On Saturday 17 January 2026 13:43:02 Pali Rohár wrote:
> On Tuesday 13 January 2026 13:39:23 LIU Hao wrote:
> > 在 2025-12-29 02:01, Pali Rohár 写道:
> > > This special case does not throw an exception from invalid parameter
> > > handler.
> > > ---
> > > mingw-w64-crt/secapi/getenv_s.c | 6 +++---
> > > mingw-w64-libraries/winstorecompat/src/getenv_s.c | 6 +++---
> > > 2 files changed, 6 insertions(+), 6 deletions(-)
> > >
> > > diff --git a/mingw-w64-crt/secapi/getenv_s.c
> > > b/mingw-w64-crt/secapi/getenv_s.c
> > > index b2731830df98..0b31f418833c 100644
> > > --- a/mingw-w64-crt/secapi/getenv_s.c
> > > +++ b/mingw-w64-crt/secapi/getenv_s.c
> > > @@ -23,15 +23,15 @@ static errno_t __cdecl emu_getenv_s(size_t
> > > *pReturnValue, char *dstBuf, rsize_t
> > > return errno = EINVAL;
> > > }
> > > + /* After passing parameter validation, the errno is not changed. */
> > > +
> > > if (!varName) {
> > > *pReturnValue = 0;
> > > if (dstBuf)
> > > dstBuf[0] = '\0';
> > > - return errno = EINVAL;
> > > + return 0;
> > I don't think this change is correct. Microsoft doc says that `getenv_s()`
> > returns `EINVAL` (and not zero) in this case:
> > https://learn.microsoft.com/en-us/cpp/c-runtime-library/reference/getenv-s-wgetenv-s?view=msvc-170#error-conditions
>
> Here is simple test which proves that it returns 0 in this case:
> https://godbolt.org/z/8sTeWY4dv
>
> I can send a test file with all of those special combinations, so would
> be sure how it really behaves.
Do you still think that this change is really incorrect?
Or do you need some other tests from me?
_______________________________________________
Mingw-w64-public mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/mingw-w64-public
