GNAT, g77, PowerPC & AIX V4..

GNAT, g77, PowerPC & AIX V4..

[Henri Sundelin]

Hello,

does anybody happen to have a full working GNU compiler set
(GNAT, gcc, g77, g++) for a PowerPC IBM RS/6000 / AIX 4.x?

I cannot compile it myself since the current GNAT binary
distribution for RS6K seems to generate POWER instruction set
based assembly - which PowerPC 'as' just rejects.

Cheers, 
HS

PS. Is anybody at ACT working on AIX V4 port of GNAT with full
os level thread support?
 
-- 
Henri.Sundelin@hut.fi
"Mechanicalize something idiosyncratic"
                   -Brian Eno, Oblique Strategies

Re: GNAT, g77, PowerPC & AIX V4..

[jim hopper]

In article <hycloa7c0pd.fsf@vipunen.hut.fi>
Henri Sundelin <hsundeli@vipunen.hut.fi> writes:

> Hello,
> 
> does anybody happen to have a full working GNU compiler set
> (GNAT, gcc, g77, g++) for a PowerPC IBM RS/6000 / AIX 4.x?
> 
> I cannot compile it myself since the current GNAT binary
> distribution for RS6K seems to generate POWER instruction set
> based assembly - which PowerPC 'as' just rejects.
> 


I have the powermac version of this working which is basically the aix
port.  if you look in config/rs6000 folder check out r6000.h it has
version options defined that you can use to specify various
architectures/instructions.  they are used on the command line as

-mpowerpc
-mno-power

etc.  i have not used them but they look like what you want.

best jim

Re: GNAT, g77, PowerPC & AIX V4..

[Meskovic Zlatko]

Henri Sundelin wrote:
> 
> Hello,
> 
> does anybody happen to have a full working GNU compiler set
> (GNAT, gcc, g77, g++) for a PowerPC IBM RS/6000 / AIX 4.x?

> --
> Henri.Sundelin@hut.fi
> "Mechanicalize something idiosyncratic"
>                    -Brian Eno, Oblique Strategies


Hello,

Yes, I have a part of your requested compilers:
  gnat, gcc and g++
for PowerPC running AIX 4.1.4
with running threads...

zlatko

-----------------------------
zlatko b. meskovic
union bank of switzerland
CH-8021 zuerich
tel +41-1-236 82 50
fax +41-1-236 94 40
email zlatko.meskovic@ubs.com

Re: GNAT, g77, PowerPC & AIX V4..

[Robert Dewar]

Stephen said

"I think that's -gnatmpowerpc, but I haven't tried it."

No, only options for GNAT itself are prefixed by -gnat, not gcc options.

By the way, i suspect that some of this discussion on building GNAT and
building cross compilers in particular would be more fruitful on the
chat@gnat.com mailing list, where a lot of people live who know a lot
about GNAT, but do not read CLA!

Re: GNAT, g77, PowerPC & AIX V4..

[Stephen Leake]

Robert Dewar wrote:
> By the way, i suspect that some of this discussion on building GNAT and
> building cross compilers in particular would be more fruitful on the
> chat@gnat.com mailing list, where a lot of people live who know a lot
> about GNAT, but do not read CLA!

In that case, I vote for promoting chat@gnat.com to a newsgroup; I find
it MUCH easier to browse newsgroups than to browse mail, especially when
I want to be able to ignore some of the messages. Maybe that's a
deficiency in my mail reader (currently Netscape); can anyone recommend
a better one?

I have no idea what's involved in creating a newsgroup, so I await
enlightenment.

-- 
- Stephe

comp.lang.ada.gnat? (was: GNAT, g77, PowerPC & AIX V4..)

[Thomas Koenig]

In comp.lang.ada, Stephen.Leake@gsfc.nasa.gov wrote:

>In that case, I vote for promoting chat@gnat.com to a newsgroup; I find
>it MUCH easier to browse newsgroups than to browse mail, especially when
>I want to be able to ignore some of the messages. Maybe that's a
>deficiency in my mail reader (currently Netscape); can anyone recommend
>a better one?

What's the current volume and number of subscribers of chat@gnat.com?
If it's high enough, and people want it, I'd suggest making a
'comp.lang.ada.gnat' newsgroup, and possibly renaming c.l.a to
comp.lang.ada.misc.

There's bound to be some crossposting between the two newsgroups, though.

Re: GNAT, g77, PowerPC & AIX V4..

[Robert Dewar]

Stephen says

"In that case, I vote for promoting chat@gnat.com to a newsgroup; I find
it MUCH easier to browse newsgroups than to browse mail, especially when
I want to be able to ignore some of the messages. Maybe that's a
deficiency in my mail reader (currently Netscape); can anyone recommend
a better one?"


Note that the contents of the chat mailing list messages are acrchived on
the gnat home page, and can be browsed at www.gnat.com/chat. You may well
find this gives you what you need!

Re: comp.lang.ada.gnat? (was: GNAT, g77, PowerPC & AIX V4..)

[Robert Dewar]

"What's the current volume and number of subscribers of chat@gnat.com?
If it's high enough, and people want it, I'd suggest making a
'comp.lang.ada.gnat' newsgroup, and possibly renaming c.l.a to
comp.lang.ada.misc."

Well we certainly intend to retain chat@gnat.com in any case, because
lots of people DO find a mailing list more convenient, and it is certainly
not prone to the kind of junk mail that is now appearing on all newsgroups.
The advantage of the mailing list is that, since the signal to noise ratio
is high, more of the GNAT folks feel like participating (almost all now
find CLA not worth reading -- I am an exception :-)

So it is fine certainly to start a GNAT newsgroup, but I doubt that anyone
from ACT would read it.

Re: GNAT, g77, PowerPC & AIX V4..

[Ronald Cole]

dewar@merv.cs.nyu.edu (Robert Dewar) writes:
> Note that the contents of the chat mailing list messages are acrchived on
> the gnat home page, and can be browsed at www.gnat.com/chat. You may well
> find this gives you what you need!

With a newsgroup, you can use a high-technology newsreader, like GNUS,
that supports threading and article scoring.  If I had a vote, I'd
vote for gnu.gnat and gnu.gnat.bug newsgroups which are repeated from
the appropriate mailing lists as the bug-gcc and bug-g++ lists are.

-- 
Forte International, P.O. Box 1412, Ridgecrest, CA  93556-1412
Ronald Cole <ronald@ridgecrest.ca.us>    Phone: (619) 499-9142
President, CEO                             Fax: (619) 499-9152
My PGP fingerprint: E9 A8 E3 68 61 88 EF 43  56 2B CE 3E E9 8F 3F 2B

Re: comp.lang.ada.gnat? (was: GNAT, g77, PowerPC & AIX V4..)

[Andrew Lynch]

Robert Dewar wrote:
> 
> "What's the current volume and number of subscribers of chat@gnat.com?
> If it's high enough, and people want it, I'd suggest making a
> 'comp.lang.ada.gnat' newsgroup, and possibly renaming c.l.a to
> comp.lang.ada.misc."
> 
> Well we certainly intend to retain chat@gnat.com in any case, because
> lots of people DO find a mailing list more convenient, and it is certainly
> not prone to the kind of junk mail that is now appearing on all newsgroups.
> The advantage of the mailing list is that, since the signal to noise ratio
> is high, more of the GNAT folks feel like participating (almost all now
> find CLA not worth reading -- I am an exception :-)

I have been told that the Plan9 mailing list is gatewayed into the
comp.os.plan9 newsgroup, but not vice versa. Maybe a similar scheme
could be adopted for a GNAT group, ie.

mails sent to chat@gnat.com appear in eg. comp.lang.ada.gnat
but articles posted in c.l.a.g stay only in the newsgroup and do
not go out to all chat@gnat.com subscribers.

This would keep junk off the mailing list while at the same time
sharing the information on chat@gnat.com with a wider audience
(who might not want to subscribe or access a web site)

Just my DM 0.02 worth!
Andrew.

Mirroring GNAT mail to newsgroups (was: GNAT, g77, PowerPC & AIX V4)

[Larry Kilgallen]

In article <m2681517af.fsf@devo.ridgecrest.ca.us>, Ronald Cole <ronald@ridgecrest.ca.us> writes:

> With a newsgroup, you can use a high-technology newsreader, like GNUS,
> that supports threading and article scoring.  If I had a vote, I'd
> vote for gnu.gnat and gnu.gnat.bug newsgroups which are repeated from
> the appropriate mailing lists as the bug-gcc and bug-g++ lists are.

I think a gnu hierarchy would be a better place for any extension,
because I have seen the mess comp.lang.pascal got into by being
extended. Presumably people who get to a gnu hierarchy have some
idea for finding their way around the hierarchy, but I think the
comp.lang attracts more confused folks.

I believe the rules require that addition of comp.lang.ada.whatever
result also in relocation of comp.lang.ada to comp.lang.ada.something.
I would hate to see mainstream Ada discussions demoted to a ".misc".

Larry Kilgallen