This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Status of PINE compilation on Solaris bug?
- To: mark at codesourcery dot com (Mark Mitchell)
- Subject: Re: Status of PINE compilation on Solaris bug?
- From: Joe Buck <jbuck at synopsys dot COM>
- Date: Mon, 21 May 2001 13:08:54 -0700 (PDT)
- Cc: jbuck at synopsys dot COM, jsm28 at cam dot ac dot uk, gcc at gcc dot gnu dot org
> >>>>> "Joe" == Joe Buck <jbuck@synopsys.COM> writes:
>
> Joe> If a regression test is created for this failure, then, if it
> Joe> is not fixed in the trunk, it will show up as a regression
> Joe> against 3.0 when we go to ship 3.1. If that's none, it
> Joe> should be safe to close the GNATS report.
>
> I think you're suggesting an extension to our normal policy.
> Normally, we check in test-cases with fixes. In this case, the
> problem was already fixed. Traditionally, we've assumed that this
> meant there was likely already a test-case for whatever the problem
> was.
>
> Do you think that we should add a test-case even for a bug that is
> already fixed when we go to look at it?
My suggestion applies only in cases like this where we know that we have
multiple branches (counting the trunk as a branch) and that a bug exists
in some but not in others. Adding a test means we won't forget about the
problem.