[llvm] [clang] [flang] [Flang] Add code-object-version option (PR #72638)

2023-11-28 Thread Dominik Adamski via cfe-commits 

https://github.com/DominikAdamski closed 
https://github.com/llvm/llvm-project/pull/72638
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[llvm] [clang] [flang] [Flang] Add code-object-version option (PR #72638)

2023-11-28 Thread Jan Leyonberg via cfe-commits 

https://github.com/jsjodin approved this pull request.

LGTM

https://github.com/llvm/llvm-project/pull/72638
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[llvm] [clang] [flang] [Flang] Add code-object-version option (PR #72638)

2023-11-23 Thread Andrzej Warzyński via cfe-commits 

https://github.com/banach-space edited 
https://github.com/llvm/llvm-project/pull/72638
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[llvm] [clang] [flang] [Flang] Add code-object-version option (PR #72638)

2023-11-22 Thread Dominik Adamski via cfe-commits 


@@ -264,6 +263,37 @@ static void addDepdendentLibs(mlir::ModuleOp ,
   }
 }
 
+// Add to MLIR code target specific items which are dependent on target
+// configuration specified by the user
+static void addTargetSpecificMLIRItems(mlir::ModuleOp ,
+   CompilerInstance ) {
+  const TargetOptions  = ci.getInvocation().getTargetOpts();
+  const llvm::Triple triple(targetOpts.triple);
+  if (triple.isAMDGPU()) {
+unsigned oclcABIVERsion;
+const unsigned defaultOclcABIVERsion = 400;
+mlir::OpBuilder builder(mlirModule.getContext());
+const CodeGenOptions  = ci.getInvocation().getCodeGenOpts();
+if (codeGenOpts.CodeObjectVersion ==
+CodeGenOptions::CodeObjectVersionKind::COV_None)
+  oclcABIVERsion = defaultOclcABIVERsion;
+else
+  oclcABIVERsion = static_cast(codeGenOpts.CodeObjectVersion);
+
+auto int32Type = builder.getI32Type();
+auto covInfo = builder.create(
+mlirModule.getLoc(), int32Type, true, mlir::LLVM::Linkage::WeakODR,
+"__oclc_ABI_version",
+builder.getIntegerAttr(int32Type, oclcABIVERsion));
+covInfo.setUnnamedAddr(mlir::LLVM::UnnamedAddr::Local);
+covInfo.setAddrSpace(4);
+covInfo.setVisibility_(mlir::LLVM::Visibility::Hidden);
+builder.setInsertionPointToStart(mlirModule.getBody());
+builder.insert(covInfo);
+  }
+  addDependentLibs(mlirModule, ci);

DominikAdamski wrote:

done

https://github.com/llvm/llvm-project/pull/72638
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[llvm] [clang] [flang] [Flang] Add code-object-version option (PR #72638)

2023-11-22 Thread Dominik Adamski via cfe-commits 


@@ -85,6 +85,19 @@ class CodeGenOptions : public CodeGenOptionsBase {
 RK_WithPattern, // Remark pattern specified via '-Rgroup=regexp'.
   };
 
+  /// \brief Enumeration value for AMDGPU code object version, which is the
+  /// code object version times 100.
+  enum class CodeObjectVersionKind {
+COV_None,
+COV_2 = 200, // Unsupported.
+COV_3 = 300, // Unsupported.
+COV_4 = 400,
+COV_5 = 500,
+  };

DominikAdamski wrote:

I moved it.

https://github.com/llvm/llvm-project/pull/72638
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[llvm] [clang] [flang] [Flang] Add code-object-version option (PR #72638)

2023-11-22 Thread Dominik Adamski via cfe-commits 


@@ -264,6 +263,37 @@ static void addDepdendentLibs(mlir::ModuleOp ,
   }
 }
 
+// Add to MLIR code target specific items which are dependent on target
+// configuration specified by the user
+static void addTargetSpecificMLIRItems(mlir::ModuleOp ,
+   CompilerInstance ) {
+  const TargetOptions  = ci.getInvocation().getTargetOpts();
+  const llvm::Triple triple(targetOpts.triple);
+  if (triple.isAMDGPU()) {

DominikAdamski wrote:

done

https://github.com/llvm/llvm-project/pull/72638
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [flang] [Flang] Add code-object-version option (PR #72638)

2023-11-19 Thread Andrzej Warzyński via cfe-commits 


@@ -264,6 +263,37 @@ static void addDepdendentLibs(mlir::ModuleOp ,
   }
 }
 
+// Add to MLIR code target specific items which are dependent on target
+// configuration specified by the user
+static void addTargetSpecificMLIRItems(mlir::ModuleOp ,
+   CompilerInstance ) {
+  const TargetOptions  = ci.getInvocation().getTargetOpts();
+  const llvm::Triple triple(targetOpts.triple);
+  if (triple.isAMDGPU()) {

banach-space wrote:

Please avoid indentation where possible 
(https://llvm.org/docs/CodingStandards.html#use-early-exits-and-continue-to-simplify-code):

```suggestion
  if (! triple.isAMDGPU())
return
```

https://github.com/llvm/llvm-project/pull/72638
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [flang] [Flang] Add code-object-version option (PR #72638)

2023-11-17 Thread Jan Leyonberg via cfe-commits 

https://github.com/jsjodin requested changes to this pull request.


https://github.com/llvm/llvm-project/pull/72638
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [flang] [Flang] Add code-object-version option (PR #72638)

2023-11-17 Thread Jan Leyonberg via cfe-commits 


@@ -264,6 +263,37 @@ static void addDepdendentLibs(mlir::ModuleOp ,
   }
 }
 
+// Add to MLIR code target specific items which are dependent on target
+// configuration specified by the user
+static void addTargetSpecificMLIRItems(mlir::ModuleOp ,
+   CompilerInstance ) {
+  const TargetOptions  = ci.getInvocation().getTargetOpts();

jsjodin wrote:

Create a separate function for adding the code object version like 
addDependentLibs is a separate function.

https://github.com/llvm/llvm-project/pull/72638
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [flang] [Flang] Add code-object-version option (PR #72638)

2023-11-17 Thread Sergio Afonso via cfe-commits 

https://github.com/skatrak edited 
https://github.com/llvm/llvm-project/pull/72638
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [flang] [Flang] Add code-object-version option (PR #72638)

2023-11-17 Thread Dominik Adamski via cfe-commits 

https://github.com/DominikAdamski created 
https://github.com/llvm/llvm-project/pull/72638

Information about code object version can be configured by the user for AMD GPU 
target and it needs to be placed in LLVM IR generated by Flang.

Information about code object version in MLIR generated by the parser can be 
reused by other tools. There is no need to specify extra flags if we want to 
invoke MLIR tools (like fir-opt) separately.

>From eb2710b0f736860dac62cc2ff8907fcefc64a8d6 Mon Sep 17 00:00:00 2001
From: Dominik Adamski 
Date: Fri, 17 Nov 2023 03:02:49 -0600
Subject: [PATCH] [Flang] Add code-object-version option

Information about code object version can be configured by the user
for AMD GPU target and it needs to be placed in LLVM IR generated
by Flang.

Information about code object version in MLIR generated by the
parser can be reused by other tools. There is no need to specify
extra flags if we want to invoke MLIR tools separately.
---
 clang/include/clang/Driver/Options.td |  2 +-
 clang/lib/Driver/ToolChains/Flang.cpp | 11 ++
 clang/lib/Driver/ToolChains/Flang.h   |  7 
 flang/include/flang/Frontend/CodeGenOptions.h | 13 +++
 flang/lib/Frontend/CompilerInvocation.cpp |  9 +
 flang/lib/Frontend/FrontendActions.cpp| 38 +--
 flang/test/Driver/code-object-version.f90 |  8 
 flang/test/Driver/driver-help-hidden.f90  |  2 +
 flang/test/Driver/driver-help.f90 |  4 ++
 flang/test/Lower/AMD/code_object_version.f90  | 11 ++
 10 files changed, 101 insertions(+), 4 deletions(-)
 create mode 100644 flang/test/Driver/code-object-version.f90
 create mode 100644 flang/test/Lower/AMD/code_object_version.f90

diff --git a/clang/include/clang/Driver/Options.td 
b/clang/include/clang/Driver/Options.td
index 811550416110b3d..e7eb94d174e75f8 100644
--- a/clang/include/clang/Driver/Options.td
+++ b/clang/include/clang/Driver/Options.td
@@ -4688,7 +4688,7 @@ defm amdgpu_ieee : BoolOption<"m", "amdgpu-ieee",
 
 def mcode_object_version_EQ : Joined<["-"], "mcode-object-version=">, 
Group,
   HelpText<"Specify code object ABI version. Defaults to 4. (AMDGPU only)">,
-  Visibility<[ClangOption, CC1Option]>,
+  Visibility<[ClangOption, FlangOption, CC1Option, FC1Option]>,
   Values<"none,4,5">,
   NormalizedValuesScope<"TargetOptions">,
   NormalizedValues<["COV_None", "COV_4", "COV_5"]>,
diff --git a/clang/lib/Driver/ToolChains/Flang.cpp 
b/clang/lib/Driver/ToolChains/Flang.cpp
index 8bdd920c3dcbb79..e60c11bfbe8e38b 100644
--- a/clang/lib/Driver/ToolChains/Flang.cpp
+++ b/clang/lib/Driver/ToolChains/Flang.cpp
@@ -204,6 +204,14 @@ void Flang::AddAArch64TargetArgs(const ArgList ,
   }
 }
 
+void Flang::AddAMDGPUTargetArgs(const ArgList ,
+ArgStringList ) const {
+  if (Arg *A = Args.getLastArg(options::OPT_mcode_object_version_EQ)) {
+StringRef Val = A->getValue();
+CmdArgs.push_back(Args.MakeArgString("-mcode-object-version=" + Val));
+  }
+}
+
 void Flang::addTargetOptions(const ArgList ,
  ArgStringList ) const {
   const ToolChain  = getToolChain();
@@ -227,6 +235,9 @@ void Flang::addTargetOptions(const ArgList ,
 
   case llvm::Triple::r600:
   case llvm::Triple::amdgcn:
+getTargetFeatures(D, Triple, Args, CmdArgs, /*ForAs*/ false);
+AddAMDGPUTargetArgs(Args, CmdArgs);
+break;
   case llvm::Triple::riscv64:
   case llvm::Triple::x86_64:
 getTargetFeatures(D, Triple, Args, CmdArgs, /*ForAs*/ false);
diff --git a/clang/lib/Driver/ToolChains/Flang.h 
b/clang/lib/Driver/ToolChains/Flang.h
index 0141240b5d3ac90..8d35080e1c0c88b 100644
--- a/clang/lib/Driver/ToolChains/Flang.h
+++ b/clang/lib/Driver/ToolChains/Flang.h
@@ -63,6 +63,13 @@ class LLVM_LIBRARY_VISIBILITY Flang : public Tool {
   void AddAArch64TargetArgs(const llvm::opt::ArgList ,
 llvm::opt::ArgStringList ) const;
 
+  /// Add specific options for AMDGPU target.
+  ///
+  /// \param [in] Args The list of input driver arguments
+  /// \param [out] CmdArgs The list of output command arguments
+  void AddAMDGPUTargetArgs(const llvm::opt::ArgList ,
+   llvm::opt::ArgStringList ) const;
+
   /// Extract offload options from the driver arguments and add them to
   /// the command arguments.
   /// \param [in] C The current compilation for the driver invocation
diff --git a/flang/include/flang/Frontend/CodeGenOptions.h 
b/flang/include/flang/Frontend/CodeGenOptions.h
index b86bb88610a9a4a..8d938c361a0aa23 100644
--- a/flang/include/flang/Frontend/CodeGenOptions.h
+++ b/flang/include/flang/Frontend/CodeGenOptions.h
@@ -85,6 +85,19 @@ class CodeGenOptions : public CodeGenOptionsBase {
 RK_WithPattern, // Remark pattern specified via '-Rgroup=regexp'.
   };
 
+  /// \brief Enumeration value for AMDGPU code object version, which is the
+  /// code object version times 100.
+  enum class CodeObjectVersionKind {
+COV_None,
+COV_2