On 10/25/2017 11:54 AM, Reid Kleckner wrote:
On Wed, Oct 25, 2017 at 10:56 AM, Friedman, Eli
<efrie...@codeaurora.org <mailto:efrie...@codeaurora.org>> wrote:
On 10/6/2017 4:09 PM, Saleem Abdulrasool via cfe-commits wrote:
Modified: cfe/trunk/lib/Driver/ToolChains/Clang.cpp
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Driver/ToolChains/Clang.cpp?rev=315126&r1=315125&r2=315126&view=diff
<http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Driver/ToolChains/Clang.cpp?rev=315126&r1=315125&r2=315126&view=diff>
==============================================================================
--- cfe/trunk/lib/Driver/ToolChains/Clang.cpp (original)
+++ cfe/trunk/lib/Driver/ToolChains/Clang.cpp Fri Oct 6
16:09:55 2017
@@ -2601,6 +2601,33 @@ static void RenderModulesOptions(Compila
Args.AddLastArg(CmdArgs,
options::OPT_fmodules_disable_diagnostic_validation);
}
+static void RenderCharacterOptions(const ArgList &Args,
const llvm::Triple &T,
+ ArgStringList &CmdArgs) {
+ // -fsigned-char is default.
+ if (const Arg *A = Args.getLastArg(options::OPT_fsigned_char,
+ options::OPT_fno_signed_char,
+ options::OPT_funsigned_char,
+ options::OPT_fno_unsigned_char)) {
+ if (A->getOption().matches(options::OPT_funsigned_char) ||
+ A->getOption().matches(options::OPT_fno_signed_char)) {
+ CmdArgs.push_back("-fno-signed-char");
+ }
+ } else if (!isSignedCharDefault(T)) {
+ CmdArgs.push_back("-fno-signed-char");
+ }
+
+ if (const Arg *A = Args.getLastArg(options::OPT_fshort_wchar,
+ options::OPT_fno_short_wchar)) {
+ if (A->getOption().matches(options::OPT_fshort_wchar)) {
+ CmdArgs.push_back("-fwchar-type=short");
+ CmdArgs.push_back("-fno-signed-wchar");
+ } else {
+ CmdArgs.push_back("-fwchar-type=int");
+ CmdArgs.push_back("-fsigned-wchar");
+ }
+ }
+}
It looks like this changes the behavior of "-fno-short-wchar".
Specifically, on targets where the default wchar_t type is
"unsigned int" (like ARM AAPCS), "-fno-short-wchar" used to be a
no-op, but now it changes the sign of wchar_t. Why are we
overriding the default here?
Overriding the default was more or less the intention, but it was for
bitwidth, not signedness. The idea was that if you pass
-fno-short-wchar, then you want the 4 byte one.
Users probably don't care about signedness when using 4 byte wchars,
so we could probably remove the -fsigned-wchar flag from the else
block and take the original sign of wchar_t when overriding the width
in Basic. They probably want an 'unsigned short' when -fshort-wchar is
passed, though.
Maybe we should just ignore -fno-short-wchar, instead? I think that's
what gcc and released versions of clang do (that means -fno-short-wchar
doesn't do anything for Windows targets, but that's probably fine).
-Eli
--
Employee of Qualcomm Innovation Center, Inc.
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, a Linux
Foundation Collaborative Project
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
