comp.lang.ada
 help / color / mirror / Atom feed
From: Ludovic Brenta <ludovic.brenta@insalien.org>
Subject: Re: A bug in gnat/gcc 3.3.3?
Date: Thu, 06 Jan 2005 20:07:10 +0100
Date: 2005-01-06T20:07:11+01:00	[thread overview]
Message-ID: <87sm5eml5t.fsf@insalien.org> (raw)
In-Reply-To: 1104972862.577700.268560@f14g2000cwb.googlegroups.com

"Adrian Hoe" writes:
> Florian Weimer wrote:
>> * Ludovic Brenta:
>>
>> >> Yes! 3.4 is currently the best option. 3.15 is old, 3.3 is buggy
> and
>> >> 4.0 is still fighting with the tree-saa extentsion.
>> >
>> > I differ with this.  The quality of GCC 3.4 is similar to, but not
>> > significantly better than that of GNAT 3.15p.  Also, 3.4 is much
>> > slower and memory-hungry than 3.15p (it may or may not matter to
> you
>> > but it's a fact).
>>
>> But this comparison is not fair because you are backporting fixes to
>> GNAT 3.15p.  No one is doing this for GNAT in GCC 3.4. 8-)
>>
>> Due to the NPTL problem, plain GNAT 3.15p is probably not so useful
>> for many people with recent GNU/Linux distributions.
>
> Could you elaborate further your last statement? Why do you think
> gnat3.15p is not so useful with recent GNU/Linux distribution?

This issue has been discussed at length here.  It is summarised in the
corresponding Debian bug report:

http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=245312

The workaround is to declare the environment variable
LD_ASSUME_KERNEL=2.4.1

or any other value less than 2.6.  The problem is on linux >= 2.6 with
glibc >= 2.3, and on some 2.4 kernels provided by Red Hat.  The patch
that fixes this problem in GNAT 3.15p is quite small.

> Also, does gnat3.15p work well with gcc3.3.3?
>
> If I'm going to bootstrap gnat 3.15p with gnat/gcc3.3.3, wouldn't the
> buggy 3.x affect the compilation? Has anyone tried to bootstrap
> gnat3.15p with gnat/gcc3.x before?

I have done that without problems.  The bootstrapping has three
stages; the bootstrap compiler is used only in the first stage, to
build a non-optimised new compiler.  Then the new compiler recompiles
itself, with optimisations, two times (stages 2 and 3) and compares
the output.  If the two recompiles gave the same results, the
bootstrap is considered successful.

-- 
Ludovic Brenta.


This patch makes it possible to use the New POSIX Thread Library on
Linux >= 2.6 and glibc >= 2.3.  It also works on linux 2.4.  I
backported it from GCC 3.4.

I omitted irrelevant parts of this patch to keep it minimal.

-- 
Ludovic Brenta.



  parent reply	other threads:[~2005-01-06 19:07 UTC|newest]

Thread overview: 31+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2005-01-04 15:58 A bug in gnat/gcc 3.3.3? Adrian Hoe
2005-01-04 16:13 ` Duncan Sands
2005-01-04 16:32   ` Adrian Hoe
2005-01-06 15:01     ` Martin Krischik
2005-01-04 18:01 ` Martin Krischik
2005-01-05  4:45   ` Adrian Hoe
2005-01-05  4:59     ` Bobby D. Bryant
2005-01-31  4:45       ` Adrian Hoe
2005-01-05  8:44     ` Martin Krischik
2005-01-05 21:22       ` Ludovic Brenta
2005-01-05 21:47         ` Duncan Sands
2005-01-05 22:22         ` Florian Weimer
2005-01-06  0:54           ` Adrian Hoe
2005-01-06 10:26             ` Martin Krischik
2005-01-06 19:07             ` Ludovic Brenta [this message]
2005-01-06 19:17           ` Ludovic Brenta
2005-01-06  1:54         ` Bobby D. Bryant
2005-01-06 10:28         ` Martin Krischik
2005-01-06 20:22           ` Ludovic Brenta
2005-01-06 21:29             ` Jerome Hugues
2005-01-06 23:04               ` Ludovic Brenta
2005-01-07  8:40                 ` Jerome Hugues
2005-01-07 16:15             ` Martin Krischik
2005-01-05 12:20     ` Pascal Obry
2005-01-05 21:21       ` Ludovic Brenta
2005-01-04 19:16 ` Ludovic Brenta
2005-01-05  4:25   ` Adrian Hoe
2005-01-05 21:19     ` Ludovic Brenta
2005-01-07 16:22 ` Adrian Hoe
2005-01-07 18:13   ` Ludovic Brenta
2005-01-07 23:32   ` Ludovic Brenta
replies disabled

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