[clang] [llvm] [AMDGPU] Define new targets gfx1200 and gfx1201 (PR #73133)
via cfe-commits
cfe-commits at lists.llvm.org
Wed Nov 22 07:10:00 PST 2023
llvmbot wrote:
<!--LLVM PR SUMMARY COMMENT-->
@llvm/pr-subscribers-clang-codegen
Author: Jay Foad (jayfoad)
<details>
<summary>Changes</summary>
Define target names and ELF numbers for new GFX12 targets gfx1200 and
gfx1201. For now they behave identically to GFX11.
---
Patch is 40.39 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/73133.diff
30 Files Affected:
- (modified) clang/include/clang/Basic/Cuda.h (+2)
- (modified) clang/lib/Basic/Cuda.cpp (+2)
- (modified) clang/lib/Basic/Targets/NVPTX.cpp (+2)
- (modified) clang/lib/CodeGen/CGOpenMPRuntimeGPU.cpp (+2)
- (modified) clang/test/CodeGenOpenCL/amdgpu-features.cl (+4)
- (modified) clang/test/Driver/amdgpu-macros.cl (+2)
- (modified) clang/test/Driver/amdgpu-mcpu.cl (+4)
- (modified) clang/test/Misc/target-invalid-cpu-note.c (+2-2)
- (modified) llvm/docs/AMDGPUUsage.rst (+17-1)
- (modified) llvm/include/llvm/BinaryFormat/ELF.h (+4-2)
- (modified) llvm/include/llvm/TargetParser/TargetParser.h (+4-1)
- (modified) llvm/lib/Object/ELFObjectFile.cpp (+6)
- (modified) llvm/lib/ObjectYAML/ELFYAML.cpp (+2)
- (modified) llvm/lib/Target/AMDGPU/AMDGPU.td (+74-1)
- (modified) llvm/lib/Target/AMDGPU/AMDGPUSubtarget.h (+2-1)
- (modified) llvm/lib/Target/AMDGPU/GCNProcessors.td (+12)
- (modified) llvm/lib/Target/AMDGPU/GCNSubtarget.h (+1)
- (modified) llvm/lib/Target/AMDGPU/MCTargetDesc/AMDGPUTargetStreamer.cpp (+4)
- (modified) llvm/lib/Target/AMDGPU/SIDefines.h (+1)
- (modified) llvm/lib/Target/AMDGPU/SIInstrInfo.cpp (+8)
- (modified) llvm/lib/Target/AMDGPU/SIInstrInfo.td (+3-1)
- (modified) llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.cpp (+25-1)
- (modified) llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.h (+6)
- (modified) llvm/lib/TargetParser/TargetParser.cpp (+27)
- (modified) llvm/test/CodeGen/AMDGPU/directive-amdgcn-target.ll (+4)
- (modified) llvm/test/CodeGen/AMDGPU/elf-header-flags-mach.ll (+4)
- (modified) llvm/test/Object/AMDGPU/elf-header-flags-mach.yaml (+14)
- (modified) llvm/test/tools/llvm-objdump/ELF/AMDGPU/subtarget.ll (+12)
- (modified) llvm/test/tools/llvm-readobj/ELF/amdgpu-elf-headers.test (+18)
- (modified) llvm/tools/llvm-readobj/ELFDumper.cpp (+4)
``````````diff
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() >= AMDGPUS...
[truncated]
``````````
</details>
https://github.com/llvm/llvm-project/pull/73133
More information about the cfe-commits
mailing list