Re: test fails
- From: Ilya Zakharevich <nospam-abuse@xxxxxxxxx>
- Date: Sat, 17 May 2008 12:27:05 +0000 (UTC)
[A complimentary Cc of this posting was sent to
sisyphus
<sisyphus359@xxxxxxxxx>], who wrote in article <b1f6dc92-a95c-4cc7-8b25-1508a84cc632@xxxxxxxxxxxxxxxxxxxxxxxxxxxx>:
On May 12, 3:38=A0am, George Weinberg <eorgeweinb...@xxxxxxxxxxxxx>
wrote:
sisyphus <sisyphus...@xxxxxxxxx> wrote in news:9fed7dcc-6b35-444a-8abe-=2E
8c73295c9...@xxxxxxxxxxxxxxxxxxxxxxxxxxx:
=2E
# Looks like your test died just after 29.
However, I was expecting to see a segfault or at least an error
message (as usually happens when there's a problem with the cleanup).
Not too sure what to make of that.
Global cleanup (or END) may have (explicitly) set $?, or called
exit(). Either from Perl code, or from XS.
Catching a call to exit() from Perl is easy, via overwriting
CORE::GLOBAL::exit(). Catching $? - I do not know, but one could put
END {warn "Value of \$? is $?" if $?}
at the start of the test script. And/or at least
END {warn "Last call to END{} reached..."}
And if it prints this, this happens during global destruction...
Hope this helps,
Ilya
.
- References:
- test fails
- From: George Weinberg
- Re: test fails
- From: sisyphus
- Re: test fails
- From: George Weinberg
- Re: test fails
- From: sisyphus
- test fails
- Prev by Date: install modules error
- Next by Date: Re: Convert MS-Word to plain text
- Previous by thread: Re: test fails
- Next by thread: Perl module for xslt parser
- Index(es):