On Sun, 13 Apr 2025, Pali Rohár wrote:
This patch series is fixing stat functions as was discussed in the long
email thread "stat, _stat, and _stat64i32 issues (was: dirent changes)".
I have not included in this patch series Lasse's followup changes.
In v2 are fixed also ftw and nftw functions which directly calls stat.
And there are lot of other fixes to ensure that the compatibility
functions for older CRT libs behaves like it is expected.
Pali Rohár (13):
crt: Provide emulation of _fstat64, _stat64 and _wstat64 functions
crt: Fix mingw-w64 emulation of _fstat64i32, _stat64i32 and
_wstat64i32 functions
crt: Provide emulation of _fstat32i64, _stat32i64 and _wstat32i64
functions
headers: Remove broken inline stat() and fstat() functions
headers: Remove inline definition of _fstat64i32() and _stat64i32()
functions
headers: Add missing _CRTIMP for _fstat64i32/_stat64i32/_wstat64i32
functions
crt: Fix ABI of mingw-w64 emulation of POSIX stat* functions
crt: Fix ABI of POSIX ftw() and nftw() functions
headers: Fix declaration of POSIX stat/fstat and LFS stat64/fstat64
functions
headers: Fix declaration of POSIX ftw/nftw functions
crt: Fix POSIX stat() with 32-bit st_size for pre-msvcr100 build
crt: Provide emulation of _time64 function
headers: Add support for -D_TIME_BITS=64
All in all, this patchset looks good to me now; the only potential issue
is that compilation fails after patch 9, which gets fixed after patch 10.
It seems possible to just swap the order of the two, though - this at
least compiles successfully for me. Pali, does that work, or is there a
reason why they must come in this order?
Other than that, I'm ok with this patchset and could push it with fixes
for the missing includes. Let me know if I should flip the order of those
two patches or not, before pushing.
// Martin
_______________________________________________
Mingw-w64-public mailing list
Mingw-w64-public@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/mingw-w64-public
