[flang-commits] [llvm] [flang] [compiler-rt] [libcxx] [mlir] [clang] [libc] [AMDGPU] Define new targets gfx1200 and gfx1201 (PR #73133)
Jay Foad via flang-commits
flang-commits at lists.llvm.org
Thu Nov 23 02:04:55 PST 2023
https://github.com/jayfoad updated https://github.com/llvm/llvm-project/pull/73133
>From 1011b8e7da1741aaaa46dfb4c9a4bf54468ea021 Mon Sep 17 00:00:00 2001
From: Jay Foad <jay.foad at amd.com>
Date: Tue, 21 Nov 2023 15:46:04 +0000
Subject: [PATCH 1/2] [AMDGPU] Define new targets gfx1200 and gfx1201
Define target names and ELF numbers for new GFX12 targets gfx1200 and
gfx1201. For now they behave identically to GFX11.
---
clang/include/clang/Basic/Cuda.h | 2 +
clang/lib/Basic/Cuda.cpp | 2 +
clang/lib/Basic/Targets/NVPTX.cpp | 2 +
clang/lib/CodeGen/CGOpenMPRuntimeGPU.cpp | 2 +
clang/test/CodeGenOpenCL/amdgpu-features.cl | 4 +
clang/test/Driver/amdgpu-macros.cl | 2 +
clang/test/Driver/amdgpu-mcpu.cl | 4 +
clang/test/Misc/target-invalid-cpu-note.c | 4 +-
llvm/docs/AMDGPUUsage.rst | 18 ++++-
llvm/include/llvm/BinaryFormat/ELF.h | 6 +-
llvm/include/llvm/TargetParser/TargetParser.h | 5 +-
llvm/lib/Object/ELFObjectFile.cpp | 6 ++
llvm/lib/ObjectYAML/ELFYAML.cpp | 2 +
llvm/lib/Target/AMDGPU/AMDGPU.td | 75 ++++++++++++++++++-
llvm/lib/Target/AMDGPU/AMDGPUSubtarget.h | 3 +-
llvm/lib/Target/AMDGPU/GCNProcessors.td | 12 +++
llvm/lib/Target/AMDGPU/GCNSubtarget.h | 1 +
.../MCTargetDesc/AMDGPUTargetStreamer.cpp | 4 +
llvm/lib/Target/AMDGPU/SIDefines.h | 1 +
llvm/lib/Target/AMDGPU/SIInstrInfo.cpp | 8 ++
llvm/lib/Target/AMDGPU/SIInstrInfo.td | 4 +-
.../Target/AMDGPU/Utils/AMDGPUBaseInfo.cpp | 26 ++++++-
llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.h | 6 ++
llvm/lib/TargetParser/TargetParser.cpp | 27 +++++++
.../CodeGen/AMDGPU/directive-amdgcn-target.ll | 4 +
.../CodeGen/AMDGPU/elf-header-flags-mach.ll | 4 +
.../Object/AMDGPU/elf-header-flags-mach.yaml | 14 ++++
.../llvm-objdump/ELF/AMDGPU/subtarget.ll | 12 +++
.../llvm-readobj/ELF/amdgpu-elf-headers.test | 18 +++++
llvm/tools/llvm-readobj/ELFDumper.cpp | 4 +
30 files changed, 272 insertions(+), 10 deletions(-)
diff --git a/clang/include/clang/Basic/Cuda.h b/clang/include/clang/Basic/Cuda.h
index 878f8d70f90c0a9..2d912bdbbd1bc59 100644
--- a/clang/include/clang/Basic/Cuda.h
+++ b/clang/include/clang/Basic/Cuda.h
@@ -113,6 +113,8 @@ enum class CudaArch {
GFX1103,
GFX1150,
GFX1151,
+ GFX1200,
+ GFX1201,
Generic, // A processor model named 'generic' if the target backend defines a
// public one.
LAST,
diff --git a/clang/lib/Basic/Cuda.cpp b/clang/lib/Basic/Cuda.cpp
index 2307352bd3becef..65840b9f20252b6 100644
--- a/clang/lib/Basic/Cuda.cpp
+++ b/clang/lib/Basic/Cuda.cpp
@@ -135,6 +135,8 @@ static const CudaArchToStringMap arch_names[] = {
GFX(1103), // gfx1103
GFX(1150), // gfx1150
GFX(1151), // gfx1151
+ GFX(1200), // gfx1200
+ GFX(1201), // gfx1201
{CudaArch::Generic, "generic", ""},
// clang-format on
};
diff --git a/clang/lib/Basic/Targets/NVPTX.cpp b/clang/lib/Basic/Targets/NVPTX.cpp
index a9fc88295700b89..3a4a75b0348f209 100644
--- a/clang/lib/Basic/Targets/NVPTX.cpp
+++ b/clang/lib/Basic/Targets/NVPTX.cpp
@@ -214,6 +214,8 @@ void NVPTXTargetInfo::getTargetDefines(const LangOptions &Opts,
case CudaArch::GFX1103:
case CudaArch::GFX1150:
case CudaArch::GFX1151:
+ case CudaArch::GFX1200:
+ case CudaArch::GFX1201:
case CudaArch::Generic:
case CudaArch::LAST:
break;
diff --git a/clang/lib/CodeGen/CGOpenMPRuntimeGPU.cpp b/clang/lib/CodeGen/CGOpenMPRuntimeGPU.cpp
index 2f7dd83bd2d65c9..9b8fbbdf8046787 100644
--- a/clang/lib/CodeGen/CGOpenMPRuntimeGPU.cpp
+++ b/clang/lib/CodeGen/CGOpenMPRuntimeGPU.cpp
@@ -3540,6 +3540,8 @@ void CGOpenMPRuntimeGPU::processRequiresDirective(
case CudaArch::GFX1103:
case CudaArch::GFX1150:
case CudaArch::GFX1151:
+ case CudaArch::GFX1200:
+ case CudaArch::GFX1201:
case CudaArch::Generic:
case CudaArch::UNUSED:
case CudaArch::UNKNOWN:
diff --git a/clang/test/CodeGenOpenCL/amdgpu-features.cl b/clang/test/CodeGenOpenCL/amdgpu-features.cl
index 03c20ae46faaa46..8959634572b44e9 100644
--- a/clang/test/CodeGenOpenCL/amdgpu-features.cl
+++ b/clang/test/CodeGenOpenCL/amdgpu-features.cl
@@ -49,6 +49,8 @@
// RUN: %clang_cc1 -triple amdgcn -target-cpu gfx1103 -S -emit-llvm -o - %s | FileCheck --check-prefix=GFX1103 %s
// RUN: %clang_cc1 -triple amdgcn -target-cpu gfx1150 -S -emit-llvm -o - %s | FileCheck --check-prefix=GFX1150 %s
// RUN: %clang_cc1 -triple amdgcn -target-cpu gfx1151 -S -emit-llvm -o - %s | FileCheck --check-prefix=GFX1151 %s
+// RUN: %clang_cc1 -triple amdgcn -target-cpu gfx1200 -S -emit-llvm -o - %s | FileCheck --check-prefix=GFX1200 %s
+// RUN: %clang_cc1 -triple amdgcn -target-cpu gfx1201 -S -emit-llvm -o - %s | FileCheck --check-prefix=GFX1201 %s
// RUN: %clang_cc1 -triple amdgcn -target-cpu gfx1103 -target-feature +wavefrontsize64 -S -emit-llvm -o - %s | FileCheck --check-prefix=GFX1103-W64 %s
@@ -98,6 +100,8 @@
// GFX1103: "target-features"="+16-bit-insts,+atomic-fadd-rtn-insts,+ci-insts,+dl-insts,+dot10-insts,+dot5-insts,+dot7-insts,+dot8-insts,+dot9-insts,+dpp,+gfx10-3-insts,+gfx10-insts,+gfx11-insts,+gfx8-insts,+gfx9-insts,+wavefrontsize32"
// GFX1150: "target-features"="+16-bit-insts,+atomic-fadd-rtn-insts,+ci-insts,+dl-insts,+dot10-insts,+dot5-insts,+dot7-insts,+dot8-insts,+dot9-insts,+dpp,+gfx10-3-insts,+gfx10-insts,+gfx11-insts,+gfx8-insts,+gfx9-insts,+wavefrontsize32"
// GFX1151: "target-features"="+16-bit-insts,+atomic-fadd-rtn-insts,+ci-insts,+dl-insts,+dot10-insts,+dot5-insts,+dot7-insts,+dot8-insts,+dot9-insts,+dpp,+gfx10-3-insts,+gfx10-insts,+gfx11-insts,+gfx8-insts,+gfx9-insts,+wavefrontsize32"
+// GFX1200: "target-features"="+16-bit-insts,+atomic-fadd-rtn-insts,+ci-insts,+dl-insts,+dot10-insts,+dot5-insts,+dot7-insts,+dot8-insts,+dot9-insts,+dpp,+gfx10-3-insts,+gfx10-insts,+gfx11-insts,+gfx12-insts,+gfx8-insts,+gfx9-insts,+wavefrontsize32"
+// GFX1201: "target-features"="+16-bit-insts,+atomic-fadd-rtn-insts,+ci-insts,+dl-insts,+dot10-insts,+dot5-insts,+dot7-insts,+dot8-insts,+dot9-insts,+dpp,+gfx10-3-insts,+gfx10-insts,+gfx11-insts,+gfx12-insts,+gfx8-insts,+gfx9-insts,+wavefrontsize32"
// GFX1103-W64: "target-features"="+16-bit-insts,+atomic-fadd-rtn-insts,+ci-insts,+dl-insts,+dot10-insts,+dot5-insts,+dot7-insts,+dot8-insts,+dot9-insts,+dpp,+gfx10-3-insts,+gfx10-insts,+gfx11-insts,+gfx8-insts,+gfx9-insts,+wavefrontsize64"
diff --git a/clang/test/Driver/amdgpu-macros.cl b/clang/test/Driver/amdgpu-macros.cl
index dc12a348334b6a1..81c22af460d12d0 100644
--- a/clang/test/Driver/amdgpu-macros.cl
+++ b/clang/test/Driver/amdgpu-macros.cl
@@ -128,6 +128,8 @@
// RUN: %clang -E -dM -target amdgcn -mcpu=gfx1103 %s 2>&1 | FileCheck --check-prefixes=ARCH-GCN,FAST_FMAF %s -DWAVEFRONT_SIZE=32 -DCPU=gfx1103 -DFAMILY=GFX11
// RUN: %clang -E -dM -target amdgcn -mcpu=gfx1150 %s 2>&1 | FileCheck --check-prefixes=ARCH-GCN,FAST_FMAF %s -DWAVEFRONT_SIZE=32 -DCPU=gfx1150 -DFAMILY=GFX11
// RUN: %clang -E -dM -target amdgcn -mcpu=gfx1151 %s 2>&1 | FileCheck --check-prefixes=ARCH-GCN,FAST_FMAF %s -DWAVEFRONT_SIZE=32 -DCPU=gfx1151 -DFAMILY=GFX11
+// RUN: %clang -E -dM -target amdgcn -mcpu=gfx1200 %s 2>&1 | FileCheck --check-prefixes=ARCH-GCN,FAST_FMAF %s -DWAVEFRONT_SIZE=32 -DCPU=gfx1200 -DFAMILY=GFX12
+// RUN: %clang -E -dM -target amdgcn -mcpu=gfx1201 %s 2>&1 | FileCheck --check-prefixes=ARCH-GCN,FAST_FMAF %s -DWAVEFRONT_SIZE=32 -DCPU=gfx1201 -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 cde48e570ae028d..eeb16ae98ebad7b 100644
--- a/clang/test/Driver/amdgpu-mcpu.cl
+++ b/clang/test/Driver/amdgpu-mcpu.cl
@@ -112,6 +112,8 @@
// RUN: %clang -### -target amdgcn -mcpu=gfx1103 %s 2>&1 | FileCheck --check-prefix=GFX1103 %s
// RUN: %clang -### -target amdgcn -mcpu=gfx1150 %s 2>&1 | FileCheck --check-prefix=GFX1150 %s
// RUN: %clang -### -target amdgcn -mcpu=gfx1151 %s 2>&1 | FileCheck --check-prefix=GFX1151 %s
+// RUN: %clang -### -target amdgcn -mcpu=gfx1200 %s 2>&1 | FileCheck --check-prefix=GFX1200 %s
+// RUN: %clang -### -target amdgcn -mcpu=gfx1201 %s 2>&1 | FileCheck --check-prefix=GFX1201 %s
// GCNDEFAULT-NOT: -target-cpu
// GFX600: "-target-cpu" "gfx600"
@@ -156,3 +158,5 @@
// GFX1103: "-target-cpu" "gfx1103"
// GFX1150: "-target-cpu" "gfx1150"
// GFX1151: "-target-cpu" "gfx1151"
+// GFX1200: "-target-cpu" "gfx1200"
+// GFX1201: "-target-cpu" "gfx1201"
diff --git a/clang/test/Misc/target-invalid-cpu-note.c b/clang/test/Misc/target-invalid-cpu-note.c
index 25ff51e071b69b3..693f47a78b7fa57 100644
--- a/clang/test/Misc/target-invalid-cpu-note.c
+++ b/clang/test/Misc/target-invalid-cpu-note.c
@@ -29,7 +29,7 @@
// 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'
-// NVPTX-NEXT: note: valid target CPU values are: sm_20, sm_21, sm_30, sm_32, sm_35, sm_37, sm_50, sm_52, sm_53, sm_60, sm_61, sm_62, sm_70, sm_72, sm_75, sm_80, sm_86, sm_87, sm_89, sm_90, gfx600, gfx601, gfx602, gfx700, gfx701, gfx702, gfx703, gfx704, gfx705, gfx801, gfx802, gfx803, gfx805, gfx810, 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{{$}}
+// NVPTX-NEXT: note: valid target CPU values are: sm_20, sm_21, sm_30, sm_32, sm_35, sm_37, sm_50, sm_52, sm_53, sm_60, sm_61, sm_62, sm_70, sm_72, sm_75, sm_80, sm_86, sm_87, sm_89, sm_90, gfx600, gfx601, gfx602, gfx700, gfx701, gfx702, gfx703, gfx704, gfx705, gfx801, gfx802, gfx803, gfx805, gfx810, gfx900, gfx902, gfx904, gfx906, gfx908, gfx909, gfx90a, gfx90c, gfx940, gfx941, gfx942, gfx1010, gfx1011, gfx1012, gfx1013, gfx1030, gfx1031, gfx1032, gfx1033, gfx1034, gfx1035, gfx1036, gfx1100, gfx1101, gfx1102, gfx1103, gfx1150, gfx1151, gfx1200, gfx1201{{$}}
// RUN: not %clang_cc1 -triple r600--- -target-cpu not-a-cpu -fsyntax-only %s 2>&1 | FileCheck %s --check-prefix R600
// R600: error: unknown target CPU 'not-a-cpu'
@@ -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{{$}}
+// AMDGCN-NEXT: note: valid target CPU values are: gfx600, tahiti, gfx601, pitcairn, verde, gfx602, hainan, oland, gfx700, kaveri, gfx701, hawaii, gfx702, gfx703, kabini, mullins, gfx704, bonaire, gfx705, gfx801, carrizo, gfx802, iceland, tonga, gfx803, fiji, polaris10, polaris11, gfx805, tongapro, gfx810, stoney, gfx900, gfx902, gfx904, gfx906, gfx908, gfx909, gfx90a, gfx90c, gfx940, gfx941, gfx942, gfx1010, gfx1011, gfx1012, gfx1013, gfx1030, gfx1031, gfx1032, gfx1033, gfx1034, gfx1035, gfx1036, gfx1100, gfx1101, gfx1102, gfx1103, gfx1150, gfx1151, gfx1200, gfx1201{{$}}
// 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 3148d4bebb96b5f..2d58ea86087e926 100644
--- a/llvm/docs/AMDGPUUsage.rst
+++ b/llvm/docs/AMDGPUUsage.rst
@@ -504,6 +504,20 @@ Every processor supports every OS ABI (see :ref:`amdgpu-os`) with the following
work-item Add product
IDs names.
+ ``gfx1200`` ``amdgcn`` dGPU - cumode - Architected *TBA*
+ - wavefrontsize64 flat
+ scratch .. TODO::
+ - Packed
+ work-item Add product
+ IDs names.
+
+ ``gfx1201`` ``amdgcn`` dGPU - cumode - Architected *TBA*
+ - wavefrontsize64 flat
+ scratch .. TODO::
+ - Packed
+ work-item Add product
+ IDs names.
+
=========== =============== ============ ===== ================= =============== =============== ======================
.. _amdgpu-target-features:
@@ -1667,11 +1681,13 @@ The AMDGPU backend uses the following ELF header:
``EF_AMDGPU_MACH_AMDGCN_GFX1036`` 0x045 ``gfx1036``
``EF_AMDGPU_MACH_AMDGCN_GFX1101`` 0x046 ``gfx1101``
``EF_AMDGPU_MACH_AMDGCN_GFX1102`` 0x047 ``gfx1102``
- *reserved* 0x048 Reserved.
+ ``EF_AMDGPU_MACH_AMDGCN_GFX1200`` 0x048 ``gfx1200``
*reserved* 0x049 Reserved.
``EF_AMDGPU_MACH_AMDGCN_GFX1151`` 0x04a ``gfx1151``
``EF_AMDGPU_MACH_AMDGCN_GFX941`` 0x04b ``gfx941``
``EF_AMDGPU_MACH_AMDGCN_GFX942`` 0x04c ``gfx942``
+ *reserved* 0x04d Reserved.
+ ``EF_AMDGPU_MACH_AMDGCN_GFX1201`` 0x04e ``gfx1201``
==================================== ========== =============================
Sections
diff --git a/llvm/include/llvm/BinaryFormat/ELF.h b/llvm/include/llvm/BinaryFormat/ELF.h
index 3596174f74dde80..e2346b821033347 100644
--- a/llvm/include/llvm/BinaryFormat/ELF.h
+++ b/llvm/include/llvm/BinaryFormat/ELF.h
@@ -778,15 +778,17 @@ enum : unsigned {
EF_AMDGPU_MACH_AMDGCN_GFX1036 = 0x045,
EF_AMDGPU_MACH_AMDGCN_GFX1101 = 0x046,
EF_AMDGPU_MACH_AMDGCN_GFX1102 = 0x047,
- EF_AMDGPU_MACH_AMDGCN_RESERVED_0X48 = 0x048,
+ EF_AMDGPU_MACH_AMDGCN_GFX1200 = 0x048,
EF_AMDGPU_MACH_AMDGCN_RESERVED_0X49 = 0x049,
EF_AMDGPU_MACH_AMDGCN_GFX1151 = 0x04a,
EF_AMDGPU_MACH_AMDGCN_GFX941 = 0x04b,
EF_AMDGPU_MACH_AMDGCN_GFX942 = 0x04c,
+ EF_AMDGPU_MACH_AMDGCN_RESERVED_0X4D = 0x04d,
+ EF_AMDGPU_MACH_AMDGCN_GFX1201 = 0x04e,
// First/last AMDGCN-based processors.
EF_AMDGPU_MACH_AMDGCN_FIRST = EF_AMDGPU_MACH_AMDGCN_GFX600,
- EF_AMDGPU_MACH_AMDGCN_LAST = EF_AMDGPU_MACH_AMDGCN_GFX942,
+ EF_AMDGPU_MACH_AMDGCN_LAST = EF_AMDGPU_MACH_AMDGCN_GFX1201,
// 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 a40599c88f28eb0..6464285980f00a1 100644
--- a/llvm/include/llvm/TargetParser/TargetParser.h
+++ b/llvm/include/llvm/TargetParser/TargetParser.h
@@ -106,8 +106,11 @@ enum GPUKind : uint32_t {
GK_GFX1150 = 94,
GK_GFX1151 = 95,
+ GK_GFX1200 = 100,
+ GK_GFX1201 = 101,
+
GK_AMDGCN_FIRST = GK_GFX600,
- GK_AMDGCN_LAST = GK_GFX1151,
+ GK_AMDGCN_LAST = GK_GFX1201,
};
/// Instruction set architecture version.
diff --git a/llvm/lib/Object/ELFObjectFile.cpp b/llvm/lib/Object/ELFObjectFile.cpp
index 143f9d37849d238..25dbcbd68431f58 100644
--- a/llvm/lib/Object/ELFObjectFile.cpp
+++ b/llvm/lib/Object/ELFObjectFile.cpp
@@ -506,6 +506,12 @@ StringRef ELFObjectFileBase::getAMDGPUCPUName() const {
return "gfx1150";
case ELF::EF_AMDGPU_MACH_AMDGCN_GFX1151:
return "gfx1151";
+
+ // AMDGCN GFX12.
+ case ELF::EF_AMDGPU_MACH_AMDGCN_GFX1200:
+ return "gfx1200";
+ case ELF::EF_AMDGPU_MACH_AMDGCN_GFX1201:
+ return "gfx1201";
default:
llvm_unreachable("Unknown EF_AMDGPU_MACH value");
}
diff --git a/llvm/lib/ObjectYAML/ELFYAML.cpp b/llvm/lib/ObjectYAML/ELFYAML.cpp
index 872b89420a9e7a7..1da4ea4e3edc919 100644
--- a/llvm/lib/ObjectYAML/ELFYAML.cpp
+++ b/llvm/lib/ObjectYAML/ELFYAML.cpp
@@ -610,6 +610,8 @@ void ScalarBitSetTraits<ELFYAML::ELF_EF>::bitset(IO &IO,
BCaseMask(EF_AMDGPU_MACH_AMDGCN_GFX1103, EF_AMDGPU_MACH);
BCaseMask(EF_AMDGPU_MACH_AMDGCN_GFX1150, EF_AMDGPU_MACH);
BCaseMask(EF_AMDGPU_MACH_AMDGCN_GFX1151, EF_AMDGPU_MACH);
+ BCaseMask(EF_AMDGPU_MACH_AMDGCN_GFX1200, EF_AMDGPU_MACH);
+ BCaseMask(EF_AMDGPU_MACH_AMDGCN_GFX1201, 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 bf244bb024a7240..f7315ecb9fa6495 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPU.td
+++ b/llvm/lib/Target/AMDGPU/AMDGPU.td
@@ -345,6 +345,12 @@ def FeatureGFX11Insts : SubtargetFeature<"gfx11-insts",
"Additional instructions for GFX11+"
>;
+def FeatureGFX12Insts : SubtargetFeature<"gfx12-insts",
+ "GFX12Insts",
+ "true",
+ "Additional instructions for GFX12+"
+>;
+
def FeatureGFX10_3Insts : SubtargetFeature<"gfx10-3-insts",
"GFX10_3Insts",
"true",
@@ -1061,6 +1067,26 @@ def FeatureGFX11 : GCNSubtargetFeatureGeneration<"GFX11",
]
>;
+def FeatureGFX12 : GCNSubtargetFeatureGeneration<"GFX12",
+ "gfx12",
+ [FeatureFP64, FeatureLocalMemorySize65536, FeatureMIMG_R128,
+ FeatureFlatAddressSpace, Feature16BitInsts,
+ FeatureInv2PiInlineImm, FeatureApertureRegs,
+ FeatureCIInsts, FeatureGFX8Insts, FeatureGFX9Insts, FeatureGFX10Insts,
+ FeatureGFX10_AEncoding, FeatureGFX10_BEncoding, FeatureGFX10_3Insts,
+ FeatureGFX11Insts, FeatureGFX12Insts, FeatureVOP3P, FeatureVOPD,
+ FeatureMovrel, FeatureFastFMAF32, FeatureDPP, FeatureIntClamp,
+ FeatureFlatInstOffsets, FeatureFlatGlobalInsts, FeatureFlatScratchInsts,
+ FeatureAddNoCarryInsts, FeatureFmaMixInsts,
+ FeatureNoSdstCMPX, FeatureVscnt,
+ FeatureVOP3Literal, FeatureDPP8,
+ FeatureNoDataDepHazard, FeaturePkFmacF16Inst,
+ FeatureA16, FeatureFastDenormalF32, FeatureG16,
+ FeatureUnalignedBufferAccess, FeatureUnalignedDSAccess, FeatureGDS,
+ FeatureGWS, FeatureTrue16BitInsts
+ ]
+>;
+
//===----------------------------------------------------------------------===//
class FeatureSet<list<SubtargetFeature> Features_> {
@@ -1413,6 +1439,29 @@ def FeatureISAVersion11_5_1 : FeatureSet<
FeatureVGPRSingleUseHintInsts,
FeatureGFX11FullVGPRs])>;
+def FeatureISAVersion12 : FeatureSet<
+ [FeatureGFX12,
+ FeatureLDSBankCount32,
+ FeatureDLInsts,
+ FeatureDot5Insts,
+ FeatureDot7Insts,
+ FeatureDot8Insts,
+ FeatureDot9Insts,
+ FeatureDot10Insts,
+ FeatureNSAEncoding,
+ FeaturePartialNSAEncoding,
+ FeatureWavefrontSize32,
+ FeatureShaderCyclesRegister,
+ FeatureArchitectedFlatScratch,
+ FeatureAtomicFaddRtnInsts,
+ FeatureAtomicFaddNoRtnInsts,
+ FeatureFlatAtomicFaddF32Inst,
+ FeatureImageInsts,
+ FeatureExtendedImageInsts,
+ FeaturePackedTID,
+ FeatureVcmpxPermlaneHazard,
+ FeatureMADIntraFwdBug]>;
+
//===----------------------------------------------------------------------===//
def AMDGPUInstrInfo : InstrInfo {
@@ -1567,6 +1616,10 @@ def isGFX6GFX7GFX8GFX9GFX10 :
"Subtarget->getGeneration() == AMDGPUSubtarget::GFX10">,
AssemblerPredicate<(all_of (not FeatureGFX11Insts))>;
+def isNotGFX12Plus :
+ Predicate<"Subtarget->getGeneration() <= AMDGPUSubtarget::GFX11">,
+ AssemblerPredicate<(all_of (not FeatureGFX12Insts))>;
+
def isGFX7GFX8GFX9GFX10 :
Predicate<"Subtarget->getGeneration() == AMDGPUSubtarget::SEA_ISLANDS ||"
"Subtarget->getGeneration() == AMDGPUSubtarget::VOLCANIC_ISLANDS ||"
@@ -1574,6 +1627,13 @@ def isGFX7GFX8GFX9GFX10 :
"Subtarget->getGeneration() == AMDGPUSubtarget::GFX10">,
AssemblerPredicate<(all_of FeatureCIInsts, (not FeatureGFX11Insts))>;
+def isGFX8GFX9GFX10GFX11 :
+ Predicate<"Subtarget->getGeneration() == AMDGPUSubtarget::VOLCANIC_ISLANDS ||"
+ "Subtarget->getGeneration() == AMDGPUSubtarget::GFX9 ||"
+ "Subtarget->getGeneration() == AMDGPUSubtarget::GFX10 ||"
+ "Subtarget->getGeneration() == AMDGPUSubtarget::GFX11">,
+ AssemblerPredicate<(all_of FeatureGFX8Insts, (not FeatureGFX12Insts))>;
+
def isGFX7Plus :
Predicate<"Subtarget->getGeneration() >= AMDGPUSubtarget::SEA_ISLANDS">,
AssemblerPredicate<(all_of FeatureCIInsts)>;
@@ -1646,6 +1706,11 @@ def isGFX10Plus :
Predicate<"Subtarget->getGeneration() >= AMDGPUSubtarget::GFX10">,
AssemblerPredicate<(all_of FeatureGFX10Insts)>;
+def isGFX10GFX11 :
+ Predicate<"Subtarget->getGeneration() == AMDGPUSubtarget::GFX10 ||"
+ "Subtarget->getGeneration() == AMDGPUSubtarget::GFX11">,
+ AssemblerPredicate<(all_of FeatureGFX10Insts, (not FeatureGFX12Insts))>;
+
def isGFX10Before1030 :
Predicate<"Subtarget->getGeneration() == AMDGPUSubtarget::GFX10 &&"
"!Subtarget->hasGFX10_3Insts()">,
@@ -1664,12 +1729,20 @@ def isGFX8GFX9GFX10 :
def isGFX11Only :
Predicate<"Subtarget->getGeneration() == AMDGPUSubtarget::GFX11">,
- AssemblerPredicate<(all_of FeatureGFX11Insts)>;
+ AssemblerPredicate<(all_of FeatureGFX11Insts, (not FeatureGFX12Insts))>;
def isGFX11Plus :
Predicate<"Subtarget->getGeneration() >= AMDGPUSubtarget::GFX11">,
AssemblerPredicate<(all_of FeatureGFX11Insts)>;
+def isGFX12Only :
+ Predicate<"Subtarget->getGeneration() == AMDGPUSubtarget::GFX12">,
+ AssemblerPredicate<(all_of FeatureGFX12Insts)>;
+
+def isGFX12Plus :
+ Predicate<"Subtarget->getGeneration() >= AMDGPUSubtarget::GFX12">,
+ AssemblerPredicate<(all_of FeatureGFX12Insts)>;
+
def HasFlatAddressSpace : Predicate<"Subtarget->hasFlatAddressSpace()">,
AssemblerPredicate<(all_of FeatureFlatAddressSpace)>;
diff --git a/llvm/lib/Target/AMDGPU/AMDGPUSubtarget.h b/llvm/lib/Target/AMDGPU/AMDGPUSubtarget.h
index 71acce809a1494f..b72697973be7a11 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPUSubtarget.h
+++ b/llvm/lib/Target/AMDGPU/AMDGPUSubtarget.h
@@ -39,7 +39,8 @@ class AMDGPUSubtarget {
VOLCANIC_ISLANDS = 7,
GFX9 = 8,
GFX10 = 9,
- GFX11 = 10
+ GFX11 = 10,
+ GFX12 = 11,
};
private:
diff --git a/llvm/lib/Target/AMDGPU/GCNProcessors.td b/llvm/lib/Target/AMDGPU/GCNProcessors.td
index b4ae3a7a08e4e8d..80669c04f2c6886 100644
--- a/llvm/lib/Target/AMDGPU/GCNProcessors.td
+++ b/llvm/lib/Target/AMDGPU/GCNProcessors.td
@@ -279,3 +279,15 @@ def : ProcessorModel<"gfx1150", GFX11SpeedModel,
def : ProcessorModel<"gfx1151", GFX11SpeedModel,
FeatureISAVersion11_5_1.Features
>;
+
+//===----------------------------------------------------------------------===//
+// GCN GFX12.
+//===----------------------------------------------------------------------===//
+
+def : ProcessorModel<"gfx1200", GFX11SpeedModel,
+ FeatureISAVersion12.Features
+>;
+
+def : ProcessorModel<"gfx1201", GFX11SpeedModel,
+ FeatureISAVersion12.Features
+>;
diff --git a/llvm/lib/Target/AMDGPU/GCNSubtarget.h b/llvm/lib/Target/AMDGPU/GCNSubtarget.h
index 4c624ed59427c95..0074d1632161d18 100644
--- a/llvm/lib/Target/AMDGPU/GCNSubtarget.h
+++ b/llvm/lib/Target/AMDGPU/GCNSubtarget.h
@@ -107,6 +107,7 @@ class GCNSubtarget final : public AMDGPUGenSubtargetInfo,
bool GFX940Insts = false;
bool GFX10Insts = false;
bool GFX11Insts = false;
+ bool GFX12Insts = false;
bool GFX10_3Insts = false;
bool GFX7GFX8GFX9Insts = false;
bool SGPRInitBug = false;
diff --git a/llvm/lib/Target/AMDGPU/MCTargetDesc/AMDGPUTargetStreamer.cpp b/llvm/lib/Target/AMDGPU/MCTargetDesc/AMDGPUTargetStreamer.cpp
index e635a9182fc08a1..7523eb13afc7b77 100644
--- a/llvm/lib/Target/AMDGPU/MCTargetDesc/AMDGPUTargetStreamer.cpp
+++ b/llvm/lib/Target/AMDGPU/MCTargetDesc/AMDGPUTargetStreamer.cpp
@@ -125,6 +125,8 @@ StringRef AMDGPUTargetStreamer::getArchNameFromElfMach(unsigned ElfMach) {
case ELF::EF_AMDGPU_MACH_AMDGCN_GFX1103: AK = GK_GFX1103; break;
case ELF::EF_AMDGPU_MACH_AMDGCN_GFX1150: AK = GK_GFX1150; break;
case ELF::EF_AMDGPU_MACH_AMDGCN_GFX1151: AK = GK_GFX1151; break;
+ case ELF::EF_AMDGPU_MACH_AMDGCN_GFX1200: AK = GK_GFX1200; break;
+ case ELF::EF_AMDGPU_MACH_AMDGCN_GFX1201: AK = GK_GFX1201; break;
case ELF::EF_AMDGPU_MACH_NONE: AK = GK_NONE; break;
default: AK = GK_NONE; break;
}
@@ -199,6 +201,8 @@ unsigned AMDGPUTargetStreamer::getElfMach(StringRef GPU) {
case GK_GFX1103: return ELF::EF_AMDGPU_MACH_AMDGCN_GFX1103;
case GK_GFX1150: return ELF::EF_AMDGPU_MACH_AMDGCN_GFX1150;
case GK_GFX1151: return ELF::EF_AMDGPU_MACH_AMDGCN_GFX1151;
+ case GK_GFX1200: return ELF::EF_AMDGPU_MACH_AMDGCN_GFX1200;
+ case GK_GFX1201: return ELF::EF_AMDGPU_MACH_AMDGCN_GFX1201;
case GK_NONE: return ELF::EF_AMDGPU_MACH_NONE;
}
diff --git a/llvm/lib/Target/AMDGPU/SIDefines.h b/llvm/lib/Target/AMDGPU/SIDefines.h
index a5a46d63f4a1010..a3c409ee5bad12c 100644
--- a/llvm/lib/Target/AMDGPU/SIDefines.h
+++ b/llvm/lib/Target/AMDGPU/SIDefines.h
@@ -44,6 +44,7 @@ enum {
GFX90A = 8,
GFX940 = 9,
GFX11 = 10,
+ GFX12 = 11,
};
}
diff --git a/llvm/lib/Target/AMDGPU/SIInstrInfo.cpp b/llvm/lib/Target/AMDGPU/SIInstrInfo.cpp
index c4baabcd9232b56..a8c5e879cb75332 100644
--- a/llvm/lib/Target/AMDGPU/SIInstrInfo.cpp
+++ b/llvm/lib/Target/AMDGPU/SIInstrInfo.cpp
@@ -8749,6 +8749,8 @@ static unsigned subtargetEncodingFamily(const GCNSubtarget &ST) {
return SIEncodingFamily::GFX10;
case AMDGPUSubtarget::GFX11:
return SIEncodingFamily::GFX11;
+ case AMDGPUSubtarget::GFX12:
+ return SIEncodingFamily::GFX12;
}
llvm_unreachable("Unknown subtarget generation!");
}
@@ -8808,6 +8810,12 @@ int SIInstrInfo::pseudoToMCOpcode(int Opcode) const {
int MCOp = AMDGPU::getMCOpcode(Opcode, Gen);
+ // TODO-GFX12: Remove this.
+ // Hack to allow some GFX12 codegen tests to run before all the encodings are
+ // implemented.
+ if (MCOp == (uint16_t)-1 && Gen == SIEncodingFamily::GFX12)
+ MCOp = AMDGPU::getMCOpcode(Opcode, SIEncodingFamily::GFX11);
+
// -1 means that Opcode is already a native instruction.
if (MCOp == -1)
return Opcode;
diff --git a/llvm/lib/Target/AMDGPU/SIInstrInfo.td b/llvm/lib/Target/AMDGPU/SIInstrInfo.td
index 02c769bf21ac3ea..79b2d8e47712c91 100644
--- a/llvm/lib/Target/AMDGPU/SIInstrInfo.td
+++ b/llvm/lib/Target/AMDGPU/SIInstrInfo.td
@@ -32,6 +32,7 @@ def SIEncodingFamily {
int GFX90A = 8;
int GFX940 = 9;
int GFX11 = 10;
+ int GFX12 = 11;
}
//===----------------------------------------------------------------------===//
@@ -2756,7 +2757,8 @@ def getMCOpcodeGen : InstrMapping {
[!cast<string>(SIEncodingFamily.SDWA10)],
[!cast<string>(SIEncodingFamily.GFX90A)],
[!cast<string>(SIEncodingFamily.GFX940)],
- [!cast<string>(SIEncodingFamily.GFX11)]];
+ [!cast<string>(SIEncodingFamily.GFX11)],
+ [!cast<string>(SIEncodingFamily.GFX12)]];
}
// Get equivalent SOPK instruction.
diff --git a/llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.cpp b/llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.cpp
index fdc59281c50d0b3..8782c3875a2f158 100644
--- a/llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.cpp
+++ b/llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.cpp
@@ -1988,6 +1988,10 @@ bool isGFX9_GFX10(const MCSubtargetInfo &STI) {
return isGFX9(STI) || isGFX10(STI);
}
+bool isGFX9_GFX10_GFX11(const MCSubtargetInfo &STI) {
+ return isGFX9(STI) || isGFX10(STI) || isGFX11(STI);
+}
+
bool isGFX8_GFX9_GFX10(const MCSubtargetInfo &STI) {
return isVI(STI) || isGFX9(STI) || isGFX10(STI);
}
@@ -2004,6 +2008,10 @@ bool isGFX10(const MCSubtargetInfo &STI) {
return STI.hasFeature(AMDGPU::FeatureGFX10);
}
+bool isGFX10_GFX11(const MCSubtargetInfo &STI) {
+ return isGFX10(STI) || isGFX11(STI);
+}
+
bool isGFX10Plus(const MCSubtargetInfo &STI) {
return isGFX10(STI) || isGFX11Plus(STI);
}
@@ -2013,7 +2021,19 @@ bool isGFX11(const MCSubtargetInfo &STI) {
}
bool isGFX11Plus(const MCSubtargetInfo &STI) {
- return isGFX11(STI);
+ return isGFX11(STI) || isGFX12Plus(STI);
+}
+
+bool isGFX12(const MCSubtargetInfo &STI) {
+ return STI.getFeatureBits()[AMDGPU::FeatureGFX12];
+}
+
+bool isGFX12Plus(const MCSubtargetInfo &STI) {
+ return isGFX12(STI);
+}
+
+bool isNotGFX12Plus(const MCSubtargetInfo &STI) {
+ return !isGFX12Plus(STI);
}
bool isNotGFX11Plus(const MCSubtargetInfo &STI) {
@@ -2044,6 +2064,10 @@ bool hasGFX10_3Insts(const MCSubtargetInfo &STI) {
return STI.hasFeature(AMDGPU::FeatureGFX10_3Insts);
}
+bool isGFX10_3_GFX11(const MCSubtargetInfo &STI) {
+ return isGFX10_BEncoding(STI) && !isGFX12Plus(STI);
+}
+
bool isGFX90A(const MCSubtargetInfo &STI) {
return STI.hasFeature(AMDGPU::FeatureGFX90AInsts);
}
diff --git a/llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.h b/llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.h
index 1e0994d0862cf5d..b60c0fcd361199b 100644
--- a/llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.h
+++ b/llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.h
@@ -1143,20 +1143,26 @@ bool isCI(const MCSubtargetInfo &STI);
bool isVI(const MCSubtargetInfo &STI);
bool isGFX9(const MCSubtargetInfo &STI);
bool isGFX9_GFX10(const MCSubtargetInfo &STI);
+bool isGFX9_GFX10_GFX11(const MCSubtargetInfo &STI);
bool isGFX8_GFX9_GFX10(const MCSubtargetInfo &STI);
bool isGFX8Plus(const MCSubtargetInfo &STI);
bool isGFX9Plus(const MCSubtargetInfo &STI);
bool isGFX10(const MCSubtargetInfo &STI);
+bool isGFX10_GFX11(const MCSubtargetInfo &STI);
bool isGFX10Plus(const MCSubtargetInfo &STI);
bool isNotGFX10Plus(const MCSubtargetInfo &STI);
bool isGFX10Before1030(const MCSubtargetInfo &STI);
bool isGFX11(const MCSubtargetInfo &STI);
bool isGFX11Plus(const MCSubtargetInfo &STI);
+bool isGFX12(const MCSubtargetInfo &STI);
+bool isGFX12Plus(const MCSubtargetInfo &STI);
+bool isNotGFX12Plus(const MCSubtargetInfo &STI);
bool isNotGFX11Plus(const MCSubtargetInfo &STI);
bool isGCN3Encoding(const MCSubtargetInfo &STI);
bool isGFX10_AEncoding(const MCSubtargetInfo &STI);
bool isGFX10_BEncoding(const MCSubtargetInfo &STI);
bool hasGFX10_3Insts(const MCSubtargetInfo &STI);
+bool isGFX10_3_GFX11(const MCSubtargetInfo &STI);
bool isGFX90A(const MCSubtargetInfo &STI);
bool isGFX940(const MCSubtargetInfo &STI);
bool hasArchitectedFlatScratch(const MCSubtargetInfo &STI);
diff --git a/llvm/lib/TargetParser/TargetParser.cpp b/llvm/lib/TargetParser/TargetParser.cpp
index 8ab48825d1b96a3..4dbc77ad2fc1cfa 100644
--- a/llvm/lib/TargetParser/TargetParser.cpp
+++ b/llvm/lib/TargetParser/TargetParser.cpp
@@ -123,6 +123,8 @@ constexpr GPUInfo AMDGCNGPUs[] = {
{{"gfx1103"}, {"gfx1103"}, GK_GFX1103, FEATURE_FAST_FMA_F32|FEATURE_FAST_DENORMAL_F32|FEATURE_WAVE32|FEATURE_WGP},
{{"gfx1150"}, {"gfx1150"}, GK_GFX1150, FEATURE_FAST_FMA_F32|FEATURE_FAST_DENORMAL_F32|FEATURE_WAVE32|FEATURE_WGP},
{{"gfx1151"}, {"gfx1151"}, GK_GFX1151, FEATURE_FAST_FMA_F32|FEATURE_FAST_DENORMAL_F32|FEATURE_WAVE32|FEATURE_WGP},
+ {{"gfx1200"}, {"gfx1200"}, GK_GFX1200, FEATURE_FAST_FMA_F32|FEATURE_FAST_DENORMAL_F32|FEATURE_WAVE32|FEATURE_WGP},
+ {{"gfx1201"}, {"gfx1201"}, GK_GFX1201, FEATURE_FAST_FMA_F32|FEATURE_FAST_DENORMAL_F32|FEATURE_WAVE32|FEATURE_WGP},
};
const GPUInfo *getArchEntry(AMDGPU::GPUKind AK, ArrayRef<GPUInfo> Table) {
@@ -246,6 +248,8 @@ AMDGPU::IsaVersion AMDGPU::getIsaVersion(StringRef GPU) {
case GK_GFX1103: return {11, 0, 3};
case GK_GFX1150: return {11, 5, 0};
case GK_GFX1151: return {11, 5, 1};
+ case GK_GFX1200: return {12, 0, 0};
+ case GK_GFX1201: return {12, 0, 1};
default: return {0, 0, 0};
}
}
@@ -264,6 +268,27 @@ void AMDGPU::fillAMDGPUFeatureMap(StringRef GPU, const Triple &T,
// XXX - What does the member GPU mean if device name string passed here?
if (T.isAMDGCN()) {
switch (parseArchAMDGCN(GPU)) {
+ case GK_GFX1201:
+ case GK_GFX1200:
+ Features["ci-insts"] = true;
+ Features["dot5-insts"] = true;
+ Features["dot7-insts"] = true;
+ Features["dot8-insts"] = true;
+ Features["dot9-insts"] = true;
+ Features["dot10-insts"] = true;
+ Features["dl-insts"] = true;
+ Features["16-bit-insts"] = true;
+ Features["dpp"] = true;
+ Features["gfx8-insts"] = true;
+ Features["gfx9-insts"] = true;
+ Features["gfx10-insts"] = true;
+ Features["gfx10-3-insts"] = true;
+ Features["gfx11-insts"] = true;
+ Features["gfx12-insts"] = true;
+ Features["atomic-fadd-rtn-insts"] = true;
+ Features["image-insts"] = true;
+ Features["gws"] = true;
+ break;
case GK_GFX1151:
case GK_GFX1150:
case GK_GFX1103:
@@ -458,6 +483,8 @@ static bool isWave32Capable(StringRef GPU, const Triple &T) {
// XXX - What does the member GPU mean if device name string passed here?
if (T.isAMDGCN()) {
switch (parseArchAMDGCN(GPU)) {
+ case GK_GFX1201:
+ case GK_GFX1200:
case GK_GFX1151:
case GK_GFX1150:
case GK_GFX1103:
diff --git a/llvm/test/CodeGen/AMDGPU/directive-amdgcn-target.ll b/llvm/test/CodeGen/AMDGPU/directive-amdgcn-target.ll
index 5c84f3298616e33..357fcf8ef156178 100644
--- a/llvm/test/CodeGen/AMDGPU/directive-amdgcn-target.ll
+++ b/llvm/test/CodeGen/AMDGPU/directive-amdgcn-target.ll
@@ -105,6 +105,8 @@
; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1103 < %s | FileCheck --check-prefixes=GFX1103 %s
; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1150 < %s | FileCheck --check-prefixes=GFX1150 %s
; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1151 < %s | FileCheck --check-prefixes=GFX1151 %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1200 < %s | FileCheck --check-prefixes=GFX1200 %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1201 < %s | FileCheck --check-prefixes=GFX1201 %s
; GFX600: .amdgcn_target "amdgcn-amd-amdhsa--gfx600"
; GFX601: .amdgcn_target "amdgcn-amd-amdhsa--gfx601"
@@ -191,6 +193,8 @@
; GFX1103: .amdgcn_target "amdgcn-amd-amdhsa--gfx1103"
; GFX1150: .amdgcn_target "amdgcn-amd-amdhsa--gfx1150"
; GFX1151: .amdgcn_target "amdgcn-amd-amdhsa--gfx1151"
+; GFX1200: .amdgcn_target "amdgcn-amd-amdhsa--gfx1200"
+; GFX1201: .amdgcn_target "amdgcn-amd-amdhsa--gfx1201"
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 64f4ad5c72990d4..380439d8cd9c657 100644
--- a/llvm/test/CodeGen/AMDGPU/elf-header-flags-mach.ll
+++ b/llvm/test/CodeGen/AMDGPU/elf-header-flags-mach.ll
@@ -74,6 +74,8 @@
; RUN: llc -filetype=obj -mtriple=amdgcn -mcpu=gfx1103 < %s | llvm-readobj --file-header - | FileCheck --check-prefixes=ALL,ARCH-GCN,GFX1103 %s
; RUN: llc -filetype=obj -mtriple=amdgcn -mcpu=gfx1150 < %s | llvm-readobj --file-header - | FileCheck --check-prefixes=ALL,ARCH-GCN,GFX1150 %s
; RUN: llc -filetype=obj -mtriple=amdgcn -mcpu=gfx1151 < %s | llvm-readobj --file-header - | FileCheck --check-prefixes=ALL,ARCH-GCN,GFX1151 %s
+; RUN: llc -filetype=obj -mtriple=amdgcn -mcpu=gfx1200 < %s | llvm-readobj --file-header - | FileCheck --check-prefixes=ALL,ARCH-GCN,GFX1200 %s
+; RUN: llc -filetype=obj -mtriple=amdgcn -mcpu=gfx1201 < %s | llvm-readobj --file-header - | FileCheck --check-prefixes=ALL,ARCH-GCN,GFX1201 %s
; FIXME: With the default attributes the eflags are not accurate for
; xnack and sramecc. Subsequent Target-ID patches will address this.
@@ -145,6 +147,8 @@
; GFX1103: EF_AMDGPU_MACH_AMDGCN_GFX1103 (0x44)
; GFX1150: EF_AMDGPU_MACH_AMDGCN_GFX1150 (0x43)
; GFX1151: EF_AMDGPU_MACH_AMDGCN_GFX1151 (0x4A)
+; GFX1200: EF_AMDGPU_MACH_AMDGCN_GFX1200 (0x48)
+; GFX1201: EF_AMDGPU_MACH_AMDGCN_GFX1201 (0x4E)
; ALL: ]
define amdgpu_kernel void @elf_header() {
diff --git a/llvm/test/Object/AMDGPU/elf-header-flags-mach.yaml b/llvm/test/Object/AMDGPU/elf-header-flags-mach.yaml
index 7c6d4b588f28228..7fb33ca662b1906 100644
--- a/llvm/test/Object/AMDGPU/elf-header-flags-mach.yaml
+++ b/llvm/test/Object/AMDGPU/elf-header-flags-mach.yaml
@@ -230,6 +230,14 @@
# RUN: llvm-readobj -S --file-headers %t.o.AMDGCN_GFX1151 | FileCheck --check-prefixes=ELF-AMDGCN-ALL,ELF-AMDGCN-GFX1151 %s
# RUN: obj2yaml %t.o.AMDGCN_GFX1151 | FileCheck --check-prefixes=YAML-AMDGCN-ALL,YAML-AMDGCN-GFX1151 %s
+# RUN: sed -e 's/<BITS>/64/' -e 's/<MACH>/AMDGCN_GFX1200/' %s | yaml2obj -o %t.o.AMDGCN_GFX1200
+# RUN: llvm-readobj -S --file-headers %t.o.AMDGCN_GFX1200 | FileCheck --check-prefixes=ELF-AMDGCN-ALL,ELF-AMDGCN-GFX1200 %s
+# RUN: obj2yaml %t.o.AMDGCN_GFX1200 | FileCheck --check-prefixes=YAML-AMDGCN-ALL,YAML-AMDGCN-GFX1200 %s
+
+# RUN: sed -e 's/<BITS>/64/' -e 's/<MACH>/AMDGCN_GFX1201/' %s | yaml2obj -o %t.o.AMDGCN_GFX1201
+# RUN: llvm-readobj -S --file-headers %t.o.AMDGCN_GFX1201 | FileCheck --check-prefixes=ELF-AMDGCN-ALL,ELF-AMDGCN-GFX1201 %s
+# RUN: obj2yaml %t.o.AMDGCN_GFX1201 | FileCheck --check-prefixes=YAML-AMDGCN-ALL,YAML-AMDGCN-GFX1201 %s
+
# ELF-R600-ALL: Format: elf32-amdgpu
# ELF-R600-ALL: Arch: r600
# ELF-R600-ALL: AddressSize: 32bit
@@ -421,6 +429,12 @@
# ELF-AMDGCN-GFX1151: EF_AMDGPU_MACH_AMDGCN_GFX1151 (0x4A)
# YAML-AMDGCN-GFX1151: Flags: [ EF_AMDGPU_MACH_AMDGCN_GFX1151 ]
+# ELF-AMDGCN-GFX1200: EF_AMDGPU_MACH_AMDGCN_GFX1200 (0x48)
+# YAML-AMDGCN-GFX1200: Flags: [ EF_AMDGPU_MACH_AMDGCN_GFX1200 ]
+
+# ELF-AMDGCN-GFX1201: EF_AMDGPU_MACH_AMDGCN_GFX1201 (0x4E)
+# YAML-AMDGCN-GFX1201: Flags: [ EF_AMDGPU_MACH_AMDGCN_GFX1201 ]
+
# 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 dc0c28f49ed037a..e296d7fb1fc8f2f 100644
--- a/llvm/test/tools/llvm-objdump/ELF/AMDGPU/subtarget.ll
+++ b/llvm/test/tools/llvm-objdump/ELF/AMDGPU/subtarget.ll
@@ -4,6 +4,18 @@ define amdgpu_kernel void @test_kernel() {
; Test subtarget detection. Disassembly is only supported for GFX8 and beyond.
;
+; ----------------------------------GFX12--------------------------------------
+;
+; 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
+; RUN: diff %t-specify.txt %t-detect.txt
+;
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1200 -filetype=obj -O0 -o %t.o %s
+; RUN: llvm-objdump -D --arch-name=amdgcn --mcpu=gfx1200 %t.o > %t-specify.txt
+; RUN: llvm-objdump -D %t.o > %t-detect.txt
+; RUN: diff %t-specify.txt %t-detect.txt
+
; ----------------------------------GFX11--------------------------------------
;
; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1151 -filetype=obj -O0 -o %t.o %s
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 b13b4e840201cc2..e2266d81d1a5972 100644
--- a/llvm/test/tools/llvm-readobj/ELF/amdgpu-elf-headers.test
+++ b/llvm/test/tools/llvm-readobj/ELF/amdgpu-elf-headers.test
@@ -409,6 +409,24 @@
# RUN: yaml2obj %s -o %t -DABI_VERSION=2 -DFLAG_NAME=EF_AMDGPU_MACH_AMDGCN_GFX1151
# 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_GFX1151 -DFLAG_VALUE=0x4A
+# RUN: yaml2obj %s -o %t -DABI_VERSION=0 -DFLAG_NAME=EF_AMDGPU_MACH_AMDGCN_GFX1200
+# RUN: llvm-readobj -h %t | FileCheck %s --check-prefixes=ALL,KNOWN-ABI-VERSION,SINGLE-FLAG --match-full-lines -DABI_VERSION=0 -DFILE=%t -DFLAG_NAME=EF_AMDGPU_MACH_AMDGCN_GFX1200 -DFLAG_VALUE=0x48
+
+# RUN: yaml2obj %s -o %t -DABI_VERSION=1 -DFLAG_NAME=EF_AMDGPU_MACH_AMDGCN_GFX1200
+# RUN: llvm-readobj -h %t | FileCheck %s --check-prefixes=ALL,KNOWN-ABI-VERSION,SINGLE-FLAG --match-full-lines -DABI_VERSION=1 -DFILE=%t -DFLAG_NAME=EF_AMDGPU_MACH_AMDGCN_GFX1200 -DFLAG_VALUE=0x48
+
+# RUN: yaml2obj %s -o %t -DABI_VERSION=2 -DFLAG_NAME=EF_AMDGPU_MACH_AMDGCN_GFX1200
+# 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_GFX1200 -DFLAG_VALUE=0x48
+
+# RUN: yaml2obj %s -o %t -DABI_VERSION=0 -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=0 -DFILE=%t -DFLAG_NAME=EF_AMDGPU_MACH_AMDGCN_GFX1201 -DFLAG_VALUE=0x4E
+
+# RUN: yaml2obj %s -o %t -DABI_VERSION=1 -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=1 -DFILE=%t -DFLAG_NAME=EF_AMDGPU_MACH_AMDGCN_GFX1201 -DFLAG_VALUE=0x4E
+
+# 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
+
--- !ELF
FileHeader:
Class: ELFCLASS64
diff --git a/llvm/tools/llvm-readobj/ELFDumper.cpp b/llvm/tools/llvm-readobj/ELFDumper.cpp
index 98cf20a3b97fcfb..ab73d0b76b3dd4e 100644
--- a/llvm/tools/llvm-readobj/ELFDumper.cpp
+++ b/llvm/tools/llvm-readobj/ELFDumper.cpp
@@ -1616,6 +1616,8 @@ const EnumEntry<unsigned> ElfHeaderAMDGPUFlagsABIVersion3[] = {
LLVM_READOBJ_ENUM_ENT(ELF, EF_AMDGPU_MACH_AMDGCN_GFX1103),
LLVM_READOBJ_ENUM_ENT(ELF, EF_AMDGPU_MACH_AMDGCN_GFX1150),
LLVM_READOBJ_ENUM_ENT(ELF, EF_AMDGPU_MACH_AMDGCN_GFX1151),
+ LLVM_READOBJ_ENUM_ENT(ELF, EF_AMDGPU_MACH_AMDGCN_GFX1200),
+ LLVM_READOBJ_ENUM_ENT(ELF, EF_AMDGPU_MACH_AMDGCN_GFX1201),
LLVM_READOBJ_ENUM_ENT(ELF, EF_AMDGPU_FEATURE_XNACK_V3),
LLVM_READOBJ_ENUM_ENT(ELF, EF_AMDGPU_FEATURE_SRAMECC_V3)
};
@@ -1680,6 +1682,8 @@ const EnumEntry<unsigned> ElfHeaderAMDGPUFlagsABIVersion4[] = {
LLVM_READOBJ_ENUM_ENT(ELF, EF_AMDGPU_MACH_AMDGCN_GFX1103),
LLVM_READOBJ_ENUM_ENT(ELF, EF_AMDGPU_MACH_AMDGCN_GFX1150),
LLVM_READOBJ_ENUM_ENT(ELF, EF_AMDGPU_MACH_AMDGCN_GFX1151),
+ LLVM_READOBJ_ENUM_ENT(ELF, EF_AMDGPU_MACH_AMDGCN_GFX1200),
+ LLVM_READOBJ_ENUM_ENT(ELF, EF_AMDGPU_MACH_AMDGCN_GFX1201),
LLVM_READOBJ_ENUM_ENT(ELF, EF_AMDGPU_FEATURE_XNACK_ANY_V4),
LLVM_READOBJ_ENUM_ENT(ELF, EF_AMDGPU_FEATURE_XNACK_OFF_V4),
LLVM_READOBJ_ENUM_ENT(ELF, EF_AMDGPU_FEATURE_XNACK_ON_V4),
>From a6975a735ef4259270f286c902867b9cf16cf505 Mon Sep 17 00:00:00 2001
From: Jay Foad <jay.foad at amd.com>
Date: Wed, 22 Nov 2023 15:41:15 +0000
Subject: [PATCH 2/2] clang-format
---
llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.cpp | 8 ++------
1 file changed, 2 insertions(+), 6 deletions(-)
diff --git a/llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.cpp b/llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.cpp
index 8782c3875a2f158..bc5cedf70362bbf 100644
--- a/llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.cpp
+++ b/llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.cpp
@@ -2028,13 +2028,9 @@ bool isGFX12(const MCSubtargetInfo &STI) {
return STI.getFeatureBits()[AMDGPU::FeatureGFX12];
}
-bool isGFX12Plus(const MCSubtargetInfo &STI) {
- return isGFX12(STI);
-}
+bool isGFX12Plus(const MCSubtargetInfo &STI) { return isGFX12(STI); }
-bool isNotGFX12Plus(const MCSubtargetInfo &STI) {
- return !isGFX12Plus(STI);
-}
+bool isNotGFX12Plus(const MCSubtargetInfo &STI) { return !isGFX12Plus(STI); }
bool isNotGFX11Plus(const MCSubtargetInfo &STI) {
return !isGFX11Plus(STI);
More information about the flang-commits
mailing list