[PATCH 1/2] Revert "stdlib: Support malloc-managed environ arrays for compatibility"

H.J. Lu hjl.tools@gmail.com
Fri Jan 24 22:19:26 GMT 2025


On Fri, Jan 24, 2025 at 7:31 PM Florian Weimer <fweimer@redhat.com> wrote:
>
> * H. J. Lu:
>
> > The BZ #32588 test should be standalone, independent of
> > the second patch.   Here is the v2 patch with getenv.
>
> Fair enough.  I've pushed my revert (only).
>
> > +#include <stdlib.h>
> > +#include <support/check.h>
> > +
> > +extern char **environ;
> > +
> > +int
> > +do_test (void)
> > +{
> > +  char *valp;
> > +  static char *dummy_environ[] = { NULL };
> > +  environ = dummy_environ;
> > +  setenv ("A", "1", 0);
> > +  valp = getenv ("A");
> > +  TEST_VERIFY_EXIT (valp[0] == '1' && valp[1] == '\0');
> > +  return 0;
> > +}
> > +
> > +#include <support/test-driver.c>
>
> This looks okay, although it's not clear to me why you do not use
> TEST_COMPARE_STRING.  If you want to minimize includes (not unreasonable
> for this test), use
>
>   return valp[0] == '1' && valp[1] == '\0';

Did you mean

return valp[0] == '1' && valp[1] == '\0' ? 0 : 1;

But stdlib/tst-setenv-environ.out will be empty when it fails.

TEST_VERIFY_EXIT (valp[0] == '1' && valp[1] == '\0');

tells you what failed.

> and drop <support/check.h>.
>
> Patchwork prefers if patches are posted to new threads, I think.

Will do.

> Thanks,
> Florian
>


-- 
H.J.


More information about the Libc-alpha mailing list