I already have a diff from brad for this in my queue

-- 
  Sent from a phone, apologies for poor formatting.

On 18 September 2026 22:47:39 Thomas Frohwein <tfrohwein@fastmail.com> wrote:

> Hi,
>
> Found that mariadb also could make use of getexecpath here instead of
> defaulting to returning 1 for failure. Slight wrinkle that with this
> 1-liner, failure returns -1 now, but this is still within the spec:
>
> @return 0, for successful filling of buf, non-zero for failure. */
>
> ok?
>
>
> Index: Makefile
> ===================================================================
> RCS file: /cvs/ports/databases/mariadb/Makefile,v
> diff -u -p -r1.169 Makefile
> --- Makefile	9 Sep 2026 12:22:59 -0000	1.169
> +++ Makefile	18 Sep 2026 21:43:06 -0000
> @@ -8,6 +8,7 @@ VERSION=	12.3.3
> DISTNAME=	mariadb-${VERSION}
> PKGNAME-main=	mariadb-client-${VERSION}
> EPOCH=		1
> +REVISION=	0
>
> CATEGORIES=	databases
> SITES=		https://archive.mariadb.org/${DISTNAME}/source/ \
> Index: patches/patch-mysys_my_getexe_c
> ===================================================================
> RCS file: /cvs/ports/databases/mariadb/patches/patch-mysys_my_getexe_c,v
> diff -u -p -r1.1 patch-mysys_my_getexe_c
> --- patches/patch-mysys_my_getexe_c	6 Mar 2025 11:47:18 -0000	1.1
> +++ patches/patch-mysys_my_getexe_c	18 Sep 2026 21:43:06 -0000
> @@ -1,5 +1,3 @@
> -Do not try getting the executable path of our own pid.
> -
> Index: mysys/my_getexe.c
> --- mysys/my_getexe.c.orig
> +++ mysys/my_getexe.c
> @@ -8,7 +6,7 @@ Index: mysys/my_getexe.c
>    if (sysctl(mib, 4, buf, &size, NULL, 0) == 0)
>      return 0;
> +#elif defined(__OpenBSD__)
> -+  return 1;
> ++  return getexecpath(buf, size);
>  #endif
>
>    if (argv0)