On 2025/06/26 09:26, Stuart Henderson wrote:
> On 2025/06/25 22:04, Volker Schlecht wrote:
> > NodeJS 22.17.0 needs it to build ... ok?
>
> > Index: Makefile
> > ===================================================================
> > RCS file: /cvs/ports/devel/libuv/Makefile,v
> > retrieving revision 1.30
> > diff -u -p -r1.30 Makefile
> > --- Makefile 19 May 2025 12:26:26 -0000 1.30
> > +++ Makefile 25 Jun 2025 20:00:07 -0000
> > @@ -1,12 +1,11 @@
> > COMMENT = multi-platform library for asynchronous I/O
> >
> > -VER = 1.50.0
> > +VER = 1.51.0
> > DISTNAME = libuv-v${VER}
> > PKGNAME = libuv-${VER}
> > CATEGORIES = devel
> > -REVISION = 2
> >
> > -SHARED_LIBS += uv 5.1 # 1.0
> > +SHARED_LIBS += uv 5.2 # 1.0
>
> why bump?
Other than that, cmake and nvim seem happy, and I verified that the
"unix,sunos: prefer SO_REUSEPORT for load balancing" does only apply to
(newer) Solaris - so apart from the bump I think it's ok.
> > HOMEPAGE = https://libuv.org/
> >
> > Index: distinfo
> > ===================================================================
> > RCS file: /cvs/ports/devel/libuv/distinfo,v
> > retrieving revision 1.11
> > diff -u -p -r1.11 distinfo
> > --- distinfo 19 Feb 2025 17:16:32 -0000 1.11
> > +++ distinfo 25 Jun 2025 20:00:07 -0000
> > @@ -1,2 +1,2 @@
> > -SHA256 (libuv-v1.50.0.tar.gz) =
> > bcKzmqf6ZsBOLgFb9Hc4uE4ca4a4mH1X3UMtFJ137iU=
> > -SIZE (libuv-v1.50.0.tar.gz) = 1340083
> > +SHA256 (libuv-v1.51.0.tar.gz) =
> > XwVXuQsRBt5xlRo8OTHeXgQw142h2aECh+vHo/eO+Os=
> > +SIZE (libuv-v1.51.0.tar.gz) = 1343638
> > Index: patches/patch-src_unix_openbsd_c
> > ===================================================================
> > RCS file: patches/patch-src_unix_openbsd_c
> > diff -N patches/patch-src_unix_openbsd_c
> > --- patches/patch-src_unix_openbsd_c 2 Mar 2025 20:50:12 -0000
> > 1.4
> > +++ /dev/null 1 Jan 1970 00:00:00 -0000
> > @@ -1,21 +0,0 @@
> > -Index: src/unix/openbsd.c
> > ---- src/unix/openbsd.c.orig
> > -+++ src/unix/openbsd.c
> > -@@ -211,8 +211,16 @@ int uv_cpu_info(uv_cpu_info_t** cpu_infos, int* count)
> > -
> > - which[1] = HW_CPUSPEED;
> > - size = sizeof(cpuspeed);
> > -- if (sysctl(which, ARRAY_SIZE(which), &cpuspeed, &size, NULL, 0))
> > -+ cpuspeed = 0;
> > -+ /*
> > -+ * HW_CPUSPEED can return EOPNOTSUPP if cpuspeed is 0,
> > -+ * so ignore that and continue the flow, because we
> > -+ * still care about the rest of the CPU info.
> > -+ */
> > -+ if (sysctl(which, ARRAY_SIZE(which), &cpuspeed, &size, NULL, 0) &&
> > -+ (errno != EOPNOTSUPP)) {
> > - goto error;
> > -+ }
> > -
> > - size = sizeof(info);
> > - for (i = 0; i < numcpus; i++) {
>
