[Openmp-commits] [PATCH] D14031: [OpenMP] Initial implementation of OpenMP offloading library - libomptarget.

[Jonas Hahnfeld via Openmp-commits]

Hahnfeld added a comment.

I'm fine otherwise - but definitely not the one for any final decision!


================
Comment at: libomptarget/cmake/Modules/LibomptargetGetDependencies.cmake:62-77
@@ +61,18 @@
+  
+################################################################################
+# Looking for libffi...
+################################################################################
+
+find_path (
+  LIBOMPTARGET_DEP_LIBFFI_INCLUDE_DIR
+  NAMES
+    ffi.h
+  PATHS
+    /usr/include
+    /usr/local/include
+    /opt/local/include
+    /sw/include
+    ENV CPATH
+  PATH_SUFFIXES
+    libffi)
+
----------------
In previous versions you have been using PkgConfig - is there a particular reason that this was removed?

(I'm asking because RHEL 6 is using a rather weird path: `/usr/lib64/libffi-3.0.5/include/ffi.h`)

================
Comment at: libomptarget/src/omptarget.cpp:197-198
@@ +196,4 @@
+
+  for (HostDataToTargetListTy::iterator ii = HostDataToTargetMap.begin(),
+                                        ie = HostDataToTargetMap.end();
+       ii != ie; ++ii) {
----------------
`auto` and range-based loop?

================
Comment at: libomptarget/src/omptarget.cpp:227-228
@@ +226,4 @@
+  // Check if the pointer is contained
+  for (HostDataToTargetListTy::iterator ii = HostDataToTargetMap.begin(),
+                                        ie = HostDataToTargetMap.end();
+       ii != ie; ++ii) {
----------------
`auto` and range-based loop?

================
Comment at: libomptarget/src/omptarget.cpp:257-258
@@ +256,4 @@
+  // Check if the pointer is contained in any sub-nodes
+  for (HostDataToTargetListTy::iterator ii = HostDataToTargetMap.begin(),
+                                        ie = HostDataToTargetMap.end();
+       ii != ie; ++ii) {
----------------
`auto` and range-based loop?

================
Comment at: libomptarget/src/omptarget.cpp:1022-1029
@@ +1021,9 @@
+
+////////////////////////////////////////////////////////////////////////////////
+// temporary for debugging (matching the ones in omptarget-nvptx
+
+EXTERN void __kmpc_kernel_print(char *title) { DP(" %s\n", title); }
+
+EXTERN void __kmpc_kernel_print_int8(char *title, int64_t data) {
+  DP(" %s val=%lld\n", title, (long long)data);
+}
----------------
Should this be committed for development?


http://reviews.llvm.org/D14031