r323345 - [CUDA] Disable PGO and coverage instrumentation in NVPTX.

[Artem Belevich via cfe-commits]

Author: tra
Date: Wed Jan 24 09:41:02 2018
New Revision: 323345

URL: http://llvm.org/viewvc/llvm-project?rev=323345&view=rev
Log:
[CUDA] Disable PGO and coverage instrumentation in NVPTX.

NVPTX does not have runtime support necessary for profiling to work
and even call arc collection is prohibitively expensive. Furthermore,
there's no easy way to collect the samples. NVPTX also does not
support global constructors that clang generates if sample/arc collection
is enabled.

Differential Revision: https://reviews.llvm.org/D42452

Added:
    cfe/trunk/test/Driver/cuda-no-pgo-or-coverage.cu
Modified:
    cfe/trunk/lib/Driver/ToolChains/Clang.cpp

Modified: cfe/trunk/lib/Driver/ToolChains/Clang.cpp
URL: http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Driver/ToolChains/Clang.cpp?rev=323345&r1=323344&r2=323345&view=diff
==============================================================================
--- cfe/trunk/lib/Driver/ToolChains/Clang.cpp (original)
+++ cfe/trunk/lib/Driver/ToolChains/Clang.cpp Wed Jan 24 09:41:02 2018
@@ -3625,7 +3625,11 @@ void Clang::ConstructJob(Compilation &C,
       options::OPT_finstrument_function_entry_bare))
     A->render(Args, CmdArgs);
 
-  addPGOAndCoverageFlags(C, D, Output, Args, CmdArgs);
+  // NVPTX doesn't support PGO or coverage. There's no runtime support for
+  // sampling, overhead of call arc collection is way too high and there's no
+  // way to collect the output.
+  if (!Triple.isNVPTX())
+    addPGOAndCoverageFlags(C, D, Output, Args, CmdArgs);
 
   if (auto *ABICompatArg = Args.getLastArg(options::OPT_fclang_abi_compat_EQ))
     ABICompatArg->render(Args, CmdArgs);

Added: cfe/trunk/test/Driver/cuda-no-pgo-or-coverage.cu
URL: http://llvm.org/viewvc/llvm-project/cfe/trunk/test/Driver/cuda-no-pgo-or-coverage.cu?rev=323345&view=auto
==============================================================================
--- cfe/trunk/test/Driver/cuda-no-pgo-or-coverage.cu (added)
+++ cfe/trunk/test/Driver/cuda-no-pgo-or-coverage.cu Wed Jan 24 09:41:02 2018
@@ -0,0 +1,34 @@
+// Check that profiling/coverage arguments doen't get passed down to device-side
+// compilation.
+//
+// REQUIRES: clang-driver
+//
+// XRUN: %clang -### -target x86_64-linux-gnu -c --cuda-gpu-arch=sm_20 \
+// XRUN:   -fprofile-generate %s 2>&1 | \
+// XRUN: FileCheck --check-prefixes=CHECK,PROF %s
+//
+// RUN: %clang -### -target x86_64-linux-gnu -c --cuda-gpu-arch=sm_20 \
+// RUN:   -fprofile-instr-generate %s 2>&1 | \
+// RUN: FileCheck --check-prefixes=CHECK,PROF %s
+//
+// RUN: %clang -### -target x86_64-linux-gnu -c --cuda-gpu-arch=sm_20 \
+// RUN:   -coverage %s 2>&1 | \
+// RUN: FileCheck --check-prefixes=CHECK,GCOV %s
+//
+// RUN: %clang -### -target x86_64-linux-gnu -c --cuda-gpu-arch=sm_20 \
+// RUN:   -ftest-coverage %s 2>&1 | \
+// RUN: FileCheck --check-prefixes=CHECK,GCOV %s
+//
+// RUN: %clang -### -target x86_64-linux-gnu -c --cuda-gpu-arch=sm_20   \
+// RUN:   -fprofile-instr-generate -fcoverage-mapping %s 2>&1 | \
+// RUN: FileCheck --check-prefixes=CHECK,PROF,GCOV %s
+//
+//
+// CHECK-NOT: error: unsupported option '-fprofile
+// CHECK-NOT: error: invalid argument
+// CHECK-DAG: "-fcuda-is-device"
+// CHECK-NOT: "-f{{[^"]*coverage.*}}"
+// CHECK-NOT: "-fprofile{{[^"]*}}"
+// CHECK: "-triple" "x86_64--linux-gnu"
+// PROF-DAG: "-fprofile{{.*}}"
+// GCOV-DAG: "-f{{(coverage|emit-coverage).*}}"