gcc bugs

gcc bugs

[Timothy Bedding]

Martin Loewis wrote:
> I guess this would help for about every bug in gcc - although proper
> patches help even more.

Proper patches that are large would require copyright issues to be
addressed I think. Descriptions of how to fix bugs avoid that.


So how do I go about finding a bug to work on?
I see there is a gcc-prs mailing list.

Could I just choose a problem mentioned on that list?


Do I need to worry about duplicating effort?

I see that some emails have a link to the GNATS system.

Is there a different front ends for GNATS?


It would be good to have a web page detailing the bug fixing
procedures that someone like myself should follow.

Cordially
Tim

Re: gcc bugs

[Martin v. Loewis]

> So how do I go about finding a bug to work on?
> I see there is a gcc-prs mailing list.
> 
> Could I just choose a problem mentioned on that list?

That, or the Web interface at http://gcc.gnu.org/cgi-bin/gnatsweb.pl

> Do I need to worry about duplicating effort?

Unlikely - especially if the bug is older. You can also look at the
expected failures in the test suites; nobody is going to work on any
of these any time soon.

If you really want to be cautious, you could announce that you are
going to look at a specific problem before you start doing so.

> I see that some emails have a link to the GNATS system.
> 
> Is there a different front ends for GNATS?

Different from what? There is the Web interface, and an email
interface. There is also a set of command line tools, but they only
work on the local machine, so nobody is using them.

> It would be good to have a web page detailing the bug fixing
> procedures that someone like myself should follow.

That would be good. Do you want to write one?

If you have any questions, don't hesitate to ask (just as you
currently do). There aren't that many new contributors every week, so
we can handle all questions that pop up. GCC internals questions
usually get a higher priority from maintainers than people that just
ask FAQs.

Looking forward to your patches,
Martin

Re: gcc bugs

[Joern Rennecke]

> Unlikely - especially if the bug is older. You can also look at the
> expected failures in the test suites; nobody is going to work on any
> of these any time soon.

I beg to differ.  There are some old bugs where the failure is well
understood.

Re: gcc bugs

[Gerald Pfeifer]

On Mon, 27 Mar 2000, Martin v. Loewis wrote:
>> It would be good to have a web page detailing the bug fixing
>> procedures that someone like myself should follow.
> That would be good. Do you want to write one?
> [...]
> Looking forward to your patches,

Note that Timothy does not have a copyright assignment, but as far as I
understood we are not that strict concerning that when it comes to the
web pages as we are when it comes to actual code, right?

Gerald
-- 
Gerald "Jerry" pfeifer@dbai.tuwien.ac.at http://www.dbai.tuwien.ac.at/~pfeifer/

Re: gcc bugs

[Jeffrey A Law]

  In message < Pine.BSF.4.21.0003280109500.86674-100000@deneb.dbai.tuwien.ac.at >
you write:
  > On Mon, 27 Mar 2000, Martin v. Loewis wrote:
  > >> It would be good to have a web page detailing the bug fixing
  > >> procedures that someone like myself should follow.
  > > That would be good. Do you want to write one?
  > > [...]
  > > Looking forward to your patches,
  > 
  > Note that Timothy does not have a copyright assignment, but as far as I
  > understood we are not that strict concerning that when it comes to the
  > web pages as we are when it comes to actual code, right?
Actually, we need to apply the same rules to the web pages too.  I believe
I've got a web page assignment around here somewhere.  Hmmmmm..

jeff

Re: gcc bugs

[Gerald Pfeifer]

On Mon, 27 Mar 2000, Jeffrey A Law wrote:
> Actually, we need to apply the same rules to the web pages too.  I
> believe I've got a web page assignment around here somewhere.  
> Hmmmmm..

Ah, okay. Thanks for the clarification!

Is there any guideline that I can follow that tells me when an assignment
is required and when a patch is small enough not to need one?

And how can I check whether someone has an assignment on file?

Gerald
-- 
Gerald "Jerry" pfeifer@dbai.tuwien.ac.at http://www.dbai.tuwien.ac.at/~pfeifer/

Re: gcc bugs

[Jeffrey A Law]

  In message < Pine.GSO.4.21.0003281707180.25688-100000@nunki.dbai.tuwien.ac.at >
you write:
  > On Mon, 27 Mar 2000, Jeffrey A Law wrote:
  > > Actually, we need to apply the same rules to the web pages too.  I
  > > believe I've got a web page assignment around here somewhere.  
  > > Hmmmmm..
  > 
  > Ah, okay. Thanks for the clarification!
  > 
  > Is there any guideline that I can follow that tells me when an assignment
  > is required and when a patch is small enough not to need one?
Yes there is.

Start with this URL:

http://www.gnu.org/prep/maintain_toc.html

jeff

Re: gcc bugs

[Tom Tromey]

>>>>> "Martin" == Martin v Loewis <martin@loewis.home.cs.tu-berlin.de> writes:

Martin> There is also a set of command line tools, but they only work
Martin> on the local machine, so nobody is using them.

They can work remotely as well.
I don't know if they are set up this way on the machine though.

Tom

gcc bugs

[Timothy Bedding]

Martin wrote:
> That, or the Web interface at http://gcc.gnu.org/cgi-bin/gnatsweb.pl

Thanks for this. I suggest a link to this on the
bugs.html page.



> Looking forward to your patches,

Well, I guess I would be providing a description of how
to fix the bug rather than the actual patch because of
copyright issues.

Does anyone see a problem with this approach? I think
I read somewhere that providing a description of how to
fix a problem does not involve copyright.


Joern wrote:
> I beg to differ.  There are some old bugs where the failure is well
> understood.

How would you advise me to proceed? Does this apply to any GCC bugs stored
in the GNATS system?

Cordially
Tim