Re: pthread_setname_np API is bad

[Jaromír Doleček <jaromir.dolecek%gmail.com@localhost>]

I'd change this immediatelly using the normal rewrite logic in headers
we use for standard versioning. It would be really good to push this
into netbsd-9 too.

Jaromir

Le ven. 9 août 2019 à 14:14, Christos Zoulas <christos%astron.com@localhost> a écrit :
>
> In article <CAEdQ38FPBEFhCTTBKvZFjQbz-B9=gzzw=tuhfPrDiPoqj3PB3w%mail.gmail.com@localhost>,
> Matt Turner  <mattst88%gmail.com@localhost> wrote:
> >On Tue, Aug 6, 2019 at 7:17 AM Kamil Rytarowski <n54%gmx.com@localhost> wrote:
> >>
> >> On 06.08.2019 07:19, Matt Turner wrote:
> >> > On Mon, Aug 5, 2019 at 10:06 PM Thor Lancelot Simon <tls%panix.com@localhost> wrote:
> >> >>
> >> >> On Fri, Aug 02, 2019 at 09:29:27PM -0700, Matt Turner wrote:
> >> >>>
> >> >>> So great, you made your API incompatible with everyone else for zero
> >> >>> gain. /o\
> >> >>
> >> >> And so forth.
> >> >>
> >> >> Is there some reason for the snotty attitude?  Whether you're right or
> >> >> wrong you're not likely to persuade anyone that way.
> >> >
> >> > Well, yes, actually. It's a waste of my time to deal with useless API
> >> > incompatibilities.
> >> >
> >> > Feel free to disregard my whole email based on the perceived attitude,
> >> > if you're unhappy with it.
> >> >
> >> > I was just trying to do the right thing, and I don't intend to expend
> >> > any more effort advocating for this change. It doesn't matter to me
> >> > whether you take the suggestion or even whether Mesa works on NetBSD,
> >> > but it might matter to you. So, ¯\_(ツ)_/¯
> >> >
> >>
> >> I propose to change the function signatures on the next ABI bump to:
> >>
> >> int pthread_setname_np(pthread_t, const char * restrict format, ...)
> >> __printflike(2, 3);
> >>
> >> int pthread_attr_setname_np(pthread_attr_t *, const char * restrict
> >> format, ...*) __printflike(2, 3);
> >>
> >> This way we will keep API compat for all the current users and keep
> >> compiler checks for safe usage.
> >>
> >> Personally, I find it convenient to use it like pthread_setname_np(t[i],
> >> "thread %d, i) and I would like to keep using it.
> >
> >FWIW, I think that's a better suggestion than either of mine. I support that.
>
> Do we have to wait for an API bump for this? Seems pretty harmless. Although
> it is probably better to have:
>
>     int pthread_setname_np(pthread_t, const char *);
> and
>     int pthread_fmtname_np(pthread_t, const char * restrict format, ...)
>         __printflike(2, 3);
>
> christos
>