[PATCH] glob: Simplify the interface for the GLOB_ALTDIRFUNC callback gl_readdir

[Paul Eggert]

On 04/11/2016 02:19 PM, Roland McGrath wrote:
> I'm not sure whether there
> was ever a standard for readdir that said anything about d_ino different
> from what the current POSIX.1 says (which indeed disallows d_ino==0
> entries).

I don't think there was either.

> My recollection is that early versions of 1003.1 did not specify
> d_ino/d_fileno at all.

Yes, that's right. I just checked, and 1003.1-1988 and 1003.1b-1993 both 
specify only d_name.