comp.lang.ada
 help / color / mirror / Atom feed
From: Martin Krischik <martin@krischik.com>
Subject: Re: [Announce] ASIS for GNAT gcc-3.4.4-20041123 released
Date: Mon, 29 Nov 2004 17:42:13 +0100
Date: 2004-11-29T17:42:13+01:00	[thread overview]
Message-ID: <1379629.3TST3z19Ev@linux1.krischik.com> (raw)
In-Reply-To: pan.2004.11.29.11.05.09.319617@n_o_p_o_r_k_a_n_d_h_a_m.abyss2.demon.co.uk

Luke A. Guest wrote:

> On Mon, 29 Nov 2004 09:25:19 +0100, Martin Krischik wrote:
> 
>> To the correct position?
> 
> Yes, but it doesn't work.
> 
>> echo $GNATSRC
>> /work/gnu_3.4/gcc/gcc/ada
>> 
>> It's not my choice! adabrowse and GLADE use GNATSRC as well.
>> 
>>> make gnat
>>> make -e -C gnat
>>> make[1]: Entering directory
>>> `/usr/src/gcc-builds/gcc-341/asis-3_4_4-20041123.src/gnat'
>>> make[1]: *** No rule to make target `version.c', needed by `gnat'. 
>>> Stop. make[1]: Leaving directory
>>> `/usr/src/gcc-builds/gcc-341/asis-3_4_4-20041123.src/gnat' make: ***
>>> [gnat] Error 2
>>> 
>>> I copied version.c/h manually.
>> 
>> You should not need that:
> 
> Why is it in the makefile then?

I meant: You should not need to copy the files since the makefile is
instructed to do for you.

>> -- snip gnat/Makefile
>> -- snap gnat/Makefile
> 
> What is that supposed to show?

As I said: That the d*mm makefile should have copied it for you. And I can't
see why it did not.
  
>>> After "make install"
>>> 
>>> gnatbind -aO./ -aO/opt/ada/include/asis -static -I- -x
>>> gnatelim-driver.ali error: "g-string.ali" not found, "g-string.adb" must
>>> be compiled gnatmake: *** bind failed.
>>> make[1]: *** [gnatelim] Error 4
>>> make[1]: Leaving directory
>>> `/usr/src/gcc-builds/gcc-341/asis-3_4_4-20041123.src/tools/gnatelim'
>>> make: *** [install] Error 2
>> 
>> Funny, g-string is the last one in the list,
> 
> Yes and if you look in the GNAT_Body, you have g-string.ads NOT
> g-string.adb, but if I change the makefile, it does not copy it across.

How very true. My mistake.
 
>>> Again, it's the same errors I emailed you about.
>> 
>> Yes, thats why I tested the system and made a release. Otherwise I would
>> have been to lazy. Of course I cleared out the "gnat/" directory and
>> made everything new. The "gnat/Makefile" copied everything into the
>> "gnat/" what is needed.
>> 
>> For the release I copied the files into "gnat/gcc_3.4.4/".
>> 
>> BTW: I always look for a maintainer for the gcc releases I don't use.
> 
> Well, I've tried a whole load of different versions and none work, this
> project just doesn't compile adn the flaky makefiles just dont help
> either.

There are indeed very flaky. I don't know how AdaCore can work with them.

> Personally, I'd prefer it if AdaCore got their act together and 
> released ASIS to the FSF then it'd compile as part of the gnattools
> target.

That's what I would hope for too. But if you look at there web side you will
see that ASIS is a adverticed as a seperate product - I wonder how much you
have to pay extra for it.

With Regards

Martin

-- 
mailto://krischik@users.sourceforge.net
http://www.ada.krischik.com



  reply	other threads:[~2004-11-29 16:42 UTC|newest]

Thread overview: 9+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2004-11-28 18:47 [Announce] ASIS for GNAT gcc-3.4.4-20041123 released Martin Krischik
2004-11-28 19:34 ` Luke A. Guest
2004-11-28 19:50   ` Luke A. Guest
2004-11-29  8:25     ` Martin Krischik
2004-11-29 11:05       ` Luke A. Guest
2004-11-29 16:42         ` Martin Krischik [this message]
2004-11-29 17:22           ` Luke A. Guest
2004-11-29 20:20             ` Martin Krischik
2004-11-29  8:08   ` Martin Krischik
replies disabled

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