[clang] [llvm] [CUDA] Mark CUDA-12.5 as supported and introduce ptx 8.5. (PR #94113)

[via llvm-commits]

llvmbot wrote:


<!--LLVM PR SUMMARY COMMENT-->

@llvm/pr-subscribers-clang

Author: Andrey Portnoy (andportnoy)

<details>
<summary>Changes</summary>

This PR is based on https://github.com/llvm/llvm-project/pull/91516.

---
Full diff: https://github.com/llvm/llvm-project/pull/94113.diff


6 Files Affected:

- (modified) clang/docs/ReleaseNotes.rst (+1-1) 
- (modified) clang/include/clang/Basic/BuiltinsNVPTX.def (+4-1) 
- (modified) clang/include/clang/Basic/Cuda.h (+2-1) 
- (modified) clang/lib/Basic/Cuda.cpp (+1) 
- (modified) clang/lib/Driver/ToolChains/Cuda.cpp (+3) 
- (modified) llvm/lib/Target/NVPTX/NVPTX.td (+2-1) 


``````````diff
diff --git a/clang/docs/ReleaseNotes.rst b/clang/docs/ReleaseNotes.rst
index 0c700d23257bf..61269e6db93e5 100644
--- a/clang/docs/ReleaseNotes.rst
+++ b/clang/docs/ReleaseNotes.rst
@@ -934,7 +934,7 @@ CUDA/HIP Language Changes
 
 CUDA Support
 ^^^^^^^^^^^^
-- Clang now supports CUDA SDK up to 12.4
+- Clang now supports CUDA SDK up to 12.5
 
 AIX Support
 ^^^^^^^^^^^
diff --git a/clang/include/clang/Basic/BuiltinsNVPTX.def b/clang/include/clang/Basic/BuiltinsNVPTX.def
index 9e243d740ed7a..504314d8d96e9 100644
--- a/clang/include/clang/Basic/BuiltinsNVPTX.def
+++ b/clang/include/clang/Basic/BuiltinsNVPTX.def
@@ -62,7 +62,9 @@
 #pragma push_macro("PTX82")
 #pragma push_macro("PTX83")
 #pragma push_macro("PTX84")
-#define PTX84 "ptx84"
+#pragma push_macro("PTX85")
+#define PTX85 "ptx85"
+#define PTX84 "ptx84|" PTX85
 #define PTX83 "ptx83|" PTX84
 #define PTX82 "ptx82|" PTX83
 #define PTX81 "ptx81|" PTX82
@@ -1094,3 +1096,4 @@ TARGET_BUILTIN(__nvvm_getctarank_shared_cluster, "iv*3", "", AND(SM_90,PTX78))
 #pragma pop_macro("PTX82")
 #pragma pop_macro("PTX83")
 #pragma pop_macro("PTX84")
+#pragma pop_macro("PTX85")
diff --git a/clang/include/clang/Basic/Cuda.h b/clang/include/clang/Basic/Cuda.h
index 2d67c4181d129..961ddaf57fc00 100644
--- a/clang/include/clang/Basic/Cuda.h
+++ b/clang/include/clang/Basic/Cuda.h
@@ -42,9 +42,10 @@ enum class CudaVersion {
   CUDA_122,
   CUDA_123,
   CUDA_124,
+  CUDA_125,
   FULLY_SUPPORTED = CUDA_123,
   PARTIALLY_SUPPORTED =
-      CUDA_124, // Partially supported. Proceed with a warning.
+      CUDA_125, // Partially supported. Proceed with a warning.
   NEW = 10000,  // Too new. Issue a warning, but allow using it.
 };
 const char *CudaVersionToString(CudaVersion V);
diff --git a/clang/lib/Basic/Cuda.cpp b/clang/lib/Basic/Cuda.cpp
index e8ce15eb0decb..2ee17f0ad4ca0 100644
--- a/clang/lib/Basic/Cuda.cpp
+++ b/clang/lib/Basic/Cuda.cpp
@@ -42,6 +42,7 @@ static const CudaVersionMapEntry CudaNameVersionMap[] = {
     CUDA_ENTRY(12, 2),
     CUDA_ENTRY(12, 3),
     CUDA_ENTRY(12, 4),
+    CUDA_ENTRY(12, 5),
     {"", CudaVersion::NEW, llvm::VersionTuple(std::numeric_limits<int>::max())},
     {"unknown", CudaVersion::UNKNOWN, {}} // End of list tombstone.
 };
diff --git a/clang/lib/Driver/ToolChains/Cuda.cpp b/clang/lib/Driver/ToolChains/Cuda.cpp
index d5f93c9c830fa..bbc8be91fd70b 100644
--- a/clang/lib/Driver/ToolChains/Cuda.cpp
+++ b/clang/lib/Driver/ToolChains/Cuda.cpp
@@ -84,6 +84,8 @@ CudaVersion getCudaVersion(uint32_t raw_version) {
     return CudaVersion::CUDA_123;
   if (raw_version < 12050)
     return CudaVersion::CUDA_124;
+  if (raw_version < 12060)
+    return CudaVersion::CUDA_125;
   return CudaVersion::NEW;
 }
 
@@ -690,6 +692,7 @@ void NVPTX::getNVPTXTargetFeatures(const Driver &D, const llvm::Triple &Triple,
   case CudaVersion::CUDA_##CUDA_VER:                                           \
     PtxFeature = "+ptx" #PTX_VER;                                              \
     break;
+    CASE_CUDA_VERSION(125, 85);
     CASE_CUDA_VERSION(124, 84);
     CASE_CUDA_VERSION(123, 83);
     CASE_CUDA_VERSION(122, 82);
diff --git a/llvm/lib/Target/NVPTX/NVPTX.td b/llvm/lib/Target/NVPTX/NVPTX.td
index 05457c71cd392..bb4549a5e6078 100644
--- a/llvm/lib/Target/NVPTX/NVPTX.td
+++ b/llvm/lib/Target/NVPTX/NVPTX.td
@@ -41,7 +41,8 @@ foreach sm = [20, 21, 30, 32, 35, 37, 50, 52, 53,
 def SM90a: FeatureSM<"90a", 901>;
 
 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] in
+                   70, 71, 72, 73, 74, 75, 76, 77, 78,
+                   80, 81, 82, 83, 84, 85] in
   def PTX#version: FeaturePTX<version>;
 
 //===----------------------------------------------------------------------===//

``````````

</details>


https://github.com/llvm/llvm-project/pull/94113