Mirrors/asmjit
1
0
Fork 0
mirror of https://github.com/asmjit/asmjit.git synced 2025-12-17 12:34:35 +03:00
Code Issues Packages Projects Releases Wiki Activity

[abi] AsmJit v1.17 - cumulative & breaking changes

Browse Source 
* Reworked register operands - all vector registers are now
    platform::Vec deriving from UniVec (universal vector operand),
    additionally, there is no platform::Reg, instead asmjit::Reg
    provides all necessary features to make it a base register for
    each target architecture
  * Reworked casting between registers - now architecture agnostic
    names are preferred - use Gp32 instead of Gpd or GpW, Gp64
    instead of Gpq and GpX, etc...
  * Reworked vector registers and their names - architecture
    agnostic naming is now preferred Vec32, Vec64, Vec128, etc...
  * Reworked naming conventions used across AsmJit - for clarity
    Identifiers are now prefixed with the type, like sectionId(),
    labelId(), etc...
  * Reworked how Zone and ZoneAllocator are used across AsmJit,
    prefering Zone in most cases and ZoneAllocator only for
    containers - this change alone achieves around 5% better
    performance of Builder and Compiler
  * Reworked LabelEntry - decreased the size of the base entry
    to 16 bytes for anonymous and unnamed labels. Avoided an
    indirection when using labelEntries() - LabelEntry is now
    a value and not a pointer
  * Renamed LabelLink to Fixup
  * Added a new header <asmjit/host.h> which would include
    <asmjit/core.h> + target tools for the host architecture,
    if enabled and supported
  * Added new AArch64 instructions (BTI, CSSC, CHKFEAT)
  * Added a mvn_ alternative of mvn instruction (fix for Windows
    ARM64 SDK)
  * Added more AArch64 CPU features to CpuInfo
  * Added better support for Apple CPU detection (Apple M3, M4)
  * Added a new benchmarking tool asmjit_bench_overhead, which
    benchmarks the overhead of CodeHolder::init()/reset() and
    creating/attaching emitters to it. Thanks to the benchmark the
    most common code-paths were optimized
  * Added a new benchmarking tool asmjit_bench_regalloc, which
    aims to benchmark the cost and complexity of register allocation.
  * Renamed asmjit_test_perf to asmjit_bench_codegen to make it
    clear what is a test and what is a benchmark
This commit is contained in:
kobalicek
2025-06-15 16:33:41 +02:00
parent f8e7f67b01
commit 2ff454d415
194 changed files with 9604 additions and 6913 deletions
Download Patch File Download Diff File Expand all files Collapse all files

90
test/asmjit_test_compiler_x86.cpp
View File

@@ -1,6 +1,6 @@
// This file is part of AsmJit project <https://asmjit.com>
//
// See asmjit.h or LICENSE.md for license and copyright information
// See <asmjit/core.h> or LICENSE.md for license and copyright information
// SPDX-License-Identifier: Zlib
#include <asmjit/core.h>
@@ -575,8 +575,8 @@ public:
}
virtual void compile(x86::Compiler& cc) {
x86::Xmm a = cc.newXmmSs("a");
x86::Xmm b = cc.newXmmSs("b");
x86::Vec a = cc.newXmmSs("a");
x86::Vec b = cc.newXmmSs("b");
x86::Gp op = cc.newUInt32("op");
x86::Gp target = cc.newIntPtr("target");
x86::Gp offset = cc.newIntPtr("offset");
@@ -1191,7 +1191,7 @@ public:
FuncNode* funcNode = cc.addFunc(FuncSignature::build<void, float, float, float, float, float, float, float, void*>());
x86::Gp p = cc.newIntPtr("p");
x86::Xmm xv[7];
x86::Vec xv[7];
for (uint32_t i = 0; i < 7; i++) {
xv[i] = cc.newXmmSs("xv%u", i);
@@ -1242,7 +1242,7 @@ public:
FuncNode* funcNode = cc.addFunc(FuncSignature::build<void, double, double, double, double, double, double, double, void*>());
x86::Gp p = cc.newIntPtr("p");
x86::Xmm xv[7];
x86::Vec xv[7];
for (uint32_t i = 0; i < 7; i++) {
xv[i] = cc.newXmmSd("xv%u", i);
@@ -1296,10 +1296,10 @@ public:
}
virtual void compile(x86::Compiler& cc) {
x86::Xmm a = cc.newXmm("aXmm");
x86::Xmm b = cc.newXmm("bXmm");
x86::Vec a = cc.newXmm("aXmm");
x86::Vec b = cc.newXmm("bXmm");
FuncNode* funcNode = cc.addFunc(FuncSignature::build<x86::Xmm, x86::Xmm, x86::Xmm>());
FuncNode* funcNode = cc.addFunc(FuncSignature::build<Type::Vec128, Type::Vec128, Type::Vec128>());
funcNode->setArg(0, a);
funcNode->setArg(1, b);
@@ -1345,7 +1345,7 @@ public:
}
virtual void compile(x86::Compiler& cc) {
x86::Xmm x = cc.newXmmSs("x");
x86::Vec x = cc.newXmmSs("x");
FuncNode* funcNode = cc.addFunc(FuncSignature::build<float, float>());
funcNode->setArg(0, x);
@@ -1380,8 +1380,8 @@ public:
}
virtual void compile(x86::Compiler& cc) {
x86::Xmm x = cc.newXmmSs("x");
x86::Xmm y = cc.newXmmSs("y");
x86::Vec x = cc.newXmmSs("x");
x86::Vec y = cc.newXmmSs("y");
FuncNode* funcNode = cc.addFunc(FuncSignature::build<float, float, float>());
funcNode->setArg(0, x);
@@ -1419,7 +1419,7 @@ public:
}
virtual void compile(x86::Compiler& cc) {
x86::Xmm x = cc.newXmmSd("x");
x86::Vec x = cc.newXmmSd("x");
FuncNode* funcNode = cc.addFunc(FuncSignature::build<double, double>());
funcNode->setArg(0, x);
@@ -1454,8 +1454,8 @@ public:
}
virtual void compile(x86::Compiler& cc) {
x86::Xmm x = cc.newXmmSd("x");
x86::Xmm y = cc.newXmmSd("y");
x86::Vec x = cc.newXmmSd("x");
x86::Vec y = cc.newXmmSd("y");
FuncNode* funcNode = cc.addFunc(FuncSignature::build<double, double, double>());
funcNode->setArg(0, x);
@@ -2495,7 +2495,7 @@ public:
using Func = void (*)(void*);
Func func = ptr_as_func<Func>(_func);
uint32_t out[16];
uint32_t out[16] {};
func(out);
result.assign("{");
@@ -2825,8 +2825,8 @@ public:
x86::Gp bPtr = cc.newIntPtr("bPtr");
x86::Gp pFn = cc.newIntPtr("pFn");
x86::Xmm aXmm = cc.newXmm("aXmm");
x86::Xmm bXmm = cc.newXmm("bXmm");
x86::Vec aXmm = cc.newXmm("aXmm");
x86::Vec bXmm = cc.newXmm("bXmm");
funcNode->setArg(0, resultPtr);
funcNode->setArg(1, aPtr);
@@ -2847,7 +2847,7 @@ public:
cc.movdqu(bXmm, x86::ptr(bPtr));
InvokeNode* invokeNode;
cc.invoke(&invokeNode, pFn, FuncSignature::build<x86::Xmm, x86::Xmm, x86::Xmm>(ccId));
cc.invoke(&invokeNode, pFn, FuncSignature::build<Type::Vec128, Type::Vec128, Type::Vec128>(ccId));
invokeNode->setArg(0, aXmm);
invokeNode->setArg(1, bXmm);
@@ -2901,7 +2901,7 @@ public:
virtual void compile(x86::Compiler& cc) {
FuncSignature f1Sig = FuncSignature::build<void, const void*, const void*, const void*, const void*, void*>();
FuncSignature f2Sig = FuncSignature::build<x86::Xmm, x86::Xmm, x86::Xmm>(CallConvId::kLightCall2);
FuncSignature f2Sig = FuncSignature::build<Type::Vec128, Type::Vec128, Type::Vec128>(CallConvId::kLightCall2);
FuncNode* f1Node = cc.newFunc(f1Sig);
FuncNode* f2Node = cc.newFunc(f2Sig);
@@ -2913,10 +2913,10 @@ public:
x86::Gp dPtr = cc.newIntPtr("dPtr");
x86::Gp pOut = cc.newIntPtr("pOut");
x86::Xmm aXmm = cc.newXmm("aXmm");
x86::Xmm bXmm = cc.newXmm("bXmm");
x86::Xmm cXmm = cc.newXmm("cXmm");
x86::Xmm dXmm = cc.newXmm("dXmm");
x86::Vec aXmm = cc.newXmm("aXmm");
x86::Vec bXmm = cc.newXmm("bXmm");
x86::Vec cXmm = cc.newXmm("cXmm");
x86::Vec dXmm = cc.newXmm("dXmm");
cc.addFunc(f1Node);
f1Node->setArg(0, aPtr);
@@ -2930,8 +2930,8 @@ public:
cc.movups(cXmm, x86::ptr(cPtr));
cc.movups(dXmm, x86::ptr(dPtr));
x86::Xmm xXmm = cc.newXmm("xXmm");
x86::Xmm yXmm = cc.newXmm("yXmm");
x86::Vec xXmm = cc.newXmm("xXmm");
x86::Vec yXmm = cc.newXmm("yXmm");
InvokeNode* invokeNode;
@@ -2952,8 +2952,8 @@ public:
}
{
x86::Xmm aXmm = cc.newXmm("aXmm");
x86::Xmm bXmm = cc.newXmm("bXmm");
x86::Vec aXmm = cc.newXmm("aXmm");
x86::Vec bXmm = cc.newXmm("bXmm");
cc.addFunc(f2Node);
f2Node->setArg(0, aXmm);
@@ -3322,9 +3322,9 @@ public:
virtual void compile(x86::Compiler& cc) {
FuncNode* funcNode = cc.addFunc(FuncSignature::build<float, float, float>());
x86::Xmm a = cc.newXmmSs("a");
x86::Xmm b = cc.newXmmSs("b");
x86::Xmm ret = cc.newXmmSs("ret");
x86::Vec a = cc.newXmmSs("a");
x86::Vec b = cc.newXmmSs("b");
x86::Vec ret = cc.newXmmSs("ret");
funcNode->setArg(0, a);
funcNode->setArg(1, b);
@@ -3372,9 +3372,9 @@ public:
virtual void compile(x86::Compiler& cc) {
FuncNode* funcNode = cc.addFunc(FuncSignature::build<double, double, double>());
x86::Xmm a = cc.newXmmSd("a");
x86::Xmm b = cc.newXmmSd("b");
x86::Xmm ret = cc.newXmmSd("ret");
x86::Vec a = cc.newXmmSd("a");
x86::Vec b = cc.newXmmSd("b");
x86::Vec ret = cc.newXmmSd("ret");
funcNode->setArg(0, a);
funcNode->setArg(1, b);
@@ -3692,10 +3692,10 @@ public:
virtual void compile(x86::Compiler& cc) {
FuncNode* funcNode = cc.addFunc(FuncSignature::build<double, double, double, double, double>());
x86::Xmm a0 = cc.newXmmSd("a0");
x86::Xmm a1 = cc.newXmmSd("a1");
x86::Xmm a2 = cc.newXmmSd("a2");
x86::Xmm a3 = cc.newXmmSd("a3");
x86::Vec a0 = cc.newXmmSd("a0");
x86::Vec a1 = cc.newXmmSd("a1");
x86::Vec a2 = cc.newXmmSd("a2");
x86::Vec a3 = cc.newXmmSd("a3");
funcNode->setArg(0, a0);
funcNode->setArg(1, a1);
@@ -3872,8 +3872,8 @@ public:
FuncNode* funcNode = cc.addFunc(FuncSignature::build<double, const double*>());
x86::Gp p = cc.newIntPtr("p");
x86::Xmm arg = cc.newXmmSd("arg");
x86::Xmm ret = cc.newXmmSd("ret");
x86::Vec arg = cc.newXmmSd("arg");
x86::Vec ret = cc.newXmmSd("ret");
funcNode->setArg(0, p);
cc.movsd(arg, x86::ptr(p));
@@ -3922,8 +3922,8 @@ public:
FuncNode* funcNode = cc.addFunc(FuncSignature::build<double, const double*>());
x86::Gp p = cc.newIntPtr("p");
x86::Xmm arg = cc.newXmmSd("arg");
x86::Xmm ret = cc.newXmmSd("ret");
x86::Vec arg = cc.newXmmSd("arg");
x86::Vec ret = cc.newXmmSd("ret");
funcNode->setArg(0, p);
cc.movsd(arg, x86::ptr(p));
@@ -3984,7 +3984,7 @@ public:
invokeSignature.setRet(TypeId::kFloat64);
cc.invoke(&invokeNode, imm((void*)calledFunc), invokeSignature);
x86::Xmm ret = cc.newXmmSd("ret");
x86::Vec ret = cc.newXmmSd("ret");
invokeNode->setRet(0, ret);
cc.ret(ret);
@@ -4157,7 +4157,7 @@ public:
x86::Gp aPtr = cc.newIntPtr("aPtr");
x86::Gp bPtr = cc.newIntPtr("bPtr");
x86::Gp tPtr = cc.newIntPtr("tPtr");
x86::Ymm acc[8];
x86::Vec acc[8];
x86::Mem stack = cc.newStack(32, 1, "stack");
mainFunc->setArg(0, dPtr);
@@ -4197,7 +4197,7 @@ public:
helperFunc->setArg(1, aPtr);
x86::Gp tmp = cc.newIntPtr("tmp");
x86::Ymm acc = cc.newYmm("acc");
x86::Vec acc = cc.newYmm("acc");
cc.mov(tmp, 1);
cc.vmovd(acc.xmm(), tmp);
@@ -4262,7 +4262,7 @@ public:
x86::Gp x = cc.newInt32("x");
x86::Gp t = cc.newInt32("t");
x86::Xmm v[kVecCount];
x86::Vec v[kVecCount];
func->setArg(0, x);