when (not) use bugzilla for GCC?

[Joseph S. Myers]

On Sun, 25 Oct 2009, Basile STARYNKEVITCH wrote:

> I cannot understand when should I use or not bugzilla. More precisely, I have
> several examples of "bugs" but I didn't use bugzilla for them

If something is a bug in trunk then you should report it, but do not 
submit reports that do not contain sufficient information to convince the 
reader that there is a bug.  For example, a report should not rely on 
source files on a third-party website, installed system headers, or on the 
hypothetical correctness of some unreduced code (if you are claiming wrong 
code, reduce it to a single function or a minimal set of functions).

For development branches, it's up to the branch maintainer.  If they wish 
to use Bugzilla for bugs in them, they should add appropriate versions so 
the bugs can be accurately reported as bugs against the branch.

> 2. a couple of weeks ago, I submitted a patch regarding gengtype for plugins
> (the patch was greatly improved by Rafael Espindola espindola@google.com) and
> the patch did correct a bug.
> http://gcc.gnu.org/ml/gcc-patches/2009-10/msg00379.html
> but I never bothered registering it (since making the initial patch was easier
> than reporting a bug). This is also the case of
> http://gcc.gnu.org/ml/gcc-patches/2009-10/msg00809.html
> (fatal error when failing to register a pass)

If you want a regression to be tracked as such, it needs a Bugzilla entry, 
even if there is also a patch submission.  Any bug with a patch can be 
filed if desired, but do not use Bugzilla as a patch submission mechanism; 
send patches to gcc-patches and use the "patch URL" field in the bugs.  
Ping patches on gcc-patches weekly until they are reviewed.  If you attach 
a patch in progress to Bugzilla, make clear that it is a work in progress 
and that a final version will be submitted to gcc-patches.

If you stop working on a bug, do not leave yourself assigned to it; 
ASSIGNED should only be for bugs where you are genuinely working on them.

> 3. On the MELT branch, I regularily find bugs & correct them, and it did
> happen that someone registered a MELT bug on the bugzilla; I did correct it
> but AFAIR was not permitted to close the bug. Sometimes I am sad of not being
> able to report MELT bugs on bugzilla and to close them when I feel I have
> covered them.

Log in to Bugzilla with your gcc.gnu.org address to get access to close 
bugs, add versions and milestones, etc.

-- 
Joseph S. Myers
joseph@codesourcery.com