Joining GCC for version 3.4

Joining GCC for version 3.4

[Tal Agmon]

Hi,

My team is in a process of contributing a new architecture to GCC. Can
someone please tell me what is the last date for adding a it to 3.4
version?

Regards,

   Tal Agmon
   National Semiconductor

caddr_t undefined

[y2bismil]

Hi all,

I'm trying to compile a program, and i'm getting the error of caddr_t being
undefined.  I know its just a char* or void*, but I'm thinking this is an error
I should not have to work around.

The error actually occurs in: /usr/include/cygwin/if.h

But, looking at if.h, I see it includes sys/types.h.  This is turn includes a
definition for caddr_t:  typedef	char *	caddr_t;

Anyone got any ideas?  I'm using the Cygwin environment if that's an issue

Yamin




----------------------------------------
This mail sent through www.mywaterloo.ca

Re: Joining GCC for version 3.4

[Daniel Jacobowitz]

On Tue, Oct 07, 2003 at 07:06:23PM +0300, Tal Agmon wrote:
> Hi,
> 
> My team is in a process of contributing a new architecture to GCC. Can
> someone please tell me what is the last date for adding a it to 3.4
> version?
> 
> Regards,
> 
>    Tal Agmon
>    National Semiconductor

It's probably too late.  Of course, if the port doesn't require any
changes outside a new config/ directory then the release manager's more
likely to make an exception.

Take a look at the gcc.gnu.org web site, where there is a future
development timeline.

-- 
Daniel Jacobowitz
MontaVista Software                         Debian GNU/Linux Developer

Re: caddr_t undefined

[y2bismil]

Quoting y2bismil@engmail.uwaterloo.ca:

> Hi all,
> 
> I'm trying to compile a program, and i'm getting the error of caddr_t being
> undefined.  I know its just a char* or void*, but I'm thinking this is an
> error
> I should not have to work around.
> 
> The error actually occurs in: /usr/include/cygwin/if.h
> 
> But, looking at if.h, I see it includes sys/types.h.  This is turn includes
> a
> definition for caddr_t:  typedef	char *	caddr_t;
> 
> Anyone got any ideas?  I'm using the Cygwin environment if that's an issue
> 
> Yamin
> 

Okay, 
I think this is something to do with the header search order.  It does not look
like sys/types is being included.  I think something in the ming directories is
being included.  Anyone know how to add include directories to be searched AFTER
the standard directories are searched?

Yamin



----------------------------------------
This mail sent through www.mywaterloo.ca

Re: Joining GCC for version 3.4

[Joe Buck]

On Tue, Oct 07, 2003 at 07:06:23PM +0300, Tal Agmon wrote:
> > My team is in a process of contributing a new architecture to GCC. Can
> > someone please tell me what is the last date for adding a it to 3.4
> > version?
 
On Tue, Oct 07, 2003 at 02:16:39PM -0400, Daniel Jacobowitz wrote:
> It's probably too late.  Of course, if the port doesn't require any
> changes outside a new config/ directory then the release manager's more
> likely to make an exception.

We even allowed the Tensilica port into a point release, because it
was completely non-intrusive (couldn't break anything else).  So, as
Daniel says, it depends on whether the port needs modification to the
processor-independent parts of GCC.

Re: caddr_t undefined

[Phil Edwards]

On Tue, Oct 07, 2003 at 02:35:29PM -0400, y2bismil@engmail.uwaterloo.ca wrote:
> Okay, 
> I think this is something to do with the header search order.  It does not look
> like sys/types is being included.  I think something in the ming directories is
> being included.  Anyone know how to add include directories to be searched AFTER
> the standard directories are searched?

RTFM.  This question is off-topic for these mailing lists, which are about
the development /of/ GCC, not the basic usage /of/ GCC.  It's taken me
longer to type this email than it would have taken you to go to gcc.gnu.org
and click on the documentation link.

-- 
Debugging is twice as hard as writing the code in the first place.
Therefore, if you write the code as cleverly as possible, you are,
by definition, not smart enough to debug it.
    - Brian W. Kernighan

Re: Joining GCC for version 3.4

[Janis Johnson]

On Thu, Oct 16, 2003 at 03:55:37PM +0300, Tal Agmon wrote:
> 
> If I understand right I can't use --enable-languages because I still want
> cc1plus to be created.

See http://gcc.gnu.org/bugs/reghunt.html for how to build just cc1plus;
building any other compiler without its runtime libraries is similar.

Janis

Re: Joining GCC for version 3.4

[Tal Agmon]

If I understand right I can't use --enable-languages because I still want
cc1plus to be created.



                                                                                                                                     
                      "Andreas Schwab"                                                                                               
                      <schwab@suse.de>         To:       "Tal Agmon" <Tal.Agmon@nsc.com>                                             
                                               cc:       gcc@gcc.gnu.org                                                             
                      10/16/03 02:52 PM        Subject:  Re: Joining GCC for version 3.4                                             
                                                                                                                                     
                                                                                                                                     




"Tal Agmon" <Tal.Agmon@nsc.com> writes:

> On 15 Oct, Hans-Peter Nilsson wrote:
>>I will try to answer anyway, presuming that you do not refer to
>>the GCC intrinsics library libgcc (or for that matter the
>>language-specific libraries which indeed can be disabled) but
>>rather to a system library like glibc or newlib.
>
> Actually, I meant both.
> How should I disable libgcc? I tried writing in t-cr16c file the empty
> initializations: LIBGCC =
>                  INSTALL_LIBGCC =
> And it didn't work.
>
> How to disable libstdc++? Can I avoid getting to target-libstdc++-v3?

--enable-languages=... (don't include c++ here).

Andreas.

--
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE Linux AG, Deutschherrnstr. 15-19, D-90429 Nürnberg
Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Re: Joining GCC for version 3.4

[Andreas Schwab]

"Tal Agmon" <Tal.Agmon@nsc.com> writes:

> On 15 Oct, Hans-Peter Nilsson wrote:
>>I will try to answer anyway, presuming that you do not refer to
>>the GCC intrinsics library libgcc (or for that matter the
>>language-specific libraries which indeed can be disabled) but
>>rather to a system library like glibc or newlib.
>
> Actually, I meant both.
> How should I disable libgcc? I tried writing in t-cr16c file the empty
> initializations: LIBGCC =
>                  INSTALL_LIBGCC =
> And it didn't work.
>
> How to disable libstdc++? Can I avoid getting to target-libstdc++-v3?

--enable-languages=... (don't include c++ here).

Andreas.

-- 
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE Linux AG, Deutschherrnstr. 15-19, D-90429 Nürnberg
Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Re: Joining GCC for version 3.4

[Tal Agmon]

On 15 Oct, Hans-Peter Nilsson wrote:
>I will try to answer anyway, presuming that you do not refer to
>the GCC intrinsics library libgcc (or for that matter the
>language-specific libraries which indeed can be disabled) but
>rather to a system library like glibc or newlib.

Actually, I meant both.
How should I disable libgcc? I tried writing in t-cr16c file the empty
initializations: LIBGCC =
                 INSTALL_LIBGCC =
And it didn't work.

How to disable libstdc++? Can I avoid getting to target-libstdc++-v3?

Thanks,

  Tal.

Re: Joining GCC for version 3.4

[Tal Agmon]

I'm not sure what are the requirements from a new port, can you give me
some details or pointers?

Thanks,

   Tal.



                                                                                                                                     
                      "Joe Buck"                                                                                                     
                      <jbuck@synopsys.c        To:       "Tal Agmon" <Tal.Agmon@nsc.com>                                             
                      om>                      cc:                                                                                   
                                               Subject:  Re: Joining GCC for version 3.4                                             
                      10/09/03 06:47 PM                                                                                              
                                                                                                                                     
                                                                                                                                     




On Thu, Oct 09, 2003 at 11:21:42AM +0300, Tal Agmon wrote:
> If our port is completely non-intrusive, what is the last date in which
> there's still a chance to accept it?

I think that would be up to Mark Mitchell as release manager.  We would
also need any legal paperwork (copyright assignments to the FSF) to be
done first, so if you haven't done that yet, start ASAP.

Re: Joining GCC for version 3.4

[Tal Agmon]

Hi,

If our port is completely non-intrusive, what is the last date in which
there's still a chance to accept it?

I didn't see dates for version 3.5 in the timeline in gcc site, can you
tell me when is it planned to be
released and what is the last date for us to contribute to it? In short,
when is the next last date for us to offer our port?

Regards,

    Tal.




On Tue, Oct 07, 2003 at 02:16:39PM -0400, Daniel Jacobowitz wrote:
> It's probably too late.  Of course, if the port doesn't require any
> changes outside a new config/ directory then the release manager's more
> likely to make an exception.

We even allowed the Tensilica port into a point release, because it
was completely non-intrusive (couldn't break anything else).  So, as
Daniel says, it depends on whether the port needs modification to the
processor-independent parts of GCC.