[gmsh] bump version to 4.12.2

[maltezfaria]

No description provided.

[giordano]

Error on 32-bit windows is

[07:36:47] /workspace/srcdir/gmsh-4.12.1-source/src/common/GmshMessage.cpp:93:12: error: redefinition of ‘int vsnprintf(char*, size_t, const char*, va_list)’
[07:36:47]    93 | static int vsnprintf(char *str, size_t size, const char *fmt, va_list ap)
[07:36:47]       |            ^~~~~~~~~
[07:36:47] In file included from /opt/i686-w64-mingw32/i686-w64-mingw32/sys-root/include/locale.h:12,
[07:36:47]                  from /opt/i686-w64-mingw32/i686-w64-mingw32/include/c++/9.1.0/clocale:42,
[07:36:47]                  from /workspace/srcdir/gmsh-4.12.1-source/src/common/GmshMessage.cpp:16:
[07:36:47] /opt/i686-w64-mingw32/i686-w64-mingw32/sys-root/include/stdio.h:444:5: note: ‘int vsnprintf(char*, size_t, const char*, char*)’ previously defined here
[07:36:47]   444 | int vsnprintf (char *__stream, size_t __n, const char *__format, __builtin_va_list __local_argv)
[07:36:47]       |     ^~~~~~~~~
[07:36:47] /workspace/srcdir/gmsh-4.12.1-source/src/common/GmshMessage.cpp:93:12: warning: ‘int vsnprintf(char*, size_t, const char*, va_list)’ defined but not used [-Wunused-function]
[07:36:47]    93 | static int vsnprintf(char *str, size_t size, const char *fmt, va_list ap)
[07:36:47]       |            ^~~~~~~~~

Relevant source code is

#if defined(_MSC_VER) && (_MSC_VER >= 1310) //NET 2003
#define vsnprintf _vsnprintf
#else
#if defined(HAVE_NO_VSNPRINTF)
static int vsnprintf(char *str, size_t size, const char *fmt, va_list ap)
{
  if (strlen(fmt) > size - 1) { // just copy the format
    strncpy(str, fmt, size - 1);
    str[size - 1] = '\0';
    return size;
  }
  return vsprintf(str, fmt, ap);
}
#endif
#endif

They're redefining vsnprintf and they're doing it wrong. There's also the question of why HAVE_NO_VSNPRINTF is defined, sounds like the relevant test is failing incorrectly.

[maltezfaria]

Thanks for looking into this! I opened an issue on the Gmsh end.

[maltezfaria]

@giordano I tried to fix some of the issues upstream in this branch, but I am not sure how to test that it works here. I tried using a GitSource in my last commit but all builds failed....

Thanks for the help!

[maltezfaria]

@giordano I tried to fix some of the issues upstream in this branch, but I am not sure how to test that it works here. I tried using a GitSource in my last commit but all builds failed....

Thanks for the help!

Sorry for the noise, it was a silly mistake on my side.

[maltezfaria]

Tests on 32-bit windows pass now after the patch in gmsh version 4.12.2. I think this is ready for a review/merge.

[giordano]

Thanks for your contribution!