[clang] [CIR] Add X86 prefetch builtins (PR #168051)
Hendrik Hübner via cfe-commits
cfe-commits at lists.llvm.org
Tue Nov 25 05:13:47 PST 2025
https://github.com/HendrikHuebner updated https://github.com/llvm/llvm-project/pull/168051
>From 67c3c53a0a2ea68492639cccd120eaba59e1678e Mon Sep 17 00:00:00 2001
From: hhuebner <hendrik.huebner18 at gmail.com>
Date: Tue, 25 Nov 2025 14:07:48 +0100
Subject: [PATCH 1/3] feedback
---
clang/include/clang/CIR/Dialect/IR/CIROps.td | 2 +-
clang/lib/CIR/CodeGen/CIRGenBuiltin.cpp | 22 +++++------
clang/lib/CIR/CodeGen/CIRGenBuiltinX86.cpp | 39 +++++++++++++++++++
.../test/CIR/CodeGen/X86/prefetchw-builtin.c | 35 +++++++++++++++++
clang/test/CIR/CodeGen/X86/sse-builtins.c | 28 +++++++++++++
5 files changed, 113 insertions(+), 13 deletions(-)
create mode 100644 clang/test/CIR/CodeGen/X86/prefetchw-builtin.c
diff --git a/clang/include/clang/CIR/Dialect/IR/CIROps.td b/clang/include/clang/CIR/Dialect/IR/CIROps.td
index a19c4f951fff9..eb0ab81c26be0 100644
--- a/clang/include/clang/CIR/Dialect/IR/CIROps.td
+++ b/clang/include/clang/CIR/Dialect/IR/CIROps.td
@@ -4431,7 +4431,7 @@ def CIR_PrefetchOp : CIR_Op<"prefetch"> {
$locality is a temporal locality specifier ranging from (0) - no locality,
to (3) - extremely local, keep in cache. If $locality is not present, the
default value is 3.
-
+
$isWrite specifies whether the prefetch is for a 'read' or 'write'. If
$isWrite is not specified, it means that prefetch is prepared for 'read'.
}];
diff --git a/clang/lib/CIR/CodeGen/CIRGenBuiltin.cpp b/clang/lib/CIR/CodeGen/CIRGenBuiltin.cpp
index d220fdf4dc8a7..f5c26b2d8507c 100644
--- a/clang/lib/CIR/CodeGen/CIRGenBuiltin.cpp
+++ b/clang/lib/CIR/CodeGen/CIRGenBuiltin.cpp
@@ -1258,10 +1258,11 @@ RValue CIRGenFunction::emitBuiltinExpr(const GlobalDecl &gd, unsigned builtinID,
// Now see if we can emit a target-specific builtin.
if (mlir::Value v = emitTargetBuiltinExpr(builtinID, e, returnValue)) {
+ if (mlir::isa<cir::VoidType>(v.getType()))
+ return RValue::get(nullptr);
+
switch (evalKind) {
case cir::TEK_Scalar:
- if (mlir::isa<cir::VoidType>(v.getType()))
- return RValue::get(nullptr);
return RValue::get(v);
case cir::TEK_Aggregate:
cgm.errorNYI(e->getSourceRange(), "aggregate return value from builtin");
@@ -1272,9 +1273,6 @@ RValue CIRGenFunction::emitBuiltinExpr(const GlobalDecl &gd, unsigned builtinID,
llvm_unreachable("Bad evaluation kind in EmitBuiltinExpr");
}
- cgm.errorNYI(e->getSourceRange(),
- std::string("unimplemented builtin call: ") +
- getContext().BuiltinInfo.getName(builtinID));
return getUndefRValue(e->getType());
}
@@ -1303,9 +1301,7 @@ static mlir::Value emitTargetArchBuiltinExpr(CIRGenFunction *cgf,
case llvm::Triple::aarch64_be:
case llvm::Triple::bpfeb:
case llvm::Triple::bpfel:
- // These are actually NYI, but that will be reported by emitBuiltinExpr.
- // At this point, we don't even know that the builtin is target-specific.
- return nullptr;
+ break;
case llvm::Triple::x86:
case llvm::Triple::x86_64:
@@ -1325,12 +1321,14 @@ static mlir::Value emitTargetArchBuiltinExpr(CIRGenFunction *cgf,
case llvm::Triple::hexagon:
case llvm::Triple::riscv32:
case llvm::Triple::riscv64:
- // These are actually NYI, but that will be reported by emitBuiltinExpr.
- // At this point, we don't even know that the builtin is target-specific.
- return {};
default:
- return {};
+ break;
}
+
+ cgf->cgm.errorNYI(e->getSourceRange(),
+ "target-specific builtin not implemented on this architecture: " +
+ cgf->getContext().BuiltinInfo.getName(builtinID));
+ return {};
}
mlir::Value
diff --git a/clang/lib/CIR/CodeGen/CIRGenBuiltinX86.cpp b/clang/lib/CIR/CodeGen/CIRGenBuiltinX86.cpp
index 978fee7dbec9d..40bccce010b85 100644
--- a/clang/lib/CIR/CodeGen/CIRGenBuiltinX86.cpp
+++ b/clang/lib/CIR/CodeGen/CIRGenBuiltinX86.cpp
@@ -16,10 +16,16 @@
#include "clang/Basic/Builtins.h"
#include "clang/Basic/TargetBuiltins.h"
#include "clang/CIR/MissingFeatures.h"
+#include "mlir/Dialect/LLVMIR/LLVMDialect.h"
using namespace clang;
using namespace clang::CIRGen;
+/// Get integer from a mlir::Value that is an int constant or a constant op.
+static int64_t getIntValueFromConstOp(mlir::Value val) {
+ return val.getDefiningOp<cir::ConstantOp>().getIntValue().getSExtValue();
+}
+
template <typename... Operands>
static mlir::Value emitIntrinsicCallOp(CIRGenFunction &cgf, const CallExpr *e,
const std::string &str,
@@ -68,6 +74,33 @@ static mlir::Value emitVectorFCmp(CIRGenBuilderTy &builder,
return bitCast;
}
+static mlir::Value emitPrefetch(CIRGenFunction &cgf, unsigned builtinID,
+ const CallExpr *e,
+ const llvm::SmallVector<mlir::Value> &ops) {
+
+ assert(builtinID == X86::BI_mm_prefetch || builtinID == X86::BI_m_prefetchw ||
+ builtinID == X86::BI_m_prefetch && "Expected prefetch builtin");
+ CIRGenBuilderTy &builder = cgf.getBuilder();
+ mlir::Location location = cgf.getLoc(e->getExprLoc());
+ mlir::Type voidTy = builder.getVoidTy();
+ mlir::Value addr = ops[0];
+ mlir::Value address = builder.createPtrBitcast(addr, voidTy);
+ bool isWrite{};
+ int locality{};
+
+ if (builtinID == X86::BI_mm_prefetch) {
+ int64_t hint = getIntValueFromConstOp(ops[1]);
+ isWrite = (hint >> 2) & 0x1;
+ locality = hint & 0x3;
+ } else {
+ isWrite = (builtinID == X86::BI_m_prefetchw);
+ locality = 0x3;
+ }
+
+ cir::PrefetchOp::create(builder, location, address, locality, isWrite);
+ return {};
+}
+
mlir::Value CIRGenFunction::emitX86BuiltinExpr(unsigned builtinID,
const CallExpr *expr) {
if (builtinID == Builtin::BI__builtin_cpu_is) {
@@ -108,6 +141,9 @@ mlir::Value CIRGenFunction::emitX86BuiltinExpr(unsigned builtinID,
switch (builtinID) {
default:
+ cgm.errorNYI(expr->getSourceRange(),
+ std::string("unimplemented X86 builtin call: ") +
+ getContext().BuiltinInfo.getName(builtinID));
return {};
case X86::BI_mm_clflush:
return emitIntrinsicCallOp(*this, expr, "x86.sse2.clflush", voidTy, ops[0]);
@@ -120,6 +156,9 @@ mlir::Value CIRGenFunction::emitX86BuiltinExpr(unsigned builtinID,
case X86::BI_mm_sfence:
return emitIntrinsicCallOp(*this, expr, "x86.sse.sfence", voidTy);
case X86::BI_mm_prefetch:
+ case X86::BI_m_prefetch:
+ case X86::BI_m_prefetchw:
+ return emitPrefetch(*this, builtinID, expr, ops);
case X86::BI__rdtsc:
case X86::BI__builtin_ia32_rdtscp: {
cgm.errorNYI(expr->getSourceRange(),
diff --git a/clang/test/CIR/CodeGen/X86/prefetchw-builtin.c b/clang/test/CIR/CodeGen/X86/prefetchw-builtin.c
new file mode 100644
index 0000000000000..78c6b26b5c2a6
--- /dev/null
+++ b/clang/test/CIR/CodeGen/X86/prefetchw-builtin.c
@@ -0,0 +1,35 @@
+// RUN: %clang_cc1 -x c -flax-vector-conversions=none -ffreestanding %s -triple=x86_64-unknown-linux -target-feature +sse -fclangir -emit-cir -o %t.cir -Wall -Werror
+// RUN: FileCheck --check-prefix=CIR --input-file=%t.cir %s
+// RUN: %clang_cc1 -x c -flax-vector-conversions=none -ffreestanding %s -triple=x86_64-unknown-linux -target-feature +sse -fclangir -emit-llvm -o %t.ll -Wall -Werror
+// RUN: FileCheck --check-prefixes=LLVM --input-file=%t.ll %s
+
+// RUN: %clang_cc1 -x c++ -flax-vector-conversions=none -ffreestanding %s -triple=x86_64-unknown-linux -target-feature +sse -fno-signed-char -fclangir -emit-cir -o %t.cir -Wall -Werror
+// RUN: FileCheck --check-prefix=CIR --input-file=%t.cir %s
+// RUN: %clang_cc1 -x c++ -flax-vector-conversions=none -ffreestanding %s -triple=x86_64-unknown-linux -target-feature +sse -fclangir -emit-llvm -o %t.ll -Wall -Werror
+// RUN: FileCheck --check-prefixes=LLVM --input-file=%t.ll %s
+
+// RUN: %clang_cc1 -x c -flax-vector-conversions=none -ffreestanding %s -triple=x86_64-unknown-linux -target-feature +sse -emit-llvm -o - -Wall -Werror | FileCheck %s -check-prefix=OGCG
+// RUN: %clang_cc1 -x c++ -flax-vector-conversions=none -ffreestanding %s -triple=x86_64-unknown-linux -target-feature +sse -emit-llvm -o - -Wall -Werror | FileCheck %s -check-prefix=OGCG
+
+
+#include <x86intrin.h>
+
+void test_m_prefetch_w(void *p) {
+ // CIR-LABEL: test_m_prefetch_w
+ // LLVM-LABEL: test_m_prefetch_w
+ // OGCG-LABEL: test_m_prefetch_w
+ return _m_prefetchw(p);
+ // CIR: cir.prefetch write locality(3) %{{.*}} : !cir.ptr<!void>
+ // LLVM: call void @llvm.prefetch.p0(ptr {{.*}}, i32 1, i32 3, i32 1)
+ // OGCG: call void @llvm.prefetch.p0(ptr {{.*}}, i32 1, i32 3, i32 1)
+}
+
+void test_m_prefetch(void *p) {
+ // CIR-LABEL: test_m_prefetch
+ // LLVM-LABEL: test_m_prefetch
+ // OGCG-LABEL: test_m_prefetch
+ return _m_prefetch(p);
+ // CIR: cir.prefetch read locality(3) %{{.*}} : !cir.ptr<!void>
+ // LLVM: call void @llvm.prefetch.p0(ptr {{.*}}, i32 0, i32 3, i32 1)
+ // OGCG: call void @llvm.prefetch.p0(ptr {{.*}}, i32 0, i32 3, i32 1)
+}
diff --git a/clang/test/CIR/CodeGen/X86/sse-builtins.c b/clang/test/CIR/CodeGen/X86/sse-builtins.c
index c893859b297cc..543dd64cc5ad8 100644
--- a/clang/test/CIR/CodeGen/X86/sse-builtins.c
+++ b/clang/test/CIR/CodeGen/X86/sse-builtins.c
@@ -71,3 +71,31 @@ __m128 test_mm_undefined_ps(void) {
// OGCG: ret <4 x float> zeroinitializer
return _mm_undefined_ps();
}
+
+void test_mm_prefetch(char const* p) {
+ // CIR-LABEL: test_mm_prefetch
+ // LLVM-LABEL: test_mm_prefetch
+ _mm_prefetch(p, 0);
+ // CIR: cir.prefetch read locality(0) %{{.*}} : !cir.ptr<!void>
+ // LLVM: call void @llvm.prefetch.p0(ptr {{.*}}, i32 0, i32 0, i32 1)
+ // OGCG: call void @llvm.prefetch.p0(ptr {{.*}}, i32 0, i32 0, i32 1)
+}
+
+void test_mm_prefetch_local(char const* p) {
+ // CIR-LABEL: test_mm_prefetch_local
+ // LLVM-LABEL: test_mm_prefetch_local
+ _mm_prefetch(p, 3);
+ // CIR: cir.prefetch read locality(3) %{{.*}} : !cir.ptr<!void>
+ // LLVM: call void @llvm.prefetch.p0(ptr {{.*}}, i32 0, i32 3, i32 1)
+ // OGCG: call void @llvm.prefetch.p0(ptr {{.*}}, i32 0, i32 3, i32 1)
+}
+
+void test_mm_prefetch_write(char const* p) {
+ // CIR-LABEL: test_mm_prefetch_write
+ // LLVM-LABEL: test_mm_prefetch_write
+ // OGCG-LABEL: test_mm_prefetch_write
+ _mm_prefetch(p, 7);
+ // CIR: cir.prefetch write locality(3) %{{.*}} : !cir.ptr<!void>
+ // LLVM: call void @llvm.prefetch.p0(ptr {{.*}}, i32 1, i32 3, i32 1)
+ // OGCG: call void @llvm.prefetch.p0(ptr {{.*}}, i32 1, i32 3, i32 1)
+}
>From c9229fa1a1a78b558abd85ba0f173fb0b12bbd3b Mon Sep 17 00:00:00 2001
From: hhuebner <hendrik.huebner18 at gmail.com>
Date: Tue, 25 Nov 2025 14:07:58 +0100
Subject: [PATCH 2/3] fmt
---
clang/lib/CIR/CodeGen/CIRGenBuiltin.cpp | 7 ++++---
clang/lib/CIR/CodeGen/CIRGenBuiltinX86.cpp | 4 ++--
2 files changed, 6 insertions(+), 5 deletions(-)
diff --git a/clang/lib/CIR/CodeGen/CIRGenBuiltin.cpp b/clang/lib/CIR/CodeGen/CIRGenBuiltin.cpp
index f5c26b2d8507c..40750b049d7d9 100644
--- a/clang/lib/CIR/CodeGen/CIRGenBuiltin.cpp
+++ b/clang/lib/CIR/CodeGen/CIRGenBuiltin.cpp
@@ -1325,9 +1325,10 @@ static mlir::Value emitTargetArchBuiltinExpr(CIRGenFunction *cgf,
break;
}
- cgf->cgm.errorNYI(e->getSourceRange(),
- "target-specific builtin not implemented on this architecture: " +
- cgf->getContext().BuiltinInfo.getName(builtinID));
+ cgf->cgm.errorNYI(
+ e->getSourceRange(),
+ "target-specific builtin not implemented on this architecture: " +
+ cgf->getContext().BuiltinInfo.getName(builtinID));
return {};
}
diff --git a/clang/lib/CIR/CodeGen/CIRGenBuiltinX86.cpp b/clang/lib/CIR/CodeGen/CIRGenBuiltinX86.cpp
index 40bccce010b85..febcfac822d6b 100644
--- a/clang/lib/CIR/CodeGen/CIRGenBuiltinX86.cpp
+++ b/clang/lib/CIR/CodeGen/CIRGenBuiltinX86.cpp
@@ -13,10 +13,10 @@
#include "CIRGenFunction.h"
#include "CIRGenModule.h"
+#include "mlir/Dialect/LLVMIR/LLVMDialect.h"
#include "clang/Basic/Builtins.h"
#include "clang/Basic/TargetBuiltins.h"
#include "clang/CIR/MissingFeatures.h"
-#include "mlir/Dialect/LLVMIR/LLVMDialect.h"
using namespace clang;
using namespace clang::CIRGen;
@@ -158,7 +158,7 @@ mlir::Value CIRGenFunction::emitX86BuiltinExpr(unsigned builtinID,
case X86::BI_mm_prefetch:
case X86::BI_m_prefetch:
case X86::BI_m_prefetchw:
- return emitPrefetch(*this, builtinID, expr, ops);
+ return emitPrefetch(*this, builtinID, expr, ops);
case X86::BI__rdtsc:
case X86::BI__builtin_ia32_rdtscp: {
cgm.errorNYI(expr->getSourceRange(),
>From 8ed31b5729e0852ff35b4cc1b86b732acc46eb40 Mon Sep 17 00:00:00 2001
From: hhuebner <hendrik.huebner18 at gmail.com>
Date: Tue, 25 Nov 2025 14:11:17 +0100
Subject: [PATCH 3/3] fix
---
clang/include/clang/CIR/Dialect/IR/CIROps.td | 2 +-
clang/lib/CIR/CodeGen/CIRGenBuiltin.cpp | 5 ++---
2 files changed, 3 insertions(+), 4 deletions(-)
diff --git a/clang/include/clang/CIR/Dialect/IR/CIROps.td b/clang/include/clang/CIR/Dialect/IR/CIROps.td
index eb0ab81c26be0..a19c4f951fff9 100644
--- a/clang/include/clang/CIR/Dialect/IR/CIROps.td
+++ b/clang/include/clang/CIR/Dialect/IR/CIROps.td
@@ -4431,7 +4431,7 @@ def CIR_PrefetchOp : CIR_Op<"prefetch"> {
$locality is a temporal locality specifier ranging from (0) - no locality,
to (3) - extremely local, keep in cache. If $locality is not present, the
default value is 3.
-
+
$isWrite specifies whether the prefetch is for a 'read' or 'write'. If
$isWrite is not specified, it means that prefetch is prepared for 'read'.
}];
diff --git a/clang/lib/CIR/CodeGen/CIRGenBuiltin.cpp b/clang/lib/CIR/CodeGen/CIRGenBuiltin.cpp
index 40750b049d7d9..56006f40274ec 100644
--- a/clang/lib/CIR/CodeGen/CIRGenBuiltin.cpp
+++ b/clang/lib/CIR/CodeGen/CIRGenBuiltin.cpp
@@ -1258,11 +1258,10 @@ RValue CIRGenFunction::emitBuiltinExpr(const GlobalDecl &gd, unsigned builtinID,
// Now see if we can emit a target-specific builtin.
if (mlir::Value v = emitTargetBuiltinExpr(builtinID, e, returnValue)) {
- if (mlir::isa<cir::VoidType>(v.getType()))
- return RValue::get(nullptr);
-
switch (evalKind) {
case cir::TEK_Scalar:
+ if (mlir::isa<cir::VoidType>(v.getType()))
+ return RValue::get(nullptr);
return RValue::get(v);
case cir::TEK_Aggregate:
cgm.errorNYI(e->getSourceRange(), "aggregate return value from builtin");
More information about the cfe-commits
mailing list