Re: robust chdir and openat (works with names longer than PATH_MAX)

[Paul Eggert]

"Alfred M. Szmidt" <address@hidden> writes:

> Is this code only used if the real chdir() implementation is incapable
> of handling infintly long directory names?  By using a autoconf check
> or whatever?

Yes, that's the current status of the code, and none of the proposed
changes should affect this.

Of course there could be bugs in the current autoconf test -- which
I've never tested on the Hurd -- but they should be fixable if they
are present.