Marking select(2) as restrict

[Eitan Adler]

After this entire thread here is the summary. If I've misrepresented
you here please let me know.

Stefan Blachmann - concerned about possible breakages.
kevlo - no comment
julian at freebsd.org - no comment
imp@ - concerned about lack of warnings; objected to original fix of devd
Gary Jennejohn  - pselect already has the modifiers. no comment.
kib@ - no benefit; concerned fallout could be hard to observe
Garrett Wollman  - positive due to following the standard
Tijil  - glibc already uses it. limited fallout expected
Eric van Gyzen - would prefer "make check" in the exp-run as well
cem@ - concerned about warnings

My thoughts:

The main concern is that things /might/ break in undetectable was.
However most other operating already have these marked as restrict,
and no concrete brokenness for FreeBSD have been shown. I'd rather aim
for correctness than avoiding speculative fallout. This is a tradeoff,
but one I think is good here.

Final action:

I am planning on committing a modified version of the original patch
after an exp-run.

I will ask for port tests to be run if possible.

This will be posted to phab for review for technical correctness. This
will not be merged to stable/11. In the event of observable fallout it
can be reverted.

On 22 February 2018 at 23:26, Eitan Adler <lists at eitanadler.com> wrote:
> On 22 February 2018 at 13:27, Jilles Tjoelker <jilles at stack.nl> wrote:
>> On Wed, Feb 21, 2018 at 03:27:21PM -0500, Garrett Wollman wrote:
>>> <<On Wed, 21 Feb 2018 22:10:02 +0200, Konstantin Belousov <kostikbel at gmail.com> said:
> ...
>
> FWIW
>
> glibc: already done -
> https://github.com/bminor/glibc/blob/master/misc/sys/select.h#L101
> openbsd: already done
> https://github.com/openbsd/src/blob/master/sys/sys/select.h#L128
> dragonflyBSD: alredy done:
> https://github.com/dragonflybsd/dragonflybsd/blob/master/sys/sys/select.h#L50
> netbsd: already done:
> https://github.com/NetBSD/src/blob/trunk/sys/sys/select.h#L69
>
> The patch that started this thread was incomplete, as it didn't
> include the timeval but I'll have an updated version in a phab soon.
>
> --
> Eitan Adler



-- 
Eitan Adler