Re: [PATCH 4/5] Cygwin: add fast-path for posix_spawn(p)

On Sat, Mar 14, 2026 at 4:30=E2=80=AFAM Mark Geisert <[email protected]> wrot=
e:
>
> On Fri, 13 Mar 2026, Sebastian Feld wrote:
>
> > On Thu, Mar 12, 2026 at 3:45 PM Jon Turney <[email protected]=
>
> wrote:
> >>
> >> On 09/03/2026 09:54, Sebastian Feld wrote:
> >>> Was this work ever merged into Cygwin1.dll?
> >>
> >> Unfortunately, not.  And Jeremy seems to have moved on to other ways t=
o
> >> apply his talents.
> >>
> >> It would be ideal if someone else would pick up that work and get it
> >> finished off.
> >
> > That would require a cygwin.dll expert beyond my skill set.
> >
> > What about adding the current work as build option?
>
> If/when "someone" can be found for this work, it would be far better to
> finish the work so it can be tested and merged.
>
> I can't imagine providing a build option for an unfinished, unsupported,
> branch of the Cygwin DLL to "release" a lightning-rod feature to users wh=
o
> won't know how to make use of the incomplete code.  That just sounds like
> more future work for us, to be honest.
>
> You've reminded us of this unfinished work so it's again visible to us.
> Thank you for that.

Monthly reminder that the posix_spawn() work is still not finished.
but needed. because the fork(),exec() performance of Cygwin is very
very bad.

Sebi
--=20
Sebastian Feld - IT security consultant