[X86] Support -march=graniterapids-d and update -march=graniterapids

Reviewed By: pengfei, RKSimon, skan

Differential Revision: https://reviews.llvm.org/D155798

diff --git a/clang/docs/ReleaseNotes.rst b/clang/docs/ReleaseNotes.rst
index 89707a658f46bd1af459d98714fa5b229291b80a..ae2a298d7f439a9d5f02d2c29d3186fc0dfee4c4 100644 (file)
--- a/clang/docs/ReleaseNotes.rst
+++ b/clang/docs/ReleaseNotes.rst
@@ -860,6 +860,7 @@ X86 Support
   * Support intrinsic of ``_mm(256)_dpwsud(s)_epi32``.
   * Support intrinsic of ``_mm(256)_dpwusd(s)_epi32``.
   * Support intrinsic of ``_mm(256)_dpwuud(s)_epi32``.
+- ``-march=graniterapids-d`` is now supported.
 
 Arm and AArch64 Support
 ^^^^^^^^^^^^^^^^^^^^^^^

diff --git a/clang/lib/Basic/Targets/X86.cpp b/clang/lib/Basic/Targets/X86.cpp
index 5e28d20be3d3f58ba0a3b2ef0a3fe1a37d5657f6..11cb2dda15ccd8ab369dd2a24d64f6e28827ca13 100644 (file)
--- a/clang/lib/Basic/Targets/X86.cpp
+++ b/clang/lib/Basic/Targets/X86.cpp
@@ -548,6 +548,7 @@ void X86TargetInfo::getTargetDefines(const LangOptions &Opts,
   case CK_Sierraforest:
   case CK_Grandridge:
   case CK_Graniterapids:
+  case CK_GraniterapidsD:
   case CK_Emeraldrapids:
     // FIXME: Historically, we defined this legacy name, it would be nice to
     // remove it at some point. We've never exposed fine-grained names for
@@ -1439,6 +1440,7 @@ std::optional<unsigned> X86TargetInfo::getCPUCacheLineSize() const {
     case CK_Sierraforest:
     case CK_Grandridge:
     case CK_Graniterapids:
+    case CK_GraniterapidsD:
     case CK_Emeraldrapids:
     case CK_KNL:
     case CK_KNM:

diff --git a/clang/test/CodeGen/attr-cpuspecific-cpus.c b/clang/test/CodeGen/attr-cpuspecific-cpus.c
index 07cd3a56bcf41246ec923c60245b43cacbf367c1..14ae4af5fa415473f441c42c76f5d3dc542ff8b3 100644 (file)
--- a/clang/test/CodeGen/attr-cpuspecific-cpus.c
+++ b/clang/test/CodeGen/attr-cpuspecific-cpus.c
@@ -79,3 +79,4 @@ ATTR(cpu_specific(sierraforest)) void CPU30(void){}
 ATTR(cpu_specific(grandridge)) void CPU31(void){}
 ATTR(cpu_specific(graniterapids)) void CPU32(void){}
 ATTR(cpu_specific(emeraldrapids)) void CPU33(void){}
+ATTR(cpu_specific(graniterapids_d)) void CPU34(void){}

diff --git a/clang/test/CodeGen/attr-target-mv.c b/clang/test/CodeGen/attr-target-mv.c
index d10f75d58dbdcaf654a135b149f44f3257bfd707..9d348fd771c17f6e0cac5981662c5fafbf231ff3 100644 (file)
--- a/clang/test/CodeGen/attr-target-mv.c
+++ b/clang/test/CodeGen/attr-target-mv.c
@@ -21,6 +21,7 @@ int __attribute__((target("arch=sierraforest"))) foo(void) {return 16;}
 int __attribute__((target("arch=grandridge"))) foo(void) {return 17;}
 int __attribute__((target("arch=graniterapids"))) foo(void) {return 18;}
 int __attribute__((target("arch=emeraldrapids"))) foo(void) {return 19;}
+int __attribute__((target("arch=graniterapids-d"))) foo(void) {return 20;}
 int __attribute__((target("default"))) foo(void) { return 2; }
 
 int bar(void) {
@@ -167,6 +168,8 @@ void calls_pr50025c(void) { pr50025c(); }
 // LINUX: ret i32 18
 // LINUX: define{{.*}} i32 @foo.arch_emeraldrapids()
 // LINUX: ret i32 19
+// LINUX: define{{.*}} i32 @foo.arch_graniterapids-d()
+// LINUX: ret i32 20
 // LINUX: define{{.*}} i32 @foo()
 // LINUX: ret i32 2
 // LINUX: define{{.*}} i32 @bar()
@@ -210,6 +213,8 @@ void calls_pr50025c(void) { pr50025c(); }
 // WINDOWS: ret i32 18
 // WINDOWS: define dso_local i32 @foo.arch_emeraldrapids()
 // WINDOWS: ret i32 19
+// WINDOWS: define dso_local i32 @foo.arch_graniterapids-d()
+// WINDOWS: ret i32 20
 // 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
index ac50da36844a4bc62f2b7790bd775ce78b5594b8..cc6b90b9da9eccb17a9e34ac2ec653a21886a7c6 100644 (file)
--- a/clang/test/CodeGen/target-builtin-noerror.c
+++ b/clang/test/CodeGen/target-builtin-noerror.c
@@ -110,6 +110,7 @@ void verifycpustrings(void) {
   (void)__builtin_cpu_is("grandridge");
   (void)__builtin_cpu_is("graniterapids");
   (void)__builtin_cpu_is("emeraldrapids");
+  (void)__builtin_cpu_is("graniterapids-d");
   (void)__builtin_cpu_is("haswell");
   (void)__builtin_cpu_is("icelake-client");
   (void)__builtin_cpu_is("icelake-server");

diff --git a/clang/test/Driver/x86-march.c b/clang/test/Driver/x86-march.c
index 8e86be6c76cd793261fb7bbd021748aaedc1e8e9..fc3c0c8a983698a0c110809f2782461a52d71b0b 100644 (file)
--- a/clang/test/Driver/x86-march.c
+++ b/clang/test/Driver/x86-march.c
@@ -131,6 +131,10 @@
 // RUN:   | FileCheck %s -check-prefix=graniterapids
 // graniterapids: "-target-cpu" "graniterapids"
 //
+// RUN: %clang --target=x86_64-unknown-unknown -c -### %s -march=graniterapids-d 2>&1 \
+// RUN:   | FileCheck %s -check-prefix=graniterapids-d
+// graniterapids-d: "-target-cpu" "graniterapids-d"
+//
 // RUN: %clang -target x86_64-unknown-unknown -c -### %s -march=emeraldrapids 2>&1 \
 // RUN:   | FileCheck %s -check-prefix=emeraldrapids
 // emeraldrapids: "-target-cpu" "emeraldrapids"

diff --git a/clang/test/Misc/target-invalid-cpu-note.c b/clang/test/Misc/target-invalid-cpu-note.c
index 9a6b0479bf7b77aa9ca8db2d8b3b56678106632a..cd1b0bc157cc9fea1e13eaf4ace15206c3e0f5be 100644 (file)
--- 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, sierraforest, grandridge, graniterapids, 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{{$}}
+// X86-NEXT: note: valid target CPU values are: i386, i486, winchip-c6, winchip2, c3, i586, pentium, pentium-mmx, pentiumpro, i686, pentium2, pentium3, pentium3m, pentium-m, c3-2, yonah, pentium4, pentium4m, prescott, nocona, core2, penryn, bonnell, atom, silvermont, slm, goldmont, goldmont-plus, tremont, nehalem, corei7, westmere, sandybridge, corei7-avx, ivybridge, core-avx-i, haswell, core-avx2, broadwell, skylake, skylake-avx512, skx, cascadelake, cooperlake, cannonlake, icelake-client, rocketlake, icelake-server, tigerlake, sapphirerapids, alderlake, raptorlake, meteorlake, sierraforest, grandridge, graniterapids, graniterapids-d, emeraldrapids, knl, knm, lakemont, k6, k6-2, k6-3, athlon, athlon-tbird, athlon-xp, athlon-mp, athlon-4, k8, athlon64, athlon-fx, opteron, k8-sse3, athlon64-sse3, opteron-sse3, amdfam10, barcelona, btver1, btver2, bdver1, bdver2, bdver3, bdver4, znver1, znver2, znver3, znver4, x86-64, x86-64-v2, x86-64-v3, x86-64-v4, geode{{$}}
 
 // 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, sierraforest, grandridge, graniterapids, 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{{$}}
+// X86_64-NEXT: note: valid target CPU values are: nocona, core2, penryn, bonnell, atom, silvermont, slm, goldmont, goldmont-plus, tremont, nehalem, corei7, westmere, sandybridge, corei7-avx, ivybridge, core-avx-i, haswell, core-avx2, broadwell, skylake, skylake-avx512, skx, cascadelake, cooperlake, cannonlake, icelake-client, rocketlake, icelake-server, tigerlake, sapphirerapids, alderlake, raptorlake, meteorlake, sierraforest, grandridge, graniterapids, graniterapids-d, emeraldrapids, knl, knm, k8, athlon64, athlon-fx, opteron, k8-sse3, athlon64-sse3, opteron-sse3, amdfam10, barcelona, btver1, btver2, bdver1, bdver2, bdver3, bdver4, znver1, znver2, znver3, znver4, x86-64, x86-64-v2, x86-64-v3, x86-64-v4{{$}}
 
 // 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, sierraforest, grandridge, graniterapids, 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{{$}}
+// TUNE_X86-NEXT: note: valid target CPU values are: i386, i486, winchip-c6, winchip2, c3, i586, pentium, pentium-mmx, pentiumpro, i686, pentium2, pentium3, pentium3m, pentium-m, c3-2, yonah, pentium4, pentium4m, prescott, nocona, core2, penryn, bonnell, atom, silvermont, slm, goldmont, goldmont-plus, tremont, nehalem, corei7, westmere, sandybridge, corei7-avx, ivybridge, core-avx-i, haswell, core-avx2, broadwell, skylake, skylake-avx512, skx, cascadelake, cooperlake, cannonlake, icelake-client, rocketlake, icelake-server, tigerlake, sapphirerapids, alderlake, raptorlake, meteorlake, sierraforest, grandridge, graniterapids, graniterapids-d, emeraldrapids, knl, knm, lakemont, k6, k6-2, k6-3, athlon, athlon-tbird, athlon-xp, athlon-mp, athlon-4, k8, athlon64, athlon-fx, opteron, k8-sse3, athlon64-sse3, opteron-sse3, amdfam10, barcelona, btver1, btver2, bdver1, bdver2, bdver3, bdver4, znver1, znver2, znver3, znver4, x86-64, geode{{$}}
 
 // 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, sierraforest, grandridge, graniterapids, 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{{$}}
+// TUNE_X86_64-NEXT: note: valid target CPU values are: i386, i486, winchip-c6, winchip2, c3, i586, pentium, pentium-mmx, pentiumpro, i686, pentium2, pentium3, pentium3m, pentium-m, c3-2, yonah, pentium4, pentium4m, prescott, nocona, core2, penryn, bonnell, atom, silvermont, slm, goldmont, goldmont-plus, tremont, nehalem, corei7, westmere, sandybridge, corei7-avx, ivybridge, core-avx-i, haswell, core-avx2, broadwell, skylake, skylake-avx512, skx, cascadelake, cooperlake, cannonlake, icelake-client, rocketlake, icelake-server, tigerlake, sapphirerapids, alderlake, raptorlake, meteorlake, sierraforest, grandridge, graniterapids, graniterapids-d, emeraldrapids, knl, knm, lakemont, k6, k6-2, k6-3, athlon, athlon-tbird, athlon-xp, athlon-mp, athlon-4, k8, athlon64, athlon-fx, opteron, k8-sse3, athlon64-sse3, opteron-sse3, amdfam10, barcelona, btver1, btver2, bdver1, bdver2, bdver3, bdver4, znver1, znver2, znver3, znver4, x86-64, geode{{$}}
 
 // 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
index 17ea99f91e9ffb7a5af72adddb62e3108dfdfc6b..13851ec11f6430e34750da9ddf8e4f5ee058287f 100644 (file)
--- a/clang/test/Preprocessor/predefined-arch-macros.c
+++ b/clang/test/Preprocessor/predefined-arch-macros.c
@@ -1838,9 +1838,13 @@
 // RUN: %clang -march=graniterapids -m32 -E -dM %s -o - 2>&1 \
 // RUN:     --target=i386 \
 // RUN:   | FileCheck -match-full-lines %s -check-prefix=CHECK_GNR_M32
+// RUN: %clang -march=graniterapids-d -m32 -E -dM %s -o - 2>&1 \
+// RUN:     --target=i386 \
+// RUN:   | FileCheck -match-full-lines %s -check-prefixes=CHECK_GNR_M32,CHECK_GNRD_M32
 // CHECK_GNR_M32: #define __AES__ 1
 // CHECK_GNR_M32: #define __AMX_BF16__ 1
-// CHECK_GNR_M32: #define __AMX_COMPLEX__ 1
+// CHECK_GNR_M32-NOT: #define __AMX_COMPLEX__ 1
+// CHECK_GNRD_M32: #define __AMX_COMPLEX__ 1
 // CHECK_GNR_M32: #define __AMX_FP16__ 1
 // CHECK_GNR_M32: #define __AMX_INT8__ 1
 // CHECK_GNR_M32: #define __AMX_TILE__ 1
@@ -1913,9 +1917,13 @@
 // RUN: %clang -march=graniterapids -m64 -E -dM %s -o - 2>&1 \
 // RUN:     --target=x86_64 \
 // RUN:   | FileCheck -match-full-lines %s -check-prefix=CHECK_GNR_M64
+// RUN: %clang -march=graniterapids-d -m64 -E -dM %s -o - 2>&1 \
+// RUN:     --target=x86_64 \
+// RUN:   | FileCheck -match-full-lines %s -check-prefixes=CHECK_GNR_M64,CHECK_GNRD_M64
 // CHECK_GNR_M64: #define __AES__ 1
 // CHECK_GNR_M64: #define __AMX_BF16__ 1
-// CHECK_GNR_M64: #define __AMX_COMPLEX__ 1
+// CHECK_GNR_M64-NOT: #define __AMX_COMPLEX__ 1
+// CHECK_GNRD_M64: #define __AMX_COMPLEX__ 1
 // CHECK_GNR_M64: #define __AMX_FP16__ 1
 // CHECK_GNR_M64: #define __AMX_INT8__ 1
 // CHECK_GNR_M64: #define __AMX_TILE__ 1

diff --git a/compiler-rt/lib/builtins/cpu_model.c b/compiler-rt/lib/builtins/cpu_model.c
index 20a03e83ca2f44ac522ed5a21bf8135e27dc890a..36eb696c39eef9e9e94c289a11c3820e2c5711b8 100644 (file)
--- a/compiler-rt/lib/builtins/cpu_model.c
+++ b/compiler-rt/lib/builtins/cpu_model.c
@@ -113,6 +113,7 @@ enum ProcessorSubtypes {
   ZHAOXIN_FAM7H_LUJIAZUI,
   AMDFAM19H_ZNVER4,
   INTEL_COREI7_GRANITERAPIDS,
+  INTEL_COREI7_GRANITERAPIDS_D,
   CPU_SUBTYPE_MAX
 };
 
@@ -474,13 +475,19 @@ getIntelProcessorTypeAndSubtype(unsigned Family, unsigned Model,
       break;
 
     // Granite Rapids:
-    case 0xae:
     case 0xad:
       CPU = "graniterapids";
       *Type = INTEL_COREI7;
       *Subtype = INTEL_COREI7_GRANITERAPIDS;
       break;
 
+    // Granite Rapids D:
+    case 0xae:
+      CPU = "graniterapids-d";
+      *Type = INTEL_COREI7;
+      *Subtype = INTEL_COREI7_GRANITERAPIDS_D;
+      break;
+
     case 0x1c: // Most 45 nm Intel Atom processors
     case 0x26: // 45 nm Atom Lincroft
     case 0x27: // 32 nm Atom Medfield

diff --git a/llvm/docs/ReleaseNotes.rst b/llvm/docs/ReleaseNotes.rst
index 04ecaa42e5d319d0b3bd6bd3e1328f76db317f3d..a1b8ff734d41a023e92f71c9744a3a623f8f1ec7 100644 (file)
--- a/llvm/docs/ReleaseNotes.rst
+++ b/llvm/docs/ReleaseNotes.rst
@@ -298,6 +298,7 @@ Changes to the X86 Backend
 * Support ISA of ``SM3``.
 * Support ISA of ``SM4``.
 * Support ISA of ``AVX-VNNI-INT16``.
+* ``-mcpu=graniterapids-d`` is now supported.
 
 Changes to the OCaml bindings
 -----------------------------

diff --git a/llvm/include/llvm/TargetParser/X86TargetParser.def b/llvm/include/llvm/TargetParser/X86TargetParser.def
index 7964353045258f1ffbcd29e8222d55f15419d39a..7f874cfac4afe007df83cbb5aa4f446c4245acf0 100644 (file)
--- a/llvm/include/llvm/TargetParser/X86TargetParser.def
+++ b/llvm/include/llvm/TargetParser/X86TargetParser.def
@@ -99,6 +99,7 @@ X86_CPU_SUBTYPE(INTEL_COREI7_ROCKETLAKE,     "rocketlake")
 X86_CPU_SUBTYPE(ZHAOXIN_FAM7H_LUJIAZUI,      "zhaoxin_fam7h_lujiazui")
 X86_CPU_SUBTYPE(AMDFAM19H_ZNVER4,            "znver4")
 X86_CPU_SUBTYPE(INTEL_COREI7_GRANITERAPIDS,  "graniterapids")
+X86_CPU_SUBTYPE(INTEL_COREI7_GRANITERAPIDS_D,"graniterapids-d")
 
 // Alternate names supported by __builtin_cpu_is and target multiversioning.
 X86_CPU_SUBTYPE_ALIAS(INTEL_COREI7_ALDERLAKE, "raptorlake")

diff --git a/llvm/include/llvm/TargetParser/X86TargetParser.h b/llvm/include/llvm/TargetParser/X86TargetParser.h
index 7a302c3d6454176c614690a0e915f65a9d58c494..bae31891e24b8bb8a95ebc7df0ca118ffc8a98a1 100644 (file)
--- a/llvm/include/llvm/TargetParser/X86TargetParser.h
+++ b/llvm/include/llvm/TargetParser/X86TargetParser.h
@@ -109,6 +109,7 @@ enum CPUKind {
   CK_Sierraforest,
   CK_Grandridge,
   CK_Graniterapids,
+  CK_GraniterapidsD,
   CK_Emeraldrapids,
   CK_KNL,
   CK_KNM,

diff --git a/llvm/lib/Target/X86/X86.td b/llvm/lib/Target/X86/X86.td
index cc88af17ab55c63bc368478234d7dc99d833c1d4..0f677b8a4afc14e29e01094ec8363349186e2cc8 100644 (file)
--- a/llvm/lib/Target/X86/X86.td
+++ b/llvm/lib/Target/X86/X86.td
@@ -1078,11 +1078,15 @@ def ProcessorFeatures {
 
   // Graniterapids
   list<SubtargetFeature> GNRAdditionalFeatures = [FeatureAMXFP16,
-                                                  FeaturePREFETCHI,
-                                                  FeatureAMXCOMPLEX];
+                                                  FeaturePREFETCHI];
   list<SubtargetFeature> GNRFeatures =
     !listconcat(SPRFeatures, GNRAdditionalFeatures);
 
+  // Graniterapids D
+  list<SubtargetFeature> GNRDAdditionalFeatures = [FeatureAMXCOMPLEX];
+  list<SubtargetFeature> GNRDFeatures =
+    !listconcat(GNRFeatures, GNRDAdditionalFeatures);
+
   // Atom
   list<SubtargetFeature> AtomFeatures = [FeatureX87,
                                          FeatureCX8,
@@ -1717,6 +1721,10 @@ def : ProcModel<"graniterapids", SapphireRapidsModel,
                 ProcessorFeatures.GNRFeatures, ProcessorFeatures.SPRTuning>;
 def : ProcModel<"emeraldrapids", SapphireRapidsModel,
                 ProcessorFeatures.SPRFeatures, ProcessorFeatures.SPRTuning>;
+foreach P = ["graniterapids-d", "graniterapids_d"] in {
+def : ProcModel<P, SapphireRapidsModel,
+                ProcessorFeatures.GNRDFeatures, ProcessorFeatures.SPRTuning>;
+}
 
 // AMD CPUs.
 

diff --git a/llvm/lib/TargetParser/Host.cpp b/llvm/lib/TargetParser/Host.cpp
index a1cedbb0308f8ef43b15b0cc7d6e8c4b38b8b0b5..518c859b11cc441eb0258adc85bb6dfbd2c491e8 100644 (file)
--- a/llvm/lib/TargetParser/Host.cpp
+++ b/llvm/lib/TargetParser/Host.cpp
@@ -833,13 +833,19 @@ getIntelProcessorTypeAndSubtype(unsigned Family, unsigned Model,
       break;
 
     // Graniterapids:
-    case 0xae:
     case 0xad:
       CPU = "graniterapids";
       *Type = X86::INTEL_COREI7;
       *Subtype = X86::INTEL_COREI7_GRANITERAPIDS;
       break;
 
+    // Granite Rapids D:
+    case 0xae:
+      CPU = "graniterapids-d";
+      *Type = X86::INTEL_COREI7;
+      *Subtype = X86::INTEL_COREI7_GRANITERAPIDS_D;
+      break;
+
     // Icelake Xeon:
     case 0x6a:
     case 0x6c:

diff --git a/llvm/lib/TargetParser/X86TargetParser.cpp b/llvm/lib/TargetParser/X86TargetParser.cpp
index 6d2c9033fa68da7fbf1f9fdd524cc27aa4e67ed2..8bd063116cf6d31293cd6750abb23bf748eea456 100644 (file)
--- a/llvm/lib/TargetParser/X86TargetParser.cpp
+++ b/llvm/lib/TargetParser/X86TargetParser.cpp
@@ -210,8 +210,7 @@ constexpr FeatureBitset FeaturesSapphireRapids =
     FeatureSERIALIZE | FeatureSHSTK | FeatureTSXLDTRK | FeatureUINTR |
     FeatureWAITPKG;
 constexpr FeatureBitset FeaturesGraniteRapids =
-    FeaturesSapphireRapids | FeatureAMX_FP16 | FeaturePREFETCHI |
-    FeatureAMX_COMPLEX;
+    FeaturesSapphireRapids | FeatureAMX_FP16 | FeaturePREFETCHI;
 
 // Intel Atom processors.
 // Bonnell has feature parity with Core2 and adds MOVBE.
@@ -429,6 +428,9 @@ constexpr ProcInfo Processors[] = {
   { {"grandridge"}, CK_Grandridge, FEATURE_AVX2, FeaturesGrandridge, 'p', false },
   // Granite Rapids microarchitecture based processors.
   { {"graniterapids"}, CK_Graniterapids, FEATURE_AVX512BF16, FeaturesGraniteRapids, 'n', false },
+  // Granite Rapids D microarchitecture based processors.
+  { {"graniterapids-d"}, CK_GraniterapidsD, FEATURE_AVX512BF16, FeaturesGraniteRapids | FeatureAMX_COMPLEX, '\0', false },
+  { {"graniterapids_d"}, CK_GraniterapidsD, FEATURE_AVX512BF16, FeaturesGraniteRapids | FeatureAMX_COMPLEX, 'n', true },
   // Emerald Rapids microarchitecture based processors.
   { {"emeraldrapids"}, CK_Emeraldrapids, FEATURE_AVX512BF16, FeaturesSapphireRapids, 'n', false },
   // Knights Landing processor.

diff --git a/llvm/test/CodeGen/X86/cpus-intel.ll b/llvm/test/CodeGen/X86/cpus-intel.ll
index 6856fccf9f42290149a2123c87318d8c2ce12521..6c7703c6e1394fb6f2c9d968b2b97a93e811200a 100644 (file)
--- a/llvm/test/CodeGen/X86/cpus-intel.ll
+++ b/llvm/test/CodeGen/X86/cpus-intel.ll
@@ -30,6 +30,7 @@
 ; RUN: llc < %s -o /dev/null -mtriple=i686-unknown-unknown -mcpu=sierraforest 2>&1 | FileCheck %s --check-prefix=CHECK-NO-ERROR --allow-empty
 ; RUN: llc < %s -o /dev/null -mtriple=i686-unknown-unknown -mcpu=grandridge 2>&1 | FileCheck %s --check-prefix=CHECK-NO-ERROR --allow-empty
 ; RUN: llc < %s -o /dev/null -mtriple=i686-unknown-unknown -mcpu=graniterapids 2>&1 | FileCheck %s --check-prefix=CHECK-NO-ERROR --allow-empty
+; RUN: llc < %s -o /dev/null -mtriple=i686-unknown-unknown -mcpu=graniterapids-d 2>&1 | FileCheck %s --check-prefix=CHECK-NO-ERROR --allow-empty
 ; RUN: llc < %s -o /dev/null -mtriple=i686-unknown-unknown -mcpu=emeraldrapids 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
@@ -88,6 +89,7 @@
 ; RUN: llc < %s -o /dev/null -mtriple=x86_64-unknown-unknown -mcpu=sierraforest 2>&1 | FileCheck %s --check-prefix=CHECK-NO-ERROR --allow-empty
 ; RUN: llc < %s -o /dev/null -mtriple=x86_64-unknown-unknown -mcpu=grandridge 2>&1 | FileCheck %s --check-prefix=CHECK-NO-ERROR --allow-empty
 ; RUN: llc < %s -o /dev/null -mtriple=x86_64-unknown-unknown -mcpu=graniterapids 2>&1 | FileCheck %s --check-prefix=CHECK-NO-ERROR --allow-empty
+; RUN: llc < %s -o /dev/null -mtriple=x86_64-unknown-unknown -mcpu=graniterapids-d 2>&1 | FileCheck %s --check-prefix=CHECK-NO-ERROR --allow-empty
 
 define void @foo() {
   ret void