Date: Wed, 10 Feb 2021 23:05:18 +0100
Joseph,
on Wed, 10 Feb 2021 20:33:24 +0000 you (Joseph Myers
<joseph_at_[hidden]>) wrote:
> My concern with those changes was not with adding new operations,
good to hear, but at that time I perceived your feedback on that aspect
also as quite discouraging
> it
> was that there were changes included in N2389 "Clean up atomics,
> non-normative changes" that seemed unambiguously normative to me
> (disallowing a trap from atomic integer /= 0, for example, so
> requiring implementations to generate more complicated code around
> such integer compound assignment to deal with such cases - anything
> requiring implementation changes is surely a normative change). (I
> also think such additional requirements on existing operations are of
> questionable merit.)
All of that was not my invention, and they were not meant as normative
changes. In the C standard we have a totally bogus phrase that "atomic
operations will not have undefined behavior" (or so) and I only tried
to coin this double negation into some realistic and comprehensible
normative text.
I interpreted "not having undefined behavior" as "has only defined
behavior" and so for me such operations are not supposed to produce
traps, for example. I have described this reasoning in the papers and
in session at several occasions. There was not even consensus what
these parts of the standard meant, and the least how we could improve
the situation.
> Other committee members may of course have had other concerns with
> those papers.
Yes, sure. Unfortunately, many committee members rediscovered their
own concerns each time newly and freshly when we discussed these
papers in session and they were not able or willing to give structured
feedback such we could move foreward.
Reviewing something as complex as the model for atomics seems outside
of the current capacities of WG14. We simply don't have the
necessary work-force and commitment of our members to pursue such an
enterprise.
Jens
on Wed, 10 Feb 2021 20:33:24 +0000 you (Joseph Myers
<joseph_at_[hidden]>) wrote:
> My concern with those changes was not with adding new operations,
good to hear, but at that time I perceived your feedback on that aspect
also as quite discouraging
> it
> was that there were changes included in N2389 "Clean up atomics,
> non-normative changes" that seemed unambiguously normative to me
> (disallowing a trap from atomic integer /= 0, for example, so
> requiring implementations to generate more complicated code around
> such integer compound assignment to deal with such cases - anything
> requiring implementation changes is surely a normative change). (I
> also think such additional requirements on existing operations are of
> questionable merit.)
All of that was not my invention, and they were not meant as normative
changes. In the C standard we have a totally bogus phrase that "atomic
operations will not have undefined behavior" (or so) and I only tried
to coin this double negation into some realistic and comprehensible
normative text.
I interpreted "not having undefined behavior" as "has only defined
behavior" and so for me such operations are not supposed to produce
traps, for example. I have described this reasoning in the papers and
in session at several occasions. There was not even consensus what
these parts of the standard meant, and the least how we could improve
the situation.
> Other committee members may of course have had other concerns with
> those papers.
Yes, sure. Unfortunately, many committee members rediscovered their
own concerns each time newly and freshly when we discussed these
papers in session and they were not able or willing to give structured
feedback such we could move foreward.
Reviewing something as complex as the model for atomics seems outside
of the current capacities of WG14. We simply don't have the
necessary work-force and commitment of our members to pursue such an
enterprise.
Jens
-- :: INRIA Nancy Grand Est ::: Camus ::::::: ICube/ICPS ::: :: ::::::::::::::: office Strasbourg : +33 368854536 :: :: :::::::::::::::::::::: gsm France : +33 651400183 :: :: ::::::::::::::: gsm international : +49 15737185122 :: :: http://icube-icps.unistra.fr/index.php/Jens_Gustedt ::
Received on 2021-02-10 16:05:26