[TargetParser][AArch64] Add support for RDM feature in the target parser.

[mcrosier]

Summary

Unfamiliar territory, so please take a look.

Chad

[samparker]

Hi Chad,

Just a couple of small test suggestions from me.

thanks,
sam

unittests/Support/TargetParserTest.cpp

816–817

Would be good to add the feature here.

846–848

and here.

[SjoerdMeijer]

Looks good to me, was wondering if you're also planning to add a Clang patch to test +rdm?

[mcrosier]

Address Sam's feedback by adding a few more tests.

[mcrosier]

In D37081#851111, @SjoerdMeijer wrote:

Looks good to me, was wondering if you're also planning to add a Clang patch to test +rdm?

Yes, we should definitely add those. Coming up shortly.

[SjoerdMeijer]

lgtm

[mcrosier]

-Add +rdm to llvm::AArch64::getExtensionFeatures. This was exposed while trying to write clang/driver tests. Thanks, Sjoerd.

[mcrosier]

In D37081#851345, @mcrosier wrote:

In D37081#851111, @SjoerdMeijer wrote:

Looks good to me, was wondering if you're also planning to add a Clang patch to test +rdm?

Yes, we should definitely add those. Coming up shortly.

Please see: https://reviews.llvm.org/D37106

[mcrosier]

unittests/Support/TargetParserTest.cpp

846–848

I also added the missing LSE test.

include/llvm/Support/AArch64TargetParser.def

 AARCH64_ARCH("invalid", INVALID, "", "",
              ARMBuildAttrs::CPUArch::v8_A, FK_NONE, AArch64::AEK_NONE)
 AARCH64_ARCH("armv8-a", ARMV8A, "8-A", "v8", ARMBuildAttrs::CPUArch::v8_A,
              FK_CRYPTO_NEON_FP_ARMV8,
              (AArch64::AEK_CRYPTO | AArch64::AEK_FP | AArch64::AEK_SIMD))
 AARCH64_ARCH("armv8.1-a", ARMV8_1A, "8.1-A", "v8.1a",
              ARMBuildAttrs::CPUArch::v8_A, FK_CRYPTO_NEON_FP_ARMV8,
              (AArch64::AEK_CRC | AArch64::AEK_CRYPTO | AArch64::AEK_FP |
-              AArch64::AEK_SIMD | AArch64::AEK_LSE))
+              AArch64::AEK_SIMD | AArch64::AEK_LSE | AArch64::AEK_RDM))
 AARCH64_ARCH("armv8.2-a", ARMV8_2A, "8.2-A", "v8.2a",
              ARMBuildAttrs::CPUArch::v8_A, FK_CRYPTO_NEON_FP_ARMV8,
              (AArch64::AEK_CRC | AArch64::AEK_CRYPTO | AArch64::AEK_FP |
-              AArch64::AEK_SIMD | AArch64::AEK_RAS | AArch64::AEK_LSE))
+              AArch64::AEK_SIMD | AArch64::AEK_RAS | AArch64::AEK_LSE |
+              AArch64::AEK_RDM))
 AARCH64_ARCH("armv8.3-a", ARMV8_3A, "8.3-A", "v8.3a",
              ARMBuildAttrs::CPUArch::v8_A, FK_CRYPTO_NEON_FP_ARMV8,
              (AArch64::AEK_CRC | AArch64::AEK_CRYPTO | AArch64::AEK_FP |
               AArch64::AEK_SIMD | AArch64::AEK_RAS | AArch64::AEK_LSE |
-              AArch64::AEK_RCPC))
+              AArch64::AEK_RDM | AArch64::AEK_RCPC))
 #undef AARCH64_ARCH
 
 #ifndef AARCH64_ARCH_EXT_NAME
 #define AARCH64_ARCH_EXT_NAME(NAME, ID, FEATURE, NEGFEATURE)
 #endif
 // FIXME: This would be nicer were it tablegen
 AARCH64_ARCH_EXT_NAME("invalid",  AArch64::AEK_INVALID,  nullptr,  nullptr)
 AARCH64_ARCH_EXT_NAME("none",     AArch64::AEK_NONE,     nullptr,  nullptr)
 AARCH64_ARCH_EXT_NAME("crc",      AArch64::AEK_CRC,      "+crc",   "-crc")
 AARCH64_ARCH_EXT_NAME("lse",      AArch64::AEK_LSE,      "+lse",   "-lse")
+AARCH64_ARCH_EXT_NAME("rdm",      AArch64::AEK_RDM,      "+rdm",   "-rdm")
 AARCH64_ARCH_EXT_NAME("crypto",   AArch64::AEK_CRYPTO,   "+crypto","-crypto")
 AARCH64_ARCH_EXT_NAME("dotprod",  AArch64::AEK_DOTPROD,  "+dotprod","-dotprod")
 AARCH64_ARCH_EXT_NAME("fp",       AArch64::AEK_FP,       "+fp-armv8",  "-fp-armv8")
 AARCH64_ARCH_EXT_NAME("simd",     AArch64::AEK_SIMD,     "+neon",  "-neon")
 AARCH64_ARCH_EXT_NAME("fp16",     AArch64::AEK_FP16,     "+fullfp16",  "-fullfp16")
 AARCH64_ARCH_EXT_NAME("profile",  AArch64::AEK_PROFILE,  "+spe",  "-spe")
 AARCH64_ARCH_EXT_NAME("ras",      AArch64::AEK_RAS,      "+ras",  "-ras")
 AARCH64_ARCH_EXT_NAME("sve",      AArch64::AEK_SVE,      "+sve",  "-sve")
AARCH64_CPU_NAME("cyclone", ARMV8A, FK_CRYPTO_NEON_FP_ARMV8, false,
                 (AArch64::AEK_NONE))
 AARCH64_CPU_NAME("exynos-m1", ARMV8A, FK_CRYPTO_NEON_FP_ARMV8, false,
                 (AArch64::AEK_CRC))
 AARCH64_CPU_NAME("exynos-m2", ARMV8A, FK_CRYPTO_NEON_FP_ARMV8, false,
                 (AArch64::AEK_CRC))
 AARCH64_CPU_NAME("exynos-m3", ARMV8A, FK_CRYPTO_NEON_FP_ARMV8, false,
                 (AArch64::AEK_CRC))
 AARCH64_CPU_NAME("falkor", ARMV8A, FK_CRYPTO_NEON_FP_ARMV8, false,
-                (AArch64::AEK_CRC))
+                (AArch64::AEK_CRC | AArch64::AEK_RDM))
 AARCH64_CPU_NAME("kryo", ARMV8A, FK_CRYPTO_NEON_FP_ARMV8, false,
                 (AArch64::AEK_CRC))
 AARCH64_CPU_NAME("thunderx2t99", ARMV8_1A, FK_CRYPTO_NEON_FP_ARMV8, false,
                 (AArch64::AEK_NONE))
 AARCH64_CPU_NAME("thunderx", ARMV8A, FK_CRYPTO_NEON_FP_ARMV8, false,
                 (AArch64::AEK_CRC | AArch64::AEK_PROFILE))
 AARCH64_CPU_NAME("thunderxt88", ARMV8A, FK_CRYPTO_NEON_FP_ARMV8, false,
                 (AArch64::AEK_CRC | AArch64::AEK_PROFILE))
 AARCH64_CPU_NAME("thunderxt81", ARMV8A, FK_CRYPTO_NEON_FP_ARMV8, false,
                 (AArch64::AEK_CRC | AArch64::AEK_PROFILE))
 AARCH64_CPU_NAME("thunderxt83", ARMV8A, FK_CRYPTO_NEON_FP_ARMV8, false,
                 (AArch64::AEK_CRC | AArch64::AEK_PROFILE))
 // Invalid CPU
 AARCH64_CPU_NAME("invalid", INVALID, FK_INVALID, true, AArch64::AEK_INVALID)
 #undef AARCH64_CPU_NAME

include/llvm/Support/TargetParser.h

enum ArchExtKind : unsigned {
   AEK_CRC =         1 << 1,
   AEK_CRYPTO =      1 << 2,
   AEK_FP =          1 << 3,
   AEK_SIMD =        1 << 4,
   AEK_FP16 =        1 << 5,
   AEK_PROFILE =     1 << 6,
   AEK_RAS =         1 << 7,
   AEK_LSE =         1 << 8,
-  AEK_SVE =         1 << 9,
-  AEK_DOTPROD =     1 << 10,
-  AEK_RCPC =        1 << 11
+  AEK_RDM =         1 << 9,
+  AEK_SVE =         1 << 10,
+  AEK_DOTPROD =     1 << 11,
+  AEK_RCPC =        1 << 12
 };
 
 StringRef getCanonicalArchName(StringRef Arch);
 
 // Information by ID
 StringRef getFPUName(unsigned FPUKind);
 ARM::FPUVersion getFPUVersion(unsigned FPUKind);
 ARM::NeonSupportLevel getFPUNeonSupportLevel(unsigned FPUKind);

test/CodeGen/AArch64/arm64-neon-v8.1a.ll

 ; RUN: llc < %s -verify-machineinstrs -mtriple=arm64-eabi -aarch64-neon-syntax=generic | FileCheck %s --check-prefix=CHECK-V8a
 ; RUN: llc < %s -verify-machineinstrs -mtriple=arm64-eabi -mattr=+rdm -aarch64-neon-syntax=generic | FileCheck %s --check-prefix=CHECK-V81a
+; RUN: llc < %s -verify-machineinstrs -mtriple=arm64-eabi -mcpu=falkor -aarch64-neon-syntax=generic | FileCheck %s --check-prefix=CHECK-V81a
 ; RUN: llc < %s -verify-machineinstrs -mtriple=arm64-eabi -mattr=+v8.1a -aarch64-neon-syntax=generic | FileCheck %s --check-prefix=CHECK-V81a
 ; RUN: llc < %s -verify-machineinstrs -mtriple=arm64-eabi -mattr=+v8.1a -aarch64-neon-syntax=apple | FileCheck %s --check-prefix=CHECK-V81a-apple
 
 declare <4 x i16> @llvm.aarch64.neon.sqrdmulh.v4i16(<4 x i16>, <4 x i16>)
 declare <8 x i16> @llvm.aarch64.neon.sqrdmulh.v8i16(<8 x i16>, <8 x i16>)
 declare <2 x i32> @llvm.aarch64.neon.sqrdmulh.v2i32(<2 x i32>, <2 x i32>)
 declare <4 x i32> @llvm.aarch64.neon.sqrdmulh.v4i32(<4 x i32>, <4 x i32>)
 declare i32 @llvm.aarch64.neon.sqrdmulh.i32(i32, i32)

unittests/Support/TargetParserTest.cpp

TEST(TargetParserTest, testAArch64CPU) {
   EXPECT_TRUE(testAArch64CPU(
       "cortex-a53", "armv8-a", "crypto-neon-fp-armv8",
       AArch64::AEK_CRC | AArch64::AEK_CRYPTO | AArch64::AEK_FP |
       AArch64::AEK_SIMD, "8-A"));
   EXPECT_TRUE(testAArch64CPU(
       "cortex-a55", "armv8.2-a", "crypto-neon-fp-armv8",
       AArch64::AEK_CRC | AArch64::AEK_CRYPTO | AArch64::AEK_FP |
       AArch64::AEK_SIMD | AArch64::AEK_RAS | AArch64::AEK_LSE |
-      AArch64::AEK_FP16 | AArch64::AEK_DOTPROD | AArch64::AEK_RCPC,
-      "8.2-A"));
+      AArch64::AEK_RDM | AArch64::AEK_FP16 | AArch64::AEK_DOTPROD |
+      AArch64::AEK_RCPC, "8.2-A"));
   EXPECT_TRUE(testAArch64CPU(
       "cortex-a57", "armv8-a", "crypto-neon-fp-armv8",
       AArch64::AEK_CRC | AArch64::AEK_CRYPTO | AArch64::AEK_FP |
       AArch64::AEK_SIMD, "8-A"));
   EXPECT_TRUE(testAArch64CPU(
       "cortex-a72", "armv8-a", "crypto-neon-fp-armv8",
       AArch64::AEK_CRC | AArch64::AEK_CRYPTO | AArch64::AEK_FP |
       AArch64::AEK_SIMD, "8-A"));
   EXPECT_TRUE(testAArch64CPU(
       "cortex-a73", "armv8-a", "crypto-neon-fp-armv8",
       AArch64::AEK_CRC | AArch64::AEK_CRYPTO | AArch64::AEK_FP |
       AArch64::AEK_SIMD, "8-A"));
   EXPECT_TRUE(testAArch64CPU(
       "cortex-a75", "armv8.2-a", "crypto-neon-fp-armv8",
       AArch64::AEK_CRC | AArch64::AEK_CRYPTO | AArch64::AEK_FP |
       AArch64::AEK_SIMD | AArch64::AEK_RAS | AArch64::AEK_LSE |
-      AArch64::AEK_FP16 | AArch64::AEK_DOTPROD | AArch64::AEK_RCPC,
-      "8.2-A"));
+      AArch64::AEK_RDM | AArch64::AEK_FP16 | AArch64::AEK_DOTPROD |
+      AArch64::AEK_RCPC, "8.2-A"));
   EXPECT_TRUE(testAArch64CPU(
       "cyclone", "armv8-a", "crypto-neon-fp-armv8",
       AArch64::AEK_CRYPTO | AArch64::AEK_FP | AArch64::AEK_SIMD, "8-A"));
   EXPECT_TRUE(testAArch64CPU(
       "exynos-m1", "armv8-a", "crypto-neon-fp-armv8",
       AArch64::AEK_CRC | AArch64::AEK_CRYPTO | AArch64::AEK_FP |
       AArch64::AEK_SIMD, "8-A"));
   EXPECT_TRUE(testAArch64CPU(
       "exynos-m2", "armv8-a", "crypto-neon-fp-armv8",
       AArch64::AEK_CRC | AArch64::AEK_CRYPTO | AArch64::AEK_FP |
       AArch64::AEK_SIMD, "8-A"));
   EXPECT_TRUE(testAArch64CPU(
       "exynos-m3", "armv8-a", "crypto-neon-fp-armv8",
       AArch64::AEK_CRC | AArch64::AEK_CRYPTO | AArch64::AEK_FP |
       AArch64::AEK_SIMD, "8-A"));
   EXPECT_TRUE(testAArch64CPU(
       "falkor", "armv8-a", "crypto-neon-fp-armv8",
       AArch64::AEK_CRC | AArch64::AEK_CRYPTO | AArch64::AEK_FP |
-      AArch64::AEK_SIMD, "8-A"));
+      AArch64::AEK_SIMD | AArch64::AEK_RDM, "8-A"));
   EXPECT_TRUE(testAArch64CPU(
       "kryo", "armv8-a", "crypto-neon-fp-armv8",
       AArch64::AEK_CRC | AArch64::AEK_CRYPTO | AArch64::AEK_FP |
       AArch64::AEK_SIMD, "8-A"));
   EXPECT_TRUE(testAArch64CPU(
       "thunderx2t99", "armv8.1-a", "crypto-neon-fp-armv8",
       AArch64::AEK_CRC | AArch64::AEK_CRYPTO | AArch64::AEK_LSE |
-      AArch64::AEK_FP | AArch64::AEK_SIMD, "8.1-A"));
+      AArch64::AEK_RDM | AArch64::AEK_FP | AArch64::AEK_SIMD, "8.1-A"));
   EXPECT_TRUE(testAArch64CPU(
       "thunderx", "armv8-a", "crypto-neon-fp-armv8",
       AArch64::AEK_CRC | AArch64::AEK_CRYPTO | AArch64::AEK_SIMD |
       AArch64::AEK_FP | AArch64::AEK_PROFILE,
       "8-A"));
   EXPECT_TRUE(testAArch64CPU(
       "thunderxt81", "armv8-a", "crypto-neon-fp-armv8",
       AArch64::AEK_CRC | AArch64::AEK_CRYPTO | AArch64::AEK_SIMD |
TEST(TargetParserTest, testAArch64Extension) {
   EXPECT_FALSE(testAArch64Extension("cortex-a73",
                                     AArch64::ArchKind::INVALID, "ras"));
   EXPECT_TRUE(testAArch64Extension("cortex-a75",
                                     AArch64::ArchKind::INVALID, "ras"));
   EXPECT_FALSE(testAArch64Extension("cyclone",
                                     AArch64::ArchKind::INVALID, "ras"));
   EXPECT_FALSE(testAArch64Extension("exynos-m1",
                                     AArch64::ArchKind::INVALID, "ras"));
+  EXPECT_TRUE(testAArch64Extension("falkor",
+                                   AArch64::ArchKind::INVALID, "rdm"));
   EXPECT_FALSE(testAArch64Extension("kryo",
                                     AArch64::ArchKind::INVALID, "ras"));
   EXPECT_FALSE(testAArch64Extension("thunderx2t99",
                                     AArch64::ArchKind::INVALID, "ras"));
   EXPECT_FALSE(testAArch64Extension("thunderx",
                                     AArch64::ArchKind::INVALID, "lse"));
   EXPECT_FALSE(testAArch64Extension("thunderxt81",
                                     AArch64::ArchKind::INVALID, "lse"));
EXPECT_FALSE(testAArch64Extension(
       "generic", AArch64::ArchKind::ARMV8_2A, "spe"));
 }
 
 TEST(TargetParserTest, AArch64ExtensionFeatures) {
   std::vector<StringRef> Features;
   unsigned Extensions = AArch64::AEK_CRC | AArch64::AEK_CRYPTO |
                         AArch64::AEK_FP | AArch64::AEK_SIMD |
                         AArch64::AEK_FP16 | AArch64::AEK_PROFILE |
-                        AArch64::AEK_RAS | AArch64::AEK_SVE |
+                        AArch64::AEK_RAS | AArch64::AEK_LSE |
+                        AArch64::AEK_RDM | AArch64::AEK_SVE |
                         AArch64::AEK_DOTPROD | AArch64::AEK_RCPC;
 
   for (unsigned i = 0; i <= Extensions; i++)
     EXPECT_TRUE(i == 0 ? !AArch64::getExtensionFeatures(i, Features)
                        : AArch64::getExtensionFeatures(i, Features));
 }
 
 TEST(TargetParserTest, AArch64ArchFeatures) {

 
 TEST(TargetParserTest, AArch64ArchExtFeature) {
   const char *ArchExt[][4] = {{"crc", "nocrc", "+crc", "-crc"},
                               {"crypto", "nocrypto", "+crypto", "-crypto"},
                               {"fp", "nofp", "+fp-armv8", "-fp-armv8"},
                               {"simd", "nosimd", "+neon", "-neon"},
                               {"fp16", "nofp16", "+fullfp16", "-fullfp16"},
                               {"profile", "noprofile", "+spe", "-spe"},
                               {"ras", "noras", "+ras", "-ras"},
+                              {"lse", "nolse", "+lse", "-lse"},
+                              {"rdm", "nordm", "+rdm", "-rdm"},
                               {"sve", "nosve", "+sve", "-sve"},
                               {"dotprod", "nodotprod", "+dotprod", "-dotprod"},
                               {"rcpc", "norcpc", "+rcpc", "-rcpc" }};
 
   for (unsigned i = 0; i < array_lengthof(ArchExt); i++) {
     EXPECT_EQ(StringRef(ArchExt[i][2]),
               AArch64::getArchExtFeature(ArchExt[i][0]));
     EXPECT_EQ(StringRef(ArchExt[i][3]),
               AArch64::getArchExtFeature(ArchExt[i][1]));
   }
 }
 }