[X86] Support -march=gracemont

FreddyLeaf · FreddyLeaf · commit 6acff5390d05 · 2023-08-21T08:49:01.000+08:00
gracemont has some different tuning features from alderlake. Reviewed By: RKSimon Differential Revision: https://reviews.llvm.org/D158046
diff --git a/clang/lib/Basic/Targets/X86.cpp b/clang/lib/Basic/Targets/X86.cpp
@@ -526,6 +526,8 @@ void X86TargetInfo::getTargetDefines(const LangOptions &Opts,
   case CK_Tremont:
     defineCPUMacros(Builder, "tremont");
     break;
+  // Gracemont and later atom-cores use P-core cpu macros.
+  case CK_Gracemont:
   case CK_Nehalem:
   case CK_Westmere:
   case CK_SandyBridge:
@@ -1420,6 +1422,7 @@ std::optional<unsigned> X86TargetInfo::getCPUCacheLineSize() const {
     case CK_Goldmont:
     case CK_GoldmontPlus:
     case CK_Tremont:
+    case CK_Gracemont:
 
     case CK_Westmere:
     case CK_SandyBridge:
diff --git a/clang/test/CodeGen/attr-cpuspecific-cpus.c b/clang/test/CodeGen/attr-cpuspecific-cpus.c
@@ -83,3 +83,4 @@ ATTR(cpu_specific(graniterapids_d)) void CPU34(void){}
 ATTR(cpu_specific(arrowlake)) void CPU35(void){}
 ATTR(cpu_specific(arrowlake_s)) void CPU36(void){}
 ATTR(cpu_specific(lunarlake)) void CPU37(void){}
+ATTR(cpu_specific(gracemont)) void CPU38(void){}
diff --git a/clang/test/CodeGen/attr-target-mv.c b/clang/test/CodeGen/attr-target-mv.c
@@ -25,6 +25,7 @@ int __attribute__((target("arch=graniterapids-d"))) foo(void) {return 20;}
 int __attribute__((target("arch=arrowlake"))) foo(void) {return 21;}
 int __attribute__((target("arch=arrowlake-s"))) foo(void) {return 22;}
 int __attribute__((target("arch=lunarlake"))) foo(void) {return 23;}
+int __attribute__((target("arch=gracemont"))) foo(void) {return 24;}
 int __attribute__((target("default"))) foo(void) { return 2; }
 
 int bar(void) {
@@ -179,6 +180,8 @@ void calls_pr50025c(void) { pr50025c(); }
 // LINUX: ret i32 22
 // LINUX: define{{.*}} i32 @foo.arch_lunarlake()
 // LINUX: ret i32 23
+// LINUX: define{{.*}} i32 @foo.arch_gracemont()
+// LINUX: ret i32 24
 // LINUX: define{{.*}} i32 @foo()
 // LINUX: ret i32 2
 // LINUX: define{{.*}} i32 @bar()
@@ -230,6 +233,8 @@ void calls_pr50025c(void) { pr50025c(); }
 // WINDOWS: ret i32 22
 // WINDOWS: define dso_local i32 @foo.arch_lunarlake()
 // WINDOWS: ret i32 23
+// WINDOWS: define dso_local i32 @foo.arch_gracemont()
+// WINDOWS: ret i32 24
 // WINDOWS: define dso_local i32 @foo()
 // WINDOWS: ret i32 2
 // WINDOWS: define dso_local i32 @bar()
diff --git a/clang/test/CodeGen/target-builtin-noerror.c b/clang/test/CodeGen/target-builtin-noerror.c
@@ -136,6 +136,7 @@ void verifycpustrings(void) {
   (void)__builtin_cpu_is("tigerlake");
   (void)__builtin_cpu_is("sapphirerapids");
   (void)__builtin_cpu_is("tremont");
+  (void)__builtin_cpu_is("gracemont");
   (void)__builtin_cpu_is("westmere");
   (void)__builtin_cpu_is("znver1");
   (void)__builtin_cpu_is("znver2");
diff --git a/clang/test/Driver/x86-march.c b/clang/test/Driver/x86-march.c
@@ -108,6 +108,10 @@
 // RUN:   | FileCheck %s -check-prefix=lunarlake
 // lunarlake: "-target-cpu" "lunarlake"
 //
+// RUN: %clang -target x86_64-unknown-unknown -c -### %s -march=gracemont 2>&1 \
+// RUN:   | FileCheck %s -check-prefix=gracemont
+// gracemont: "-target-cpu" "gracemont"
+//
 // RUN: %clang -target x86_64-unknown-unknown -c -### %s -march=lakemont 2>&1 \
 // RUN:   | FileCheck %s -check-prefix=lakemont
 // lakemont: "-target-cpu" "lakemont"
diff --git a/clang/test/Misc/target-invalid-cpu-note.c b/clang/test/Misc/target-invalid-cpu-note.c
@@ -13,19 +13,19 @@
 
 // RUN: not %clang_cc1 -triple i386--- -target-cpu not-a-cpu -fsyntax-only %s 2>&1 | FileCheck %s --check-prefix X86
 // X86: error: unknown target CPU 'not-a-cpu'
-// X86-NEXT: note: valid target CPU values are: i386, i486, winchip-c6, winchip2, c3, i586, pentium, pentium-mmx, pentiumpro, i686, pentium2, pentium3, pentium3m, pentium-m, c3-2, yonah, pentium4, pentium4m, prescott, nocona, core2, penryn, bonnell, atom, silvermont, slm, goldmont, goldmont-plus, tremont, nehalem, corei7, westmere, sandybridge, corei7-avx, ivybridge, core-avx-i, haswell, core-avx2, broadwell, skylake, skylake-avx512, skx, cascadelake, cooperlake, cannonlake, icelake-client, rocketlake, icelake-server, tigerlake, sapphirerapids, alderlake, raptorlake, meteorlake, arrowlake, arrowlake-s, lunarlake, sierraforest, grandridge, graniterapids, graniterapids-d, emeraldrapids, knl, knm, lakemont, k6, k6-2, k6-3, athlon, athlon-tbird, athlon-xp, athlon-mp, athlon-4, k8, athlon64, athlon-fx, opteron, k8-sse3, athlon64-sse3, opteron-sse3, amdfam10, barcelona, btver1, btver2, bdver1, bdver2, bdver3, bdver4, znver1, znver2, znver3, znver4, x86-64, x86-64-v2, x86-64-v3, x86-64-v4, geode{{$}}
+// X86-NEXT: note: valid target CPU values are: i386, i486, winchip-c6, winchip2, c3, i586, pentium, pentium-mmx, pentiumpro, i686, pentium2, pentium3, pentium3m, pentium-m, c3-2, yonah, pentium4, pentium4m, prescott, nocona, core2, penryn, bonnell, atom, silvermont, slm, goldmont, goldmont-plus, tremont, nehalem, corei7, westmere, sandybridge, corei7-avx, ivybridge, core-avx-i, haswell, core-avx2, broadwell, skylake, skylake-avx512, skx, cascadelake, cooperlake, cannonlake, icelake-client, rocketlake, icelake-server, tigerlake, sapphirerapids, alderlake, raptorlake, meteorlake, arrowlake, arrowlake-s, lunarlake, gracemont, sierraforest, grandridge, graniterapids, graniterapids-d, emeraldrapids, knl, knm, lakemont, k6, k6-2, k6-3, athlon, athlon-tbird, athlon-xp, athlon-mp, athlon-4, k8, athlon64, athlon-fx, opteron, k8-sse3, athlon64-sse3, opteron-sse3, amdfam10, barcelona, btver1, btver2, bdver1, bdver2, bdver3, bdver4, znver1, znver2, znver3, znver4, x86-64, x86-64-v2, x86-64-v3, x86-64-v4, geode{{$}}
 
 // RUN: not %clang_cc1 -triple x86_64--- -target-cpu not-a-cpu -fsyntax-only %s 2>&1 | FileCheck %s --check-prefix X86_64
 // X86_64: error: unknown target CPU 'not-a-cpu'
-// X86_64-NEXT: note: valid target CPU values are: nocona, core2, penryn, bonnell, atom, silvermont, slm, goldmont, goldmont-plus, tremont, nehalem, corei7, westmere, sandybridge, corei7-avx, ivybridge, core-avx-i, haswell, core-avx2, broadwell, skylake, skylake-avx512, skx, cascadelake, cooperlake, cannonlake, icelake-client, rocketlake, icelake-server, tigerlake, sapphirerapids, alderlake, raptorlake, meteorlake, arrowlake, arrowlake-s, lunarlake, sierraforest, grandridge, graniterapids, graniterapids-d, emeraldrapids, knl, knm, k8, athlon64, athlon-fx, opteron, k8-sse3, athlon64-sse3, opteron-sse3, amdfam10, barcelona, btver1, btver2, bdver1, bdver2, bdver3, bdver4, znver1, znver2, znver3, znver4, x86-64, x86-64-v2, x86-64-v3, x86-64-v4{{$}}
+// X86_64-NEXT: note: valid target CPU values are: nocona, core2, penryn, bonnell, atom, silvermont, slm, goldmont, goldmont-plus, tremont, nehalem, corei7, westmere, sandybridge, corei7-avx, ivybridge, core-avx-i, haswell, core-avx2, broadwell, skylake, skylake-avx512, skx, cascadelake, cooperlake, cannonlake, icelake-client, rocketlake, icelake-server, tigerlake, sapphirerapids, alderlake, raptorlake, meteorlake, arrowlake, arrowlake-s, lunarlake, gracemont, sierraforest, grandridge, graniterapids, graniterapids-d, emeraldrapids, knl, knm, k8, athlon64, athlon-fx, opteron, k8-sse3, athlon64-sse3, opteron-sse3, amdfam10, barcelona, btver1, btver2, bdver1, bdver2, bdver3, bdver4, znver1, znver2, znver3, znver4, x86-64, x86-64-v2, x86-64-v3, x86-64-v4{{$}}
 
 // RUN: not %clang_cc1 -triple i386--- -tune-cpu not-a-cpu -fsyntax-only %s 2>&1 | FileCheck %s --check-prefix TUNE_X86
 // TUNE_X86: error: unknown target CPU 'not-a-cpu'
-// TUNE_X86-NEXT: note: valid target CPU values are: i386, i486, winchip-c6, winchip2, c3, i586, pentium, pentium-mmx, pentiumpro, i686, pentium2, pentium3, pentium3m, pentium-m, c3-2, yonah, pentium4, pentium4m, prescott, nocona, core2, penryn, bonnell, atom, silvermont, slm, goldmont, goldmont-plus, tremont, nehalem, corei7, westmere, sandybridge, corei7-avx, ivybridge, core-avx-i, haswell, core-avx2, broadwell, skylake, skylake-avx512, skx, cascadelake, cooperlake, cannonlake, icelake-client, rocketlake, icelake-server, tigerlake, sapphirerapids, alderlake, raptorlake, meteorlake, arrowlake, arrowlake-s, lunarlake, sierraforest, grandridge, graniterapids, graniterapids-d, emeraldrapids, knl, knm, lakemont, k6, k6-2, k6-3, athlon, athlon-tbird, athlon-xp, athlon-mp, athlon-4, k8, athlon64, athlon-fx, opteron, k8-sse3, athlon64-sse3, opteron-sse3, amdfam10, barcelona, btver1, btver2, bdver1, bdver2, bdver3, bdver4, znver1, znver2, znver3, znver4, x86-64, geode{{$}}
+// TUNE_X86-NEXT: note: valid target CPU values are: i386, i486, winchip-c6, winchip2, c3, i586, pentium, pentium-mmx, pentiumpro, i686, pentium2, pentium3, pentium3m, pentium-m, c3-2, yonah, pentium4, pentium4m, prescott, nocona, core2, penryn, bonnell, atom, silvermont, slm, goldmont, goldmont-plus, tremont, nehalem, corei7, westmere, sandybridge, corei7-avx, ivybridge, core-avx-i, haswell, core-avx2, broadwell, skylake, skylake-avx512, skx, cascadelake, cooperlake, cannonlake, icelake-client, rocketlake, icelake-server, tigerlake, sapphirerapids, alderlake, raptorlake, meteorlake, arrowlake, arrowlake-s, lunarlake, gracemont, sierraforest, grandridge, graniterapids, graniterapids-d, emeraldrapids, knl, knm, lakemont, k6, k6-2, k6-3, athlon, athlon-tbird, athlon-xp, athlon-mp, athlon-4, k8, athlon64, athlon-fx, opteron, k8-sse3, athlon64-sse3, opteron-sse3, amdfam10, barcelona, btver1, btver2, bdver1, bdver2, bdver3, bdver4, znver1, znver2, znver3, znver4, x86-64, geode{{$}}
 
 // RUN: not %clang_cc1 -triple x86_64--- -tune-cpu not-a-cpu -fsyntax-only %s 2>&1 | FileCheck %s --check-prefix TUNE_X86_64
 // TUNE_X86_64: error: unknown target CPU 'not-a-cpu'
-// TUNE_X86_64-NEXT: note: valid target CPU values are: i386, i486, winchip-c6, winchip2, c3, i586, pentium, pentium-mmx, pentiumpro, i686, pentium2, pentium3, pentium3m, pentium-m, c3-2, yonah, pentium4, pentium4m, prescott, nocona, core2, penryn, bonnell, atom, silvermont, slm, goldmont, goldmont-plus, tremont, nehalem, corei7, westmere, sandybridge, corei7-avx, ivybridge, core-avx-i, haswell, core-avx2, broadwell, skylake, skylake-avx512, skx, cascadelake, cooperlake, cannonlake, icelake-client, rocketlake, icelake-server, tigerlake, sapphirerapids, alderlake, raptorlake, meteorlake, arrowlake, arrowlake-s, lunarlake, sierraforest, grandridge, graniterapids, graniterapids-d, emeraldrapids, knl, knm, lakemont, k6, k6-2, k6-3, athlon, athlon-tbird, athlon-xp, athlon-mp, athlon-4, k8, athlon64, athlon-fx, opteron, k8-sse3, athlon64-sse3, opteron-sse3, amdfam10, barcelona, btver1, btver2, bdver1, bdver2, bdver3, bdver4, znver1, znver2, znver3, znver4, x86-64, geode{{$}}
+// TUNE_X86_64-NEXT: note: valid target CPU values are: i386, i486, winchip-c6, winchip2, c3, i586, pentium, pentium-mmx, pentiumpro, i686, pentium2, pentium3, pentium3m, pentium-m, c3-2, yonah, pentium4, pentium4m, prescott, nocona, core2, penryn, bonnell, atom, silvermont, slm, goldmont, goldmont-plus, tremont, nehalem, corei7, westmere, sandybridge, corei7-avx, ivybridge, core-avx-i, haswell, core-avx2, broadwell, skylake, skylake-avx512, skx, cascadelake, cooperlake, cannonlake, icelake-client, rocketlake, icelake-server, tigerlake, sapphirerapids, alderlake, raptorlake, meteorlake, arrowlake, arrowlake-s, lunarlake, gracemont, sierraforest, grandridge, graniterapids, graniterapids-d, emeraldrapids, knl, knm, lakemont, k6, k6-2, k6-3, athlon, athlon-tbird, athlon-xp, athlon-mp, athlon-4, k8, athlon64, athlon-fx, opteron, k8-sse3, athlon64-sse3, opteron-sse3, amdfam10, barcelona, btver1, btver2, bdver1, bdver2, bdver3, bdver4, znver1, znver2, znver3, znver4, x86-64, geode{{$}}
 
 // RUN: not %clang_cc1 -triple nvptx--- -target-cpu not-a-cpu -fsyntax-only %s 2>&1 | FileCheck %s --check-prefix NVPTX
 // NVPTX: error: unknown target CPU 'not-a-cpu'
diff --git a/clang/test/Preprocessor/predefined-arch-macros.c b/clang/test/Preprocessor/predefined-arch-macros.c
@@ -2003,6 +2003,9 @@
 // RUN: %clang -march=meteorlake -m32 -E -dM %s -o - 2>&1 \
 // RUN:     -target i386-unknown-linux \
 // RUN:   | FileCheck -match-full-lines %s -check-prefix=CHECK_ADL_M32
+// RUN: %clang -march=gracemont -m32 -E -dM %s -o - 2>&1 \
+// RUN:     -target i386-unknown-linux \
+// RUN:   | FileCheck -match-full-lines %s -check-prefix=CHECK_ADL_M32
 // CHECK_ADL_M32: #define __ADX__ 1
 // CHECK_ADL_M32: #define __AES__ 1
 // CHECK_ADL_M32: #define __AVX2__ 1
@@ -2071,6 +2074,9 @@
 // RUN: %clang -march=meteorlake -m64 -E -dM %s -o - 2>&1 \
 // RUN:     -target i386-unknown-linux \
 // RUN:   | FileCheck -match-full-lines %s -check-prefix=CHECK_ADL_M64
+// RUN: %clang -march=gracemont -m64 -E -dM %s -o - 2>&1 \
+// RUN:     -target i386-unknown-linux \
+// RUN:   | FileCheck -match-full-lines %s -check-prefix=CHECK_ADL_M64
 // CHECK_ADL_M64: #define __ADX__ 1
 // CHECK_ADL_M64: #define __AES__ 1
 // CHECK_ADL_M64: #define __AVX2__ 1
diff --git a/compiler-rt/lib/builtins/cpu_model.c b/compiler-rt/lib/builtins/cpu_model.c
@@ -456,6 +456,8 @@ getIntelProcessorTypeAndSubtype(unsigned Family, unsigned Model,
     // Meteorlake:
     case 0xaa:
     case 0xac:
+    // Gracemont:
+    case 0xbe:
       CPU = "alderlake";
       *Type = INTEL_COREI7;
       *Subtype = INTEL_COREI7_ALDERLAKE;
diff --git a/llvm/include/llvm/TargetParser/X86TargetParser.def b/llvm/include/llvm/TargetParser/X86TargetParser.def
@@ -108,6 +108,7 @@ X86_CPU_SUBTYPE_ALIAS(INTEL_COREI7_ALDERLAKE, "raptorlake")
 X86_CPU_SUBTYPE_ALIAS(INTEL_COREI7_ALDERLAKE, "meteorlake")
 X86_CPU_SUBTYPE_ALIAS(INTEL_COREI7_SAPPHIRERAPIDS, "emeraldrapids")
 X86_CPU_SUBTYPE_ALIAS(INTEL_COREI7_ARROWLAKE_S,"lunarlake")
+X86_CPU_SUBTYPE_ALIAS(INTEL_COREI7_ALDERLAKE, "gracemont")
 
 #undef X86_CPU_SUBTYPE_ALIAS
 #undef X86_CPU_SUBTYPE
diff --git a/llvm/include/llvm/TargetParser/X86TargetParser.h b/llvm/include/llvm/TargetParser/X86TargetParser.h
@@ -87,6 +87,7 @@ enum CPUKind {
   CK_Goldmont,
   CK_GoldmontPlus,
   CK_Tremont,
+  CK_Gracemont,
   CK_Nehalem,
   CK_Westmere,
   CK_SandyBridge,
diff --git a/llvm/lib/Target/X86/X86.td b/llvm/lib/Target/X86/X86.td
@@ -1207,6 +1207,18 @@ def ProcessorFeatures {
   list<SubtargetFeature> ADLFeatures =
     !listconcat(TRMFeatures, ADLAdditionalFeatures);
 
+  // Gracemont
+  list<SubtargetFeature> GRTTuning = [TuningMacroFusion,
+                                      TuningSlow3OpsLEA,
+                                      TuningSlowDivide32,
+                                      TuningSlowDivide64,
+                                      TuningFastScalarFSQRT,
+                                      TuningFastVectorFSQRT,
+                                      TuningFast15ByteNOP,
+                                      TuningFastVariablePerLaneShuffle,
+                                      TuningPOPCNTFalseDeps,
+                                      TuningInsertVZEROUPPER];
+
   // Sierraforest
   list<SubtargetFeature> SRFAdditionalFeatures = [FeatureCMPCCXADD,
                                                   FeatureAVXIFMA,
@@ -1728,6 +1740,9 @@ def : ProcModel<"sapphirerapids", SapphireRapidsModel,
                 ProcessorFeatures.SPRFeatures, ProcessorFeatures.SPRTuning>;
 def : ProcModel<"alderlake", AlderlakePModel,
                 ProcessorFeatures.ADLFeatures, ProcessorFeatures.ADLTuning>;
+// FIXME: Use Gracemont Schedule Model when it is ready.
+def : ProcModel<"gracemont", AlderlakePModel,
+                ProcessorFeatures.ADLFeatures, ProcessorFeatures.GRTTuning>;
 def : ProcModel<"raptorlake", AlderlakePModel,
                 ProcessorFeatures.ADLFeatures, ProcessorFeatures.ADLTuning>;
 def : ProcModel<"meteorlake", AlderlakePModel,
diff --git a/llvm/lib/TargetParser/Host.cpp b/llvm/lib/TargetParser/Host.cpp
@@ -822,6 +822,8 @@ getIntelProcessorTypeAndSubtype(unsigned Family, unsigned Model,
     // Alderlake:
     case 0x97:
     case 0x9a:
+    // Gracemont
+    case 0xbe:
     // Raptorlake:
     case 0xb7:
     case 0xba:
diff --git a/llvm/lib/TargetParser/X86TargetParser.cpp b/llvm/lib/TargetParser/X86TargetParser.cpp
@@ -430,6 +430,8 @@ constexpr ProcInfo Processors[] = {
   { {"arrowlake_s"}, CK_ArrowlakeS, FEATURE_AVX2, FeaturesArrowlakeS, 'p', true },
   // Lunarlake microarchitecture based processors.
   { {"lunarlake"}, CK_Lunarlake, FEATURE_AVX2, FeaturesArrowlakeS, 'p', false },
+  // Gracemont microarchitecture based processors.
+  { {"gracemont"}, CK_Gracemont, FEATURE_AVX2, FeaturesAlderlake, 'p', false },
   // Sierraforest microarchitecture based processors.
   { {"sierraforest"}, CK_Sierraforest, FEATURE_AVX2, FeaturesSierraforest, 'p', false },
   // Grandridge microarchitecture based processors.
diff --git a/llvm/test/CodeGen/X86/cpus-intel.ll b/llvm/test/CodeGen/X86/cpus-intel.ll
@@ -36,6 +36,7 @@
 ; RUN: llc < %s -o /dev/null -mtriple=i686-unknown-unknown -mcpu=arrowlake-s 2>&1 | FileCheck %s --check-prefix=CHECK-NO-ERROR --allow-empty
 ; RUN: llc < %s -o /dev/null -mtriple=i686-unknown-unknown -mcpu=arrowlake_s 2>&1 | FileCheck %s --check-prefix=CHECK-NO-ERROR --allow-empty
 ; RUN: llc < %s -o /dev/null -mtriple=i686-unknown-unknown -mcpu=lunarlake 2>&1 | FileCheck %s --check-prefix=CHECK-NO-ERROR --allow-empty
+; RUN: llc < %s -o /dev/null -mtriple=i686-unknown-unknown -mcpu=gracemont 2>&1 | FileCheck %s --check-prefix=CHECK-NO-ERROR --allow-empty
 
 ; RUN: llc < %s -o /dev/null -mtriple=x86_64-unknown-unknown -mcpu=nocona 2>&1 | FileCheck %s --check-prefix=CHECK-NO-ERROR --allow-empty
 ; RUN: llc < %s -o /dev/null -mtriple=x86_64-unknown-unknown -mcpu=core2 2>&1 | FileCheck %s --check-prefix=CHECK-NO-ERROR --allow-empty
@@ -98,6 +99,7 @@
 ; RUN: llc < %s -o /dev/null -mtriple=x86_64-unknown-unknown -mcpu=arrowlake-s 2>&1 | FileCheck %s --check-prefix=CHECK-NO-ERROR --allow-empty
 ; RUN: llc < %s -o /dev/null -mtriple=x86_64-unknown-unknown -mcpu=arrowlake_s 2>&1 | FileCheck %s --check-prefix=CHECK-NO-ERROR --allow-empty
 ; RUN: llc < %s -o /dev/null -mtriple=x86_64-unknown-unknown -mcpu=lunarlake 2>&1 | FileCheck %s --check-prefix=CHECK-NO-ERROR --allow-empty
+; RUN: llc < %s -o /dev/null -mtriple=x86_64-unknown-unknown -mcpu=gracemont 2>&1 | FileCheck %s --check-prefix=CHECK-NO-ERROR --allow-empty
 
 define void @foo() {
   ret void
