[PATCH] D18999: [ELF/LTO] Parallel Codegen for LLD

[Rafael Ávila de Espíndola via llvm-commits]

rafael added a comment.

>From 28m10.310s to 22m27.724s when using 4 threads is a pretty small speedup.

Any idea what is slow? Is it just that each of the 1/4 sized chunks is still slow? Is the split not even in compile time?

What is the impact in output size. Even better if you have performance numbers of the linked program.


================
Comment at: ELF/LTO.cpp:40
@@ +39,3 @@
+  SmallString<128> FileName = Config->OutputFile;
+  FileName += (Many) ? utostr(I) : "";
+  FileName += ".lto.o";
----------------
Drop the () around many.

================
Comment at: ELF/LTO.cpp:142
@@ -138,1 +141,3 @@
 
+static SubtargetFeatures getFeatures(Triple &TheTriple) {
+  SubtargetFeatures Features;
----------------
Is this related to parallel LTO?


http://reviews.llvm.org/D18999