[PATCH] process the -fno-signed-zeros optimization flag (PR20870)

Sanjay Patel spatel at rotateright.com
Thu Jan 22 14:15:00 PST 2015


================
Comment at: lib/Frontend/CompilerInvocation.cpp:429-430
@@ -428,3 +428,4 @@
                        Args.hasArg(OPT_cl_fast_relaxed_math));
-  Opts.NoSignedZeros = Args.hasArg(OPT_cl_no_signed_zeros);
+  Opts.NoSignedZeros = (Args.hasArg(OPT_cl_no_signed_zeros) ||
+                        Args.hasArg(OPT_fno_signed_zeros));
   Opts.NoZeroInitializedInBSS = Args.hasArg(OPT_mno_zero_initialized_in_bss);
----------------
rsmith wrote:
> We generally try to keep our `-cc1` interface simple and minimal. Can we remove the `OPT_cl_no_signed_zeros` option, and instead create `-fno-signed-zeros` in the CL case from the driver?
Sure, that seems reasonable. It looks like we can accomplish this using an Alias in the td file. I'll update the patch.

http://reviews.llvm.org/D6873

EMAIL PREFERENCES
  http://reviews.llvm.org/settings/panel/emailpreferences/






More information about the cfe-commits mailing list