Re: bin/60016 (make(1) tests failing since sjg MaybeSubMake changes)

["Simon J. Gerraty" <sjg%juniper.net@localhost>]

Christos Zoulas <christos%zoulas.com@localhost> wrote:

> Index: job-output.mk
> ===================================================================
> RCS file: /cvsroot/src/usr.bin/make/unit-tests/job-output.mk,v
> retrieving revision 1.2
> diff -u -u -r1.2 job-output.mk
> --- job-output.mk       13 Jun 2025 06:13:20 -0000      1.2
> +++ job-output.mk       28 Feb 2026 20:24:01 -0000
> @@ -28,7 +28,7 @@
>  # expect: This is stderr.
>  # expect: only stderr:
>  # expect: end stdout-and-stderr
> -stdout-and-stderr:
> +stdout-and-stderr: .MAKE

This fixes it?
Interesting.

I would not expect .MAKE to be needed, since ${MAKE} below *should* be
recognized, and cause OP_SUBMAKE to be set.

Adding .MAKEFLAGS: -dg2 at the end we see

# all, unmade, type OP_DEPENDS|OP_PHONY|OP_HAS_COMMANDS, flags none
# empty-lines, unmade, type OP_DEPENDS|OP_PHONY|OP_HAS_COMMANDS, flags none
# stdout-and-stderr, made, type OP_DEPENDS|OP_SUBMAKE|OP_HAS_COMMANDS|OP_DEPS_FOUND|OP_MARK, flags REMAKE|DONE_WAIT|DONE_ALLSRC|DONE_SUBMAKE|DONECYCLE
#

but that's on a box where the test passes.
If you have netbsd-i386 can you see what you get for
.MAKEFLAGS: -dg2 and without the .MAKE above

Thanks!
--sjg