On Mon, 14 Jun 2021, Martin Storsjö wrote:
On Mon, 14 Jun 2021, Liu Hao wrote:
在 6/14/21 3:09 PM, Martin Storsjö 写道:
Right yes... Normal C++ e.g. declared in a class header are emitted that
way (but that comes with all the extra hairiness regarding inline
functions that might not be emitted and/or optimized out). So unless we
know we can do manual selectany on a plain non-inline C function, I guess
the thing you're proposing is the most robust choice.
Is this patch good for master?
You may propose another patch for the underscore thing.
I sent another patch set for the underscores now, I'd rather have at least
1/3 of that applied before landing this one, but then I'm ok with this one
(updated with the new symbol name).
BTW, one different aspect of this initialization; if you have a DLL that
is built with -funsigned-char, and this is loaded by an EXE that isn't
built that way, the DLL still initializes the lconv info to unsigned mode.
The same happens with MSVC too (as long as building with -MD so that it
uses a shared CRT DLL), even though it seems like MSVC has intended for it
not to override it in that case.
// Martin
_______________________________________________
Mingw-w64-public mailing list
Mingw-w64-public@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/mingw-w64-public
