24 GCC regressions, 0 new, with your patch on 2001-11-25T06:45:01Z.

[Ian Lance Taylor]

Geoff Keating <geoffk@geoffk.org> writes:

> > From: Ian Lance Taylor <ian@airs.com>
> > Date: 25 Nov 2001 10:48:30 -0800
> > User-Agent: Gnus/5.0808 (Gnus v5.8.8) Emacs/20.7
> > 
> > "GCC regression checker" <regress@maat.cygnus.com> writes:
> > 
> > > With your recent patch, GCC has some regression test failures, which
> > > used to pass.  There are 0 new failures, and 24
> > > failures that existed before and after that patch; 0 failures
> > > have been fixed.
> > 
> > This seems to say that there are no new failures.  So my patch (which
> > was only to the testsuite anyhow) didn't make things worse.  So why
> > send me this message?  I wouldn't mind getting a message if there is a
> > new failure, but I don't understand getting a message when there is no
> > new testsuite failure.
> 
> Oddly enough, this is a FAQ.  See
> < http://people.redhat.com/geoffk/gcc-regression/ >.

Oddly enough, I read that FAQ before sending my message, and it did
not answer my question.

However, now I read it again, and I guess this is meant to be covered
by the last point on the page.  I recommend that you word the mail
message and/or the web page more clearly.  I didn't interpret the mail
I received as indicating that the tester couldn't fully test my patch,
since the message did not say that.  The message said ``[w]ith your
recent patch, GCC has some regression test failures, which used to
pass.''  Technically correct, I suppose, but misleading.

For example, add a paragraph to the mail message along these lines:

    You are receiving this message because your patch may not have
    been fully tested, since there were existing test failures.  You
    will receive another message each time the list of test failures
    changes, until there are no failures [or whatever].

And/or expand the last paragraph in the FAQ:

    The tester sent you mail to warn you that it couldn't fully test
    your patch, because there were pre-existing regressions, and it's
    possible that the regressions were hiding a bug in your patch (for
    instance, if your patch is trying to do something with exception
    handling, and all the exception handling tests were failing before
    your patch went in, then your patch might have a bug but there's
    no way to tell).  This mail will include the list of pre-existing
    test failures; these may of course be completely unrelated to your
    patch, but the tester has no way to tell.  It'll only send you one
    of these until the list of regressions changes.  This may happen
    when unfixed patches are left in the tree.

Ian