[PATCH] D16082: [CUDA] Invoke ptxas and fatbinary during compilation.

Eric Christopher via cfe-commits cfe-commits at lists.llvm.org
Mon Jan 11 14:39:29 PST 2016

echristo added a comment.

One question inline, one nit, and one more question here: You've got a couple of checks inline for null names/architectures, where do you expect those to come from and can you test for them? Or, another question, is if they're multiple architectures shouldn't we be able to see all of the actions that arise from each gpu?

Or I'm missing something, either way an explanation is good. :)


Comment at: lib/Driver/Driver.cpp:1880
@@ +1879,3 @@
+    // retrieve the Input's GPU arch.
+    II.setAction(A);
+    return II;
Weren't you just adding it as part of the InputInfo constructor in 16078?

Comment at: lib/Driver/ToolChains.cpp:4261
@@ -4255,2 +4260,3 @@
-  DAL->AddJoinedArg(nullptr, Opts.getOption(options::OPT_march_EQ), BoundArch);
+  if (BoundArch) {
+    DAL->AddJoinedArg(nullptr, Opts.getOption(options::OPT_march_EQ), BoundArch);
Nit: No braces around single lines.


More information about the cfe-commits mailing list