[Libomptarget] Add RPC-based printf implementation for OpenMP #85638

Summary:
Relanding after reverting, only applies to AMDGPU for now.

This patch adds an implementation of printf that's provided by the GPU
C library runtime. This pritnf currently implemented using the same
wrapper handling that OpenMP sets up. This will be removed once we have
proper varargs support.

This printf differs from the one CUDA offers in that it is synchronous
and uses a finite size. Additionally we support pretty much every
format specifier except the %n option.

Depends on #85331

Author: jhuber6

openmp/libomptarget/DeviceRTL/CMakeLists.txt

@@ -122,6 +122,11 @@ set(clang_opt_flags -O3 -mllvm -openmp-opt-disable -DSHARED_SCRATCHPAD_SIZE=512
 set(link_opt_flags  -O3        -openmp-opt-disable -attributor-enable=module -vectorize-slp=false )
 set(link_export_flag -passes=internalize -internalize-public-api-file=${source_directory}/exports)
 
+# If the user built with the GPU C library enabled we will use that instead.
+if(${LIBOMPTARGET_GPU_LIBC_SUPPORT})
+  list(APPEND clang_opt_flags -DOMPTARGET_HAS_LIBC)
+endif()
+
 # Prepend -I to each list element
 set (LIBOMPTARGET_LLVM_INCLUDE_DIRS_DEVICERTL "${LIBOMPTARGET_LLVM_INCLUDE_DIRS}")
 list(TRANSFORM LIBOMPTARGET_LLVM_INCLUDE_DIRS_DEVICERTL PREPEND "-I")

openmp/libomptarget/DeviceRTL/src/LibC.cpp

@@ -25,13 +25,26 @@ int32_t omp_vprintf(const char *Format, void *Arguments, uint32_t) {
 } // namespace impl
 #pragma omp end declare variant
 
-// We do not have a vprintf implementation for AMD GPU yet so we use a stub.
 #pragma omp begin declare variant match(device = {arch(amdgcn)})
+
+#ifdef OMPTARGET_HAS_LIBC
+// TODO: Remove this handling once we have varargs support.
+extern "C" struct FILE *stdout;
+extern "C" int32_t rpc_fprintf(FILE *, const char *, void *, uint64_t);
+
+namespace impl {
+int32_t omp_vprintf(const char *Format, void *Arguments, uint32_t Size) {
+  return rpc_fprintf(stdout, Format, Arguments, Size);
+}
+} // namespace impl
+#else
+// We do not have a vprintf implementation for AMD GPU so we use a stub.
 namespace impl {
 int32_t omp_vprintf(const char *Format, void *Arguments, uint32_t) {
   return -1;
 }
 } // namespace impl
+#endif
 #pragma omp end declare variant
 
 extern "C" {