On 11/01/2016 04:39 PM, Bruno Haible wrote:
All the new change does was to remove this forward-caution,
right?
Yes. It's also sideways caution, in the sense that neither kernel might
be "newer" than the other.
Paul Eggert wrote:
> > Should this change be accompanied with a documentation change
> > in doc/posix-functions/utimensat.texi ?
>
> It's not strictly needed, as Gnulib still works around bugs in older
> systems, it's just that 'configure' now defaults to assuming the older
> bugs don't exist in
On 11/01/2016 11:20 AM, Bruno Haible wrote:
Should this change be accompanied with a documentation change
in doc/posix-functions/utimensat.texi ?
It's not strictly needed, as Gnulib still works around bugs in older
systems, it's just that 'configure' now defaults to assuming the older
bugs do
Hi Paul,
> We no longer need to worry
> about Linux kernel 2.6.32 when building with newer kernels.
Should this change be accompanied with a documentation change
in doc/posix-functions/utimensat.texi ?
And what about m4/futimens.m4 and doc/posix-functions/futimens.texi ?
Bruno
* lib/utimensat.c (rpl_utimensat): Update FIXME comment.
* m4/utimensat.m4 (gl_FUNC_UTIMENSAT): Belatedy do a "simplify
this in 2012" FIXME, by assuming the file system bug is absent
unless demonstrated to be present. We no longer need to worry
about Linux kernel 2.6.32 when building with newer ke
