comp.lang.ada
 help / color / mirror / Atom feed
From: Ludovic Brenta <ludovic.brenta@insalien.org>
Subject: Re: Ada compiler
Date: 15 Apr 2004 12:41:48 GMT
Date: 2004-04-15T12:41:48+00:00	[thread overview]
Message-ID: <2004415-144148-721184@foorum.com> (raw)
In-Reply-To: mailman.2.1082014724.314.comp.lang.ada@ada-france.org



[about reporting bugs in GCC 3.2.3 to bugzilla]

> I think that if you find a bug with 3.2.3, then you should check if
> the bug still exists in gcc CVS before reporting it to the gcc
> people.  What is the point of telling the gcc developers about bugs
> they already fixed?

This is indeed the policy of GCC developers; they do not support any
version of GCC but the latest release (3.3.3 as of now).

Before filing a bug to GCC, you should not only check it against the
latest release, but also make sure that this bug was not introduced by
the distribution; i.e. you should only file bugs if you use an
unmodified GCC.  Also, you should search the bug database to see if it
has been reported before.

Still, if you find a bug in 3.2.3 that is fixed in 3.3.3, I think you
should file it and close it immediately, and provide any workaround
you know of for the version in question.  The fact that this bug was
fixed is valuable information that deserves to be public.  The
workaround, if any, is even more valuable.

> There is however a point in reporting such bugs to your distribution
> (Debian etc), since it is sometimes possible to backport fixes from
> CVS to gnat 3.15p or whatever

Yes.  The distribution acts as the front line of support for all the
software therein; if you find a bug in any software that came with
your distro, you normally report it to the distro.  The main reason
for this is that distros often patch the software, and the bug may or
may not originate from upstream.

If it turns out that the bug does come from upstream, the maintainer
(for Debian that would be Matthias Klose or myself) will forward the
bug to upstream, i.e. report@gnat.com or the GCC bugzilla database.
If you *know* that the bug comes from upstream, then you help us by
submitting the bug there as well as in the distro's bug database.

Another way in which bug databases are useful is to permit tracing
patches to bugs; i.e. by looking at the bug log, you can find out
which change to GCC fixed a particular bug.  Without this information,
it is very difficult to backport fixes.

I just filed a bunch of bug reports in the Debian database for GNAT
3.15p.  There are several more that I have to file as well.  The next
steps are to try and reproduce them with GCC 3.2, 3.3 and 3.4.  Those
not fixed should then be tagged upstream and forwarded to GCC's
bugzilla.  I will appreciate any help with this; if you wish to
contribute some of your time, please add information to the bugs in
the Debian database:

http://bugs.debian.org/cgi-bin/pkgreport.cgi?pkg=gnat

> (I do this myself: I'm running 3.15p + patches).

I have already applied one of your patches to gnat in Debian (see the
changelog if you don't believe me).  But I would be very interested in
your other patches, if any.  Here again, the bug database is a useful
tool for traceability.

>> Moreover, several GNU/Linux distributions still in widespread >>
>> use ship GCC 3.2.x (e.g. Red Hat 9, Slackware 9.1, Debian Sarge,
>> and Red Hat Enterprise Linux which Sergey, the OP, uses).
>
> People compiling Ada would do better to stick to 3.15p.

Yes, but they don't necessarily know that and use whatever their
distribution chose for them.

-- 
Ludovic Brenta.

-- 
Ce message a ete poste via la plateforme Web club-Internet.fr
This message has been posted by the Web platform club-Internet.fr

http://forums.club-internet.fr/



  reply	other threads:[~2004-04-15 12:41 UTC|newest]

Thread overview: 65+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2004-04-09 14:35 Ada compiler Sergey
2004-04-09 16:50 ` Jacob Sparre Andersen
2004-04-09 16:53 ` Mark H Johnson
2004-04-12 17:08   ` Marc A. Criley
2004-04-13 13:13     ` Martin Krischik
2004-04-14  9:23     ` Sergey
2004-04-14 12:32       ` Marc A. Criley
2004-04-14 13:59       ` Ludovic Brenta
2004-04-14 14:27         ` Duncan Sands
2004-04-14 21:42           ` Ludovic Brenta
2004-04-15  7:38             ` Duncan Sands
2004-04-15 12:41               ` Ludovic Brenta [this message]
2004-04-17 12:22                 ` Duncan Sands
     [not found]                 ` <200404171422.10879.baldrick@free.fr>
2004-04-17 13:04                   ` Ludovic Brenta
     [not found]                   ` <87zn9aycvc.fsf@insalien.org>
2004-04-17 18:24                     ` Duncan Sands
2004-04-15  9:53         ` Sergey
2004-04-15 10:30           ` Martin Krischik
  -- strict thread matches above, loose matches on Subject: below --
2006-10-04 19:34 James E. Lowery
2006-10-04 21:45 ` Ludovic Brenta
2006-10-06  0:29 ` Stephen Leake
2006-10-07 16:22 ` trg
2006-10-09 11:27 ` Gerd
2003-10-15 13:12 Ada Compiler Claire Wong
2003-10-15 13:28 ` Stephane Richard
2003-10-17 14:53   ` Adrian Hoe
2003-10-16  1:52 ` Steve
2003-10-16  8:18   ` Preben Randhol
2003-10-16 16:34     ` Stephen Leake
2003-10-17  2:28     ` Steve
2002-02-28 19:12 Ada compiler Somesh Agrawal
2002-02-28 19:30 ` Dave Poirier
2002-02-28 21:19 ` Pascal Obry
1998-07-12  0:00 Ramsey Khair
1998-07-12  0:00 ` Markus Kuhn
1997-11-05  0:00 ADA Compiler pjo
1997-11-05  0:00 ` Robert Dewar
1996-12-27  0:00 ada compiler Patrick Baker
1996-12-28  0:00 ` ElecWeb
     [not found] <01bba2c2$e035d4c0$84dc11cc@default>
1996-09-14  0:00 ` ADA compiler Aron Felix Gurski
1996-09-15  0:00 ` kjprice
1996-09-19  0:00   ` William Lane
1996-09-20  0:00     ` David C. Hoos, Sr.
1996-09-20  0:00 ` Jim Dorman
1996-09-20  0:00   ` Samuel Tardieu
1996-09-14  0:00 Ada compiler M Quintin
1996-09-12  0:00 ada compiler nathan richmond
1996-09-12  0:00 ` John Howard
1996-09-16  0:00   ` Michael Feldman
1996-09-12  0:00 ` kjp
1996-07-09  0:00 Ada compiler Nazzario
1996-07-10  0:00 ` Jon S Anthony
1996-07-10  0:00 ` John Howard
1996-07-11  0:00   ` Michael Feldman
1996-07-11  0:00 ` David Wheeler
1996-03-25  0:00 ADA compiler "Mike Schofield"
1996-03-26  0:00 ` Samuel Abbe
1995-03-28 17:02 Ada Compiler Browett
1995-03-16 17:51 First Year of the HBAP Ada WWW Server Tucker Taft
1995-03-21 19:24 ` Ada compiler JG TAYLOR
1995-03-22 22:17   ` Tucker Taft
1994-11-12  1:17 Ada Compiler icsia184
1994-09-28  9:12 Tung-lin Tsai
1993-07-06 23:33 John L Hendricks
1991-01-22  5:15 ADA compiler Eric DeWald
1990-07-30 19:46 Captain Moonrock
1989-10-24 18:22 Ada Compiler Ron Baalke
replies disabled

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox