[clang] 916415b - [AArch64][SME] Make the overloaded svreinterpret_* functions streaming-compatible.

[Sander de Smalen via cfe-commits]

Author: Sander de Smalen
Date: 2023-09-04T10:15:26Z
New Revision: 916415b83732332d15125dee6d7a4b55438c8d7b

URL: https://github.com/llvm/llvm-project/commit/916415b83732332d15125dee6d7a4b55438c8d7b
DIFF: https://github.com/llvm/llvm-project/commit/916415b83732332d15125dee6d7a4b55438c8d7b.diff

LOG: [AArch64][SME] Make the overloaded svreinterpret_* functions streaming-compatible.

Otherwise these functions are not inlined when invoked from streaming
functions.

Reviewed By: paulwalker-arm

Differential Revision: https://reviews.llvm.org/D159188

Added: 
    clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_reinterpret_from_streaming_mode.c

Modified: 
    clang/include/clang/Basic/Attr.td
    clang/utils/TableGen/SveEmitter.cpp

Removed: 
    


################################################################################
diff  --git a/clang/include/clang/Basic/Attr.td b/clang/include/clang/Basic/Attr.td
index 2623d1741b9d3ba..4f281c5c8bdec3b 100644
--- a/clang/include/clang/Basic/Attr.td
+++ b/clang/include/clang/Basic/Attr.td
@@ -415,7 +415,7 @@ class TargetArch<list<string> arches> : TargetSpec {
   let Arches = arches;
 }
 def TargetARM : TargetArch<["arm", "thumb", "armeb", "thumbeb"]>;
-def TargetAArch64 : TargetArch<["aarch64"]>;
+def TargetAArch64 : TargetArch<["aarch64", "aarch64_be", "aarch64_32"]>;
 def TargetAnyArm : TargetArch<!listconcat(TargetARM.Arches, TargetAArch64.Arches)>;
 def TargetAVR : TargetArch<["avr"]>;
 def TargetBPF : TargetArch<["bpfel", "bpfeb"]>;

diff  --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_reinterpret_from_streaming_mode.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_reinterpret_from_streaming_mode.c
new file mode 100644
index 000000000000000..b430a0ba3cc75a3
--- /dev/null
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_reinterpret_from_streaming_mode.c
@@ -0,0 +1,35 @@
+// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py
+// REQUIRES: aarch64-registered-target
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve -S -O1 -Werror -Wall -emit-llvm -o - %s | FileCheck %s
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve -S -O1 -Werror -Wall -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -S -O1 -Werror -Wall -emit-llvm -o - %s | FileCheck %s
+// RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -S -O1 -Werror -Wall -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve -S -O1 -Werror -Wall -o /dev/null %s
+
+// Note: We need to run this test with '-O1' because oddly enough the svreinterpret is always inlined at -O0.
+
+#include <arm_sve.h>
+
+#ifdef SVE_OVERLOADED_FORMS
+// A simple used,unused... macro, long enough to represent any SVE builtin.
+#define SVE_ACLE_FUNC(A1,A2_UNUSED,A3,A4_UNUSED) A1##A3
+#else
+#define SVE_ACLE_FUNC(A1,A2,A3,A4) A1##A2##A3##A4
+#endif
+
+// Test that svreinterpret is inlined (because it should be streaming-compatible)
+__attribute__((target("sme")))
+// CHECK-LABEL: @test_svreinterpret_s16_s8_from_streaming_mode(
+// CHECK-NEXT:  entry:
+// CHECK-NEXT:    [[TMP0:%.*]] = bitcast <vscale x 16 x i8> [[OP:%.*]] to <vscale x 8 x i16>
+// CHECK-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z45test_svreinterpret_s16_s8_from_streaming_modeu10__SVInt8_t(
+// CPP-CHECK-NEXT:  entry:
+// CPP-CHECK-NEXT:    [[TMP0:%.*]] = bitcast <vscale x 16 x i8> [[OP:%.*]] to <vscale x 8 x i16>
+// CPP-CHECK-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
+//
+svint16_t test_svreinterpret_s16_s8_from_streaming_mode(svint8_t op) __arm_streaming {
+  return SVE_ACLE_FUNC(svreinterpret_s16,_s8,,)(op);
+}
+

diff  --git a/clang/utils/TableGen/SveEmitter.cpp b/clang/utils/TableGen/SveEmitter.cpp
index dbf5122fdf22317..f725c395400508f 100644
--- a/clang/utils/TableGen/SveEmitter.cpp
+++ b/clang/utils/TableGen/SveEmitter.cpp
@@ -1284,7 +1284,7 @@ void SVEEmitter::createHeader(raw_ostream &OS) {
         if (ShortForm) {
           OS << "__aio __attribute__((target(\"sve\"))) " << From.Type
              << " svreinterpret_" << From.Suffix;
-          OS << "(" << To.Type << " op) {\n";
+          OS << "(" << To.Type << " op) __arm_streaming_compatible {\n";
           OS << "  return __builtin_sve_reinterpret_" << From.Suffix << "_"
              << To.Suffix << "(op);\n";
           OS << "}\n\n";