[NVPTX] Add SM versions for 101 and 120 #124155
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This patch adds SM and PTX versions for SM 101, 120 and their arch-accelerated variants. All these are supported in cuda-12.8. sm120/a requires ptx8.7 and the rest require ptx8.6. Signed-off-by: Durgadoss R <durgadossr@nvidia.com>
|
@llvm/pr-subscribers-backend-nvptx Author: Durgadoss R (durga4github) ChangesThis patch adds SM and PTX versions for SM All these are supported in cuda-12.8. Full diff: https://github.com/llvm/llvm-project/pull/124155.diff 3 Files Affected:
diff --git a/llvm/lib/Target/NVPTX/NVPTX.td b/llvm/lib/Target/NVPTX/NVPTX.td
index 3ca8b4d294079c..5467ae011a2081 100644
--- a/llvm/lib/Target/NVPTX/NVPTX.td
+++ b/llvm/lib/Target/NVPTX/NVPTX.td
@@ -35,15 +35,18 @@ class FeaturePTX<int version>:
"Use PTX version " # version>;
foreach sm = [20, 21, 30, 32, 35, 37, 50, 52, 53,
- 60, 61, 62, 70, 72, 75, 80, 86, 87, 89, 90, 100] in
+ 60, 61, 62, 70, 72, 75, 80, 86, 87,
+ 89, 90, 100, 101, 120] in
def SM#sm: FeatureSM<""#sm, !mul(sm, 10)>;
def SM90a: FeatureSM<"90a", 901>;
def SM100a: FeatureSM<"100a", 1001>;
+def SM101a: FeatureSM<"101a", 1011>;
+def SM120a: FeatureSM<"120a", 1201>;
foreach version = [32, 40, 41, 42, 43, 50, 60, 61, 62, 63, 64, 65,
70, 71, 72, 73, 74, 75, 76, 77, 78,
- 80, 81, 82, 83, 84, 85, 86] in
+ 80, 81, 82, 83, 84, 85, 86, 87] in
def PTX#version: FeaturePTX<version>;
//===----------------------------------------------------------------------===//
@@ -76,6 +79,10 @@ def : Proc<"sm_90", [SM90, PTX78]>;
def : Proc<"sm_90a", [SM90a, PTX80]>;
def : Proc<"sm_100", [SM100, PTX86]>;
def : Proc<"sm_100a", [SM100a, PTX86]>;
+def : Proc<"sm_101", [SM101, PTX86]>;
+def : Proc<"sm_101a", [SM101a, PTX86]>;
+def : Proc<"sm_120", [SM120, PTX87]>;
+def : Proc<"sm_120a", [SM120a, PTX87]>;
def NVPTXInstrInfo : InstrInfo {
}
diff --git a/llvm/lib/Target/NVPTX/NVPTXInstrInfo.td b/llvm/lib/Target/NVPTX/NVPTXInstrInfo.td
index a076fde8ee7676..f17799c1300153 100644
--- a/llvm/lib/Target/NVPTX/NVPTXInstrInfo.td
+++ b/llvm/lib/Target/NVPTX/NVPTXInstrInfo.td
@@ -172,6 +172,9 @@ class hasSM<int version>: Predicate<"Subtarget->getSmVersion() >= " # version>;
// Explicit records for arch-accelerated SM versions
def hasSM90a : Predicate<"Subtarget->getFullSmVersion() == 901">;
+def hasSM100a : Predicate<"Subtarget->getFullSmVersion() == 1001">;
+def hasSM101a : Predicate<"Subtarget->getFullSmVersion() == 1011">;
+def hasSM120a : Predicate<"Subtarget->getFullSmVersion() == 1201">;
// non-sync shfl instructions are not available on sm_70+ in PTX6.4+
def hasSHFL : Predicate<"!(Subtarget->getSmVersion() >= 70"
diff --git a/llvm/test/CodeGen/NVPTX/sm-version.ll b/llvm/test/CodeGen/NVPTX/sm-version.ll
index 0e37d6e4b0d87f..ce9a1b1b161dce 100644
--- a/llvm/test/CodeGen/NVPTX/sm-version.ll
+++ b/llvm/test/CodeGen/NVPTX/sm-version.ll
@@ -16,6 +16,12 @@
; RUN: llc < %s -mtriple=nvptx -mcpu=sm_86 | FileCheck %s --check-prefix=SM86
; RUN: llc < %s -mtriple=nvptx -mcpu=sm_90 | FileCheck %s --check-prefix=SM90
; RUN: llc < %s -mtriple=nvptx -mcpu=sm_90a | FileCheck %s --check-prefix=SM90a
+; RUN: llc < %s -mtriple=nvptx -mcpu=sm_100 | FileCheck %s --check-prefix=SM100
+; RUN: llc < %s -mtriple=nvptx -mcpu=sm_100a | FileCheck %s --check-prefix=SM100a
+; RUN: llc < %s -mtriple=nvptx -mcpu=sm_101 | FileCheck %s --check-prefix=SM101
+; RUN: llc < %s -mtriple=nvptx -mcpu=sm_101a | FileCheck %s --check-prefix=SM101a
+; RUN: llc < %s -mtriple=nvptx -mcpu=sm_120 | FileCheck %s --check-prefix=SM120
+; RUN: llc < %s -mtriple=nvptx -mcpu=sm_120a | FileCheck %s --check-prefix=SM120a
; RUN: llc < %s -mtriple=nvptx64 -mcpu=sm_20 | FileCheck %s --check-prefix=SM20
; RUN: llc < %s -mtriple=nvptx64 -mcpu=sm_21 | FileCheck %s --check-prefix=SM21
@@ -35,6 +41,12 @@
; RUN: llc < %s -mtriple=nvptx64 -mcpu=sm_86 | FileCheck %s --check-prefix=SM86
; RUN: llc < %s -mtriple=nvptx64 -mcpu=sm_90 | FileCheck %s --check-prefix=SM90
; RUN: llc < %s -mtriple=nvptx64 -mcpu=sm_90a | FileCheck %s --check-prefix=SM90a
+; RUN: llc < %s -mtriple=nvptx64 -mcpu=sm_100 | FileCheck %s --check-prefix=SM100
+; RUN: llc < %s -mtriple=nvptx64 -mcpu=sm_100a | FileCheck %s --check-prefix=SM100a
+; RUN: llc < %s -mtriple=nvptx64 -mcpu=sm_101 | FileCheck %s --check-prefix=SM101
+; RUN: llc < %s -mtriple=nvptx64 -mcpu=sm_101a | FileCheck %s --check-prefix=SM101a
+; RUN: llc < %s -mtriple=nvptx64 -mcpu=sm_120 | FileCheck %s --check-prefix=SM120
+; RUN: llc < %s -mtriple=nvptx64 -mcpu=sm_120a | FileCheck %s --check-prefix=SM120a
; SM20: .version 3.2
; SM21: .version 3.2
@@ -54,6 +66,12 @@
; SM86: .version 7.1
; SM90: .version 7.8
; SM90a: .version 8.0
+; SM100: .version 8.6
+; SM100a: .version 8.6
+; SM101: .version 8.6
+; SM101a: .version 8.6
+; SM120: .version 8.7
+; SM120a: .version 8.7
; SM20: .target sm_20
; SM21: .target sm_21
@@ -73,3 +91,9 @@
; SM86: .target sm_86
; SM90: .target sm_90
; SM90a: .target sm_90a
+; SM100: .target sm_100
+; SM100a: .target sm_100a
+; SM101: .target sm_101
+; SM101a: .target sm_101a
+; SM120: .target sm_120
+; SM120a: .target sm_120a
|
Artem-B
approved these changes
Jan 23, 2025
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/50/builds/9352 Here is the relevant piece of the build log for the reference |
copybara-service bot
pushed a commit
to openxla/xla
that referenced
this pull request
Jan 28, 2025
Imported from GitHub PR #21822 Created `ShouldUsePtxExtension` helper for the extension suffix (this will also be used for sm120, etc). CUDA 12.8 was recently released, which supports PTX 8.7, but that is not supported by the integrated LLVM (support added in llvm/llvm-project#124155), so leaving the association with PTX 8.6 - this doesn't raise warnings during compilation. Copybara import of the project: -- 267cf74 by Sergey Kozub <skozub@nvidia.com>: Add support for SM100a architecture (Blackwell) Merging this change closes #21822 FUTURE_COPYBARA_INTEGRATE_REVIEW=#21822 from openxla:devel/sm100a 267cf74 PiperOrigin-RevId: 720655796
copybara-service bot
pushed a commit
to tensorflow/tensorflow
that referenced
this pull request
Jan 28, 2025
Imported from GitHub PR openxla/xla#21822 Created `ShouldUsePtxExtension` helper for the extension suffix (this will also be used for sm120, etc). CUDA 12.8 was recently released, which supports PTX 8.7, but that is not supported by the integrated LLVM (support added in llvm/llvm-project#124155), so leaving the association with PTX 8.6 - this doesn't raise warnings during compilation. Copybara import of the project: -- 267cf74a084c933e532a622da2485befdc47f8ce by Sergey Kozub <skozub@nvidia.com>: Add support for SM100a architecture (Blackwell) Merging this change closes #21822 FUTURE_COPYBARA_INTEGRATE_REVIEW=openxla/xla#21822 from openxla:devel/sm100a 267cf74a084c933e532a622da2485befdc47f8ce PiperOrigin-RevId: 720655796
copybara-service bot
pushed a commit
to openxla/xla
that referenced
this pull request
Jan 29, 2025
Imported from GitHub PR #21822 Created `ShouldUsePtxExtension` helper for the extension suffix (this will also be used for sm120, etc). CUDA 12.8 was recently released, which supports PTX 8.7, but that is not supported by the integrated LLVM (support added in llvm/llvm-project#124155), so leaving the association with PTX 8.6 - this doesn't raise warnings during compilation. Copybara import of the project: -- 267cf74 by Sergey Kozub <skozub@nvidia.com>: Add support for SM100a architecture (Blackwell) Merging this change closes #21822 FUTURE_COPYBARA_INTEGRATE_REVIEW=#21822 from openxla:devel/sm100a 267cf74 PiperOrigin-RevId: 720655796
copybara-service bot
pushed a commit
to tensorflow/tensorflow
that referenced
this pull request
Jan 29, 2025
Imported from GitHub PR openxla/xla#21822 Created `ShouldUsePtxExtension` helper for the extension suffix (this will also be used for sm120, etc). CUDA 12.8 was recently released, which supports PTX 8.7, but that is not supported by the integrated LLVM (support added in llvm/llvm-project#124155), so leaving the association with PTX 8.6 - this doesn't raise warnings during compilation. Copybara import of the project: -- 267cf74a084c933e532a622da2485befdc47f8ce by Sergey Kozub <skozub@nvidia.com>: Add support for SM100a architecture (Blackwell) Merging this change closes #21822 FUTURE_COPYBARA_INTEGRATE_REVIEW=openxla/xla#21822 from openxla:devel/sm100a 267cf74a084c933e532a622da2485befdc47f8ce PiperOrigin-RevId: 720655796
copybara-service bot
pushed a commit
to openxla/xla
that referenced
this pull request
Jan 29, 2025
Imported from GitHub PR #21822 Created `ShouldUsePtxExtension` helper for the extension suffix (this will also be used for sm120, etc). CUDA 12.8 was recently released, which supports PTX 8.7, but that is not supported by the integrated LLVM (support added in llvm/llvm-project#124155), so leaving the association with PTX 8.6 - this doesn't raise warnings during compilation. Copybara import of the project: -- 267cf74 by Sergey Kozub <skozub@nvidia.com>: Add support for SM100a architecture (Blackwell) Merging this change closes #21822 FUTURE_COPYBARA_INTEGRATE_REVIEW=#21822 from openxla:devel/sm100a 267cf74 PiperOrigin-RevId: 720655796
copybara-service bot
pushed a commit
to openxla/xla
that referenced
this pull request
Jan 29, 2025
Imported from GitHub PR #21822 Created `ShouldUsePtxExtension` helper for the extension suffix (this will also be used for sm120, etc). CUDA 12.8 was recently released, which supports PTX 8.7, but that is not supported by the integrated LLVM (support added in llvm/llvm-project#124155), so leaving the association with PTX 8.6 - this doesn't raise warnings during compilation. Copybara import of the project: -- 267cf74 by Sergey Kozub <skozub@nvidia.com>: Add support for SM100a architecture (Blackwell) Merging this change closes #21822 COPYBARA_INTEGRATE_REVIEW=#21822 from openxla:devel/sm100a 267cf74 PiperOrigin-RevId: 720806648
copybara-service bot
pushed a commit
to tensorflow/tensorflow
that referenced
this pull request
Jan 29, 2025
Imported from GitHub PR openxla/xla#21822 Created `ShouldUsePtxExtension` helper for the extension suffix (this will also be used for sm120, etc). CUDA 12.8 was recently released, which supports PTX 8.7, but that is not supported by the integrated LLVM (support added in llvm/llvm-project#124155), so leaving the association with PTX 8.6 - this doesn't raise warnings during compilation. Copybara import of the project: -- 267cf74a084c933e532a622da2485befdc47f8ce by Sergey Kozub <skozub@nvidia.com>: Add support for SM100a architecture (Blackwell) Merging this change closes #21822 PiperOrigin-RevId: 720806648
copybara-service bot
pushed a commit
to openxla/xla
that referenced
this pull request
Jan 29, 2025
…(Blackwell) Imported from GitHub PR #22029 In addition to SM120a, also add SM101a mentioned in the PTX 8.7 spec (https://docs.nvidia.com/cuda/parallel-thread-execution/#release-notes), which is a slight variation of SM100a. Bumping the max supported PTX version to 8.7, as the LLVM PR (llvm/llvm-project#124155) adding the support is now integrated to OpenXLA. Copybara import of the project: -- be59b7a by Sergey Kozub <skozub@nvidia.com>: [XLA:GPU] Add support for SM101a and SM120a architectures (Blackwell) Merging this change closes #22029 FUTURE_COPYBARA_INTEGRATE_REVIEW=#22029 from openxla:devel/sm120a be59b7a PiperOrigin-RevId: 721049239
copybara-service bot
pushed a commit
to openxla/xla
that referenced
this pull request
Jan 29, 2025
…(Blackwell) Imported from GitHub PR #22029 In addition to SM120a, also add SM101a mentioned in the PTX 8.7 spec (https://docs.nvidia.com/cuda/parallel-thread-execution/#release-notes), which is a slight variation of SM100a. Bumping the max supported PTX version to 8.7, as the LLVM PR (llvm/llvm-project#124155) adding the support is now integrated to OpenXLA. Copybara import of the project: -- be59b7a by Sergey Kozub <skozub@nvidia.com>: [XLA:GPU] Add support for SM101a and SM120a architectures (Blackwell) Merging this change closes #22029 FUTURE_COPYBARA_INTEGRATE_REVIEW=#22029 from openxla:devel/sm120a be59b7a PiperOrigin-RevId: 721049239
copybara-service bot
pushed a commit
to tensorflow/tensorflow
that referenced
this pull request
Jan 29, 2025
…(Blackwell) Imported from GitHub PR openxla/xla#22029 In addition to SM120a, also add SM101a mentioned in the PTX 8.7 spec (https://docs.nvidia.com/cuda/parallel-thread-execution/#release-notes), which is a slight variation of SM100a. Bumping the max supported PTX version to 8.7, as the LLVM PR (llvm/llvm-project#124155) adding the support is now integrated to OpenXLA. Copybara import of the project: -- be59b7a51721637d880207e7adb69a18c3a92bea by Sergey Kozub <skozub@nvidia.com>: [XLA:GPU] Add support for SM101a and SM120a architectures (Blackwell) Merging this change closes #22029 FUTURE_COPYBARA_INTEGRATE_REVIEW=openxla/xla#22029 from openxla:devel/sm120a be59b7a51721637d880207e7adb69a18c3a92bea PiperOrigin-RevId: 721049239
copybara-service bot
pushed a commit
to openxla/xla
that referenced
this pull request
Jan 29, 2025
…(Blackwell) Imported from GitHub PR #22029 In addition to SM120a, also add SM101a mentioned in the PTX 8.7 spec (https://docs.nvidia.com/cuda/parallel-thread-execution/#release-notes), which is a slight variation of SM100a. Bumping the max supported PTX version to 8.7, as the LLVM PR (llvm/llvm-project#124155) adding the support is now integrated to OpenXLA. Copybara import of the project: -- be59b7a by Sergey Kozub <skozub@nvidia.com>: [XLA:GPU] Add support for SM101a and SM120a architectures (Blackwell) Merging this change closes #22029 FUTURE_COPYBARA_INTEGRATE_REVIEW=#22029 from openxla:devel/sm120a be59b7a PiperOrigin-RevId: 721049239
copybara-service bot
pushed a commit
to tensorflow/tensorflow
that referenced
this pull request
Jan 29, 2025
…(Blackwell) Imported from GitHub PR openxla/xla#22029 In addition to SM120a, also add SM101a mentioned in the PTX 8.7 spec (https://docs.nvidia.com/cuda/parallel-thread-execution/#release-notes), which is a slight variation of SM100a. Bumping the max supported PTX version to 8.7, as the LLVM PR (llvm/llvm-project#124155) adding the support is now integrated to OpenXLA. Copybara import of the project: -- be59b7a51721637d880207e7adb69a18c3a92bea by Sergey Kozub <skozub@nvidia.com>: [XLA:GPU] Add support for SM101a and SM120a architectures (Blackwell) Merging this change closes #22029 FUTURE_COPYBARA_INTEGRATE_REVIEW=openxla/xla#22029 from openxla:devel/sm120a be59b7a51721637d880207e7adb69a18c3a92bea PiperOrigin-RevId: 721049239
copybara-service bot
pushed a commit
to openxla/xla
that referenced
this pull request
Jan 29, 2025
…(Blackwell) Imported from GitHub PR #22029 In addition to SM120a, also add SM101a mentioned in the PTX 8.7 spec (https://docs.nvidia.com/cuda/parallel-thread-execution/#release-notes), which is a slight variation of SM100a. Bumping the max supported PTX version to 8.7, as the LLVM PR (llvm/llvm-project#124155) adding the support is now integrated to OpenXLA. Copybara import of the project: -- be59b7a by Sergey Kozub <skozub@nvidia.com>: [XLA:GPU] Add support for SM101a and SM120a architectures (Blackwell) Merging this change closes #22029 COPYBARA_INTEGRATE_REVIEW=#22029 from openxla:devel/sm120a be59b7a PiperOrigin-RevId: 721088886
copybara-service bot
pushed a commit
to tensorflow/tensorflow
that referenced
this pull request
Jan 29, 2025
…(Blackwell) Imported from GitHub PR openxla/xla#22029 In addition to SM120a, also add SM101a mentioned in the PTX 8.7 spec (https://docs.nvidia.com/cuda/parallel-thread-execution/#release-notes), which is a slight variation of SM100a. Bumping the max supported PTX version to 8.7, as the LLVM PR (llvm/llvm-project#124155) adding the support is now integrated to OpenXLA. Copybara import of the project: -- be59b7a51721637d880207e7adb69a18c3a92bea by Sergey Kozub <skozub@nvidia.com>: [XLA:GPU] Add support for SM101a and SM120a architectures (Blackwell) Merging this change closes #22029 PiperOrigin-RevId: 721088886
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This patch adds SM and PTX versions for SM
101, 120 and their arch-accelerated variants.
All these are supported in cuda-12.8.
sm120/120a requires ptx8.7 and the rest require ptx8.6.