- Notifications
You must be signed in to change notification settings - Fork42
Detect CPU features with single-file
License
nihui/ruapu
Folders and files
| Name | Name | Last commit message | Last commit date | |
|---|---|---|---|---|
Repository files navigation
Detect CPU ISA features with single-file
| CPU | ✅ x86, x86-64 ✅ arm, aarch64 ✅ mips ✅ powerpc ✅ s390x ✅ loongarch ✅ risc-v ✅ openrisc | #defineRUAPU_IMPLEMENTATION#include"ruapu.h"intmain(){// initialize ruapu onceruapu_init();// now, tell me if this cpu has avx2inthas_avx2=ruapu_supports("avx2");// loop all supported featuresconstchar*const*supported=ruapu_rua();while (*supported) {fprintf(stderr,"%s\n",*supported);supported++; }return0;} |
| OS | ✅ Windows ✅ Linux ✅ macOS ✅ Android ✅ iOS ✅ FreeBSD ✅ NetBSD ✅ OpenBSD ✅ DragonflyBSD ✅ Solaris ✅ SyterKit | |
| Compiler | ✅ GCC ✅ Clang ✅ MSVC ✅ MinGW |
- Create one
ruapu.cfor your project ruapu.cisONLY#define RUAPU_IMPLEMENTATIONand#include "ruapu.h"- Other sources
#include "ruapu.h"butNO#define RUAPU_IMPLEMENTATION
- DetectCPU ISA with single-file
sse2,avx,avx512f,neon, etc. - Detectvendor extended ISA apple
amx, risc-v vendor ISA, etc. - Detectricher ISA on Windows ARM
IsProcessorFeaturePresent()returns little ISA information - Detect
x86-avx512on macOS correctly macOS hides it incpuid - Detectnew CPU's ISA on old systems they are usually not exposed in
auxvorMISA - DetectCPU hidden ISA
fma4on zen1, ISA in hypervisor, etc.
| CPU | ISA |
|---|---|
| x86 | mmxssesse2sse3ssse3sse41sse42sse4axopavxf16cfmafma4avx2avx512favx512bwavx512cdavx512dqavx512vlavx512vnniavx512bf16avx512ifmaavx512vbmiavx512vbmi2avx512fp16avx512eravx5124fmapsavx5124vnniwavxvnniavxvnniint8avxvnniint16avxifmaavxneconvertamxfp16amxbf16amxint8amxtilebmi1bmi2gfniaesnivaessha1sha256sha512sm3sm4rdrandrdseedtsx |
| arm | halfedspneonvfpv4idiv |
| aarch64 | neonvfpv4lsecpuidasimdrdmasimdhpasimddpasimdfhmbf16i8mmfrintjscvtfcmamtemte2svesve2svebf16svei8mmsvef32mmsvef64mmsmesmef16f16smef64f64smei64i64pmullcrc32aessha1sha2sha3sha512sm3sm4svepmullsvebitpermsveaessvesha3svesm4amx |
| mips | msammisxasxmsa2crypto |
| powerpc | vsx |
| s390x | zvector |
| loongarch | lsxlasx |
| risc-v | imafdcvzbazbbzbczbszbkbzbkczbkxzcbzfazfbfminzfhzfhminzicbopzicondzicsrzifenceizihintpausezmmulzvbbzvbczvfhzvfhminzvfbfminzvfbfwmazvkbzvl32bzvl64bzvl128bzvl256bzvl512bzvl1024bxtheadbaxtheadbbxtheadbsxtheadcondmovxtheadfmemidxxtheadfmvxtheadmacxtheadmemidxxtheadmempairxtheadsyncxtheadvectorxtheadvdotspacemitvmadotspacemitvmadotnspacemitvfmadot |
| openrisc | orbis32orbis64orfpx32orfpx64orvdx64 |
Compile ruapu test program # GCC / MinGWgcc main.c -o ruapu# Clangclang main.c -o ruapu# MSVCcl.exe /Fe: ruapu.exe main.c | Run ruapu in command line ./ruapummx = 1sse = 1sse2 = 1sse3 = 1ssse3 = 1sse41 = 1sse42 = 1sse4a = 1xop = 0... more lines omitted ... |
Compile and install ruapu library # from pypipip3 install ruapu# from source codepip3 install ./python | Use ruapu in python importruapuruapu.supports("avx2")# Trueruapu.supports(isa="avx2")# Trueruapu.rua()#(mmx', 'sse', 'sse2', 'sse3', 'ssse3', 'sse41', 'sse42', 'avx', 'f16c', 'fma', 'avx2') |
Compile ruapu library # from source codecd rustcargo build --release | Use ruapu in Rust externcrate ruapu;fnmain(){println!("supports neon: {}", ruapu::supports("neon").unwrap());println!("supports avx2: {}", ruapu::supports("avx2").unwrap());println!("rua: {:?}", ruapu::rua());} |
Compile ruapu library # from source codecd lua# lua binding has been tested on Lua 5.2~5.4luarocks make | Use ruapu in Lua ruapu=require"ruapu";print(ruapu.supports("mmx"));for_,extinipairs(ruapu.rua())doprint(ext);end |
Compile ruapu library % add this to deps list% in your rebar.config{ruapu,"0.1.0"} | Use ruapu in Erlang ruapu:rua().{ok,["neon","vfpv4","asimdrdm","asimdhp","asimddp","asimdfhm","bf16","i8mm","pmull","crc32","aes","sha1","sha2","sha3","sha512","amx"]}>ruapu:supports("neon").true>ruapu:supports(neon).true>ruapu:supports(<<"neon">>).true>ruapu:supports("avx2").false>ruapu:supports(avx2).false>ruapu:supports(<<"avx2">>).false |
Compile ruapu library # from source codecd fortrancmake -B buildcmake --build build | Use ruapu in Fortran program main use ruapu, only: ruapu_init, ruapu_supports, ruapu_ruaimplicit nonecharacter(len=:), allocatable:: isa_supported(:)integer:: icall ruapu_init()print*,"supports sse:", ruapu_supports("sse")print*,"supports neon:", ruapu_supports("neon") isa_supported= ruapu_rua()do i=1,size(isa_supported)print*,trim(isa_supported(i))end doend program main |
Compile ruapu library cd gogo build -o ruapu-go | Use ruapu in Golang package mainimport ("fmt""ruapu-go/ruapu""strconv")funcmain() {ruapu.Init()avx2Status:=ruapu.Supports("avx2")fmt.Println("avx2:"+strconv.Itoa(avx2Status))rua:=ruapu.Rua()fmt.Println(rua)} |
Add ruapu library to your project
| Use ruapu in Haskell importRuapu-- Ruapu.rua :: IO [String]-- Ruapu.supports :: String -> IO Boolmain=doRuapu.initRuapu.supports"mmx">>=putStrLn.showRuapu.rua>>=sequence_.mapputStrLn |
Compile ruapu library cd vlangv. | Use ruapu in Vlang modulemainimportruapufnmain() {ruapu.ruapu_init()mutavx2_status:= ruapu.ruapu_supports('avx2')ifavx2_status {println('avx2: '+avx2_status.str()) }println(ruapu.ruapu_rua())} |
Compile ruapu library cd pascalsudo apt install fpccmake.makefpc ruapu.lpr | Use ruapu in Pascal program ruapu;uses ruapu_pascal;var has_avx2: integer; supported: PPAnsiChar;begin// initialize ruapu once ruapu_init();// now, tell me if this cpu has avx2 has_avx2 := ruapu_supports('avx2');// loop all supported features supported := ruapu_rua();while supported^ <>nildobegin writeln(supported^); inc(supported);end; readln();end. |
Compile ruapu library and example ./gradlew build Run example java -cp \ ./build/libs/ruapu-1.0-SNAPSHOT.jar \ ./Example.java | Use ruapu in Java importruapu.Ruapu;importjava.util.*;classExample {publicstaticvoidmain(Stringargs[]) {Ruapuruapu =newRuapu();System.out.println("avx: " +ruapu.supports("avx"));// avx: 1System.out.println(Arrays.toString(ruapu.rua()));// [mmx, sse, sse2, sse3, ssse3, sse41, sse42, avx, f16c, fma, avx2] }} |
Compile ruapu library cd cangjiecd c-srccmake.make run example cd cangjiecjpm runor compile example cd cangjiecjpm build./target/release/bin/main | Use ruapu in cangjieimport ruapu.*main(): Int64{ruapu_init()letneon_supported=ruapu_supports("neon")println("supports neon: ${neon_supported}")letd=ruapu_rua() for(i in d){println(i)}return0} |
Compile ruapu library cd dartbash build.sh | Use ruapu in dartvoidmain() {var libraryPath= path.join(Directory.current.path,'build','libruapu.so');if (Platform.isMacOS) { libraryPath= path.join(Directory.current.path,'build','libruapu.dylib'); }if (Platform.isWindows) { libraryPath= path.join(Directory.current.path,'build','Debug','ruapu.dll'); }Ruapu ruapu=Ruapu(libraryPath); ruapu.init();List<String> isas= ruapu.rua();print("This CPU Support:");for (String isain isas) {print(isa); }print("=================");String isaToCheck='aes';bool isSupported= ruapu.supports(isaToCheck);print('Does the system support $isaToCheck? $isSupported');} |
Github-hosted runner result (Linux)
mmx = 1sse = 1sse2 = 1sse3 = 1ssse3 = 1sse41 = 1sse42 = 1sse4a = 1xop = 0avx = 1f16c = 1fma = 1avx2 = 1avx512f = 0avx512bw = 0avx512cd = 0avx512dq = 0avx512vl = 0avx512vnni = 0avx512bf16 = 0avx512ifma = 0avx512vbmi = 0avx512vbmi2 = 0avx512fp16 = 0avx512er = 0avx5124fmaps = 0avx5124vnniw = 0avxvnni = 0avxvnniint8 = 0avxifma = 0amxfp16 = 0amxbf16 = 0amxint8 = 0amxtile = 0Github-hosted runner result (macOS)
mmx = 1sse = 1sse2 = 1sse3 = 1ssse3 = 1sse41 = 1sse42 = 1sse4a = 0xop = 0avx = 1f16c = 1fma = 1avx2 = 1avx512f = 0avx512bw = 0avx512cd = 0avx512dq = 0avx512vl = 0avx512vnni = 0avx512bf16 = 0avx512ifma = 0avx512vbmi = 0avx512vbmi2 = 0avx512fp16 = 0avx512er = 0avx5124fmaps = 0avx5124vnniw = 0avxvnni = 0avxvnniint8 = 0avxifma = 0amxfp16 = 0amxbf16 = 0amxint8 = 0amxtile = 0Github-hosted runner result (macOS M1)
neon = 1vfpv4 = 1cpuid = 0asimdhp = 1asimddp = 1asimdfhm = 1bf16 = 0i8mm = 0sve = 0sve2 = 0svebf16 = 0svei8mm = 0svef32mm = 0Github-hosted runner result (Windows)
mmx = 1sse = 1sse2 = 1sse3 = 1ssse3 = 1sse41 = 1sse42 = 1sse4a = 1xop = 0avx = 1f16c = 1fma = 1avx2 = 1avx512f = 0avx512bw = 0avx512cd = 0avx512dq = 0avx512vl = 0avx512vnni = 0avx512bf16 = 0avx512ifma = 0avx512vbmi = 0avx512vbmi2 = 0avx512fp16 = 0avx512er = 0avx5124fmaps = 0avx5124vnniw = 0avxvnni = 0avxvnniint8 = 0avxifma = 0amxfp16 = 0amxbf16 = 0amxint8 = 0amxtile = 0FreeBSD/NetBSD/OpenBSD VM result (x86_64)
mmx = 1sse = 1sse2 = 1sse3 = 1ssse3 = 1sse41 = 1sse42 = 1sse4a = 1xop = 0avx = 1f16c = 1fma = 1fma4 = 0avx2 = 1avx512f = 0avx512bw = 0avx512cd = 0avx512dq = 0avx512vl = 0avx512vnni = 0avx512bf16 = 0avx512ifma = 0avx512vbmi = 0avx512vbmi2 = 0avx512fp16 = 0avx512er = 0avx5124fmaps = 0avx5124vnniw = 0avxvnni = 0avxvnniint8 = 0avxifma = 0amxfp16 = 0amxbf16 = 0amxint8 = 0amxtile = 0ruapu is implemented in C language to ensure the widest possible portability.
ruapu determines whether the CPU supports certain instruction sets by trying to execute instructions and detecting whether anIllegal Instruction exception occurs. ruapu does not rely on the cpuid instructions and registers related to the CPU architecture, nor does it rely on theMISA information and system calls of the operating system. This can help us get more detailed CPU ISA information.
ruapu is the abbreviation of rua-cpu, which means using various extended instructions to harass and amuse the CPU (rua!). Based on whether the CPU reacts violently (throws an illegal instruction exception), it is inferred whether the CPU supports a certain extended instruction set.
We consider gcc builtin functions to be good practice, saying__builtin_cpu_init() and__builtin_cpu_supports(). ruapu refers to this design, which can be a 1:1 replacement for gcc functions, and supports more operating systems and compilers, giving it better portability.
Because debuggers and simulators capture the signal and stop the ruapu signal handler function by default, we can continue execution at this time, or configure it specifically, such ashandle SIGILL nostop in gdb. ruapu technically cannot prevent programs from stopping in debuggers and emulators
Assume that the new extended instruction set is namedrua
- Add
RUAPU_INSTCODE(rua, rua-inst-hex) // rua r0,r0andRUAPU_ISAENTRY(rua)inruapu.h - Add
PRINT_ISA_SUPPORT(rua)inmain.cto print the detection result - Add entries about
ruain README.md - Create a pull request!
https://godbolt.org/ is a good helper to view the compiled binary code of instructions.
- ncnn High-performance neural network inference framework
- libllm Efficient inference of large language models
- @nihui Write the initial POC code and ruapu maintainer
- @kernelbin Implement exception handling for Windows
- @zchrissirhcz Detect x86 FMA4
- @MollySophia Fix C++ export symbol
- @strongtz Detect more aarch64 ISA
- @monkeyking Detect apple arm64 AMX
- @junchao-loongson Add loongarch support
- @ziyao233 Detect more risc-v ISA
- @dreamcmi Detect more risc-v ISA
- @cocoa-xu Add FreeBSD support, python support
- @YuzukiTsuru Add OpenRISC support
- @whyb Detect more x86 AMX*, SHA*, AVX512*, SM*
MIT License
About
Detect CPU features with single-file