GCC >= 5? (was: [patch] Removal of strfwd.h)

[Jean-Marc Lasgouttes]

Le 04/10/2024 à 15:17, Pavel Sanda a écrit :
> On Fri, Oct 04, 2024 at 12:51:56PM +0200, Enrico Forestieri wrote:
>> On Fri, Oct 04, 2024 at 11:52:01AM +0200, Pavel Sanda wrote:
>>>
>>> I am all for ditching trivstring, but it all depends on Enrico's stance,
>>> I didn't
>>> folow the previous discussion on strings and COW, so I have little idea
>>> what is the right course of action.
>>
>> Cygwin does not need trivstring. It was using the implementation in
>> src/support/trivstring.cpp simply because we wrongly decide that std::string
>> implementation uses cow when _GLIBCXX_USE_CXX11_ABI is set.
>>
>> I think that if we ditch trivstring we don't need to set
>> STD_STRING_USES_COW. In this case, the patch I proposed to check whether cow
>> is actually used or not will also not be necessary.
> 
> So are you able to compile on cygwin after these two patches applied on top
> of each other (first strfwd out, then trivstring out)?

This looks good to me. Note that this whole test program can be removed:

--- a/src/support/tests/check_trivstring.cpp
+++ b/src/support/tests/check_trivstring.cpp
@@ -1,6 +1,5 @@
  #include <config.h>

-#include "../trivstring.h"
  #include "../docstring.h"


JMarc