[PATCH] D18127: Remove compile time PreserveName in favor of a runtime cc1 -discard-value-names option

[David Blaikie via cfe-commits]

Interesting question going forwards: We usually try to make test cases
+/-Asserts agnostic (not using named values), now that we have a flag for
it, should we not bother with that & just add the command line argument to
enable names when names make testing easier? Or do we think that'll make
tests more brittle/depend on unnecessary implementation details?

On Sat, Mar 12, 2016 at 6:00 PM, Mehdi AMINI via cfe-commits <
cfe-commits at lists.llvm.org> wrote:

> joker.eph created this revision.
> joker.eph added reviewers: echristo, chandlerc.
> joker.eph added a subscriber: cfe-commits.
>
> This flag is enabled by default in the driver when NDEBUG is set. It
> is forwarded on the LLVMContext to discard all value names (but
> GlobalValue) for performance purpose.
>
> This an improved version of D18024
>
> http://reviews.llvm.org/D18127
>
> Files:
>   include/clang/Driver/CC1Options.td
>   include/clang/Frontend/CodeGenOptions.def
>   lib/CodeGen/CGBuilder.h
>   lib/CodeGen/CGCall.cpp
>   lib/CodeGen/CGExpr.cpp
>   lib/CodeGen/CodeGenFunction.cpp
>   lib/CodeGen/ModuleBuilder.cpp
>   lib/Driver/Tools.cpp
>   lib/Frontend/CompilerInvocation.cpp
>   test/CodeGen/mips-byval-arg.c
>   test/CodeGen/mips-vector-arg.c
>   test/CodeGen/mips-zero-sized-struct.c
>   test/CodeGen/mips64-class-return.cpp
>   test/CodeGen/mips64-padding-arg.c
>   test/CodeGenCXX/debug-info-class.cpp
>   test/CodeGenCXX/stack-reuse.cpp
>
>
> _______________________________________________
> cfe-commits mailing list
> cfe-commits at lists.llvm.org
> http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20160313/557b1ce5/attachment.html>