All the mail mirrored from lore.kernel.org
 help / color / mirror / Atom feed
* [Buildroot] gcc-error with latest buildroot
@ 2009-12-26 13:27 Marco Glatz
  2009-12-26 14:08 ` Michael S. Zick
  2009-12-26 17:36 ` Thomas Petazzoni
  0 siblings, 2 replies; 12+ messages in thread
From: Marco Glatz @ 2009-12-26 13:27 UTC (permalink / raw
  To: buildroot

hello,

i'm trying to make a toolchain using latest buildroot, but when compiling gcc 
i get this error:

checking for correct version of gmp.h... yes
checking for correct version of mpfr.h... no
configure: error: Building GCC requires GMP 4.1+ and MPFR 2.3.2+.
Try the --with-gmp and/or --with-mpfr options to specify their locations.
Copies of these libraries' source code can be found at their respective
hosting sites as well as at ftp://gcc.gnu.org/pub/gcc/infrastructure/.
See also http://gcc.gnu.org/install/prerequisites.html for additional info.
If you obtained GMP and/or MPFR from a vendor distribution package, make
sure that you have installed both the libraries and the header files.
They may be located in separate packages.
make: *** [/devel/buildroot/output/build/gcc-4.4.2-target/.configured] Fehler 1



regards,
marco

^ permalink raw reply	[flat|nested] 12+ messages in thread

* [Buildroot] gcc-error with latest buildroot
  2009-12-26 13:27 [Buildroot] gcc-error with latest buildroot Marco Glatz
@ 2009-12-26 14:08 ` Michael S. Zick
  2009-12-26 14:20   ` Marco Glatz
  2009-12-26 17:36 ` Thomas Petazzoni
  1 sibling, 1 reply; 12+ messages in thread
From: Michael S. Zick @ 2009-12-26 14:08 UTC (permalink / raw
  To: buildroot

On Sat December 26 2009, Marco Glatz wrote:
> hello,
> 
> i'm trying to make a toolchain using latest buildroot, but when compiling gcc 
> i get this error:
> 
> checking for correct version of gmp.h... yes
> checking for correct version of mpfr.h... no
>

This problem shipped in 2009.11 also -
Fails even if mpfr is put into the GCC tree -

Looks like it might be tied in with the logic of building Fortran -
(GCC once only used mpfr on the Fortran builds).
Haven't had time for further testing myself.

There is what might be a related bug report already posted.

Mike
> configure: error: Building GCC requires GMP 4.1+ and MPFR 2.3.2+.
> Try the --with-gmp and/or --with-mpfr options to specify their locations.
> Copies of these libraries' source code can be found at their respective
> hosting sites as well as at ftp://gcc.gnu.org/pub/gcc/infrastructure/.
> See also http://gcc.gnu.org/install/prerequisites.html for additional info.
> If you obtained GMP and/or MPFR from a vendor distribution package, make
> sure that you have installed both the libraries and the header files.
> They may be located in separate packages.
> make: *** [/devel/buildroot/output/build/gcc-4.4.2-target/.configured] Fehler 1
> 
> 
> 
> regards,
> marco
> _______________________________________________
> buildroot mailing list
> buildroot at busybox.net
> http://lists.busybox.net/mailman/listinfo/buildroot
> 
> 

^ permalink raw reply	[flat|nested] 12+ messages in thread

* [Buildroot] gcc-error with latest buildroot
  2009-12-26 14:08 ` Michael S. Zick
@ 2009-12-26 14:20   ` Marco Glatz
  2009-12-26 14:29     ` Michael S. Zick
  0 siblings, 1 reply; 12+ messages in thread
From: Marco Glatz @ 2009-12-26 14:20 UTC (permalink / raw
  To: buildroot

Am Samstag, 26. Dezember 2009 15:08:53 schrieb Michael S. Zick:
> On Sat December 26 2009, Marco Glatz wrote:
> > hello,
> >
> > i'm trying to make a toolchain using latest buildroot, but when compiling
> > gcc i get this error:
> >
> > checking for correct version of gmp.h... yes
> > checking for correct version of mpfr.h... no
> 
> This problem shipped in 2009.11 also -
> Fails even if mpfr is put into the GCC tree -
> 
> Looks like it might be tied in with the logic of building Fortran -
> (GCC once only used mpfr on the Fortran builds).
> Haven't had time for further testing myself.
> 
> There is what might be a related bug report already posted.


hi,

this helped me out:

http://www.linuxquestions.org/questions/linux-from-scratch-13/i-have-an-error-
building-gcc-requires-gmp-4.1-and-mpfr-2.3.0.-690426/

so i put  gmp and mpfr directories in the gcc-folder in the output/toolchain-
directory. then it compiled. later when linking there were a lot of errors not 
finding gmp-related functions, so i put again mpfr and gmp in output/build/gcc-
x-x-target directory. but that didn't help, still getting lot of undefined 
reference to __gmpz_* functions.

all the pages i found on google said you should configure --with-mpfr=... and 
--with-gmp=....., and there are options fpr this in the makefiles, but they do 
net get filled.


marco

^ permalink raw reply	[flat|nested] 12+ messages in thread

* [Buildroot] gcc-error with latest buildroot
  2009-12-26 14:20   ` Marco Glatz
@ 2009-12-26 14:29     ` Michael S. Zick
       [not found]       ` <200912261558.18888.marco-glatz@web.de>
  0 siblings, 1 reply; 12+ messages in thread
From: Michael S. Zick @ 2009-12-26 14:29 UTC (permalink / raw
  To: buildroot

On Sat December 26 2009, Marco Glatz wrote:
> Am Samstag, 26. Dezember 2009 15:08:53 schrieb Michael S. Zick:
> > On Sat December 26 2009, Marco Glatz wrote:
> > > hello,
> > >
> > > i'm trying to make a toolchain using latest buildroot, but when compiling
> > > gcc i get this error:
> > >
> > > checking for correct version of gmp.h... yes
> > > checking for correct version of mpfr.h... no
> > 
> > This problem shipped in 2009.11 also -
> > Fails even if mpfr is put into the GCC tree -
> > 
> > Looks like it might be tied in with the logic of building Fortran -
> > (GCC once only used mpfr on the Fortran builds).
> > Haven't had time for further testing myself.
> > 
> > There is what might be a related bug report already posted.
> 
> 
> hi,
> 
> this helped me out:
> 
> http://www.linuxquestions.org/questions/linux-from-scratch-13/i-have-an-error-
> building-gcc-requires-gmp-4.1-and-mpfr-2.3.0.-690426/
> 
> so i put  gmp and mpfr directories in the gcc-folder in the output/toolchain-
> directory. then it compiled. later when linking there were a lot of errors not 
> finding gmp-related functions, so i put again mpfr and gmp in output/build/gcc-
> x-x-target directory. but that didn't help, still getting lot of undefined 
> reference to __gmpz_* functions.
> 
> all the pages i found on google said you should configure --with-mpfr=... and 
> --with-gmp=....., and there are options fpr this in the makefiles, but they do 
> net get filled.
>

When I looked at it a month or so ago (prior to 2009.11 release) -
It looked to me as if filling those options where conditionalized
based on building Fortran.  But I have about zero Makefile knowledge.

Try enabling Fortran compiler in the build, it might "just work".

The Buildroot project uses a "tag-it, forget-it and move-on" release cycle;
they aren't staffed to support anything like a mission-critical,
stable release, version of Buildroot.
 
Mike
> 
> marco
> _______________________________________________
> buildroot mailing list
> buildroot at busybox.net
> http://lists.busybox.net/mailman/listinfo/buildroot
> 
> 

^ permalink raw reply	[flat|nested] 12+ messages in thread

* [Buildroot] gcc-error with latest buildroot
       [not found]       ` <200912261558.18888.marco-glatz@web.de>
@ 2009-12-26 15:16         ` Michael S. Zick
  0 siblings, 0 replies; 12+ messages in thread
From: Michael S. Zick @ 2009-12-26 15:16 UTC (permalink / raw
  To: buildroot

On Sat December 26 2009, Marco Glatz wrote:
> > Try enabling Fortran compiler in the build, it might "just work".
> > 
> > The Buildroot project uses a "tag-it, forget-it and move-on" release cycle;
> > they aren't staffed to support anything like a mission-critical,
> > stable release, version of Buildroot.
> 
> 
> hi,
> 
> with fortran enabled, the --with-mpfr and --with-gmp get set, but still 
> doesn'nt work.
> 
> i found this:
> 
> http://dvb-community.de/wiki/index.php/GCC-uClibC_Toolchain_erstellen
> 
> and this:
> 
> http://downloads.openwrt.org/docs/buildroot-documentation.html
> 

Been following OpenWRT (and OpenEmbedded) for several years;
I may have to switch to OpenWRT's build system myself.

I really don't need something that can build an entire distribution -
But at least those projects do have more people on the maintenance staff.

Will be postponing my own decision until I finish putting full MIPS support
into 2009.11 - I don't think I want to tackle the Makefile code churn here
to also put that support into dev-head.

Mike
> 
> 
> i will give it a try. also tried scratchbox, but that din't work either.
> 
> 
> marco
> 
> 

^ permalink raw reply	[flat|nested] 12+ messages in thread

* [Buildroot] gcc-error with latest buildroot
  2009-12-26 13:27 [Buildroot] gcc-error with latest buildroot Marco Glatz
  2009-12-26 14:08 ` Michael S. Zick
@ 2009-12-26 17:36 ` Thomas Petazzoni
  2009-12-26 17:42   ` Michael S. Zick
                     ` (2 more replies)
  1 sibling, 3 replies; 12+ messages in thread
From: Thomas Petazzoni @ 2009-12-26 17:36 UTC (permalink / raw
  To: buildroot

Hello,

Le Sat, 26 Dec 2009 14:27:53 +0100,
Marco Glatz <marco-glatz@web.de> a ?crit :

> i'm trying to make a toolchain using latest buildroot, but when
> compiling gcc i get this error:

> [/devel/buildroot/output/build/gcc-4.4.2-target/.configured] Fehler 1

Do you really need gcc on your target ?

(Note: of course this is something we should fix, but having gcc and
associated development files on the target is an unusual setup, that's
why it isn't widely tested)

Cheers,

Thomas
-- 
Thomas Petazzoni, Free Electrons
Kernel, drivers and embedded Linux development,
consulting, training and support.
http://free-electrons.com

^ permalink raw reply	[flat|nested] 12+ messages in thread

* [Buildroot] gcc-error with latest buildroot
  2009-12-26 17:36 ` Thomas Petazzoni
@ 2009-12-26 17:42   ` Michael S. Zick
  2009-12-26 19:58   ` Michael S. Zick
  2009-12-26 20:40   ` Marco Glatz
  2 siblings, 0 replies; 12+ messages in thread
From: Michael S. Zick @ 2009-12-26 17:42 UTC (permalink / raw
  To: buildroot

On Sat December 26 2009, Thomas Petazzoni wrote:
> Hello,
> 
> Le Sat, 26 Dec 2009 14:27:53 +0100,
> Marco Glatz <marco-glatz@web.de> a ?crit :
> 
> > i'm trying to make a toolchain using latest buildroot, but when
> > compiling gcc i get this error:
> 
> > [/devel/buildroot/output/build/gcc-4.4.2-target/.configured] Fehler 1
> 
> Do you really need gcc on your target ?
> 

__YES__

We do after-market product development by device owners here:
http://minimodding.com
and these little MIPS boxes have more than enough resources to self-host.


> (Note: of course this is something we should fix, but having gcc and
> associated development files on the target is an unusual setup, that's
> why it isn't widely tested)
> 

Yeah, just like sstrip, fix it or mark it broken.

Mike
> Cheers,
> 
> Thomas

^ permalink raw reply	[flat|nested] 12+ messages in thread

* [Buildroot] gcc-error with latest buildroot
  2009-12-26 17:36 ` Thomas Petazzoni
  2009-12-26 17:42   ` Michael S. Zick
@ 2009-12-26 19:58   ` Michael S. Zick
  2009-12-26 21:18     ` Thomas Petazzoni
  2009-12-26 20:40   ` Marco Glatz
  2 siblings, 1 reply; 12+ messages in thread
From: Michael S. Zick @ 2009-12-26 19:58 UTC (permalink / raw
  To: buildroot

On Sat December 26 2009, Thomas Petazzoni wrote:
> 
> (Note: of course this is something we should fix, but having gcc and
> associated development files on the target is an unusual setup, that's
> why it isn't widely tested)
> 

Say what?
How can you know that?

The request for a "test report" tag in the bug tracker has not yet
(as of a couple of hours ago) been acted on.

So unless you have some other means of tracking user submitted test 
reports how can you say that something is "not widely tested"?  you have
no way for people to report good/bad/pass/fail/broke.

Which means the maintainer(s) have no way to know what to mark as "tested",
"untested" or "broken" in a tagged release.

Or is it the Buildroot policy to throw all of the testing work onto the
back of the maintainer(s)?  Why should they be loaded up with that work?

Mike
> Cheers,
> 
> Thomas

^ permalink raw reply	[flat|nested] 12+ messages in thread

* [Buildroot] gcc-error with latest buildroot
  2009-12-26 17:36 ` Thomas Petazzoni
  2009-12-26 17:42   ` Michael S. Zick
  2009-12-26 19:58   ` Michael S. Zick
@ 2009-12-26 20:40   ` Marco Glatz
  2 siblings, 0 replies; 12+ messages in thread
From: Marco Glatz @ 2009-12-26 20:40 UTC (permalink / raw
  To: buildroot

Thomas Petazzoni schrieb:
> Do you really need gcc on your target ?
>   

no, only the toolchain so i can link against uclibs to build small 
linux-systems fpr embedded-systems

marco

^ permalink raw reply	[flat|nested] 12+ messages in thread

* [Buildroot] gcc-error with latest buildroot
  2009-12-26 19:58   ` Michael S. Zick
@ 2009-12-26 21:18     ` Thomas Petazzoni
  2009-12-26 21:39       ` Michael S. Zick
  0 siblings, 1 reply; 12+ messages in thread
From: Thomas Petazzoni @ 2009-12-26 21:18 UTC (permalink / raw
  To: buildroot

Hello,

First of all, could you please stop being aggressive and offensive ?

Le Sat, 26 Dec 2009 13:58:06 -0600,
"Michael S. Zick" <minimod@morethan.org> a ?crit :

> > (Note: of course this is something we should fix, but having gcc and
> > associated development files on the target is an unusual setup,
> > that's why it isn't widely tested)
> > 
> 
> Say what?
> How can you know that?

Because the feature has existed for years and until a few days/weeks
ago, no one reported the issue and no one stepped in to fix it. Which
means that no one was seriously using it.

> So unless you have some other means of tracking user submitted test 
> reports how can you say that something is "not widely tested"?

Users facing issues are invited to report them in the bugzilla.
However, we cannot guarantee any specific response time, Buildroot is a
community effort. Bugs are fixed depending on developer
interest, availability and patches provided by other developers.

If you want guarantees, please contract a commercial support. I'd be
happy to provide one through the company I'm working for (Free
Electrons).

Please note that I'm not happy to see features broken in Buildroot and
I very much would like the gcc on target issue fixed. But this is not a
reason for being offensive.

Cheers,

Thomas
-- 
Thomas Petazzoni, Free Electrons
Kernel, drivers and embedded Linux development,
consulting, training and support.
http://free-electrons.com

^ permalink raw reply	[flat|nested] 12+ messages in thread

* [Buildroot] gcc-error with latest buildroot
  2009-12-26 21:18     ` Thomas Petazzoni
@ 2009-12-26 21:39       ` Michael S. Zick
  2009-12-26 22:11         ` Michael S. Zick
  0 siblings, 1 reply; 12+ messages in thread
From: Michael S. Zick @ 2009-12-26 21:39 UTC (permalink / raw
  To: buildroot

On Sat December 26 2009, Thomas Petazzoni wrote:
> Hello,
> 
> First of all, could you please stop being aggressive and offensive ?
> 
> Le Sat, 26 Dec 2009 13:58:06 -0600,
> "Michael S. Zick" <minimod@morethan.org> a ?crit :
> 
> > > (Note: of course this is something we should fix, but having gcc and
> > > associated development files on the target is an unusual setup,
> > > that's why it isn't widely tested)
> > > 
> > 
> > Say what?
> > How can you know that?
> 
> Because the feature has existed for years and until a few days/weeks
> ago, no one reported the issue and no one stepped in to fix it. Which
> means that no one was seriously using it.
> 
> > So unless you have some other means of tracking user submitted test 
> > reports how can you say that something is "not widely tested"?
> 
> Users facing issues are invited to report them in the bugzilla.
> However, we cannot guarantee any specific response time, Buildroot is a
> community effort. Bugs are fixed depending on developer
> interest, availability and patches provided by other developers.
> 
> If you want guarantees, please contract a commercial support. I'd be
> happy to provide one through the company I'm working for (Free
> Electrons).
> 
> Please note that I'm not happy to see features broken in Buildroot and
> I very much would like the gcc on target issue fixed. But this is not a
> reason for being offensive.
> 

Requesting that a posted bug, posted 2 weeks before a release gets a:
"depends on Broken" in the config file is offensive?

And that GCC bug in the bug tracker is much older than that.

I don't see how asking for a "depends on Broken" is offensive for that
or any other pending open bug.  Just responsible behavior.

As a matter of fact, that can be your first 30 day 2009.11 "bug fix" -
add "depends on Broken" to everything broken in 2009.11

If that request offends someone, then that is just a matter of personal viewpoint.

Mike
> Cheers,
> 
> Thomas

^ permalink raw reply	[flat|nested] 12+ messages in thread

* [Buildroot] gcc-error with latest buildroot
  2009-12-26 21:39       ` Michael S. Zick
@ 2009-12-26 22:11         ` Michael S. Zick
  0 siblings, 0 replies; 12+ messages in thread
From: Michael S. Zick @ 2009-12-26 22:11 UTC (permalink / raw
  To: buildroot

On Sat December 26 2009, Michael S. Zick wrote:
> On Sat December 26 2009, Thomas Petazzoni wrote:
> > Hello,
> > 
> > First of all, could you please stop being aggressive and offensive ?
> > 

I think the word you are searching for is "Assertive" in English (assertif).

Mike

^ permalink raw reply	[flat|nested] 12+ messages in thread

end of thread, other threads:[~2009-12-26 22:11 UTC | newest]

Thread overview: 12+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2009-12-26 13:27 [Buildroot] gcc-error with latest buildroot Marco Glatz
2009-12-26 14:08 ` Michael S. Zick
2009-12-26 14:20   ` Marco Glatz
2009-12-26 14:29     ` Michael S. Zick
     [not found]       ` <200912261558.18888.marco-glatz@web.de>
2009-12-26 15:16         ` Michael S. Zick
2009-12-26 17:36 ` Thomas Petazzoni
2009-12-26 17:42   ` Michael S. Zick
2009-12-26 19:58   ` Michael S. Zick
2009-12-26 21:18     ` Thomas Petazzoni
2009-12-26 21:39       ` Michael S. Zick
2009-12-26 22:11         ` Michael S. Zick
2009-12-26 20:40   ` Marco Glatz

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.