[PATCH] D19827: Do not disable completely loop unroll when optimizing for size.

[Marianne Mailhot-Sarrasin via cfe-commits]

mamai retitled this revision from "Do not disable completely loop unroll according to optimization level." to "Do not disable completely loop unroll when optimizing for size.".
mamai updated the summary for this revision.
mamai updated this revision to Diff 56133.
mamai added a comment.

Modified the patch not to affect /O1 optimization level.


Repository:
  rL LLVM

http://reviews.llvm.org/D19827

Files:
  lib/Frontend/CompilerInvocation.cpp

Index: lib/Frontend/CompilerInvocation.cpp
===================================================================
--- lib/Frontend/CompilerInvocation.cpp
+++ lib/Frontend/CompilerInvocation.cpp
@@ -513,7 +513,7 @@
     getAllNoBuiltinFuncValues(Args, Opts.NoBuiltinFuncs);
   Opts.UnrollLoops =
       Args.hasFlag(OPT_funroll_loops, OPT_fno_unroll_loops,
-                   (Opts.OptimizationLevel > 1 && !Opts.OptimizeSize));
+                   (Opts.OptimizationLevel > 1));
   Opts.RerollLoops = Args.hasArg(OPT_freroll_loops);
 
   Opts.DisableIntegratedAS = Args.hasArg(OPT_fno_integrated_as);


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D19827.56133.patch
Type: text/x-patch
Size: 599 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20160504/4263c760/attachment.bin>