[clang] [llvm] AMDGPU: Add gfx12-generic target (PR #93875)
Konstantin Zhuravlyov via llvm-commits
llvm-commits at lists.llvm.org
Thu May 30 13:56:19 PDT 2024
https://github.com/kzhuravl updated https://github.com/llvm/llvm-project/pull/93875
>From be005a9baf0e7f0ae7b28b96ef7e562158800af4 Mon Sep 17 00:00:00 2001
From: Konstantin Zhuravlyov <kzhuravl_dev at outlook.com>
Date: Thu, 30 May 2024 16:27:42 -0400
Subject: [PATCH 1/3] AMDGPU: Add gfx12-generic target
---
clang/test/Driver/amdgpu-macros.cl | 1 +
clang/test/Driver/amdgpu-mcpu.cl | 2 ++
clang/test/Misc/target-invalid-cpu-note.c | 2 +-
llvm/docs/AMDGPUUsage.rst | 8 ++++++++
llvm/include/llvm/BinaryFormat/ELF.h | 3 ++-
llvm/include/llvm/TargetParser/TargetParser.h | 3 ++-
llvm/lib/Object/ELFObjectFile.cpp | 2 ++
llvm/lib/ObjectYAML/ELFYAML.cpp | 1 +
llvm/lib/Target/AMDGPU/AMDGPU.td | 4 ++++
llvm/lib/Target/AMDGPU/GCNProcessors.td | 7 +++++--
.../Target/AMDGPU/MCTargetDesc/AMDGPUTargetStreamer.cpp | 5 +++++
llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.h | 1 +
llvm/lib/TargetParser/TargetParser.cpp | 6 ++++++
llvm/test/CodeGen/AMDGPU/directive-amdgcn-target.ll | 2 ++
llvm/test/CodeGen/AMDGPU/elf-header-flags-mach.ll | 2 ++
llvm/test/CodeGen/AMDGPU/generic-targets-require-v6.ll | 3 +++
llvm/test/CodeGen/AMDGPU/hsa-generic-target-features.ll | 2 ++
llvm/test/Object/AMDGPU/elf-header-flags-mach.yaml | 7 +++++++
llvm/test/tools/llvm-objdump/ELF/AMDGPU/subtarget.ll | 5 +++++
llvm/test/tools/llvm-readobj/ELF/AMDGPU/elf-headers.test | 3 +++
llvm/tools/llvm-readobj/ELFDumper.cpp | 3 ++-
21 files changed, 66 insertions(+), 6 deletions(-)
diff --git a/clang/test/Driver/amdgpu-macros.cl b/clang/test/Driver/amdgpu-macros.cl
index 004619321b271..3dbe10dfcf1dc 100644
--- a/clang/test/Driver/amdgpu-macros.cl
+++ b/clang/test/Driver/amdgpu-macros.cl
@@ -135,6 +135,7 @@
// RUN: %clang -E -dM -target amdgcn -mcpu=gfx10-1-generic %s 2>&1 | FileCheck --check-prefixes=ARCH-GCN,FAST_FMAF %s -DWAVEFRONT_SIZE=32 -DCPU=gfx10_1_generic -DFAMILY=GFX10
// RUN: %clang -E -dM -target amdgcn -mcpu=gfx10-3-generic %s 2>&1 | FileCheck --check-prefixes=ARCH-GCN,FAST_FMAF %s -DWAVEFRONT_SIZE=32 -DCPU=gfx10_3_generic -DFAMILY=GFX10
// RUN: %clang -E -dM -target amdgcn -mcpu=gfx11-generic %s 2>&1 | FileCheck --check-prefixes=ARCH-GCN,FAST_FMAF %s -DWAVEFRONT_SIZE=32 -DCPU=gfx11_generic -DFAMILY=GFX11
+// RUN: %clang -E -dM -target amdgcn -mcpu=gfx12-generic %s 2>&1 | FileCheck --check-prefixes=ARCH-GCN,FAST_FMAF %s -DWAVEFRONT_SIZE=32 -DCPU=gfx12_generic -DFAMILY=GFX12
// ARCH-GCN-DAG: #define FP_FAST_FMA 1
diff --git a/clang/test/Driver/amdgpu-mcpu.cl b/clang/test/Driver/amdgpu-mcpu.cl
index 915fa6473ac07..5b6a22016f043 100644
--- a/clang/test/Driver/amdgpu-mcpu.cl
+++ b/clang/test/Driver/amdgpu-mcpu.cl
@@ -119,6 +119,7 @@
// RUN: %clang -### -target amdgcn -mcpu=gfx10-1-generic %s 2>&1 | FileCheck --check-prefix=GFX10_1_GENERIC %s
// RUN: %clang -### -target amdgcn -mcpu=gfx10-3-generic %s 2>&1 | FileCheck --check-prefix=GFX10_3_GENERIC %s
// RUN: %clang -### -target amdgcn -mcpu=gfx11-generic %s 2>&1 | FileCheck --check-prefix=GFX11_GENERIC %s
+// RUN: %clang -### -target amdgcn -mcpu=gfx12-generic %s 2>&1 | FileCheck --check-prefix=GFX12_GENERIC %s
// GCNDEFAULT-NOT: -target-cpu
// GFX600: "-target-cpu" "gfx600"
@@ -170,3 +171,4 @@
// GFX10_1_GENERIC: "-target-cpu" "gfx10-1-generic"
// GFX10_3_GENERIC: "-target-cpu" "gfx10-3-generic"
// GFX11_GENERIC: "-target-cpu" "gfx11-generic"
+// GFX12_GENERIC: "-target-cpu" "gfx12-generic"
diff --git a/clang/test/Misc/target-invalid-cpu-note.c b/clang/test/Misc/target-invalid-cpu-note.c
index 768b243b04e3a..6558fd753d1d1 100644
--- a/clang/test/Misc/target-invalid-cpu-note.c
+++ b/clang/test/Misc/target-invalid-cpu-note.c
@@ -37,7 +37,7 @@
// RUN: not %clang_cc1 -triple amdgcn--- -target-cpu not-a-cpu -fsyntax-only %s 2>&1 | FileCheck %s --check-prefix AMDGCN
// AMDGCN: error: unknown target CPU 'not-a-cpu'
-// AMDGCN-NEXT: note: valid target CPU values are: gfx600, tahiti, gfx601, pitcairn, verde, gfx602, hainan, oland, gfx700, kaveri, gfx701, hawaii, gfx702, gfx703, kabini, mullins, gfx704, bonaire, gfx705, gfx801, carrizo, gfx802, iceland, tonga, gfx803, fiji, polaris10, polaris11, gfx805, tongapro, gfx810, stoney, gfx900, gfx902, gfx904, gfx906, gfx908, gfx909, gfx90a, gfx90c, gfx940, gfx941, gfx942, gfx1010, gfx1011, gfx1012, gfx1013, gfx1030, gfx1031, gfx1032, gfx1033, gfx1034, gfx1035, gfx1036, gfx1100, gfx1101, gfx1102, gfx1103, gfx1150, gfx1151, gfx1200, gfx1201, gfx9-generic, gfx10-1-generic, gfx10-3-generic, gfx11-generic{{$}}
+// AMDGCN-NEXT: note: valid target CPU values are: gfx600, tahiti, gfx601, pitcairn, verde, gfx602, hainan, oland, gfx700, kaveri, gfx701, hawaii, gfx702, gfx703, kabini, mullins, gfx704, bonaire, gfx705, gfx801, carrizo, gfx802, iceland, tonga, gfx803, fiji, polaris10, polaris11, gfx805, tongapro, gfx810, stoney, gfx900, gfx902, gfx904, gfx906, gfx908, gfx909, gfx90a, gfx90c, gfx940, gfx941, gfx942, gfx1010, gfx1011, gfx1012, gfx1013, gfx1030, gfx1031, gfx1032, gfx1033, gfx1034, gfx1035, gfx1036, gfx1100, gfx1101, gfx1102, gfx1103, gfx1150, gfx1151, gfx1200, gfx1201, gfx9-generic, gfx10-1-generic, gfx10-3-generic, gfx11-generic, gfx12-generic{{$}}
// RUN: not %clang_cc1 -triple wasm64--- -target-cpu not-a-cpu -fsyntax-only %s 2>&1 | FileCheck %s --check-prefix WEBASM
// WEBASM: error: unknown target CPU 'not-a-cpu'
diff --git a/llvm/docs/AMDGPUUsage.rst b/llvm/docs/AMDGPUUsage.rst
index 95b54548f4fa8..77d68dd22f58b 100644
--- a/llvm/docs/AMDGPUUsage.rst
+++ b/llvm/docs/AMDGPUUsage.rst
@@ -610,6 +610,13 @@ Generic processor code objects are versioned. See :ref:`amdgpu-generic-processor
- ``gfx1150``
- ``gfx1151``
+
+
+ ``gfx12-generic`` ``amdgcn`` - ``gfx1200`` - wavefrontsize64 - Architected No restrictions.
+ - ``gfx1201`` - cumode flat scratch
+ - Packed
+ work-item
+ IDs
==================== ============== ================= ================== ================= =================================
.. _amdgpu-generic-processor-versioning:
@@ -1969,6 +1976,7 @@ The AMDGPU backend uses the following ELF header:
``EF_AMDGPU_MACH_AMDGCN_GFX10_1_GENERIC`` 0x052 ``gfx10-1-generic``
``EF_AMDGPU_MACH_AMDGCN_GFX10_3_GENERIC`` 0x053 ``gfx10-3-generic``
``EF_AMDGPU_MACH_AMDGCN_GFX11_GENERIC`` 0x054 ``gfx11-generic``
+ ``EF_AMDGPU_MACH_AMDGCN_GFX12_GENERIC`` 0x059 ``gfx12-generic``
*reserved* 0x055 Reserved.
*reserved* 0x056 Reserved.
*reserved* 0x057 Reserved.
diff --git a/llvm/include/llvm/BinaryFormat/ELF.h b/llvm/include/llvm/BinaryFormat/ELF.h
index 9a538252d9bef..c9456a3b36054 100644
--- a/llvm/include/llvm/BinaryFormat/ELF.h
+++ b/llvm/include/llvm/BinaryFormat/ELF.h
@@ -799,11 +799,12 @@ enum : unsigned {
EF_AMDGPU_MACH_AMDGCN_RESERVED_0X56 = 0x056,
EF_AMDGPU_MACH_AMDGCN_RESERVED_0X57 = 0x057,
EF_AMDGPU_MACH_AMDGCN_RESERVED_0X58 = 0x058,
+ EF_AMDGPU_MACH_AMDGCN_GFX12_GENERIC = 0x059,
// clang-format on
// First/last AMDGCN-based processors.
EF_AMDGPU_MACH_AMDGCN_FIRST = EF_AMDGPU_MACH_AMDGCN_GFX600,
- EF_AMDGPU_MACH_AMDGCN_LAST = EF_AMDGPU_MACH_AMDGCN_RESERVED_0X57,
+ EF_AMDGPU_MACH_AMDGCN_LAST = EF_AMDGPU_MACH_AMDGCN_GFX12_GENERIC,
// Indicates if the "xnack" target feature is enabled for all code contained
// in the object.
diff --git a/llvm/include/llvm/TargetParser/TargetParser.h b/llvm/include/llvm/TargetParser/TargetParser.h
index 7da11701fefb8..8fc6fa37cb9aa 100644
--- a/llvm/include/llvm/TargetParser/TargetParser.h
+++ b/llvm/include/llvm/TargetParser/TargetParser.h
@@ -116,9 +116,10 @@ enum GPUKind : uint32_t {
GK_GFX10_1_GENERIC = 193,
GK_GFX10_3_GENERIC = 194,
GK_GFX11_GENERIC = 195,
+ GK_GFX12_GENERIC = 196,
GK_AMDGCN_GENERIC_FIRST = GK_GFX9_GENERIC,
- GK_AMDGCN_GENERIC_LAST = GK_GFX11_GENERIC,
+ GK_AMDGCN_GENERIC_LAST = GK_GFX12_GENERIC,
};
/// Instruction set architecture version.
diff --git a/llvm/lib/Object/ELFObjectFile.cpp b/llvm/lib/Object/ELFObjectFile.cpp
index 24d7a7a280fd9..2b6bdbf24afa0 100644
--- a/llvm/lib/Object/ELFObjectFile.cpp
+++ b/llvm/lib/Object/ELFObjectFile.cpp
@@ -602,6 +602,8 @@ StringRef ELFObjectFileBase::getAMDGPUCPUName() const {
return "gfx10-3-generic";
case ELF::EF_AMDGPU_MACH_AMDGCN_GFX11_GENERIC:
return "gfx11-generic";
+ case ELF::EF_AMDGPU_MACH_AMDGCN_GFX12_GENERIC:
+ return "gfx12-generic";
default:
llvm_unreachable("Unknown EF_AMDGPU_MACH value");
}
diff --git a/llvm/lib/ObjectYAML/ELFYAML.cpp b/llvm/lib/ObjectYAML/ELFYAML.cpp
index 045211c44b907..8e2a9481c9227 100644
--- a/llvm/lib/ObjectYAML/ELFYAML.cpp
+++ b/llvm/lib/ObjectYAML/ELFYAML.cpp
@@ -617,6 +617,7 @@ void ScalarBitSetTraits<ELFYAML::ELF_EF>::bitset(IO &IO,
BCaseMask(EF_AMDGPU_MACH_AMDGCN_GFX10_1_GENERIC, EF_AMDGPU_MACH);
BCaseMask(EF_AMDGPU_MACH_AMDGCN_GFX10_3_GENERIC, EF_AMDGPU_MACH);
BCaseMask(EF_AMDGPU_MACH_AMDGCN_GFX11_GENERIC, EF_AMDGPU_MACH);
+ BCaseMask(EF_AMDGPU_MACH_AMDGCN_GFX12_GENERIC, EF_AMDGPU_MACH);
switch (Object->Header.ABIVersion) {
default:
// ELFOSABI_AMDGPU_PAL, ELFOSABI_AMDGPU_MESA3D support *_V3 flags.
diff --git a/llvm/lib/Target/AMDGPU/AMDGPU.td b/llvm/lib/Target/AMDGPU/AMDGPU.td
index 35b0cb439bfa6..8e302786c746e 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPU.td
+++ b/llvm/lib/Target/AMDGPU/AMDGPU.td
@@ -1568,6 +1568,10 @@ def FeatureISAVersion12 : FeatureSet<
FeatureScalarDwordx3Loads,
FeatureDPPSrc1SGPR]>;
+def FeatureISAVersion12_Generic: FeatureSet<
+ !listconcat(FeatureISAVersion12.Features,
+ [FeatureRequiresCOV6])>;
+
//===----------------------------------------------------------------------===//
def AMDGPUInstrInfo : InstrInfo {
diff --git a/llvm/lib/Target/AMDGPU/GCNProcessors.td b/llvm/lib/Target/AMDGPU/GCNProcessors.td
index f7e091904bad3..2ada981a77cd8 100644
--- a/llvm/lib/Target/AMDGPU/GCNProcessors.td
+++ b/llvm/lib/Target/AMDGPU/GCNProcessors.td
@@ -304,8 +304,6 @@ def : ProcessorModel<"gfx11-generic", GFX11SpeedModel,
// GCN GFX12.
//===----------------------------------------------------------------------===//
-// TODO: gfx12-generic ?
-
def : ProcessorModel<"gfx1200", GFX12SpeedModel,
FeatureISAVersion12.Features
>;
@@ -313,3 +311,8 @@ def : ProcessorModel<"gfx1200", GFX12SpeedModel,
def : ProcessorModel<"gfx1201", GFX12SpeedModel,
FeatureISAVersion12.Features
>;
+
+// [gfx1200, gfx1201]
+def : ProcessorModel<"gfx12-generic", GFX12SpeedModel,
+ FeatureISAVersion12_Generic.Features
+>;
diff --git a/llvm/lib/Target/AMDGPU/MCTargetDesc/AMDGPUTargetStreamer.cpp b/llvm/lib/Target/AMDGPU/MCTargetDesc/AMDGPUTargetStreamer.cpp
index 00e64e3419ba0..d7d6e00d23897 100644
--- a/llvm/lib/Target/AMDGPU/MCTargetDesc/AMDGPUTargetStreamer.cpp
+++ b/llvm/lib/Target/AMDGPU/MCTargetDesc/AMDGPUTargetStreamer.cpp
@@ -119,6 +119,7 @@ StringRef AMDGPUTargetStreamer::getArchNameFromElfMach(unsigned ElfMach) {
case ELF::EF_AMDGPU_MACH_AMDGCN_GFX10_1_GENERIC: AK = GK_GFX10_1_GENERIC; break;
case ELF::EF_AMDGPU_MACH_AMDGCN_GFX10_3_GENERIC: AK = GK_GFX10_3_GENERIC; break;
case ELF::EF_AMDGPU_MACH_AMDGCN_GFX11_GENERIC: AK = GK_GFX11_GENERIC; break;
+ case ELF::EF_AMDGPU_MACH_AMDGCN_GFX12_GENERIC: AK = GK_GFX12_GENERIC; break;
case ELF::EF_AMDGPU_MACH_NONE: AK = GK_NONE; break;
default: AK = GK_NONE; break;
}
@@ -201,6 +202,7 @@ unsigned AMDGPUTargetStreamer::getElfMach(StringRef GPU) {
case GK_GFX10_1_GENERIC: return ELF::EF_AMDGPU_MACH_AMDGCN_GFX10_1_GENERIC;
case GK_GFX10_3_GENERIC: return ELF::EF_AMDGPU_MACH_AMDGCN_GFX10_3_GENERIC;
case GK_GFX11_GENERIC: return ELF::EF_AMDGPU_MACH_AMDGCN_GFX11_GENERIC;
+ case GK_GFX12_GENERIC: return ELF::EF_AMDGPU_MACH_AMDGCN_GFX12_GENERIC;
case GK_NONE: return ELF::EF_AMDGPU_MACH_NONE;
}
// clang-format on
@@ -768,6 +770,9 @@ unsigned AMDGPUTargetELFStreamer::getEFlagsV6() {
case AMDGPU::GK_GFX11_GENERIC:
Version = GenericVersion::GFX11;
break;
+ case AMDGPU::GK_GFX12_GENERIC:
+ Version = GenericVersion::GFX12;
+ break;
default:
break;
}
diff --git a/llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.h b/llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.h
index 3cfc42a7d24d5..cf8236b8e23b7 100644
--- a/llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.h
+++ b/llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.h
@@ -49,6 +49,7 @@ static constexpr unsigned GFX9 = 1;
static constexpr unsigned GFX10_1 = 1;
static constexpr unsigned GFX10_3 = 1;
static constexpr unsigned GFX11 = 1;
+static constexpr unsigned GFX12 = 1;
} // namespace GenericVersion
enum { AMDHSA_COV4 = 4, AMDHSA_COV5 = 5, AMDHSA_COV6 = 6 };
diff --git a/llvm/lib/TargetParser/TargetParser.cpp b/llvm/lib/TargetParser/TargetParser.cpp
index 0d784a79e5bac..7464237d26d48 100644
--- a/llvm/lib/TargetParser/TargetParser.cpp
+++ b/llvm/lib/TargetParser/TargetParser.cpp
@@ -131,6 +131,7 @@ constexpr GPUInfo AMDGCNGPUs[] = {
{{"gfx10-1-generic"}, {"gfx10-1-generic"}, GK_GFX10_1_GENERIC, FEATURE_FAST_FMA_F32|FEATURE_FAST_DENORMAL_F32|FEATURE_WAVE32|FEATURE_XNACK|FEATURE_WGP},
{{"gfx10-3-generic"}, {"gfx10-3-generic"}, GK_GFX10_3_GENERIC, FEATURE_FAST_FMA_F32|FEATURE_FAST_DENORMAL_F32|FEATURE_WAVE32|FEATURE_WGP},
{{"gfx11-generic"}, {"gfx11-generic"}, GK_GFX11_GENERIC, FEATURE_FAST_FMA_F32|FEATURE_FAST_DENORMAL_F32|FEATURE_WAVE32|FEATURE_WGP},
+ {{"gfx12-generic"}, {"gfx12-generic"}, GK_GFX12_GENERIC, FEATURE_FAST_FMA_F32|FEATURE_FAST_DENORMAL_F32|FEATURE_WAVE32|FEATURE_WGP},
// clang-format on
};
@@ -158,6 +159,8 @@ StringRef llvm::AMDGPU::getArchFamilyNameAMDGCN(GPUKind AK) {
return "gfx10";
case AMDGPU::GK_GFX11_GENERIC:
return "gfx11";
+ case AMDGPU::GK_GFX12_GENERIC:
+ return "gfx12";
default: {
StringRef ArchName = getArchNameAMDGCN(AK);
return ArchName.empty() ? "" : ArchName.drop_back(2);
@@ -292,6 +295,7 @@ AMDGPU::IsaVersion AMDGPU::getIsaVersion(StringRef GPU) {
case GK_GFX10_1_GENERIC: return {10, 1, 0};
case GK_GFX10_3_GENERIC: return {10, 3, 0};
case GK_GFX11_GENERIC: return {11, 0, 3};
+ case GK_GFX12_GENERIC: return {12, 0, 0};
default: return {0, 0, 0};
}
// clang-format on
@@ -313,6 +317,7 @@ void AMDGPU::fillAMDGPUFeatureMap(StringRef GPU, const Triple &T,
switch (parseArchAMDGCN(GPU)) {
case GK_GFX1201:
case GK_GFX1200:
+ case GK_GFX12_GENERIC:
Features["ci-insts"] = true;
Features["dot7-insts"] = true;
Features["dot8-insts"] = true;
@@ -554,6 +559,7 @@ static bool isWave32Capable(StringRef GPU, const Triple &T) {
case GK_GFX1011:
case GK_GFX1013:
case GK_GFX1010:
+ case GK_GFX12_GENERIC:
case GK_GFX11_GENERIC:
case GK_GFX10_3_GENERIC:
case GK_GFX10_1_GENERIC:
diff --git a/llvm/test/CodeGen/AMDGPU/directive-amdgcn-target.ll b/llvm/test/CodeGen/AMDGPU/directive-amdgcn-target.ll
index 2979b52d00e2e..c347437c30828 100644
--- a/llvm/test/CodeGen/AMDGPU/directive-amdgcn-target.ll
+++ b/llvm/test/CodeGen/AMDGPU/directive-amdgcn-target.ll
@@ -114,6 +114,7 @@
; RUN: llc --amdhsa-code-object-version=6 -mtriple=amdgcn-amd-amdhsa -mcpu=gfx10-1-generic -mattr=+xnack < %s | FileCheck --check-prefixes=GFX10_1_GENERIC_XNACK %s
; RUN: llc --amdhsa-code-object-version=6 -mtriple=amdgcn-amd-amdhsa -mcpu=gfx10-3-generic < %s | FileCheck --check-prefixes=GFX10_3_GENERIC %s
; RUN: llc --amdhsa-code-object-version=6 -mtriple=amdgcn-amd-amdhsa -mcpu=gfx11-generic < %s | FileCheck --check-prefixes=GFX11_GENERIC %s
+; RUN: llc --amdhsa-code-object-version=6 -mtriple=amdgcn-amd-amdhsa -mcpu=gfx12-generic < %s | FileCheck --check-prefixes=GFX12_GENERIC %s
; GFX600: .amdgcn_target "amdgcn-amd-amdhsa--gfx600"
; GFX601: .amdgcn_target "amdgcn-amd-amdhsa--gfx601"
@@ -209,6 +210,7 @@
; GFX10_1_GENERIC_XNACK: .amdgcn_target "amdgcn-amd-amdhsa--gfx10-1-generic:xnack+"
; GFX10_3_GENERIC: .amdgcn_target "amdgcn-amd-amdhsa--gfx10-3-generic"
; GFX11_GENERIC: .amdgcn_target "amdgcn-amd-amdhsa--gfx11-generic"
+; GFX12_GENERIC: .amdgcn_target "amdgcn-amd-amdhsa--gfx12-generic"
define amdgpu_kernel void @directive_amdgcn_target() {
ret void
diff --git a/llvm/test/CodeGen/AMDGPU/elf-header-flags-mach.ll b/llvm/test/CodeGen/AMDGPU/elf-header-flags-mach.ll
index 5f4bfe7ea9d5f..edc20153ffd77 100644
--- a/llvm/test/CodeGen/AMDGPU/elf-header-flags-mach.ll
+++ b/llvm/test/CodeGen/AMDGPU/elf-header-flags-mach.ll
@@ -81,6 +81,7 @@
; RUN: llc -filetype=obj --amdhsa-code-object-version=6 -mtriple=amdgcn -mcpu=gfx10-1-generic < %s | llvm-readobj --file-header - | FileCheck --check-prefixes=ALL,ARCH-GCN,GFX10_1_GENERIC %s
; RUN: llc -filetype=obj --amdhsa-code-object-version=6 -mtriple=amdgcn -mcpu=gfx10-3-generic < %s | llvm-readobj --file-header - | FileCheck --check-prefixes=ALL,ARCH-GCN,GFX10_3_GENERIC %s
; RUN: llc -filetype=obj --amdhsa-code-object-version=6 -mtriple=amdgcn -mcpu=gfx11-generic < %s | llvm-readobj --file-header - | FileCheck --check-prefixes=ALL,ARCH-GCN,GFX11_GENERIC %s
+; RUN: llc -filetype=obj --amdhsa-code-object-version=6 -mtriple=amdgcn -mcpu=gfx12-generic < %s | llvm-readobj --file-header - | FileCheck --check-prefixes=ALL,ARCH-GCN,GFX12_GENERIC %s
; FIXME: With the default attributes the eflags are not accurate for
; xnack and sramecc. Subsequent Target-ID patches will address this.
@@ -159,6 +160,7 @@
; GFX10_1_GENERIC: EF_AMDGPU_MACH_AMDGCN_GFX10_1_GENERIC (0x52)
; GFX10_3_GENERIC: EF_AMDGPU_MACH_AMDGCN_GFX10_3_GENERIC (0x53)
; GFX11_GENERIC: EF_AMDGPU_MACH_AMDGCN_GFX11_GENERIC (0x54)
+; GFX12_GENERIC: EF_AMDGPU_MACH_AMDGCN_GFX12_GENERIC (0x59)
; ALL: ]
define amdgpu_kernel void @elf_header() {
diff --git a/llvm/test/CodeGen/AMDGPU/generic-targets-require-v6.ll b/llvm/test/CodeGen/AMDGPU/generic-targets-require-v6.ll
index 15a696bb3af09..d30cf1531a06b 100644
--- a/llvm/test/CodeGen/AMDGPU/generic-targets-require-v6.ll
+++ b/llvm/test/CodeGen/AMDGPU/generic-targets-require-v6.ll
@@ -2,16 +2,19 @@
; RUN: not llc -mtriple=amdgcn -mcpu=gfx10-1-generic --amdhsa-code-object-version=5 -o - %s 2>&1 | FileCheck --check-prefix=GFX101-V5 %s
; RUN: not llc -mtriple=amdgcn -mcpu=gfx10-3-generic --amdhsa-code-object-version=5 -o - %s 2>&1 | FileCheck --check-prefix=GFX103-V5 %s
; RUN: not llc -mtriple=amdgcn -mcpu=gfx11-generic --amdhsa-code-object-version=5 -o - %s 2>&1 | FileCheck --check-prefix=GFX11-V5 %s
+; RUN: not llc -mtriple=amdgcn -mcpu=gfx12-generic --amdhsa-code-object-version=5 -o - %s 2>&1 | FileCheck --check-prefix=GFX12-V5 %s
; RUN: llc -mtriple=amdgcn -mcpu=gfx9-generic --amdhsa-code-object-version=6 -o - %s
; RUN: llc -mtriple=amdgcn -mcpu=gfx10-1-generic --amdhsa-code-object-version=6 -o - %s
; RUN: llc -mtriple=amdgcn -mcpu=gfx10-3-generic --amdhsa-code-object-version=6 -o - %s
; RUN: llc -mtriple=amdgcn -mcpu=gfx11-generic --amdhsa-code-object-version=6 -o - %s
+; RUN: llc -mtriple=amdgcn -mcpu=gfx12-generic --amdhsa-code-object-version=6 -o - %s
; GFX9-V5: gfx9-generic is only available on code object version 6 or better
; GFX101-V5: gfx10-1-generic is only available on code object version 6 or better
; GFX103-V5: gfx10-3-generic is only available on code object version 6 or better
; GFX11-V5: gfx11-generic is only available on code object version 6 or better
+; GFX12-V5: gfx12-generic is only available on code object version 6 or better
define void @foo() {
ret void
diff --git a/llvm/test/CodeGen/AMDGPU/hsa-generic-target-features.ll b/llvm/test/CodeGen/AMDGPU/hsa-generic-target-features.ll
index b69fd884ee4a3..1cb27dc6fc45c 100644
--- a/llvm/test/CodeGen/AMDGPU/hsa-generic-target-features.ll
+++ b/llvm/test/CodeGen/AMDGPU/hsa-generic-target-features.ll
@@ -4,6 +4,7 @@
; RUN: llc --amdhsa-code-object-version=6 -mtriple=amdgcn-amd-amdhsa -mcpu=gfx10-3-generic < %s | FileCheck -check-prefix=CU %s
; RUN: llc --amdhsa-code-object-version=6 -mtriple=amdgcn-amd-amdhsa -mcpu=gfx11-generic -mattr=+cumode < %s | FileCheck -check-prefix=NOCU %s
; RUN: llc --amdhsa-code-object-version=6 -mtriple=amdgcn-amd-amdhsa -mcpu=gfx11-generic < %s | FileCheck -check-prefix=CU %s
+; RUN: llc --amdhsa-code-object-version=6 -mtriple=amdgcn-amd-amdhsa -mcpu=gfx12-generic < %s | FileCheck -check-prefix=CU %s
; RUN: llc --amdhsa-code-object-version=6 -mtriple=amdgcn-amd-amdhsa -mcpu=gfx10-1-generic -mattr=+wavefrontsize32,-wavefrontsize64 < %s | FileCheck -check-prefix=W32 %s
; RUN: llc --amdhsa-code-object-version=6 -mtriple=amdgcn-amd-amdhsa -mcpu=gfx10-1-generic -mattr=-wavefrontsize32,+wavefrontsize64 < %s | FileCheck -check-prefix=W64 %s
@@ -11,6 +12,7 @@
; RUN: llc --amdhsa-code-object-version=6 -mtriple=amdgcn-amd-amdhsa -mcpu=gfx10-3-generic -mattr=-wavefrontsize32,+wavefrontsize64 < %s | FileCheck -check-prefix=W64 %s
; RUN: llc --amdhsa-code-object-version=6 -mtriple=amdgcn-amd-amdhsa -mcpu=gfx11-generic -mattr=+wavefrontsize32,-wavefrontsize64 < %s | FileCheck -check-prefix=W32 %s
; RUN: llc --amdhsa-code-object-version=6 -mtriple=amdgcn-amd-amdhsa -mcpu=gfx11-generic -mattr=-wavefrontsize32,+wavefrontsize64 < %s | FileCheck -check-prefix=W64 %s
+; RUN: llc --amdhsa-code-object-version=6 -mtriple=amdgcn-amd-amdhsa -mcpu=gfx12-generic -mattr=-wavefrontsize32,+wavefrontsize64 < %s | FileCheck -check-prefix=W64 %s
; Checks 10.1, 10.3 and 11 generic targets allow cumode/wave64.
diff --git a/llvm/test/Object/AMDGPU/elf-header-flags-mach.yaml b/llvm/test/Object/AMDGPU/elf-header-flags-mach.yaml
index 4c2b4479a4aa3..7f36795aa6a9e 100644
--- a/llvm/test/Object/AMDGPU/elf-header-flags-mach.yaml
+++ b/llvm/test/Object/AMDGPU/elf-header-flags-mach.yaml
@@ -254,6 +254,10 @@
# RUN: llvm-readobj -S --file-headers %t.o.AMDGCN_GFX11_GENERIC | FileCheck --check-prefixes=ELF-AMDGCN-ALL,ELF-AMDGCN-GFX11_GENERIC %s
# RUN: obj2yaml %t.o.AMDGCN_GFX11_GENERIC | FileCheck --check-prefixes=YAML-AMDGCN-ALL,YAML-AMDGCN-GFX11_GENERIC %s
+# RUN: sed -e 's/<BITS>/64/' -e 's/<MACH>/AMDGCN_GFX12_GENERIC/' %s | yaml2obj -o %t.o.AMDGCN_GFX12_GENERIC
+# RUN: llvm-readobj -S --file-headers %t.o.AMDGCN_GFX12_GENERIC | FileCheck --check-prefixes=ELF-AMDGCN-ALL,ELF-AMDGCN-GFX12_GENERIC %s
+# RUN: obj2yaml %t.o.AMDGCN_GFX12_GENERIC | FileCheck --check-prefixes=YAML-AMDGCN-ALL,YAML-AMDGCN-GFX12_GENERIC %s
+
# ELF-R600-ALL: Format: elf32-amdgpu
# ELF-R600-ALL: Arch: r600
@@ -464,6 +468,9 @@
# ELF-AMDGCN-GFX11_GENERIC: EF_AMDGPU_MACH_AMDGCN_GFX11_GENERIC (0x54)
# YAML-AMDGCN-GFX11_GENERIC: Flags: [ EF_AMDGPU_MACH_AMDGCN_GFX11_GENERIC ]
+# ELF-AMDGCN-GFX12_GENERIC: EF_AMDGPU_MACH_AMDGCN_GFX12_GENERIC (0x59)
+# YAML-AMDGCN-GFX12_GENERIC: Flags: [ EF_AMDGPU_MACH_AMDGCN_GFX12_GENERIC ]
+
# ELF-AMDGCN-ALL: ]
diff --git a/llvm/test/tools/llvm-objdump/ELF/AMDGPU/subtarget.ll b/llvm/test/tools/llvm-objdump/ELF/AMDGPU/subtarget.ll
index bcec6ce9ed376..f1262c5a2fc2a 100644
--- a/llvm/test/tools/llvm-objdump/ELF/AMDGPU/subtarget.ll
+++ b/llvm/test/tools/llvm-objdump/ELF/AMDGPU/subtarget.ll
@@ -6,6 +6,11 @@ define amdgpu_kernel void @test_kernel() {
;
; ----------------------------------GFX12--------------------------------------
;
+; RUN: llc -mtriple=amdgcn-amd-amdhsa --amdhsa-code-object-version=6 -mcpu=gfx12-generic -filetype=obj -O0 -o %t.o %s
+; RUN: llvm-objdump -D --arch-name=amdgcn -mllvm --amdhsa-code-object-version=6 --mcpu=gfx12-generic %t.o > %t-specify.txt
+; RUN: llvm-objdump -D -mllvm --amdhsa-code-object-version=6 %t.o > %t-detect.txt
+; RUN: diff %t-specify.txt %t-detect.txt
+;
; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1201 -filetype=obj -O0 -o %t.o %s
; RUN: llvm-objdump -D --arch-name=amdgcn --mcpu=gfx1201 %t.o > %t-specify.txt
; RUN: llvm-objdump -D %t.o > %t-detect.txt
diff --git a/llvm/test/tools/llvm-readobj/ELF/AMDGPU/elf-headers.test b/llvm/test/tools/llvm-readobj/ELF/AMDGPU/elf-headers.test
index 7fbf4aae6c246..f9e90e21400ea 100644
--- a/llvm/test/tools/llvm-readobj/ELF/AMDGPU/elf-headers.test
+++ b/llvm/test/tools/llvm-readobj/ELF/AMDGPU/elf-headers.test
@@ -439,6 +439,9 @@
# RUN: yaml2obj %s -o %t -DABI_VERSION=2 -DFLAG_NAME=EF_AMDGPU_MACH_AMDGCN_GFX1201
# RUN: llvm-readobj -h %t | FileCheck %s --check-prefixes=ALL,KNOWN-ABI-VERSION,SINGLE-FLAG --match-full-lines -DABI_VERSION=2 -DFILE=%t -DFLAG_NAME=EF_AMDGPU_MACH_AMDGCN_GFX1201 -DFLAG_VALUE=0x4E
+# RUN: yaml2obj %s -o %t -DABI_VERSION=4 -DFLAG_NAME=EF_AMDGPU_MACH_AMDGCN_GFX12_GENERIC
+# RUN: llvm-readobj -h %t | FileCheck %s --check-prefixes=ALL,KNOWN-ABI-VERSION,SINGLE-FLAG --match-full-lines -DABI_VERSION=4 -DFILE=%t -DFLAG_NAME=EF_AMDGPU_MACH_AMDGCN_GFX12_GENERIC -DFLAG_VALUE=0x59
+
--- !ELF
FileHeader:
Class: ELFCLASS64
diff --git a/llvm/tools/llvm-readobj/ELFDumper.cpp b/llvm/tools/llvm-readobj/ELFDumper.cpp
index 966531ef56099..40494da8b21bd 100644
--- a/llvm/tools/llvm-readobj/ELFDumper.cpp
+++ b/llvm/tools/llvm-readobj/ELFDumper.cpp
@@ -1636,7 +1636,8 @@ const EnumEntry<unsigned> ElfHeaderMipsFlags[] = {
ENUM_ENT(EF_AMDGPU_MACH_AMDGCN_GFX9_GENERIC, "gfx9-generic"), \
ENUM_ENT(EF_AMDGPU_MACH_AMDGCN_GFX10_1_GENERIC, "gfx10-1-generic"), \
ENUM_ENT(EF_AMDGPU_MACH_AMDGCN_GFX10_3_GENERIC, "gfx10-3-generic"), \
- ENUM_ENT(EF_AMDGPU_MACH_AMDGCN_GFX11_GENERIC, "gfx11-generic")
+ ENUM_ENT(EF_AMDGPU_MACH_AMDGCN_GFX11_GENERIC, "gfx11-generic"), \
+ ENUM_ENT(EF_AMDGPU_MACH_AMDGCN_GFX12_GENERIC, "gfx12-generic")
// clang-format on
const EnumEntry<unsigned> ElfHeaderAMDGPUFlagsABIVersion3[] = {
>From 469b197e9104794ec9d19e728c690a0309e3b0ee Mon Sep 17 00:00:00 2001
From: Konstantin Zhuravlyov <kzhuravl_dev at outlook.com>
Date: Thu, 30 May 2024 16:29:57 -0400
Subject: [PATCH 2/3] AMDGPU/Docs: Sort MACHs based on value
---
llvm/docs/AMDGPUUsage.rst | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/llvm/docs/AMDGPUUsage.rst b/llvm/docs/AMDGPUUsage.rst
index 77d68dd22f58b..8a787d3160eb9 100644
--- a/llvm/docs/AMDGPUUsage.rst
+++ b/llvm/docs/AMDGPUUsage.rst
@@ -1976,11 +1976,11 @@ The AMDGPU backend uses the following ELF header:
``EF_AMDGPU_MACH_AMDGCN_GFX10_1_GENERIC`` 0x052 ``gfx10-1-generic``
``EF_AMDGPU_MACH_AMDGCN_GFX10_3_GENERIC`` 0x053 ``gfx10-3-generic``
``EF_AMDGPU_MACH_AMDGCN_GFX11_GENERIC`` 0x054 ``gfx11-generic``
- ``EF_AMDGPU_MACH_AMDGCN_GFX12_GENERIC`` 0x059 ``gfx12-generic``
*reserved* 0x055 Reserved.
*reserved* 0x056 Reserved.
*reserved* 0x057 Reserved.
*reserved* 0x058 Reserved.
+ ``EF_AMDGPU_MACH_AMDGCN_GFX12_GENERIC`` 0x059 ``gfx12-generic``
========================================== ========== =============================
Sections
>From 2d9ba5742305c9e022b5071cbd10f017848062b8 Mon Sep 17 00:00:00 2001
From: Konstantin Zhuravlyov <kzhuravl_dev at outlook.com>
Date: Thu, 30 May 2024 16:55:41 -0400
Subject: [PATCH 3/3] Formatting
---
llvm/include/llvm/BinaryFormat/ELF.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/llvm/include/llvm/BinaryFormat/ELF.h b/llvm/include/llvm/BinaryFormat/ELF.h
index c9456a3b36054..746501a370f35 100644
--- a/llvm/include/llvm/BinaryFormat/ELF.h
+++ b/llvm/include/llvm/BinaryFormat/ELF.h
@@ -799,7 +799,7 @@ enum : unsigned {
EF_AMDGPU_MACH_AMDGCN_RESERVED_0X56 = 0x056,
EF_AMDGPU_MACH_AMDGCN_RESERVED_0X57 = 0x057,
EF_AMDGPU_MACH_AMDGCN_RESERVED_0X58 = 0x058,
- EF_AMDGPU_MACH_AMDGCN_GFX12_GENERIC = 0x059,
+ EF_AMDGPU_MACH_AMDGCN_GFX12_GENERIC = 0x059,
// clang-format on
// First/last AMDGCN-based processors.
More information about the llvm-commits
mailing list