Committed, but how come this doesn't trip PLIST_DB?
On 2026/10/18 11:43, Marc Espie wrote:
> It's noticeable on any debug package. A very small one is
> sysutils/random_run
>
> Somehow, I'm emitting the directories holding the debug info
> *after* the debug info proper.
>
> Could this be fixed, please ?
>
>
> Index: build-debug-info
> ===================================================================
> RCS file: /build/data/openbsd/cvs/ports/infrastructure/bin/build-debug-info,v
> diff -u -p -r1.50 build-debug-info
> --- build-debug-info 2 Jul 2026 15:38:45 -0000 1.50
> +++ build-debug-info 16 Oct 2026 09:08:08 -0000
> @@ -140,11 +140,11 @@ sub write_debug_info($self, $base, $fh,
>
> sub write_debug_entry($self, $fh, $dir, $dbg, $o)
> {
> - print $fh $dbg, "\n";
> if (!exists $o->{dirstash}{$dir}) {
> print $fh $dir, "\n";
> $o->{dirstash}{$dir} = 1;
> }
> + print $fh $dbg, "\n";
> $o->{ostash}{$dbg} = 1;
> delete $o->{empty};
> }
>