Re: [jamboost] Regression test: link-fail problem?

From: David Abrahams (david.abrahams_at_[hidden])
Date: 2002-07-08 20:28:01

This is exactly the sort of thing I was afraid would arise. AFAICT there's
no simple way to put this in the build system without making core Jam
changes.

The basic problem looks like this. Given a dependency chain:

a->b->c->d->e

when building "e" fails, you want something to happen with the
.test/.success/.failure files. These are associated with A. However, the
build system just stops working on "a" the instant any of its dependencies
fails to build. In Boost.Build v2, it will be much easier to deal with
this, because we can traverse the dependency graph generated for A
ourselves, and attach actions to erase the result files at all the leaves.

----- Original Message -----
From: "Beman Dawes" <bdawes_at_[hidden]>
To: <jamboost_at_[hidden]>
Sent: Monday, July 08, 2002 8:38 PM
Subject: [jamboost] Regression test: link-fail problem?

> In trying to track down some of the broken "Fail" links in
> http://www.esva.net/~beman/jam_regr.html, I realized that
> class_concept_fail_expected.cpp is a "link-fail" test, so it isn't
> surprising that there is nothing in the log file. After all, there is an
> up-to-date .exe file. Of course, on a link-fail test, that isn't good
> news. I'll deal with the problem by adding some boilerplate text that
says
> on a link-fail test which fails, there aren't any messages to inspect
> because everything worked (even though it shouldn't have.)
>
> But in tracking that down, I spotted a more serious problem:
> concept_check_fail_expected.cpp (which is also "link-fail") is reported
as
> succeeding (there is a .success file) even though the code failed to
> compile!
>
> The compile failure causes the test to be reported as a success (because,
> after all, failure was expected). In other words, testing.jam should be
> smart enough to figure out that a compile failure doesn't represent
success
> on a "link-fail" test.
>
> We need to do something about this one, since it is causing an incorrect
> "Pass" to be reported.
>
> --Beman
>
>
>
> To unsubscribe from this group, send an email to:
> jamboost-unsubscribe_at_[hidden]
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>