Re: Automated report: NetBSD-current/i386 test failure

Robert Elz <[email protected]> Mon, 23 Mar 2026 12:50:05 +0700
Newsgroups gmane.os.netbsd.current
Message-ID <[email protected]>
    Date:        Mon, 23 Mar 2026 01:51:38 +0000 (UTC)
    From:        NetBSD Test Fixture <[email protected]>
    Message-ID:  <[email protected]>

  | The newly failing test cases are:
  |     bin/sh/t_expand:dollar_at_empty_and_conditional
  |     bin/sh/t_fsplit:split_arith

This is behaving exactly as anticipated (planned), and is verifying
that the shell $* change is operating as intended.

I will fix those two tests (which are presuming behaviour not promised,
which previously was implemented illogically, and is now corrected) later
today.

The tests will still be checking private NetBSD shell behaviour, not
guaranteed by the standard, but they will go back to verifying what our
shell (and others which behave rationally) generate (and so will not fail
when testing the NetBSD shell any more, as they didn't before).

kre