This is the mail archive of the
mailing list for the GCC project.
GCC 3.0 Bugs: Call for Volunteers
- To: gcc at gcc dot gnu dot org
- Subject: GCC 3.0 Bugs: Call for Volunteers
- From: Mark Mitchell <mark at codesourcery dot com>
- Date: Mon, 12 Feb 2001 16:00:53 -0800
- Organization: CodeSourcery, LLC
As of now, there are 955 open bugs in the GNATS database. Before GCC
3.0, we need to go through these bugs, and examine each to determine
whether or not it needs to be fixed before GCC 3.0.
As discussed earlier, the way we will do this is to mark bugs that
need to be fixed as having "high" priority. For now, a bug should
have "high" priority if (and only if) it represents a regression
relative to GCC 2.95..
Therefore, we should apply the following algorith to each open bug:
- Check if it is really a bug. (Did the bug-reporter misunderstand
the language standard, for example?)
If not, close it.
- Check if it is already fixed.
If so, close it.
- If it is not fixed, check to see whether or not it occurs with
If so, and the priority of the bug is presently "high" reset it to
"medium". Otherwise, do nothing.
- Otherwise, the bug is not fixed, and represents a regression from
GCC 2.95.2. How embarassing!
In that case, mark the bug as having "high" priority.
This is a largely mechanical task. Only the first step requires
knowledge of the source language, and for bugs that result in compiler
crashes, even that is unnecessary.
If you have GNATS write access (which is very similar to having CVS
write access), I would very much appreciate it if you would volunteer.
Please send me (privately) the number of bugs that you are willing to
analyze in the next week or so, and I will assign you a range of the
GNATS buglist. Then, you can process that range. If there are bugs
you are unable to analyze in that range (because, for example, they
involve a machine to which you do not have access, or a language which
you do not know well enough to determine whether a bug report is
valid), please keep a record of those bugs, and when you are all done,
give me that list.
If you do not have GNATS write access, please see:
If you are interested in helping, it is likely that GNATS write access
will be set up for you.
If you can, please volunteer for at least 50 open bugs. I think that
with that level of committement, we can quickly get a sense of how
many high-priority bugs are outstanding. We will then fix these bugs,
simultaneously addressing the compile-time performance issues.
Thank you in advance for volunteering!
Mark Mitchell firstname.lastname@example.org
CodeSourcery, LLC http://www.codesourcery.com