[CIR][X86] Add support for permd builtins

[YGGkk]

Part of #167765

[github-actions]

Thank you for submitting a Pull Request (PR) to the LLVM Project!

This PR will be automatically labeled and the relevant teams will be notified.

If you wish to, you can add reviewers by using the "Reviewers" section on this page.

If this is not working for you, it is probably because you do not have write permissions for the repository. In which case you can instead tag reviewers by name in a comment by using @ followed by their GitHub username.

If you have received no comments on your PR for a week, you can request a review by "ping"ing the PR by adding a comment “Ping”. The common courtesy "ping" rate is once a week. Please remember that you are asking for valuable time from other developers.

If you have further questions, they may be answered by the LLVM GitHub User Guide.

You can also ask questions in a comment on this PR, on the LLVM Discord or on the forums.

[llvmbot]

@llvm/pr-subscribers-clangir

@llvm/pr-subscribers-clang

Author: Zhihui Yang (YGGkk)

Changes

Part of #167765

---

Full diff: https://github.com/llvm/llvm-project/pull/172132.diff

2 Files Affected:

• (modified) clang/lib/CIR/CodeGen/CIRGenBuiltinX86.cpp (+16-1)
• (added) clang/test/CIR/CodeGenBuiltins/X86/permd-builtins.c (+62)

diff --git a/clang/lib/CIR/CodeGen/CIRGenBuiltinX86.cpp b/clang/lib/CIR/CodeGen/CIRGenBuiltinX86.cpp
index 72e6bea244802..aef0eb86b6826 100644
--- a/clang/lib/CIR/CodeGen/CIRGenBuiltinX86.cpp
+++ b/clang/lib/CIR/CodeGen/CIRGenBuiltinX86.cpp
@@ -1210,7 +1210,22 @@ CIRGenFunction::emitX86BuiltinExpr(unsigned builtinID, const CallExpr *expr) {
   case X86::BI__builtin_ia32_permdi256:
   case X86::BI__builtin_ia32_permdf256:
   case X86::BI__builtin_ia32_permdi512:
-  case X86::BI__builtin_ia32_permdf512:
+  case X86::BI__builtin_ia32_permdf512: {
+    unsigned imm = ops[1].getDefiningOp<cir::ConstantOp>()
+                       .getIntValue()
+                       .getZExtValue();
+    unsigned numElts = cast<cir::VectorType>(ops[0].getType()).getSize();
+
+    // These intrinsics operate on 256-bit lanes of four 64-bit elements.
+    int64_t Indices[8];
+    
+    for (unsigned l = 0; l != numElts; l += 4)
+      for (unsigned i = 0; i != 4; ++i)
+        Indices[l + i] = l + ((imm >> (2 * i)) & 0x3);
+
+    return builder.createVecShuffle(getLoc(expr->getExprLoc()), ops[0],
+                                    ArrayRef(Indices, numElts));
+  }
   case X86::BI__builtin_ia32_palignr128:
   case X86::BI__builtin_ia32_palignr256:
   case X86::BI__builtin_ia32_palignr512:
diff --git a/clang/test/CIR/CodeGenBuiltins/X86/permd-builtins.c b/clang/test/CIR/CodeGenBuiltins/X86/permd-builtins.c
new file mode 100644
index 0000000000000..93ae8d9c5304d
--- /dev/null
+++ b/clang/test/CIR/CodeGenBuiltins/X86/permd-builtins.c
@@ -0,0 +1,62 @@
+// RUN: %clang_cc1 -x c -ffreestanding -triple x86_64-unknown-linux -target-feature +avx512vl -Wno-implicit-function-declaration -fclangir -emit-cir -o %t.cir %s
+// RUN: FileCheck --check-prefix=CIR --input-file=%t.cir %s
+// RUN: %clang_cc1 -x c++ -ffreestanding -triple x86_64-unknown-linux -target-feature +avx512vl -Wno-implicit-function-declaration -fclangir -emit-cir -o %t.cir %s
+// RUN: FileCheck --check-prefix=CIR --input-file=%t.cir %s
+
+// RUN: %clang_cc1 -x c -ffreestanding -triple x86_64-unknown-linux -target-feature +avx512vl -Wno-implicit-function-declaration -fclangir -emit-llvm -o %t.ll %s
+// RUN: FileCheck --check-prefix=LLVM --input-file=%t.ll %s
+// RUN: %clang_cc1 -x c++ -ffreestanding -triple x86_64-unknown-linux -target-feature +avx512vl -Wno-implicit-function-declaration -fclangir -emit-llvm -o %t.ll %s
+// RUN: FileCheck --check-prefix=LLVM --input-file=%t.ll %s
+
+// RUN: %clang_cc1 -x c -ffreestanding -triple=x86_64-unknown-linux -target-feature +avx512vl -emit-llvm -Wall -Werror %s -o - | FileCheck %s -check-prefix=OGCG
+// RUN: %clang_cc1 -x c++ -ffreestanding -triple=x86_64-unknown-linux -target-feature +avx512vl -emit-llvm -Wall -Werror %s -o - | FileCheck %s -check-prefix=OGCG
+
+#include <immintrin.h>
+
+__m256i test__builtin_ia32_permdi256()
+{
+    // CIR-LABEL: test__builtin_ia32_permdi256
+    // CIR: cir.vec.shuffle({{%.*}}, {{%.*}}: !cir.vector<4 x !s64i>) [#cir.int<1> : !s32i, #cir.int<0> : !s32i, #cir.int<0> : !s32i, #cir.int<0> : !s32i] : !cir.vector<4 x !s64i>
+    // LLVM-LABEL: test__builtin_ia32_permdi256
+    // LLVM: shufflevector <4 x i64> {{%.*}}, <4 x i64> poison, <4 x i32> <i32 1, i32 0, i32 0, i32 0>
+    // OGCG-LABEL: test__builtin_ia32_permdi256
+    // OGCG: shufflevector <4 x i64> {{%.*}}, <4 x i64> poison, <4 x i32> <i32 1, i32 0, i32 0, i32 0>
+    __v4di vec = {0, 1, 2, 3};
+    return __builtin_ia32_permdi256(vec, 1);
+} 
+
+__m512i test__builtin_ia32_permdi512()
+{
+    // CIR-LABEL: test__builtin_ia32_permdi512
+    // CIR: cir.vec.shuffle({{%.*}}, {{%.*}}: !cir.vector<8 x !s64i>) [#cir.int<1> : !s32i, #cir.int<0> : !s32i, #cir.int<0> : !s32i, #cir.int<0> : !s32i, #cir.int<5> : !s32i, #cir.int<4> : !s32i, #cir.int<4> : !s32i, #cir.int<4> : !s32i] : !cir.vector<8 x !s64i>
+    // LLVM-LABEL: test__builtin_ia32_permdi512
+    // LLVM: shufflevector <8 x i64> {{%.*}}, <8 x i64> poison, <8 x i32> <i32 1, i32 0, i32 0, i32 0, i32 5, i32 4, i32 4, i32 4>
+    // OGCG-LABEL: test__builtin_ia32_permdi512
+    // OGCG: shufflevector <8 x i64> {{%.*}}, <8 x i64> poison, <8 x i32> <i32 1, i32 0, i32 0, i32 0, i32 5, i32 4, i32 4, i32 4>
+    __v8di vec = {0, 1, 2, 3, 4, 5, 6, 7};
+    return __builtin_ia32_permdi512(vec, 1);
+}
+
+__m256d test__builtin_ia32_permdf256()
+{
+    // CIR-LABEL: test__builtin_ia32_permdf256
+    // CIR: cir.vec.shuffle({{%.*}}, {{%.*}}: !cir.vector<4 x !cir.double>) [#cir.int<1> : !s32i, #cir.int<0> : !s32i, #cir.int<0> : !s32i, #cir.int<0> : !s32i] : !cir.vector<4 x !cir.double>
+    // LLVM-LABEL: test__builtin_ia32_permdf256
+    // LLVM: shufflevector <4 x double> {{%.*}}, <4 x double> poison, <4 x i32> <i32 1, i32 0, i32 0, i32 0>
+    // OGCG-LABEL: test__builtin_ia32_permdf256
+    // OGCG: shufflevector <4 x double> {{%.*}}, <4 x double> poison, <4 x i32> <i32 1, i32 0, i32 0, i32 0>
+    __v4df vec = {0, 1, 2, 3};
+    return __builtin_ia32_permdf256(vec, 1);
+}   
+
+__m512d test__builtin_ia32_permdf512()
+{
+    // CIR-LABEL: test__builtin_ia32_permdf512
+    // CIR: cir.vec.shuffle({{%.*}}, {{%.*}}: !cir.vector<8 x !cir.double>) [#cir.int<1> : !s32i, #cir.int<0> : !s32i, #cir.int<0> : !s32i, #cir.int<0> : !s32i, #cir.int<5> : !s32i, #cir.int<4> : !s32i, #cir.int<4> : !s32i, #cir.int<4> : !s32i] : !cir.vector<8 x !cir.double>
+    // LLVM-LABEL: test__builtin_ia32_permdf512
+    // LLVM: shufflevector <8 x double> {{%.*}}, <8 x double> poison, <8 x i32> <i32 1, i32 0, i32 0, i32 0, i32 5, i32 4, i32 4, i32 4>
+    // OGCG-LABEL: test__builtin_ia32_permdf512
+    // OGCG: shufflevector <8 x double> {{%.*}}, <8 x double> poison, <8 x i32> <i32 1, i32 0, i32 0, i32 0, i32 5, i32 4, i32 4, i32 4>
+    __v8df vec = {0, 1, 2, 3, 4, 5, 6, 7};
+    return __builtin_ia32_permdf512(vec, 1);
+}