux-release (pull_request) Failing after 1h9m36s
ci / powerpc64le-linux-debug (pull_request) Failing after 1h14m46s
ci / s390x-linux-release (pull_request) Failing after 1h17m28s
ci / aarch64-linux-release (pull_request) Failing after 1h21m41s
ci / aarch64-linux-debug (pull_request) Failing after 2h2m39s
ci / s390x-linux-debug (pull_request) Failing after 2h17m32s
ci / x86_64-linux-release (pull_request) Failing after 3h16m41s
ci / x86_64-linux-debug-llvm (pull_request) Failing after 3h19m50s
ci / aarch64-freebsd-debug (pull_request) Has been cancelled
ci / aarch64-freebsd-release (pull_request) Has been cancelled
ci / aarch64-netbsd-debug (pull_request) Has been cancelled
ci / aarch64-netbsd-release (pull_request) Has been cancelled
ci / loongarch64-linux-debug (pull_request) Has been cancelled
ci / loongarch64-linux-release (pull_request) Has been cancelled
ci / riscv64-linux-debug (pull_request) Has been cancelled
ci / riscv64-linux-release (pull_request) Has been cancelled
2026年06月13日 17:04:51 +02:00 Compare
Author
Contributor
Copy link

@mlugg wrote in #35754 (comment):

To implement bulk memory operations, the values you store/load to/from memory must be arrays, never vectors.

Thanks for the explanation. I now managed to achieve similar performance with no @Vector.

@mlugg wrote in https://codeberg.org/ziglang/zig/pulls/35754#issuecomment-17392871: > To implement bulk memory operations, the values you store/load to/from memory must be _arrays_, never vectors. Thanks for the explanation. I now managed to achieve similar performance with no `@Vector`.
rpkak force-pushed memset from 9d69ab5195
Some checks failed
ci / aarch64-macos-release (pull_request) Failing after 15m54s
ci / aarch64-macos-debug (pull_request) Failing after 16m44s
ci / x86_64-netbsd-release (pull_request) Failing after 29m10s
ci / x86_64-freebsd-release (pull_request) Failing after 28m12s
ci / x86_64-netbsd-debug (pull_request) Failing after 39m5s
ci / x86_64-freebsd-debug (pull_request) Failing after 38m25s
ci / x86_64-openbsd-release (pull_request) Failing after 41m51s
ci / x86_64-windows-release (pull_request) Failing after 46m37s
ci / x86_64-openbsd-debug (pull_request) Failing after 48m35s
ci / x86_64-linux-debug (pull_request) Failing after 1h0m9s
ci / x86_64-windows-debug (pull_request) Failing after 1h0m6s
ci / powerpc64le-linux-release (pull_request) Failing after 1h9m36s
ci / powerpc64le-linux-debug (pull_request) Failing after 1h14m46s
ci / s390x-linux-release (pull_request) Failing after 1h17m28s
ci / aarch64-linux-release (pull_request) Failing after 1h21m41s
ci / aarch64-linux-debug (pull_request) Failing after 2h2m39s
ci / s390x-linux-debug (pull_request) Failing after 2h17m32s
ci / x86_64-linux-release (pull_request) Failing after 3h16m41s
ci / x86_64-linux-debug-llvm (pull_request) Failing after 3h19m50s
ci / aarch64-freebsd-debug (pull_request) Has been cancelled
ci / aarch64-freebsd-release (pull_request) Has been cancelled
ci / aarch64-netbsd-debug (pull_request) Has been cancelled
ci / aarch64-netbsd-release (pull_request) Has been cancelled
ci / loongarch64-linux-debug (pull_request) Has been cancelled
ci / loongarch64-linux-release (pull_request) Has been cancelled
ci / riscv64-linux-debug (pull_request) Has been cancelled
ci / riscv64-linux-release (pull_request) Has been cancelled
to 1aa2663aae
All checks were successful
ci / x86_64-netbsd-release (pull_request) Successful in 41m48s
ci / aarch64-macos-release (pull_request) Successful in 43m13s
ci / x86_64-freebsd-release (pull_request) Successful in 43m3s
ci / x86_64-windows-release (pull_request) Successful in 45m13s
ci / x86_64-freebsd-debug (pull_request) Successful in 46m29s
ci / x86_64-netbsd-debug (pull_request) Successful in 47m48s
ci / x86_64-openbsd-release (pull_request) Successful in 51m50s
ci / x86_64-openbsd-debug (pull_request) Successful in 55m28s
ci / aarch64-macos-debug (pull_request) Successful in 58m35s
ci / x86_64-linux-debug (pull_request) Successful in 1h3m59s
ci / x86_64-windows-debug (pull_request) Successful in 1h7m24s
ci / s390x-linux-release (pull_request) Successful in 1h29m49s
ci / aarch64-linux-release (pull_request) Successful in 1h36m31s
ci / powerpc64le-linux-release (pull_request) Successful in 2h13m10s
ci / aarch64-linux-debug (pull_request) Successful in 2h38m10s
ci / s390x-linux-debug (pull_request) Successful in 2h43m58s
ci / x86_64-linux-debug-llvm (pull_request) Successful in 2h49m42s
ci / x86_64-linux-release (pull_request) Successful in 3h13m59s
ci / loongarch64-linux-release (pull_request) Successful in 2h4m15s
ci / powerpc64le-linux-debug (pull_request) Successful in 5h36m7s
ci / loongarch64-linux-debug (pull_request) Successful in 3h32m53s
ci / aarch64-freebsd-debug (pull_request) Has been skipped
ci / aarch64-freebsd-release (pull_request) Has been skipped
ci / aarch64-netbsd-debug (pull_request) Has been skipped
ci / aarch64-netbsd-release (pull_request) Has been skipped
ci / riscv64-linux-debug (pull_request) Has been skipped
ci / riscv64-linux-release (pull_request) Has been skipped
2026年06月14日 14:54:52 +02:00 Compare
All checks were successful
ci / x86_64-netbsd-release (pull_request) Successful in 41m48s
Required
Details
ci / aarch64-macos-release (pull_request) Successful in 43m13s
Required
Details
ci / x86_64-freebsd-release (pull_request) Successful in 43m3s
Required
Details
ci / x86_64-windows-release (pull_request) Successful in 45m13s
Required
Details
ci / x86_64-freebsd-debug (pull_request) Successful in 46m29s
Required
Details
ci / x86_64-netbsd-debug (pull_request) Successful in 47m48s
Required
Details
ci / x86_64-openbsd-release (pull_request) Successful in 51m50s
Required
Details
ci / x86_64-openbsd-debug (pull_request) Successful in 55m28s
Required
Details
ci / aarch64-macos-debug (pull_request) Successful in 58m35s
Required
Details
ci / x86_64-linux-debug (pull_request) Successful in 1h3m59s
Required
Details
ci / x86_64-windows-debug (pull_request) Successful in 1h7m24s
Required
Details
ci / s390x-linux-release (pull_request) Successful in 1h29m49s
Required
Details
ci / aarch64-linux-release (pull_request) Successful in 1h36m31s
Required
Details
ci / powerpc64le-linux-release (pull_request) Successful in 2h13m10s
Required
Details
ci / aarch64-linux-debug (pull_request) Successful in 2h38m10s
Required
Details
ci / s390x-linux-debug (pull_request) Successful in 2h43m58s
Required
Details
ci / x86_64-linux-debug-llvm (pull_request) Successful in 2h49m42s
Required
Details
ci / x86_64-linux-release (pull_request) Successful in 3h13m59s
Required
Details
ci / loongarch64-linux-release (pull_request) Successful in 2h4m15s
ci / powerpc64le-linux-debug (pull_request) Successful in 5h36m7s
Required
Details
ci / loongarch64-linux-debug (pull_request) Successful in 3h32m53s
ci / aarch64-freebsd-debug (pull_request) Has been skipped
ci / aarch64-freebsd-release (pull_request) Has been skipped
ci / aarch64-netbsd-debug (pull_request) Has been skipped
ci / aarch64-netbsd-release (pull_request) Has been skipped
ci / riscv64-linux-debug (pull_request) Has been skipped
ci / riscv64-linux-release (pull_request) Has been skipped
This pull request can be merged automatically.
This branch is out-of-date with the base branch
You are not authorized to merge this pull request.
View command line instructions

Checkout

From your project repository, check out a new branch and test the changes.
git fetch -u memset:rpkak-memset
git switch rpkak-memset
Sign in to join this conversation.
No reviewers
Labels
Clear labels
abi/f32
abi/ilp32
abi/sf
accepted
This proposal is planned.
arch/21k
arch/6502
arch/aarch64
arch/alpha
arch/amdgcn
arch/arc
arch/arc32
arch/arc64
arch/arm
arch/avr
arch/bfin
arch/bpf
arch/colossus
arch/cris
arch/csky
arch/dlx
arch/epiphany
arch/fr30
arch/frv
arch/hexagon
arch/hppa
arch/hppa64
arch/ia64
arch/kalimba
arch/kvx
arch/lanai
arch/lm32
arch/loongarch32
arch/loongarch64
arch/m32r
arch/m68k
arch/m88k
arch/mcore
arch/microblaze
arch/mips
arch/mips64
arch/mmix
arch/moxie
arch/mrisc32
arch/msp430
arch/nds32
arch/ns32k
arch/nvptx
arch/or1k
arch/powerpc
arch/powerpc64
arch/propeller
arch/riscv32
arch/riscv64
arch/rl78
arch/rx
arch/s390x
arch/sh
arch/sparc
arch/sparc64
arch/spirv
arch/spu
arch/tricore
arch/v850
arch/vax
arch/vc4
arch/ve
arch/wasm
arch/x86
arch/x86_64
arch/xcore
arch/xtensa
autodoc
The web application for interactive documentation and generation of its assets.
backend/c
The C backend outputs C source code.
backend/llvm
The LLVM backend outputs an LLVM bitcode module.
backend/self-hosted
The self-hosted backends produce machine code directly.
binutils
Zig's included binary utilities: zig ar, zig dlltool, zig lib, zig ranlib, zig objcopy, and zig rc.
breaking
Implementing this issue could cause existing code to no longer compile or have different behavior.
build system
The Zig build system - zig build, std.Build, the build runner, and package management.
debug info
An issue related to debug information (e.g. DWARF) produced by the Zig compiler.
docs
An issue with documentation, e.g. the language reference or standard library doc comments.
error message
This issue points out an error message that is unhelpful and should be improved.
frontend
Tokenization, parsing, AstGen, ZonGen, Sema, Legalize, and Liveness.
fuzzing
An issue related to Zig's integrated fuzz testing.
incremental
Reuse of internal compiler state for faster compilation.
lib/c
This issue relates to Zig's libc implementation and/or vendored libcs.
lib/compiler-rt
This issue relates to Zig's compiler-rt library.
lib/cxx
This issue relates to Zig's vendored libc++ and/or libc++abi.
lib/std
This issue relates to Zig's standard library.
lib/tsan
This issue relates to Zig's vendored libtsan.
lib/ubsan-rt
This issue relates to Zig's ubsan-rt library.
lib/unwind
This issue relates to Zig's vendored libunwind.
linking
Zig's integrated object file and incremental linker.
miscompilation
The compiler reports success but produces semantically incorrect code.
os/android
os/contiki
os/dragonfly
os/driverkit
os/emscripten
os/freebsd
os/fuchsia
os/haiku
os/hermit
os/hurd
os/illumos
os/ios
os/linux
os/maccatalyst
os/macos
os/managarm
os/netbsd
os/ohos
os/openbsd
os/plan9
os/redox
os/rtems
os/serenity
os/tvos
os/uefi
os/visionos
os/wasi
os/watchos
os/windows
proposal
This issue suggests language modifications. If it also has the "accepted" label then it is planned.
release notes
This issue or pull request should be mentioned in the release notes.
testing
This issue is related to testing the compiler, standard library, or other parts of Zig.
zig cc
Zig as a drop-in C-family compiler.
zig fmt
The Zig source code formatter.
zig reduce
The Zig source code reduction tool.
bounty
https://ziglang.org/news/announcing-donor-bounties
bug
Observed behavior contradicts documented or intended behavior.
contributor-friendly
This issue is limited in scope and/or knowledge of project internals.
downstream
An issue with a third-party project that uses this project.
enhancement
Solving this issue will likely involve adding new logic or components to the codebase.
infra
An issue related to project infrastructure, e.g. continuous integration.
optimization
A task to improve performance and/or resource usage.
question
No questions on the issue tracker; use a community space instead.
regression
Something that used to work in a previous version stopped working
upstream
An issue with a third-party project that this project uses.
use case
Describes a real use case that is difficult or impossible, but does not propose a solution.
No labels
abi/f32
abi/ilp32
abi/sf
accepted
arch/21k
arch/6502
arch/aarch64
arch/alpha
arch/amdgcn
arch/arc
arch/arc32
arch/arc64
arch/arm
arch/avr
arch/bfin
arch/bpf
arch/colossus
arch/cris
arch/csky
arch/dlx
arch/epiphany
arch/fr30
arch/frv
arch/hexagon
arch/hppa
arch/hppa64
arch/ia64
arch/kalimba
arch/kvx
arch/lanai
arch/lm32
arch/loongarch32
arch/loongarch64
arch/m32r
arch/m68k
arch/m88k
arch/mcore
arch/microblaze
arch/mips
arch/mips64
arch/mmix
arch/moxie
arch/mrisc32
arch/msp430
arch/nds32
arch/ns32k
arch/nvptx
arch/or1k
arch/powerpc
arch/powerpc64
arch/propeller
arch/riscv32
arch/riscv64
arch/rl78
arch/rx
arch/s390x
arch/sh
arch/sparc
arch/sparc64
arch/spirv
arch/spu
arch/tricore
arch/v850
arch/vax
arch/vc4
arch/ve
arch/wasm
arch/x86
arch/x86_64
arch/xcore
arch/xtensa
autodoc
backend/c
backend/llvm
backend/self-hosted
binutils
breaking
build system
debug info
docs
error message
frontend
fuzzing
incremental
lib/c
lib/compiler-rt
lib/cxx
lib/std
lib/tsan
lib/ubsan-rt
lib/unwind
linking
miscompilation
os/android
os/contiki
os/dragonfly
os/driverkit
os/emscripten
os/freebsd
os/fuchsia
os/haiku
os/hermit
os/hurd
os/illumos
os/ios
os/linux
os/maccatalyst
os/macos
os/managarm
os/netbsd
os/ohos
os/openbsd
os/plan9
os/redox
os/rtems
os/serenity
os/tvos
os/uefi
os/visionos
os/wasi
os/watchos
os/windows
proposal
release notes
testing
zig cc
zig fmt
zig reduce
bounty
bug
contributor-friendly
downstream
enhancement
infra
optimization
question
regression
upstream
use case
Milestone
Clear milestone
No items
No milestone
Projects
Clear projects
No items
No project
Assignees
Clear assignees
No assignees
2 participants
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference
ziglang/zig!35754
Reference in a new issue
ziglang/zig
No description provided.
Delete branch "rpkak/zig:memset"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?