This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [PATCH] accept all C integer types in function parameters referenced by alloc_align (PR 88363)


On Tue, Dec 11, 2018 at 2:46 PM Martin Sebor <msebor@gmail.com> wrote:
> On 12/11/18 11:15 AM, Joseph Myers wrote:
> > On Tue, 11 Dec 2018, Martin Sebor wrote:
> >
> >> I recently brought up the question of the write w/o approval
> >> policy on the gcc list:
> >>
> >>    https://gcc.gnu.org/ml/gcc/2018-11/msg00165.html
> >>
> >> looking for clarification.  Except for Jeff's comment (which
> >> I'm afraid didn't really clarify things), didn't get any.
> >
> > I think "will the person who objects to my work the most be able to find a
> > fault with my fix?" in the policy on obviousness is clear enough.  A
> > policy decision on what is or is not part of a language extension can't be
> > obvious, and nor can determining subtle questions of exactly what the
> > definition of some internal interface is or should be.
>
> Anything that someone might find fault with is so broad as to
> remove the ability to make the judgment in any case.  Reviewers
> have been known to find fault with the slightest things, from
> trivial formatting nits, to punctuation in comments, to names
> of variables, to the location of new tests, to ChangeLogs.
>
> If the policy's intent is not to let people make judgment calls
> then it ought to be updated.  I have no proposal for changes to
> it at the moment, but I don't see how anyone can reasonably
> object to someone posting a patch and saying "if there are no
> objections I will go ahead and commit this change because I think
> it's obviously correct."  If even that is against the policy then
> change it to make that clear (though I sincerely hope that isn't
> so).

I don't think anyone objected to your mail, they were just disagreeing
that the patch was obvious.  That is also a judgment call.  IMO
there's no need to have an ironclad policy here, since the
consequences of a particular change being "wrongly" consider either
obvious or non-obvious are small.

Jason


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]