r344281 - [HIP] Replace irif library with hip.amdgcn.bc

2018-10-11 Thread Aaron Enye Shi via cfe-commits 
Author: aaronenyeshi
Date: Thu Oct 11 12:41:54 2018
New Revision: 344281

URL: http://llvm.org/viewvc/llvm-project?rev=344281&view=rev
Log:
[HIP] Replace irif library with hip.amdgcn.bc

No longer use irif amdgcn library, instead we will use the previous fence 
functions from new hip.amdgcn.bc bitcode library. Update hip-device-libs.hip 
test as well.

Added:
cfe/trunk/test/Driver/Inputs/hip_dev_lib/hip.amdgcn.bc
Modified:
cfe/trunk/lib/Driver/ToolChains/HIP.cpp
cfe/trunk/test/Driver/hip-device-libs.hip

Modified: cfe/trunk/lib/Driver/ToolChains/HIP.cpp
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Driver/ToolChains/HIP.cpp?rev=344281&r1=344280&r2=344281&view=diff
==
--- cfe/trunk/lib/Driver/ToolChains/HIP.cpp (original)
+++ cfe/trunk/lib/Driver/ToolChains/HIP.cpp Thu Oct 11 12:41:54 2018
@@ -81,8 +81,8 @@ const char *AMDGCN::Linker::constructLLV
 else
   FlushDenormalControlBC = "oclc_daz_opt_off.amdgcn.bc";
 
-BCLibs.append({"opencl.amdgcn.bc",
-   "ocml.amdgcn.bc", "ockl.amdgcn.bc", "irif.amdgcn.bc",
+BCLibs.append({"hip.amdgcn.bc", "opencl.amdgcn.bc",
+   "ocml.amdgcn.bc", "ockl.amdgcn.bc",
"oclc_finite_only_off.amdgcn.bc",
FlushDenormalControlBC,
"oclc_correctly_rounded_sqrt_on.amdgcn.bc",

Added: cfe/trunk/test/Driver/Inputs/hip_dev_lib/hip.amdgcn.bc
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/Driver/Inputs/hip_dev_lib/hip.amdgcn.bc?rev=344281&view=auto
==
(empty)

Modified: cfe/trunk/test/Driver/hip-device-libs.hip
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/Driver/hip-device-libs.hip?rev=344281&r1=344280&r2=344281&view=diff
==
--- cfe/trunk/test/Driver/hip-device-libs.hip (original)
+++ cfe/trunk/test/Driver/hip-device-libs.hip Thu Oct 11 12:41:54 2018
@@ -21,7 +21,8 @@
 
 
 // COM: [[LLVM_LINK:"*.llvm-link"]]
-// COM-SAME: {{.*}} "{{.*}}ocml.amdgcn.bc" "{{.*}}ockl.amdgcn.bc" 
"{{.*}}irif.amdgcn.bc"
+// COM-SAME: "{{.*}}hip.amdgcn.bc" "{{.*}}opencl.amdgcn.bc"
+// COM-SAME: "{{.*}}ocml.amdgcn.bc" "{{.*}}ockl.amdgcn.bc"
 // FLUSHD-SAME: {{.*}} "{{.*}}oclc_daz_opt_on.amdgcn.bc"
 // NOFLUSHD-SAME: {{.*}} "{{.*}}oclc_daz_opt_off.amdgcn.bc"
 // COM-SAME: {{.*}} "-o" "{{.*}}-gfx900-linked-{{.*bc}}"


___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

r344285 - [HIP] Remove unused irif bitcode from test

2018-10-11 Thread Aaron Enye Shi via cfe-commits 
Author: aaronenyeshi
Date: Thu Oct 11 12:52:32 2018
New Revision: 344285

URL: http://llvm.org/viewvc/llvm-project?rev=344285&view=rev
Log:
[HIP] Remove unused irif bitcode from test

This is part of previous commit [HIP] Replace irif library with hip.amdgcn.bc

Reviewers: yaxunl

Differential Revision: https://reviews.llvm.org/D52673


Removed:
cfe/trunk/test/Driver/Inputs/hip_dev_lib/irif.amdgcn.bc

Removed: cfe/trunk/test/Driver/Inputs/hip_dev_lib/irif.amdgcn.bc
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/Driver/Inputs/hip_dev_lib/irif.amdgcn.bc?rev=344284&view=auto
==
(empty)


___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

r356277 - [HIP-Clang] propagate -mllvm options to opt and llc

2019-03-15 Thread Aaron Enye Shi via cfe-commits 
Author: aaronenyeshi
Date: Fri Mar 15 10:31:51 2019
New Revision: 356277

URL: http://llvm.org/viewvc/llvm-project?rev=356277&view=rev
Log:
[HIP-Clang] propagate -mllvm options to opt and llc

Change the HIP Toolchain to pass the OPT_mllvm options into OPT and LLC stages. 
Added a lit test to verify the command args.

Reviewers: yaxunl

Differential Revision: https://reviews.llvm.org/D59316



Added:
cfe/trunk/test/Driver/hip-toolchain-mllvm.hip
Modified:
cfe/trunk/lib/Driver/ToolChains/HIP.cpp

Modified: cfe/trunk/lib/Driver/ToolChains/HIP.cpp
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Driver/ToolChains/HIP.cpp?rev=356277&r1=356276&r2=356277&view=diff
==
--- cfe/trunk/lib/Driver/ToolChains/HIP.cpp (original)
+++ cfe/trunk/lib/Driver/ToolChains/HIP.cpp Fri Mar 15 10:31:51 2019
@@ -140,6 +140,11 @@ const char *AMDGCN::Linker::constructOpt
   }
   OptArgs.push_back("-mtriple=amdgcn-amd-amdhsa");
   OptArgs.push_back(Args.MakeArgString("-mcpu=" + SubArchName));
+
+  for (const Arg *A : Args.filtered(options::OPT_mllvm)) {
+OptArgs.push_back(A->getValue(0));
+  }
+
   OptArgs.push_back("-o");
   std::string TmpFileName = C.getDriver().GetTemporaryPath(
   OutputFilePrefix.str() + "-optimized", "bc");
@@ -177,6 +182,10 @@ const char *AMDGCN::Linker::constructLlc
   if(!Features.empty())
 LlcArgs.push_back(Args.MakeArgString(MAttrString));
 
+  for (const Arg *A : Args.filtered(options::OPT_mllvm)) {
+LlcArgs.push_back(A->getValue(0));
+  }
+
   // Add output filename
   LlcArgs.push_back("-o");
   std::string LlcOutputFileName =

Added: cfe/trunk/test/Driver/hip-toolchain-mllvm.hip
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/Driver/hip-toolchain-mllvm.hip?rev=356277&view=auto
==
--- cfe/trunk/test/Driver/hip-toolchain-mllvm.hip (added)
+++ cfe/trunk/test/Driver/hip-toolchain-mllvm.hip Fri Mar 15 10:31:51 2019
@@ -0,0 +1,36 @@
+// REQUIRES: clang-driver
+// REQUIRES: x86-registered-target
+// REQUIRES: amdgpu-registered-target
+
+// RUN: %clang -### -target x86_64-linux-gnu \
+// RUN:   -x hip --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 \
+// RUN:   -mllvm -amdgpu-function-calls=0 \
+// RUN:   %s 2>&1 | FileCheck %s
+
+// CHECK: [[CLANG:".*clang.*"]] "-cc1" "-triple" "amdgcn-amd-amdhsa"
+// CHECK-SAME: "-aux-triple" "x86_64-unknown-linux-gnu" "-emit-llvm-bc"
+// CHECK-SAME: {{.*}} "-target-cpu" "gfx803"
+// CHECK-SAME: {{.*}} "-mllvm" "-amdgpu-function-calls=0" {{.*}}
+
+// CHECK: [[OPT:".*opt"]] {{".*-gfx803-linked.*bc"}} 
"-mtriple=amdgcn-amd-amdhsa"
+// CHECK-SAME: "-mcpu=gfx803" "-amdgpu-function-calls=0"
+// CHECK-SAME: "-o" [[OPT_803_BC:".*-gfx803-optimized.*bc"]]
+
+// CHECK: [[LLC: ".*llc"]] [[OPT_803_BC]]
+// CHECK-SAME: "-mtriple=amdgcn-amd-amdhsa" "-filetype=obj"
+// CHECK-SAME: {{.*}} "-mcpu=gfx803"
+// CHECK-SAME: "-amdgpu-function-calls=0" "-o" {{".*-gfx803-.*o"}}
+
+// CHECK: [[CLANG]] "-cc1" "-triple" "amdgcn-amd-amdhsa"
+// CHECK-SAME: "-aux-triple" "x86_64-unknown-linux-gnu" "-emit-llvm-bc"
+// CHECK-SAME: {{.*}} "-target-cpu" "gfx900"
+// CHECK-SAME: {{.*}} "-mllvm" "-amdgpu-function-calls=0" {{.*}}
+
+// CHECK: [[OPT]] {{".*-gfx900-linked.*bc"}} "-mtriple=amdgcn-amd-amdhsa"
+// CHECK-SAME: "-mcpu=gfx900" "-amdgpu-function-calls=0"
+// CHECK-SAME: "-o" [[OPT_900_BC:".*-gfx900-optimized.*bc"]]
+
+// CHECK: [[LLC]] [[OPT_900_BC]]
+// CHECK-SAME: "-mtriple=amdgcn-amd-amdhsa" "-filetype=obj"
+// CHECK-SAME: {{.*}} "-mcpu=gfx900"
+// CHECK-SAME: "-amdgpu-function-calls=0" "-o" {{".*-gfx900-.*o"}}


___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

r357520 - [HIP-Clang] Fat binary should not be produced for non GPU code

2019-04-02 Thread Aaron Enye Shi via cfe-commits 
Author: aaronenyeshi
Date: Tue Apr  2 13:10:18 2019
New Revision: 357520

URL: http://llvm.org/viewvc/llvm-project?rev=357520&view=rev
Log:
[HIP-Clang] Fat binary should not be produced for non GPU code

Skip producing the fat binary functions for HIP when no device code is present.

Reviewers: yaxunl

Differential Review: https://reviews.llvm.org/D60141


Modified:
cfe/trunk/lib/CodeGen/CGCUDANV.cpp
cfe/trunk/test/CodeGenCUDA/device-stub.cu

Modified: cfe/trunk/lib/CodeGen/CGCUDANV.cpp
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/CodeGen/CGCUDANV.cpp?rev=357520&r1=357519&r2=357520&view=diff
==
--- cfe/trunk/lib/CodeGen/CGCUDANV.cpp (original)
+++ cfe/trunk/lib/CodeGen/CGCUDANV.cpp Tue Apr  2 13:10:18 2019
@@ -472,6 +472,8 @@ llvm::Function *CGNVCUDARuntime::makeMod
   StringRef CudaGpuBinaryFileName = CGM.getCodeGenOpts().CudaGpuBinaryFileName;
   if (CudaGpuBinaryFileName.empty() && !IsHIP)
 return nullptr;
+  if (IsHIP && EmittedKernels.empty() && DeviceVars.empty())
+return nullptr;
 
   // void __{cuda|hip}_register_globals(void* handle);
   llvm::Function *RegisterGlobalsFunc = makeRegisterGlobalsFn();

Modified: cfe/trunk/test/CodeGenCUDA/device-stub.cu
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/CodeGenCUDA/device-stub.cu?rev=357520&r1=357519&r2=357520&view=diff
==
--- cfe/trunk/test/CodeGenCUDA/device-stub.cu (original)
+++ cfe/trunk/test/CodeGenCUDA/device-stub.cu Tue Apr  2 13:10:18 2019
@@ -228,13 +228,19 @@ void hostfunc(void) { kernelfunc<<<1, 1>
 // device-side globals, but we still need to register GPU binary.
 // Skip GPU binary string first.
 // CUDANOGLOBALS: @{{.*}} = private constant{{.*}}
-// HIPNOGLOBALS: @{{.*}} = internal constant{{.*}}
+// HIPNOGLOBALS-NOT: @{{.*}} = internal constant{{.*}}
 // NOGLOBALS-NOT: define internal void @__{{.*}}_register_globals
-// NOGLOBALS: define internal void @__[[PREFIX:cuda|hip]]_module_ctor
-// NOGLOBALS: 
call{{.*}}[[PREFIX]]RegisterFatBinary{{.*}}__[[PREFIX]]_fatbin_wrapper
+// CUDANOGLOBALS: define internal void @__[[PREFIX:cuda|hip]]_module_ctor
+// CUDANOGLOBALS: 
call{{.*}}[[PREFIX]]RegisterFatBinary{{.*}}__[[PREFIX]]_fatbin_wrapper
 // NOGLOBALS-NOT: call void @__[[PREFIX]]_register_globals
-// NOGLOBALS: define internal void @__[[PREFIX]]_module_dtor
-// NOGLOBALS: call void @__[[PREFIX]]UnregisterFatBinary
+// CUDANOGLOBALS: define internal void @__[[PREFIX]]_module_dtor
+// CUDANOGLOBALS: call void @__[[PREFIX]]UnregisterFatBinary
+
+// There should be no fat binary functions when no device-code is found for 
HIP.
+// HIPNOGLOBALS-NOT: define internal void @__[[PREFIX:cuda|hip]]_module_ctor
+// HIPNOGLOBALS-NOT: 
call{{.*}}[[PREFIX]]RegisterFatBinary{{.*}}__[[PREFIX]]_fatbin_wrapper
+// HIPNOGLOBALS-NOT: define internal void @__[[PREFIX]]_module_dtor
+// HIPNOGLOBALS-NOT: call void @__[[PREFIX]]UnregisterFatBinary
 
 // There should be no constructors/destructors if we have no GPU binary.
 // NOGPUBIN-NOT: define internal void @__[[PREFIX]]_register_globals


___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

r357526 - [HIP-Clang] Fat binary should not be produced for non GPU code 2

2019-04-02 Thread Aaron Enye Shi via cfe-commits 
Author: aaronenyeshi
Date: Tue Apr  2 13:49:41 2019
New Revision: 357526

URL: http://llvm.org/viewvc/llvm-project?rev=357526&view=rev
Log:
[HIP-Clang] Fat binary should not be produced for non GPU code 2

Also for CUDA, we need to disable producing these fat binary functions when 
there is no GPU code.

Reviewers: yaxunl, tra

Differential Revision: https://reviews.llvm.org/D60141

Modified:
cfe/trunk/lib/CodeGen/CGCUDANV.cpp
cfe/trunk/test/CodeGenCUDA/device-stub.cu

Modified: cfe/trunk/lib/CodeGen/CGCUDANV.cpp
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/CodeGen/CGCUDANV.cpp?rev=357526&r1=357525&r2=357526&view=diff
==
--- cfe/trunk/lib/CodeGen/CGCUDANV.cpp (original)
+++ cfe/trunk/lib/CodeGen/CGCUDANV.cpp Tue Apr  2 13:49:41 2019
@@ -468,11 +468,13 @@ llvm::Function *CGNVCUDARuntime::makeReg
 /// \endcode
 llvm::Function *CGNVCUDARuntime::makeModuleCtorFunction() {
   bool IsHIP = CGM.getLangOpts().HIP;
+  bool IsCUDA = CGM.getLangOpts().CUDA;
   // No need to generate ctors/dtors if there is no GPU binary.
   StringRef CudaGpuBinaryFileName = CGM.getCodeGenOpts().CudaGpuBinaryFileName;
   if (CudaGpuBinaryFileName.empty() && !IsHIP)
 return nullptr;
-  if (IsHIP && EmittedKernels.empty() && DeviceVars.empty())
+  if ( (IsHIP || (IsCUDA && !RelocatableDeviceCode) )
+   && EmittedKernels.empty() && DeviceVars.empty())
 return nullptr;
 
   // void __{cuda|hip}_register_globals(void* handle);

Modified: cfe/trunk/test/CodeGenCUDA/device-stub.cu
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/CodeGenCUDA/device-stub.cu?rev=357526&r1=357525&r2=357526&view=diff
==
--- cfe/trunk/test/CodeGenCUDA/device-stub.cu (original)
+++ cfe/trunk/test/CodeGenCUDA/device-stub.cu Tue Apr  2 13:49:41 2019
@@ -227,20 +227,14 @@ void hostfunc(void) { kernelfunc<<<1, 1>
 // There should be no __[[PREFIX]]_register_globals if we have no
 // device-side globals, but we still need to register GPU binary.
 // Skip GPU binary string first.
-// CUDANOGLOBALS: @{{.*}} = private constant{{.*}}
+// CUDANOGLOBALS-NOT: @{{.*}} = private constant{{.*}}
 // HIPNOGLOBALS-NOT: @{{.*}} = internal constant{{.*}}
 // NOGLOBALS-NOT: define internal void @__{{.*}}_register_globals
-// CUDANOGLOBALS: define internal void @__[[PREFIX:cuda|hip]]_module_ctor
-// CUDANOGLOBALS: 
call{{.*}}[[PREFIX]]RegisterFatBinary{{.*}}__[[PREFIX]]_fatbin_wrapper
+// NOGLOBALS-NOT: define internal void @__[[PREFIX:cuda|hip]]_module_ctor
+// NOGLOBALS-NOT: 
call{{.*}}[[PREFIX]]RegisterFatBinary{{.*}}__[[PREFIX]]_fatbin_wrapper
 // NOGLOBALS-NOT: call void @__[[PREFIX]]_register_globals
-// CUDANOGLOBALS: define internal void @__[[PREFIX]]_module_dtor
-// CUDANOGLOBALS: call void @__[[PREFIX]]UnregisterFatBinary
-
-// There should be no fat binary functions when no device-code is found for 
HIP.
-// HIPNOGLOBALS-NOT: define internal void @__[[PREFIX:cuda|hip]]_module_ctor
-// HIPNOGLOBALS-NOT: 
call{{.*}}[[PREFIX]]RegisterFatBinary{{.*}}__[[PREFIX]]_fatbin_wrapper
-// HIPNOGLOBALS-NOT: define internal void @__[[PREFIX]]_module_dtor
-// HIPNOGLOBALS-NOT: call void @__[[PREFIX]]UnregisterFatBinary
+// NOGLOBALS-NOT: define internal void @__[[PREFIX]]_module_dtor
+// NOGLOBALS-NOT: call void @__[[PREFIX]]UnregisterFatBinary
 
 // There should be no constructors/destructors if we have no GPU binary.
 // NOGPUBIN-NOT: define internal void @__[[PREFIX]]_register_globals


___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

r357532 - [HIP-Clang] Fat binary should not be produced for non GPU code

2019-04-02 Thread Aaron Enye Shi via cfe-commits 
Author: aaronenyeshi
Date: Tue Apr  2 14:54:41 2019
New Revision: 357532

URL: http://llvm.org/viewvc/llvm-project?rev=357532&view=rev
Log:
[HIP-Clang] Fat binary should not be produced for non GPU code

clang-format the changes to CUDA and HIP fat binary.

Reviewers: yaxunl, tra

Differential Revision: https://reviews.llvm.org/D60141

Modified:
cfe/trunk/lib/CodeGen/CGCUDANV.cpp

Modified: cfe/trunk/lib/CodeGen/CGCUDANV.cpp
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/CodeGen/CGCUDANV.cpp?rev=357532&r1=357531&r2=357532&view=diff
==
--- cfe/trunk/lib/CodeGen/CGCUDANV.cpp (original)
+++ cfe/trunk/lib/CodeGen/CGCUDANV.cpp Tue Apr  2 14:54:41 2019
@@ -473,8 +473,8 @@ llvm::Function *CGNVCUDARuntime::makeMod
   StringRef CudaGpuBinaryFileName = CGM.getCodeGenOpts().CudaGpuBinaryFileName;
   if (CudaGpuBinaryFileName.empty() && !IsHIP)
 return nullptr;
-  if ( (IsHIP || (IsCUDA && !RelocatableDeviceCode) )
-   && EmittedKernels.empty() && DeviceVars.empty())
+  if ((IsHIP || (IsCUDA && !RelocatableDeviceCode)) && EmittedKernels.empty() 
&&
+  DeviceVars.empty())
 return nullptr;
 
   // void __{cuda|hip}_register_globals(void* handle);


___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

r353880 - [HIP] Handle compile -m options and propagate into LLC

2019-02-12 Thread Aaron Enye Shi via cfe-commits 
Author: aaronenyeshi
Date: Tue Feb 12 13:21:06 2019
New Revision: 353880

URL: http://llvm.org/viewvc/llvm-project?rev=353880&view=rev
Log:
[HIP] Handle compile -m options and propagate into LLC

Allow the compile options for -m such as -mxnack/-mno-xnack, 
-msram-ecc/-mno-sram-ecc, -mcode-object-v3/-mno-code-object-v3 to propagate 
into LLC args.

Also add lit tests to verify features are properly passed.

Differential Revision: https://reviews.llvm.org/D57977

Reviewers: yaxunl, kzhuravl


Added:
cfe/trunk/test/Driver/hip-toolchain-features.hip
Modified:
cfe/trunk/lib/Driver/ToolChains/HIP.cpp

Modified: cfe/trunk/lib/Driver/ToolChains/HIP.cpp
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Driver/ToolChains/HIP.cpp?rev=353880&r1=353879&r2=353880&view=diff
==
--- cfe/trunk/lib/Driver/ToolChains/HIP.cpp (original)
+++ cfe/trunk/lib/Driver/ToolChains/HIP.cpp Tue Feb 12 13:21:06 2019
@@ -159,8 +159,25 @@ const char *AMDGCN::Linker::constructLlc
 llvm::StringRef OutputFilePrefix, const char *InputFileName) const {
   // Construct llc command.
   ArgStringList LlcArgs{InputFileName, "-mtriple=amdgcn-amd-amdhsa",
-"-filetype=obj", "-mattr=-code-object-v3",
-Args.MakeArgString("-mcpu=" + SubArchName), "-o"};
+"-filetype=obj",
+Args.MakeArgString("-mcpu=" + SubArchName)};
+
+  // Extract all the -m options
+  std::vector Features;
+  handleTargetFeaturesGroup(
+Args, Features, options::OPT_m_amdgpu_Features_Group);
+
+  // Add features to mattr such as code-object-v3 and xnack
+  std::string MAttrString = "-mattr=";
+  for(auto OneFeature : Features) {
+MAttrString.append(Args.MakeArgString(OneFeature));
+if (OneFeature != Features.back())
+  MAttrString.append(",");
+  }
+  LlcArgs.push_back(Args.MakeArgString(MAttrString));
+
+  // Add output filename
+  LlcArgs.push_back("-o");
   std::string LlcOutputFileName =
   C.getDriver().GetTemporaryPath(OutputFilePrefix, "o");
   const char *LlcOutputFile =

Added: cfe/trunk/test/Driver/hip-toolchain-features.hip
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/Driver/hip-toolchain-features.hip?rev=353880&view=auto
==
--- cfe/trunk/test/Driver/hip-toolchain-features.hip (added)
+++ cfe/trunk/test/Driver/hip-toolchain-features.hip Tue Feb 12 13:21:06 2019
@@ -0,0 +1,49 @@
+// REQUIRES: clang-driver
+// REQUIRES: x86-registered-target
+// REQUIRES: amdgpu-registered-target
+
+// RUN: %clang -### -c -target x86_64-linux-gnu -fgpu-rdc \
+// RUN:   -x hip --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s \
+// RUN:   -mcode-object-v3 2>&1 | FileCheck %s -check-prefix=COV3
+// RUN: %clang -### -c -target x86_64-linux-gnu -fgpu-rdc \
+// RUN:   -x hip --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s \
+// RUN:   -mno-code-object-v3 2>&1 | FileCheck %s -check-prefix=NOCOV3
+
+// COV3: {{.*}}clang{{.*}}"-target-feature" "+code-object-v3"
+// NOCOV3: {{.*}}clang{{.*}}"-target-feature" "-code-object-v3"
+
+
+// RUN: %clang -### -c -target x86_64-linux-gnu -fgpu-rdc \
+// RUN:   -x hip --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s \
+// RUN:   -mxnack 2>&1 | FileCheck %s -check-prefix=XNACK
+// RUN: %clang -### -c -target x86_64-linux-gnu -fgpu-rdc \
+// RUN:   -x hip --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s \
+// RUN:   -mno-xnack 2>&1 | FileCheck %s -check-prefix=NOXNACK
+
+// XNACK: {{.*}}clang{{.*}}"-target-feature" "+xnack"
+// NOXNACK: {{.*}}clang{{.*}}"-target-feature" "-xnack"
+
+
+// RUN: %clang -### -c -target x86_64-linux-gnu -fgpu-rdc \
+// RUN:   -x hip --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s \
+// RUN:   -msram-ecc 2>&1 | FileCheck %s -check-prefix=SRAM
+// RUN: %clang -### -c -target x86_64-linux-gnu -fgpu-rdc \
+// RUN:   -x hip --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s \
+// RUN:   -mno-sram-ecc 2>&1 | FileCheck %s -check-prefix=NOSRAM
+
+// SRAM: {{.*}}clang{{.*}}"-target-feature" "+sram-ecc"
+// NOSRAM: {{.*}}clang{{.*}}"-target-feature" "-sram-ecc"
+
+
+// RUN: %clang -### -c -target x86_64-linux-gnu -fgpu-rdc \
+// RUN:   -x hip --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s \
+// RUN:   -mcode-object-v3 -mxnack -msram-ecc \
+// RUN:   2>&1 | FileCheck %s -check-prefix=ALL3
+// RUN: %clang -### -c -target x86_64-linux-gnu -fgpu-rdc \
+// RUN:   -x hip --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s \
+// RUN:   -mno-code-object-v3 -mno-xnack -mno-sram-ecc \
+// RUN:   2>&1 | FileCheck %s -check-prefix=NOALL3
+
+// ALL3: {{.*}}clang{{.*}}"-target-feature" "+code-object-v3" 
"-target-feature" "+xnack" "-target-feature" "+sram-ecc"
+// NOALL3: {{.*}}clang{{.*}}"-target-feature" "-code-object-v3" 
"-target-feature" "-xnack" "-target-feature" "-sram-ecc"
+


___
cfe-commits mailing list
cfe-commits@li

r353885 - [HIP] Fix hip-toolchain-rdc tests

2019-02-12 Thread Aaron Enye Shi via cfe-commits 
Author: aaronenyeshi
Date: Tue Feb 12 14:01:19 2019
New Revision: 353885

URL: http://llvm.org/viewvc/llvm-project?rev=353885&view=rev
Log:
[HIP] Fix hip-toolchain-rdc tests

Since we removed changed the way HIP Toolchain will propagate -m options into 
LLC, we need to remove from these older tests.

This is related to rC353880.

Differential Revision: https://reviews.llvm.org/D57977


Modified:
cfe/trunk/test/Driver/hip-toolchain-no-rdc.hip
cfe/trunk/test/Driver/hip-toolchain-rdc.hip

Modified: cfe/trunk/test/Driver/hip-toolchain-no-rdc.hip
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/Driver/hip-toolchain-no-rdc.hip?rev=353885&r1=353884&r2=353885&view=diff
==
--- cfe/trunk/test/Driver/hip-toolchain-no-rdc.hip (original)
+++ cfe/trunk/test/Driver/hip-toolchain-no-rdc.hip Tue Feb 12 14:01:19 2019
@@ -34,7 +34,6 @@
 
 // CHECK: [[LLC: ".*llc"]] [[OPT_BC_DEV_A_803]] "-mtriple=amdgcn-amd-amdhsa"
 // CHECK-SAME: "-filetype=obj"
-// CHECK-SAME: "-mattr=-code-object-v3"
 // CHECK-SAME: "-mcpu=gfx803" "-o" [[OBJ_DEV_A_803:".*-gfx803-.*o"]]
 
 // CHECK: [[LLD: ".*lld"]] "-flavor" "gnu" "--no-undefined" "-shared"
@@ -62,7 +61,6 @@
 
 // CHECK: [[LLC: ".*llc"]] [[OPT_BC_DEV_A_900]] "-mtriple=amdgcn-amd-amdhsa"
 // CHECK-SAME: "-filetype=obj"
-// CHECK-SAME: "-mattr=-code-object-v3"
 // CHECK-SAME: "-mcpu=gfx900" "-o" [[OBJ_DEV_A_900:".*-gfx900-.*o"]]
 
 // CHECK: [[LLD: ".*lld"]] "-flavor" "gnu" "--no-undefined" "-shared"
@@ -105,7 +103,6 @@
 
 // CHECK: [[LLC: ".*llc"]] [[OPT_BC_DEV_B_803]] "-mtriple=amdgcn-amd-amdhsa"
 // CHECK-SAME: "-filetype=obj"
-// CHECK-SAME: "-mattr=-code-object-v3"
 // CHECK-SAME: "-mcpu=gfx803" "-o" [[OBJ_DEV_B_803:".*-gfx803-.*o"]]
 
 // CHECK: [[LLD: ".*lld"]] "-flavor" "gnu" "--no-undefined" "-shared"
@@ -133,7 +130,6 @@
 
 // CHECK: [[LLC: ".*llc"]] [[OPT_BC_DEV_B_900]] "-mtriple=amdgcn-amd-amdhsa"
 // CHECK-SAME: "-filetype=obj"
-// CHECK-SAME: "-mattr=-code-object-v3"
 // CHECK-SAME: "-mcpu=gfx900" "-o" [[OBJ_DEV_B_900:".*-gfx900-.*o"]]
 
 // CHECK: [[LLD: ".*lld"]] "-flavor" "gnu" "--no-undefined" "-shared"

Modified: cfe/trunk/test/Driver/hip-toolchain-rdc.hip
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/Driver/hip-toolchain-rdc.hip?rev=353885&r1=353884&r2=353885&view=diff
==
--- cfe/trunk/test/Driver/hip-toolchain-rdc.hip (original)
+++ cfe/trunk/test/Driver/hip-toolchain-rdc.hip Tue Feb 12 14:01:19 2019
@@ -38,7 +38,6 @@
 
 // CHECK: [[LLC: ".*llc"]] [[OPT_BC_DEV1]] "-mtriple=amdgcn-amd-amdhsa"
 // CHECK-SAME: "-filetype=obj"
-// CHECK-SAME: "-mattr=-code-object-v3"
 // CHECK-SAME: "-mcpu=gfx803" "-o" [[OBJ_DEV1:".*-gfx803-.*o"]]
 
 // CHECK: [[LLD: ".*lld"]] "-flavor" "gnu" "--no-undefined" "-shared"
@@ -66,7 +65,6 @@
 
 // CHECK: [[LLC]] [[OPT_BC_DEV2]] "-mtriple=amdgcn-amd-amdhsa"
 // CHECK-SAME: "-filetype=obj"
-// CHECK-SAME: "-mattr=-code-object-v3"
 // CHECK-SAME: "-mcpu=gfx900" "-o" [[OBJ_DEV2:".*-gfx900-.*o"]]
 
 // CHECK: [[LLD]] "-flavor" "gnu" "--no-undefined" "-shared"


___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

r353952 - [HIP] Handle compile -m options and propagate into LLC

2019-02-13 Thread Aaron Enye Shi via cfe-commits 
Author: aaronenyeshi
Date: Wed Feb 13 08:12:16 2019
New Revision: 353952

URL: http://llvm.org/viewvc/llvm-project?rev=353952&view=rev
Log:
[HIP] Handle compile -m options and propagate into LLC

Allow the compile options for -m such as -mxnack/-mno-xnack, 
-msram-ecc/-mno-sram-ecc, -mcode-object-v3/-mno-code-object-v3 to propagate 
into LLC args. Fix an issue where -mattr was pushed even when it was empty.

Also add lit tests to verify features are properly passed.

Differential Revision: https://reviews.llvm.org/D57977

Reviewers: yaxunl, kzhuravl


Added:
cfe/trunk/test/Driver/hip-toolchain-features.hip
Modified:
cfe/trunk/lib/Driver/ToolChains/HIP.cpp

Modified: cfe/trunk/lib/Driver/ToolChains/HIP.cpp
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Driver/ToolChains/HIP.cpp?rev=353952&r1=353951&r2=353952&view=diff
==
--- cfe/trunk/lib/Driver/ToolChains/HIP.cpp (original)
+++ cfe/trunk/lib/Driver/ToolChains/HIP.cpp Wed Feb 13 08:12:16 2019
@@ -159,8 +159,26 @@ const char *AMDGCN::Linker::constructLlc
 llvm::StringRef OutputFilePrefix, const char *InputFileName) const {
   // Construct llc command.
   ArgStringList LlcArgs{InputFileName, "-mtriple=amdgcn-amd-amdhsa",
-"-filetype=obj", "-mattr=-code-object-v3",
-Args.MakeArgString("-mcpu=" + SubArchName), "-o"};
+"-filetype=obj",
+Args.MakeArgString("-mcpu=" + SubArchName)};
+
+  // Extract all the -m options
+  std::vector Features;
+  handleTargetFeaturesGroup(
+Args, Features, options::OPT_m_amdgpu_Features_Group);
+
+  // Add features to mattr such as code-object-v3 and xnack
+  std::string MAttrString = "-mattr=";
+  for(auto OneFeature : Features) {
+MAttrString.append(Args.MakeArgString(OneFeature));
+if (OneFeature != Features.back())
+  MAttrString.append(",");
+  }
+  if(!Features.empty())
+LlcArgs.push_back(Args.MakeArgString(MAttrString));
+
+  // Add output filename
+  LlcArgs.push_back("-o");
   std::string LlcOutputFileName =
   C.getDriver().GetTemporaryPath(OutputFilePrefix, "o");
   const char *LlcOutputFile =

Added: cfe/trunk/test/Driver/hip-toolchain-features.hip
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/Driver/hip-toolchain-features.hip?rev=353952&view=auto
==
--- cfe/trunk/test/Driver/hip-toolchain-features.hip (added)
+++ cfe/trunk/test/Driver/hip-toolchain-features.hip Wed Feb 13 08:12:16 2019
@@ -0,0 +1,48 @@
+// REQUIRES: clang-driver
+// REQUIRES: x86-registered-target
+// REQUIRES: amdgpu-registered-target
+
+// RUN: %clang -### -c -target x86_64-linux-gnu -fgpu-rdc \
+// RUN:   -x hip --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s \
+// RUN:   -mcode-object-v3 2>&1 | FileCheck %s -check-prefix=COV3
+// RUN: %clang -### -c -target x86_64-linux-gnu -fgpu-rdc \
+// RUN:   -x hip --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s \
+// RUN:   -mno-code-object-v3 2>&1 | FileCheck %s -check-prefix=NOCOV3
+
+// COV3: {{.*}}clang{{.*}}"-target-feature" "+code-object-v3"
+// NOCOV3: {{.*}}clang{{.*}}"-target-feature" "-code-object-v3"
+
+
+// RUN: %clang -### -c -target x86_64-linux-gnu -fgpu-rdc \
+// RUN:   -x hip --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s \
+// RUN:   -mxnack 2>&1 | FileCheck %s -check-prefix=XNACK
+// RUN: %clang -### -c -target x86_64-linux-gnu -fgpu-rdc \
+// RUN:   -x hip --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s \
+// RUN:   -mno-xnack 2>&1 | FileCheck %s -check-prefix=NOXNACK
+
+// XNACK: {{.*}}clang{{.*}}"-target-feature" "+xnack"
+// NOXNACK: {{.*}}clang{{.*}}"-target-feature" "-xnack"
+
+
+// RUN: %clang -### -c -target x86_64-linux-gnu -fgpu-rdc \
+// RUN:   -x hip --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s \
+// RUN:   -msram-ecc 2>&1 | FileCheck %s -check-prefix=SRAM
+// RUN: %clang -### -c -target x86_64-linux-gnu -fgpu-rdc \
+// RUN:   -x hip --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s \
+// RUN:   -mno-sram-ecc 2>&1 | FileCheck %s -check-prefix=NOSRAM
+
+// SRAM: {{.*}}clang{{.*}}"-target-feature" "+sram-ecc"
+// NOSRAM: {{.*}}clang{{.*}}"-target-feature" "-sram-ecc"
+
+
+// RUN: %clang -### -c -target x86_64-linux-gnu -fgpu-rdc \
+// RUN:   -x hip --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s \
+// RUN:   -mcode-object-v3 -mxnack -msram-ecc \
+// RUN:   2>&1 | FileCheck %s -check-prefix=ALL3
+// RUN: %clang -### -c -target x86_64-linux-gnu -fgpu-rdc \
+// RUN:   -x hip --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s \
+// RUN:   -mno-code-object-v3 -mno-xnack -mno-sram-ecc \
+// RUN:   2>&1 | FileCheck %s -check-prefix=NOALL3
+
+// ALL3: {{.*}}clang{{.*}}"-target-feature" "+code-object-v3" 
"-target-feature" "+xnack" "-target-feature" "+sram-ecc"
+// NOALL3: {{.*}}clang{{.*}}"-target-feature" "-code-object-v3" 
"-target-feature" "-xnack" "-target-feature" "-sram-ecc"


_

r350425 - [HIP][DRIVER][OFFLOAD] Do not unbundle unsupported file types

2019-01-04 Thread Aaron Enye Shi via cfe-commits 
Author: aaronenyeshi
Date: Fri Jan  4 11:05:41 2019
New Revision: 350425

URL: http://llvm.org/viewvc/llvm-project?rev=350425&view=rev
Log:
[HIP][DRIVER][OFFLOAD] Do not unbundle unsupported file types

The offload bundler action should not unbundle the input file types that does 
not match the action type. This fixes an issue where .so files are unbundled 
when the action type is object files.

Added:
cfe/trunk/test/Driver/hip-link-shared-library.hip
Modified:
cfe/trunk/lib/Driver/Driver.cpp

Modified: cfe/trunk/lib/Driver/Driver.cpp
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Driver/Driver.cpp?rev=350425&r1=350424&r2=350425&view=diff
==
--- cfe/trunk/lib/Driver/Driver.cpp (original)
+++ cfe/trunk/lib/Driver/Driver.cpp Fri Jan  4 11:05:41 2019
@@ -2323,6 +2323,18 @@ class OffloadingActionBuilder final {
   // If this is an unbundling action use it as is for each CUDA toolchain.
   if (auto *UA = dyn_cast(HostAction)) {
 CudaDeviceActions.clear();
+auto *IA = cast(UA->getInputs().back());
+std::string FileName = IA->getInputArg().getAsString(Args);
+// Check if the type of the file is the same as the action. Do not
+// unbundle it if it is not. Do not unbundle .so files, for example,
+// which are not object files.
+if (IA->getType() == types::TY_Object &&
+(!llvm::sys::path::has_extension(FileName) ||
+ types::lookupTypeForExtension(
+ llvm::sys::path::extension(FileName).drop_front()) !=
+ types::TY_Object))
+  return ABRT_Inactive;
+
 for (auto Arch : GpuArchList) {
   CudaDeviceActions.push_back(UA);
   UA->registerDependentActionInfo(ToolChains[0], 
CudaArchToString(Arch),

Added: cfe/trunk/test/Driver/hip-link-shared-library.hip
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/Driver/hip-link-shared-library.hip?rev=350425&view=auto
==
--- cfe/trunk/test/Driver/hip-link-shared-library.hip (added)
+++ cfe/trunk/test/Driver/hip-link-shared-library.hip Fri Jan  4 11:05:41 2019
@@ -0,0 +1,11 @@
+// RUN: touch %t.o
+// RUN: %clang --hip-link -ccc-print-bindings -target x86_64-linux-gnu \
+// RUN:   --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %t.o %S/Inputs/in.so \
+// RUN: 2>&1 | FileCheck %s
+
+// CHECK: # "amdgcn-amd-amdhsa" - "offload bundler", inputs: ["[[IN:.*o]]"], 
outputs: ["[[OBJ1:.*o]]", "[[OBJ2:.*o]]", "[[OBJ3:.*o]]"]
+// CHECK: # "amdgcn-amd-amdhsa" - "AMDGCN::Linker", inputs: ["[[OBJ2]]"], 
output: "[[IMG2:.*out]]"
+// CHECK-NOT: offload bundler
+// CHECK: # "amdgcn-amd-amdhsa" - "AMDGCN::Linker", inputs: ["[[OBJ3]]"], 
output: "[[IMG3:.*out]]"
+// CHECK-NOT: offload bundler
+// CHECK: # "x86_64-unknown-linux-gnu" - "GNU::Linker", inputs: ["[[OBJ1]]", 
"{{.*}}/Inputs/in.so", "[[IMG2]]", "[[IMG3]]"], output: "a.out"


___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

r350426 - [HIP][DRIVER][OFFLOAD] Do not unbundle unsupported file types

2019-01-04 Thread Aaron Enye Shi via cfe-commits 
Author: aaronenyeshi
Date: Fri Jan  4 11:09:20 2019
New Revision: 350426

URL: http://llvm.org/viewvc/llvm-project?rev=350426&view=rev
Log:
[HIP][DRIVER][OFFLOAD] Do not unbundle unsupported file types

The offload bundler action should not unbundle the input file types that does 
not match the action type. This fixes an issue where .so files are unbundled 
when the action type is object files.

Reviewers: yaxunl

Differential Revision: https://reviews.llvm.org/D56321

Modified:
cfe/trunk/test/Driver/hip-link-shared-library.hip

Modified: cfe/trunk/test/Driver/hip-link-shared-library.hip
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/Driver/hip-link-shared-library.hip?rev=350426&r1=350425&r2=350426&view=diff
==
--- cfe/trunk/test/Driver/hip-link-shared-library.hip (original)
+++ cfe/trunk/test/Driver/hip-link-shared-library.hip Fri Jan  4 11:09:20 2019
@@ -9,3 +9,4 @@
 // CHECK: # "amdgcn-amd-amdhsa" - "AMDGCN::Linker", inputs: ["[[OBJ3]]"], 
output: "[[IMG3:.*out]]"
 // CHECK-NOT: offload bundler
 // CHECK: # "x86_64-unknown-linux-gnu" - "GNU::Linker", inputs: ["[[OBJ1]]", 
"{{.*}}/Inputs/in.so", "[[IMG2]]", "[[IMG3]]"], output: "a.out"
+


___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

r335630 - [Test] Initial test commit access

2018-06-26 Thread Aaron Enye Shi via cfe-commits 
Author: aaronenyeshi
Date: Tue Jun 26 10:12:29 2018
New Revision: 335630

URL: http://llvm.org/viewvc/llvm-project?rev=335630&view=rev
Log:
[Test] Initial test commit access

Modified:
cfe/trunk/lib/Driver/ToolChains/HIP.cpp

Modified: cfe/trunk/lib/Driver/ToolChains/HIP.cpp
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Driver/ToolChains/HIP.cpp?rev=335630&r1=335629&r2=335630&view=diff
==
--- cfe/trunk/lib/Driver/ToolChains/HIP.cpp (original)
+++ cfe/trunk/lib/Driver/ToolChains/HIP.cpp Tue Jun 26 10:12:29 2018
@@ -254,7 +254,7 @@ HIPToolChain::TranslateArgs(const llvm::
 
   for (Arg *A : Args) {
 if (A->getOption().matches(options::OPT_Xarch__)) {
-  // Skip this argument unless the architecture matches BoundArch
+  // Skip this argument unless the architecture matches BoundArch.
   if (BoundArch.empty() || A->getValue(0) != BoundArch)
 continue;
 


___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

r335634 - [HIP] Remove hip/hc.amdgcn.bc from HIP Toolchains

2018-06-26 Thread Aaron Enye Shi via cfe-commits 
Author: aaronenyeshi
Date: Tue Jun 26 10:40:36 2018
New Revision: 335634

URL: http://llvm.org/viewvc/llvm-project?rev=335634&view=rev
Log:
[HIP] Remove hip/hc.amdgcn.bc from HIP Toolchains

Summary:
The hc.amdgcn.bc and hip.amdgcn.bc are removed in VDI build and no longer 
needed.

Reviewers: yaxunl

Reviewed By: yaxunl

Subscribers: cfe-commits

Differential Revision: https://reviews.llvm.org/D48455

Modified:
cfe/trunk/lib/Driver/ToolChains/HIP.cpp

Modified: cfe/trunk/lib/Driver/ToolChains/HIP.cpp
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Driver/ToolChains/HIP.cpp?rev=335634&r1=335633&r2=335634&view=diff
==
--- cfe/trunk/lib/Driver/ToolChains/HIP.cpp (original)
+++ cfe/trunk/lib/Driver/ToolChains/HIP.cpp Tue Jun 26 10:40:36 2018
@@ -75,12 +75,12 @@ const char *AMDGCN::Linker::constructLLV
 std::string ISAVerBC =
 "oclc_isa_version_" + SubArchName.drop_front(3).str() + ".amdgcn.bc";
 
-BCLibs.append({"hip.amdgcn.bc", "hc.amdgcn.bc", "opencl.amdgcn.bc",
+BCLibs.append({"opencl.amdgcn.bc",
"ockl.amdgcn.bc", "irif.amdgcn.bc", "ocml.amdgcn.bc",
"oclc_finite_only_off.amdgcn.bc",
"oclc_daz_opt_off.amdgcn.bc",
"oclc_correctly_rounded_sqrt_on.amdgcn.bc",
-   "oclc_unsafe_math_off.amdgcn.bc", "hc.amdgcn.bc", 
ISAVerBC});
+   "oclc_unsafe_math_off.amdgcn.bc", ISAVerBC});
   }
   for (auto Lib : BCLibs)
 addBCLib(C, Args, CmdArgs, LibraryPaths, Lib);


___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

r335765 - [HIP] Support flush denormals bitcode

2018-06-27 Thread Aaron Enye Shi via cfe-commits 
Author: aaronenyeshi
Date: Wed Jun 27 11:58:55 2018
New Revision: 335765

URL: http://llvm.org/viewvc/llvm-project?rev=335765&view=rev
Log:
[HIP] Support flush denormals bitcode

Summary:
Use oclc_daz_opt_on.amdgcn.bc bitcode when option fcuda-flush-denormal-to-zero 
is enabled, otherwise use oclc_daz_opt_off.amdgcn.bc bitcode. Added lit tests 
to verify that the correct bitcode is linked when -fcuda-flush-denormal-to-zero 
option is enabled or disabled.

Reviewers: yaxunl, scchan, b-sumner

Reviewed By: yaxunl, scchan, b-sumner

Subscribers: cfe-commits, yaxunl

Differential Revision: https://reviews.llvm.org/D48493

Added:
cfe/trunk/test/Driver/Inputs/hip_dev_lib/
cfe/trunk/test/Driver/Inputs/hip_dev_lib/irif.amdgcn.bc
cfe/trunk/test/Driver/Inputs/hip_dev_lib/ockl.amdgcn.bc

cfe/trunk/test/Driver/Inputs/hip_dev_lib/oclc_correctly_rounded_sqrt_on.amdgcn.bc
cfe/trunk/test/Driver/Inputs/hip_dev_lib/oclc_daz_opt_off.amdgcn.bc
cfe/trunk/test/Driver/Inputs/hip_dev_lib/oclc_daz_opt_on.amdgcn.bc
cfe/trunk/test/Driver/Inputs/hip_dev_lib/oclc_finite_only_off.amdgcn.bc
cfe/trunk/test/Driver/Inputs/hip_dev_lib/oclc_isa_version_803.amdgcn.bc
cfe/trunk/test/Driver/Inputs/hip_dev_lib/oclc_isa_version_900.amdgcn.bc
cfe/trunk/test/Driver/Inputs/hip_dev_lib/oclc_unsafe_math_off.amdgcn.bc
cfe/trunk/test/Driver/Inputs/hip_dev_lib/ocml.amdgcn.bc
cfe/trunk/test/Driver/Inputs/hip_dev_lib/opencl.amdgcn.bc
cfe/trunk/test/Driver/hip-device-libs.hip
Modified:
cfe/trunk/lib/Driver/ToolChains/HIP.cpp

Modified: cfe/trunk/lib/Driver/ToolChains/HIP.cpp
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Driver/ToolChains/HIP.cpp?rev=335765&r1=335764&r2=335765&view=diff
==
--- cfe/trunk/lib/Driver/ToolChains/HIP.cpp (original)
+++ cfe/trunk/lib/Driver/ToolChains/HIP.cpp Wed Jun 27 11:58:55 2018
@@ -75,10 +75,16 @@ const char *AMDGCN::Linker::constructLLV
 std::string ISAVerBC =
 "oclc_isa_version_" + SubArchName.drop_front(3).str() + ".amdgcn.bc";
 
+llvm::StringRef FlushDenormalControlBC;
+if (Args.hasArg(options::OPT_fcuda_flush_denormals_to_zero))
+  FlushDenormalControlBC = "oclc_daz_opt_on.amdgcn.bc";
+else
+  FlushDenormalControlBC = "oclc_daz_opt_off.amdgcn.bc";
+
 BCLibs.append({"opencl.amdgcn.bc",
"ockl.amdgcn.bc", "irif.amdgcn.bc", "ocml.amdgcn.bc",
"oclc_finite_only_off.amdgcn.bc",
-   "oclc_daz_opt_off.amdgcn.bc",
+   FlushDenormalControlBC,
"oclc_correctly_rounded_sqrt_on.amdgcn.bc",
"oclc_unsafe_math_off.amdgcn.bc", ISAVerBC});
   }

Added: cfe/trunk/test/Driver/Inputs/hip_dev_lib/irif.amdgcn.bc
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/Driver/Inputs/hip_dev_lib/irif.amdgcn.bc?rev=335765&view=auto
==
(empty)

Added: cfe/trunk/test/Driver/Inputs/hip_dev_lib/ockl.amdgcn.bc
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/Driver/Inputs/hip_dev_lib/ockl.amdgcn.bc?rev=335765&view=auto
==
(empty)

Added: 
cfe/trunk/test/Driver/Inputs/hip_dev_lib/oclc_correctly_rounded_sqrt_on.amdgcn.bc
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/Driver/Inputs/hip_dev_lib/oclc_correctly_rounded_sqrt_on.amdgcn.bc?rev=335765&view=auto
==
(empty)

Added: cfe/trunk/test/Driver/Inputs/hip_dev_lib/oclc_daz_opt_off.amdgcn.bc
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/Driver/Inputs/hip_dev_lib/oclc_daz_opt_off.amdgcn.bc?rev=335765&view=auto
==
(empty)

Added: cfe/trunk/test/Driver/Inputs/hip_dev_lib/oclc_daz_opt_on.amdgcn.bc
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/Driver/Inputs/hip_dev_lib/oclc_daz_opt_on.amdgcn.bc?rev=335765&view=auto
==
(empty)

Added: cfe/trunk/test/Driver/Inputs/hip_dev_lib/oclc_finite_only_off.amdgcn.bc
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/Driver/Inputs/hip_dev_lib/oclc_finite_only_off.amdgcn.bc?rev=335765&view=auto
==
(empty)

Added: cfe/trunk/test/Driver/Inputs/hip_dev_lib/oclc_isa_version_803.amdgcn.bc
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/Driver/Inputs/hip_dev_lib/oclc_isa_version_803.amdgcn.bc?rev=335765&view=auto
==
(empty)

Added: cfe/trunk/test/Driver/Inputs/hip_dev_lib/oclc_isa_version_900.amdgcn.bc
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/Driver/Inputs/hip_dev_lib/oclc_isa_v

r335774 - [HIP] Fix ordering of device-libs linking

2018-06-27 Thread Aaron Enye Shi via cfe-commits 
Author: aaronenyeshi
Date: Wed Jun 27 12:51:42 2018
New Revision: 335774

URL: http://llvm.org/viewvc/llvm-project?rev=335774&view=rev
Log:
[HIP] Fix ordering of device-libs linking

Summary:
HIP should link the bitcodes with caller functions before callee functions. 
Also added lit test to check the ordering of the linked bitcodes is matches.

Reviewers: yaxunl, b-sumner

Reviewed By: yaxunl, b-sumner

Subscribers: cfe-commits, yaxunl, b-sumner, scchan

Differential Revision: https://reviews.llvm.org/D48667

Modified:
cfe/trunk/lib/Driver/ToolChains/HIP.cpp
cfe/trunk/test/Driver/hip-device-libs.hip

Modified: cfe/trunk/lib/Driver/ToolChains/HIP.cpp
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Driver/ToolChains/HIP.cpp?rev=335774&r1=335773&r2=335774&view=diff
==
--- cfe/trunk/lib/Driver/ToolChains/HIP.cpp (original)
+++ cfe/trunk/lib/Driver/ToolChains/HIP.cpp Wed Jun 27 12:51:42 2018
@@ -82,7 +82,7 @@ const char *AMDGCN::Linker::constructLLV
   FlushDenormalControlBC = "oclc_daz_opt_off.amdgcn.bc";
 
 BCLibs.append({"opencl.amdgcn.bc",
-   "ockl.amdgcn.bc", "irif.amdgcn.bc", "ocml.amdgcn.bc",
+   "ocml.amdgcn.bc", "ockl.amdgcn.bc", "irif.amdgcn.bc",
"oclc_finite_only_off.amdgcn.bc",
FlushDenormalControlBC,
"oclc_correctly_rounded_sqrt_on.amdgcn.bc",

Modified: cfe/trunk/test/Driver/hip-device-libs.hip
URL: 
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/Driver/hip-device-libs.hip?rev=335774&r1=335773&r2=335774&view=diff
==
--- cfe/trunk/test/Driver/hip-device-libs.hip (original)
+++ cfe/trunk/test/Driver/hip-device-libs.hip Wed Jun 27 12:51:42 2018
@@ -21,6 +21,7 @@
 
 
 // COM: [[LLVM_LINK:"*.llvm-link"]]
+// COM-SAME: {{.*}} "{{.*}}ocml.amdgcn.bc" "{{.*}}ockl.amdgcn.bc" 
"{{.*}}irif.amdgcn.bc"
 // FLUSHD-SAME: {{.*}} "{{.*}}oclc_daz_opt_on.amdgcn.bc"
 // NOFLUSHD-SAME: {{.*}} "{{.*}}oclc_daz_opt_off.amdgcn.bc"
 // COM-SAME: {{.*}} "-o" "{{.*}}-gfx900-linked-{{.*bc}}"


___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits