[vlc-devel] [3.0 PATCH v2 18/20] configure: Make the new value of _WIN32_WINNT available to configure tests

[Steve Lhomme]

> On 2 Apr 2020, at 16:45, Martin Storsjö <martin at martin.st> wrote:
> 
> On Thu, 2 Apr 2020, Steve Lhomme wrote:
> 
>> From: Martin Storsjö <martin at martin.st>
>> 
>> Definitions from AH_TOP aren't included in the configure checks.
>> 
>> This was the original intent of e807c8e91.
>> 
>> This makes sure configure checks and actual compilation are run
>> with the same version defined.
>> 
>> EDIT: adapted to the default for 3.0 (0x502 or XP SP2)
>> 
>> (cherry picked from commit 0a27328b5ac2bc0e417c70f048131f783b201d04)
>> ---
>> configure.ac | 11 ++++++++---
>> 1 file changed, 8 insertions(+), 3 deletions(-)
>> 
>> diff --git a/configure.ac b/configure.ac
>> index c60fd78e116..6b9325a5266 100644
>> --- a/configure.ac
>> +++ b/configure.ac
>> @@ -218,9 +218,14 @@ case "${host_os}" in
>>    CXXFLAGS="${CXXFLAGS} -static-libgcc"
>>    AC_CHECK_TOOL(WINDRES, windres, :)
>>    AC_CHECK_TOOL(OBJCOPY, objcopy, :)
>> -    AH_TOP([#if defined(_WIN32) && !defined(_WIN32_WINNT)])
>> -    AH_TOP([# define _WIN32_WINNT 0x0502 /* Windows XP SP2 */])
>> -    AH_TOP([#endif])
>> +    AC_PREPROC_IFELSE([AC_LANG_PROGRAM(
>> +      [[#ifdef _WIN32_WINNT
>> +       # error _WIN32_WINNT already defined
>> +       #endif
>> +      ]],[[;]])
>> +    ],[
>> +      AC_DEFINE([_WIN32_WINNT], 0x0502, [Define to '0x0502' for Windows XP SP2.])
>> +    ])
> 
> Most of these patches look great - but don't you need a number of more patches as well (e.g. for fixing building contrib/qt?). I have a ready patchset for backporting qt patches to make Qt 5.6 work with the very latest llvm-mingw - I can provide those patches later today if it's of any use.

For now I'm trying to fix the winrt builds. There is no Qt there. But I have a branch with Qt 5.12.7 ready to merge in 3.0 once 3.0.9 is out.

> For this patch, this ends up defining _WIN32_WINNT to 0x0502 on a toolchain that otherwise would default to a higher value.
> 
> I can send a different variant of the patch.
> 
> // Martin