Linux MIPS Architecture development
 help / color / mirror / Atom feed
* Recommended toolchain
@ 2001-03-20  3:06 Simon Gee
  2001-03-20 20:12 ` Karel van Houten
  0 siblings, 1 reply; 23+ messages in thread
From: Simon Gee @ 2001-03-20  3:06 UTC (permalink / raw)
  To: linux-mips

Hello,

Recently I've been working with various version mixes of the gnu tool
chain for a mipsel-linux target and settled on a patchy binutils
2.8.1/egcs 1.1.2/glibc 2.0.6 setup. However this lacks the functionality
that I would get from a newer toolchain for use with the linux 2.4
kernel. As a result, I was wondering if someone could recommend the
latest "stable"/recommended toolchain for a mipsel-linux target.

Thanks in advance,

Simon

^ permalink raw reply	[flat|nested] 23+ messages in thread

* Re: Recommended toolchain
  2001-03-20  3:06 Simon Gee
@ 2001-03-20 20:12 ` Karel van Houten
  2001-03-20 20:12   ` Karel van Houten
                     ` (2 more replies)
  0 siblings, 3 replies; 23+ messages in thread
From: Karel van Houten @ 2001-03-20 20:12 UTC (permalink / raw)
  To: Simon Gee; +Cc: linux-mips

Hi Simon,

You wrote:
> Recently I've been working with various version mixes of the gnu tool
> chain for a mipsel-linux target and settled on a patchy binutils
> 2.8.1/egcs 1.1.2/glibc 2.0.6 setup. However this lacks the functionality
> that I would get from a newer toolchain for use with the linux 2.4
> kernel. As a result, I was wondering if someone could recommend the
> latest "stable"/recommended toolchain for a mipsel-linux target.

Well, I'm currently using:
binutils 2.10.1
gcc 2.95.3 (with Maciej's patches)
glibc 2.2.2 (compiled with above toolchain).

This toolchain works for native compiles on my mipsel box.
One drawback: I can't compile any kernels with this setup,
For kernel compiles I use 2.8.1/egcs-2.90.27/glibc-2.0.6.

-- 
Karel van Houten
----------------------------------------------------------
The box said "Requires Windows 95 or better."
I can't understand why it won't work on my Linux computer. 
----------------------------------------------------------

^ permalink raw reply	[flat|nested] 23+ messages in thread

* Re: Recommended toolchain
  2001-03-20 20:12 ` Karel van Houten
@ 2001-03-20 20:12   ` Karel van Houten
  2001-03-20 20:37   ` Ralf Baechle
  2001-03-27 19:02   ` Maciej W. Rozycki
  2 siblings, 0 replies; 23+ messages in thread
From: Karel van Houten @ 2001-03-20 20:12 UTC (permalink / raw)
  To: Simon Gee; +Cc: linux-mips

Hi Simon,

You wrote:
> Recently I've been working with various version mixes of the gnu tool
> chain for a mipsel-linux target and settled on a patchy binutils
> 2.8.1/egcs 1.1.2/glibc 2.0.6 setup. However this lacks the functionality
> that I would get from a newer toolchain for use with the linux 2.4
> kernel. As a result, I was wondering if someone could recommend the
> latest "stable"/recommended toolchain for a mipsel-linux target.

Well, I'm currently using:
binutils 2.10.1
gcc 2.95.3 (with Maciej's patches)
glibc 2.2.2 (compiled with above toolchain).

This toolchain works for native compiles on my mipsel box.
One drawback: I can't compile any kernels with this setup,
For kernel compiles I use 2.8.1/egcs-2.90.27/glibc-2.0.6.

-- 
Karel van Houten
----------------------------------------------------------
The box said "Requires Windows 95 or better."
I can't understand why it won't work on my Linux computer. 
----------------------------------------------------------

^ permalink raw reply	[flat|nested] 23+ messages in thread

* Re: Recommended toolchain
  2001-03-20 20:12 ` Karel van Houten
  2001-03-20 20:12   ` Karel van Houten
@ 2001-03-20 20:37   ` Ralf Baechle
  2001-03-20 20:37     ` Ralf Baechle
  2001-03-27 19:02   ` Maciej W. Rozycki
  2 siblings, 1 reply; 23+ messages in thread
From: Ralf Baechle @ 2001-03-20 20:37 UTC (permalink / raw)
  To: Karel van Houten; +Cc: Simon Gee, linux-mips

On Tue, Mar 20, 2001 at 09:12:02PM +0100, Karel van Houten wrote:

> You wrote:
> > Recently I've been working with various version mixes of the gnu tool
> > chain for a mipsel-linux target and settled on a patchy binutils
> > 2.8.1/egcs 1.1.2/glibc 2.0.6 setup. However this lacks the functionality
> > that I would get from a newer toolchain for use with the linux 2.4
> > kernel. As a result, I was wondering if someone could recommend the
> > latest "stable"/recommended toolchain for a mipsel-linux target.
> 
> Well, I'm currently using:
> binutils 2.10.1
> gcc 2.95.3 (with Maciej's patches)
> glibc 2.2.2 (compiled with above toolchain).
> 
> This toolchain works for native compiles on my mipsel box.
> One drawback: I can't compile any kernels with this setup,
> For kernel compiles I use 2.8.1/egcs-2.90.27/glibc-2.0.6.

You MUST use minimum egcs-2.91.66 (1.1.2); older compilers WILL misscompile
kernels.

  Ralf

^ permalink raw reply	[flat|nested] 23+ messages in thread

* Re: Recommended toolchain
  2001-03-20 20:37   ` Ralf Baechle
@ 2001-03-20 20:37     ` Ralf Baechle
  0 siblings, 0 replies; 23+ messages in thread
From: Ralf Baechle @ 2001-03-20 20:37 UTC (permalink / raw)
  To: Karel van Houten; +Cc: Simon Gee, linux-mips

On Tue, Mar 20, 2001 at 09:12:02PM +0100, Karel van Houten wrote:

> You wrote:
> > Recently I've been working with various version mixes of the gnu tool
> > chain for a mipsel-linux target and settled on a patchy binutils
> > 2.8.1/egcs 1.1.2/glibc 2.0.6 setup. However this lacks the functionality
> > that I would get from a newer toolchain for use with the linux 2.4
> > kernel. As a result, I was wondering if someone could recommend the
> > latest "stable"/recommended toolchain for a mipsel-linux target.
> 
> Well, I'm currently using:
> binutils 2.10.1
> gcc 2.95.3 (with Maciej's patches)
> glibc 2.2.2 (compiled with above toolchain).
> 
> This toolchain works for native compiles on my mipsel box.
> One drawback: I can't compile any kernels with this setup,
> For kernel compiles I use 2.8.1/egcs-2.90.27/glibc-2.0.6.

You MUST use minimum egcs-2.91.66 (1.1.2); older compilers WILL misscompile
kernels.

  Ralf

^ permalink raw reply	[flat|nested] 23+ messages in thread

* Re: Recommended toolchain
  2001-03-21 14:00 Recommended toolchain Phil Thompson
@ 2001-03-21 12:11 ` Erik Mullinix
  2001-03-21 16:06 ` Keith M Wesolowski
  1 sibling, 0 replies; 23+ messages in thread
From: Erik Mullinix @ 2001-03-21 12:11 UTC (permalink / raw)
  To: Phil Thompson, linux-mips

What Error's are you getting?
Hesp
----- Original Message -----
From: "Phil Thompson" <Phil.Thompson@pace.co.uk>
To: <linux-mips@oss.sgi.com>
Sent: Wednesday, March 21, 2001 9:00 AM
Subject: RE: Recommended toolchain


> So, going back to the original question, what are the recommended
versions -
> particularly if you are building 2.4.x kernels?
>
> Are the recomendations in the HOWTO out of date? I had to patch va-mips.h
> from the egcs RPM and more fixes seem to be needed before I get a kernel
> that compiles.
>
> Thanks,
> Phil
>
> -----Original Message-----
> From: Ralf Baechle [mailto:ralf@oss.sgi.com]
> Sent: 20 March 2001 20:37
> To: Karel van Houten
> Cc: simong@oz.agile.tv; linux-mips@oss.sgi.com
> Subject: Re: Recommended toolchain
>
>
> On Tue, Mar 20, 2001 at 09:12:02PM +0100, Karel van Houten wrote:
>
> > You wrote:
> > > Recently I've been working with various version mixes of the gnu tool
> > > chain for a mipsel-linux target and settled on a patchy binutils
> > > 2.8.1/egcs 1.1.2/glibc 2.0.6 setup. However this lacks the
functionality
> > > that I would get from a newer toolchain for use with the linux 2.4
> > > kernel. As a result, I was wondering if someone could recommend the
> > > latest "stable"/recommended toolchain for a mipsel-linux target.
> >
> > Well, I'm currently using:
> > binutils 2.10.1
> > gcc 2.95.3 (with Maciej's patches)
> > glibc 2.2.2 (compiled with above toolchain).
> >
> > This toolchain works for native compiles on my mipsel box.
> > One drawback: I can't compile any kernels with this setup,
> > For kernel compiles I use 2.8.1/egcs-2.90.27/glibc-2.0.6.
>
> You MUST use minimum egcs-2.91.66 (1.1.2); older compilers WILL
misscompile
> kernels.
>
>   Ralf
>

^ permalink raw reply	[flat|nested] 23+ messages in thread

* RE: Recommended toolchain
@ 2001-03-21 14:00 Phil Thompson
  2001-03-21 12:11 ` Erik Mullinix
  2001-03-21 16:06 ` Keith M Wesolowski
  0 siblings, 2 replies; 23+ messages in thread
From: Phil Thompson @ 2001-03-21 14:00 UTC (permalink / raw)
  To: 'linux-mips@oss.sgi.com'

So, going back to the original question, what are the recommended versions -
particularly if you are building 2.4.x kernels?

Are the recomendations in the HOWTO out of date? I had to patch va-mips.h
from the egcs RPM and more fixes seem to be needed before I get a kernel
that compiles.

Thanks,
Phil

-----Original Message-----
From: Ralf Baechle [mailto:ralf@oss.sgi.com]
Sent: 20 March 2001 20:37
To: Karel van Houten
Cc: simong@oz.agile.tv; linux-mips@oss.sgi.com
Subject: Re: Recommended toolchain


On Tue, Mar 20, 2001 at 09:12:02PM +0100, Karel van Houten wrote:

> You wrote:
> > Recently I've been working with various version mixes of the gnu tool
> > chain for a mipsel-linux target and settled on a patchy binutils
> > 2.8.1/egcs 1.1.2/glibc 2.0.6 setup. However this lacks the functionality
> > that I would get from a newer toolchain for use with the linux 2.4
> > kernel. As a result, I was wondering if someone could recommend the
> > latest "stable"/recommended toolchain for a mipsel-linux target.
> 
> Well, I'm currently using:
> binutils 2.10.1
> gcc 2.95.3 (with Maciej's patches)
> glibc 2.2.2 (compiled with above toolchain).
> 
> This toolchain works for native compiles on my mipsel box.
> One drawback: I can't compile any kernels with this setup,
> For kernel compiles I use 2.8.1/egcs-2.90.27/glibc-2.0.6.

You MUST use minimum egcs-2.91.66 (1.1.2); older compilers WILL misscompile
kernels.

  Ralf

^ permalink raw reply	[flat|nested] 23+ messages in thread

* Re: Recommended toolchain
  2001-03-21 14:00 Recommended toolchain Phil Thompson
  2001-03-21 12:11 ` Erik Mullinix
@ 2001-03-21 16:06 ` Keith M Wesolowski
  1 sibling, 0 replies; 23+ messages in thread
From: Keith M Wesolowski @ 2001-03-21 16:06 UTC (permalink / raw)
  To: Phil Thompson; +Cc: linux-mips

On Wed, Mar 21, 2001 at 02:00:22PM -0000, Phil Thompson wrote:

> So, going back to the original question, what are the recommended versions -
> particularly if you are building 2.4.x kernels?

I don't really know that there is one.  I am using the gcc 3.0 branch
and binutils from CVS (2.11.90).  You can find the same toolchain
that's building my kernels on
ftp://oss.sgi.com/pub/linux/mips/mips-linux/simple/crossdev at all
times, including patches.  No binaries, source and patches only.

This linker has a bug in it regarding section placement, but the
current cvs kernel has a workaround for it.

-- 
Keith M Wesolowski <wesolows@foobazco.org> http://foobazco.org/~wesolows
------(( Project Foobazco Coordinator and Network Administrator ))------
"I should have crushed his marketing-addled skull with a fucking bat."

^ permalink raw reply	[flat|nested] 23+ messages in thread

* RE: Recommended toolchain
@ 2001-03-21 18:46 Phil Thompson
  2001-03-21 22:06 ` Keith M Wesolowski
  2001-03-22  7:49 ` Geert Uytterhoeven
  0 siblings, 2 replies; 23+ messages in thread
From: Phil Thompson @ 2001-03-21 18:46 UTC (permalink / raw)
  To: 'Erik Mullinix', linux-mips

I had to patch va-mips.h to #include <asm/sgidefs.h> rather than
<sgidefs.h>.

The current errors are:

- warnings about struct flock64 not being declared (it's defined in
asm-mips64/fcntl.h but not asm-mips/fcntl.h)

- compilation stops because loops_per_sec is undeclared as far as
asm-mips/delay.h is concerned (although it seems fine in
asm-mips64/delay.h).

This seems to imply that the mips architecture (as opposed to mips64) isn't
being maintained. Is this the case? Should I be using mips64 - but what
would be the point on an embedded CPU?

Thanks,
Phil

-----Original Message-----
From: Erik Mullinix [mailto:Hesp@rainworks.org]
Sent: 21 March 2001 12:12
To: Phil Thompson; linux-mips@oss.sgi.com
Subject: Re: Recommended toolchain


What Error's are you getting?
Hesp
----- Original Message -----
From: "Phil Thompson" <Phil.Thompson@pace.co.uk>
To: <linux-mips@oss.sgi.com>
Sent: Wednesday, March 21, 2001 9:00 AM
Subject: RE: Recommended toolchain


> So, going back to the original question, what are the recommended
versions -
> particularly if you are building 2.4.x kernels?
>
> Are the recomendations in the HOWTO out of date? I had to patch va-mips.h
> from the egcs RPM and more fixes seem to be needed before I get a kernel
> that compiles.
>
> Thanks,
> Phil
>
> -----Original Message-----
> From: Ralf Baechle [mailto:ralf@oss.sgi.com]
> Sent: 20 March 2001 20:37
> To: Karel van Houten
> Cc: simong@oz.agile.tv; linux-mips@oss.sgi.com
> Subject: Re: Recommended toolchain
>
>
> On Tue, Mar 20, 2001 at 09:12:02PM +0100, Karel van Houten wrote:
>
> > You wrote:
> > > Recently I've been working with various version mixes of the gnu tool
> > > chain for a mipsel-linux target and settled on a patchy binutils
> > > 2.8.1/egcs 1.1.2/glibc 2.0.6 setup. However this lacks the
functionality
> > > that I would get from a newer toolchain for use with the linux 2.4
> > > kernel. As a result, I was wondering if someone could recommend the
> > > latest "stable"/recommended toolchain for a mipsel-linux target.
> >
> > Well, I'm currently using:
> > binutils 2.10.1
> > gcc 2.95.3 (with Maciej's patches)
> > glibc 2.2.2 (compiled with above toolchain).
> >
> > This toolchain works for native compiles on my mipsel box.
> > One drawback: I can't compile any kernels with this setup,
> > For kernel compiles I use 2.8.1/egcs-2.90.27/glibc-2.0.6.
>
> You MUST use minimum egcs-2.91.66 (1.1.2); older compilers WILL
misscompile
> kernels.
>
>   Ralf
>

^ permalink raw reply	[flat|nested] 23+ messages in thread

* Re: Recommended toolchain
  2001-03-21 18:46 Phil Thompson
@ 2001-03-21 22:06 ` Keith M Wesolowski
  2001-03-22  7:49 ` Geert Uytterhoeven
  1 sibling, 0 replies; 23+ messages in thread
From: Keith M Wesolowski @ 2001-03-21 22:06 UTC (permalink / raw)
  To: Phil Thompson; +Cc: 'Erik Mullinix', linux-mips

On Wed, Mar 21, 2001 at 06:46:07PM -0000, Phil Thompson wrote:

> - compilation stops because loops_per_sec is undeclared as far as
> asm-mips/delay.h is concerned (although it seems fine in
> asm-mips64/delay.h).
> 
> This seems to imply that the mips architecture (as opposed to mips64) isn't
> being maintained. Is this the case? Should I be using mips64 - but what
> would be the point on an embedded CPU?

mips(32) is in fact being actively maintained.  It sounds like your
kernel sources may be out of date; I specifically remember removing
all references to loops_per_sec in favour of loops_per_jiffy in both
mips and mips64.

Ask again about fcntl.h after updating...

-- 
Keith M Wesolowski <wesolows@foobazco.org> http://foobazco.org/~wesolows
------(( Project Foobazco Coordinator and Network Administrator ))------
"I should have crushed his marketing-addled skull with a fucking bat."

^ permalink raw reply	[flat|nested] 23+ messages in thread

* RE: Recommended toolchain
  2001-03-21 18:46 Phil Thompson
  2001-03-21 22:06 ` Keith M Wesolowski
@ 2001-03-22  7:49 ` Geert Uytterhoeven
  1 sibling, 0 replies; 23+ messages in thread
From: Geert Uytterhoeven @ 2001-03-22  7:49 UTC (permalink / raw)
  To: Phil Thompson; +Cc: 'Erik Mullinix', linux-mips

On Wed, 21 Mar 2001, Phil Thompson wrote:
> I had to patch va-mips.h to #include <asm/sgidefs.h> rather than
> <sgidefs.h>.
> 
> The current errors are:
> 
> - warnings about struct flock64 not being declared (it's defined in
> asm-mips64/fcntl.h but not asm-mips/fcntl.h)
> 
> - compilation stops because loops_per_sec is undeclared as far as
> asm-mips/delay.h is concerned (although it seems fine in
> asm-mips64/delay.h).
> 
> This seems to imply that the mips architecture (as opposed to mips64) isn't
> being maintained. Is this the case? Should I be using mips64 - but what
> would be the point on an embedded CPU?

You're definitely not using the Linux/MIPS CVS tree, since these things were
fixed there some months ago.

Gr{oetje,eeting}s,

						Geert

--
Geert Uytterhoeven ------------- Sony Software Development Center Europe (SDCE)
Geert.Uytterhoeven@sonycom.com ------------------- Sint-Stevens-Woluwestraat 55
Voice +32-2-7248626 Fax +32-2-7262686 ---------------- B-1130 Brussels, Belgium

^ permalink raw reply	[flat|nested] 23+ messages in thread

* RE: Recommended toolchain
@ 2001-03-22  9:39 Phil Thompson
  2001-03-22  9:43 ` Geert Uytterhoeven
  0 siblings, 1 reply; 23+ messages in thread
From: Phil Thompson @ 2001-03-22  9:39 UTC (permalink / raw)
  To: 'Geert Uytterhoeven'; +Cc: linux-mips

Correct - I am using the standard 2.4.2 sources. I had (obviously
incorrectly) assumed that the standard sources would be up to date and not
"some months" out of date.

How are the differences in the two trees managed? At what point are the
latest non-MIPS changes applied to the MIPS tree? At what point are the MIPS
specific changes applied to the standard tree?

Thanks,
Phil

-----Original Message-----
From: Geert Uytterhoeven [mailto:Geert.Uytterhoeven@sonycom.com]
Sent: 22 March 2001 07:50
To: Phil Thompson
Cc: 'Erik Mullinix'; linux-mips@oss.sgi.com
Subject: RE: Recommended toolchain


On Wed, 21 Mar 2001, Phil Thompson wrote:
> I had to patch va-mips.h to #include <asm/sgidefs.h> rather than
> <sgidefs.h>.
> 
> The current errors are:
> 
> - warnings about struct flock64 not being declared (it's defined in
> asm-mips64/fcntl.h but not asm-mips/fcntl.h)
> 
> - compilation stops because loops_per_sec is undeclared as far as
> asm-mips/delay.h is concerned (although it seems fine in
> asm-mips64/delay.h).
> 
> This seems to imply that the mips architecture (as opposed to mips64)
isn't
> being maintained. Is this the case? Should I be using mips64 - but what
> would be the point on an embedded CPU?

You're definitely not using the Linux/MIPS CVS tree, since these things were
fixed there some months ago.

Gr{oetje,eeting}s,

						Geert

--
Geert Uytterhoeven ------------- Sony Software Development Center Europe
(SDCE)
Geert.Uytterhoeven@sonycom.com ------------------- Sint-Stevens-Woluwestraat
55
Voice +32-2-7248626 Fax +32-2-7262686 ---------------- B-1130 Brussels,
Belgium

^ permalink raw reply	[flat|nested] 23+ messages in thread

* RE: Recommended toolchain
  2001-03-22  9:39 Recommended toolchain Phil Thompson
@ 2001-03-22  9:43 ` Geert Uytterhoeven
  0 siblings, 0 replies; 23+ messages in thread
From: Geert Uytterhoeven @ 2001-03-22  9:43 UTC (permalink / raw)
  To: Phil Thompson; +Cc: linux-mips

On Thu, 22 Mar 2001, Phil Thompson wrote:
> Correct - I am using the standard 2.4.2 sources. I had (obviously
> incorrectly) assumed that the standard sources would be up to date and not
> "some months" out of date.

:-)

> How are the differences in the two trees managed? At what point are the
> latest non-MIPS changes applied to the MIPS tree? At what point are the MIPS

When Ralf finds some time?

> specific changes applied to the standard tree?

When Ralf finds some time and Linus is in a good mood?

Gr{oetje,eeting}s,

						Geert

--
Geert Uytterhoeven ------------- Sony Software Development Center Europe (SDCE)
Geert.Uytterhoeven@sonycom.com ------------------- Sint-Stevens-Woluwestraat 55
Voice +32-2-7248626 Fax +32-2-7262686 ---------------- B-1130 Brussels, Belgium

^ permalink raw reply	[flat|nested] 23+ messages in thread

* Re: Recommended toolchain
  2001-03-20 20:12 ` Karel van Houten
  2001-03-20 20:12   ` Karel van Houten
  2001-03-20 20:37   ` Ralf Baechle
@ 2001-03-27 19:02   ` Maciej W. Rozycki
  2001-03-29 17:28     ` Houten K.H.C. van (Karel)
  2 siblings, 1 reply; 23+ messages in thread
From: Maciej W. Rozycki @ 2001-03-27 19:02 UTC (permalink / raw)
  To: Karel van Houten; +Cc: Simon Gee, linux-mips

On Tue, 20 Mar 2001, Karel van Houten wrote:

> Well, I'm currently using:
> binutils 2.10.1
> gcc 2.95.3 (with Maciej's patches)
> glibc 2.2.2 (compiled with above toolchain).
> 
> This toolchain works for native compiles on my mipsel box.
> One drawback: I can't compile any kernels with this setup,
> For kernel compiles I use 2.8.1/egcs-2.90.27/glibc-2.0.6.

 What's wrong with the toolchain wrt the kernel now?  I've been compiling
2.4 kernels successfully till the end of January -- it's just a lack of
time and a nasty bug I want to track down that stop me from trying to
compile a new kernel these days.

-- 
+  Maciej W. Rozycki, Technical University of Gdansk, Poland   +
+--------------------------------------------------------------+
+        e-mail: macro@ds2.pg.gda.pl, PGP key available        +

^ permalink raw reply	[flat|nested] 23+ messages in thread

* Re: Recommended toolchain
  2001-03-27 19:02   ` Maciej W. Rozycki
@ 2001-03-29 17:28     ` Houten K.H.C. van (Karel)
  2001-03-29 17:28       ` Houten K.H.C. van (Karel)
                         ` (2 more replies)
  0 siblings, 3 replies; 23+ messages in thread
From: Houten K.H.C. van (Karel) @ 2001-03-29 17:28 UTC (permalink / raw)
  To: Maciej W. Rozycki
  Cc: Karel van Houten, Simon Gee, linux-mips, K.H.C.vanHouten


Hi all,

"Maciej W. Rozycki" writes:
> What's wrong with the toolchain wrt the kernel now?  I've been compiling
>2.4 kernels successfully till the end of January -- it's just a lack of
>time and a nasty bug I want to track down that stop me from trying to
>compile a new kernel these days.

This happens when I compile kernel 2.4.0-test9 with
binutils 2.10.1, gcc 2.95.3, glibc 2.2.2 on my 5000/260 (R4k)
(the same source/config compiles fine with 2.8.1/egcs-2.90.27/glibc-2.0.6):

# make boot
....
ld -static -G 0 -T arch/mips/ld.script.little -Ttext 0x80040000 arch/mips/kernel
/head.o arch/mips/kernel/init_task.o init/main.o init/version.o \
        --start-group \
        arch/mips/kernel/kernel.o arch/mips/mm/mm.o kernel/kernel.o mm/mm.o fs/f
s.o ipc/ipc.o arch/mips/dec/dec.o \
        drivers/block/block.o drivers/char/char.o drivers/misc/misc.o drivers/ne
t/net.o drivers/media/media.o drivers/parport/parport.a  drivers/scsi/scsidrv.o 
drivers/cdrom/cdrom.a drivers/tc/tc.a \
        net/network.o \
        arch/mips/lib/lib.a /usr/src/linux/lib/lib.a arch/mips/dec/prom/rexlib.a
 \
        --end-group \
        -o vmlinux
nm vmlinux | grep -v '\(compiled\)\|\(\.o$\)\|\( [aU] \)\|\(\.\.ng$\)\|\(LASH[RL
]DI\)' | sort > System.map
make[1]: Entering directory `/usr/src/linux-2.4.0t9-R4k/arch/mips/boot'
./elf2ecoff /usr/src/linux/vmlinux vmlinux.ecoff -a
wrote 20 byte file header.
wrote 56 byte a.out header.
wrote 240 bytes of section headers.
wrote 4 byte pad.
writing 30560 bytes...
Intersegment gap (-2147256160 bytes) too large.
make[1]: *** [vmlinux.ecoff] Error 1
make[1]: Leaving directory `/usr/src/linux-2.4.0t9-R4k/arch/mips/boot'
make: *** [boot] Error 2

^ permalink raw reply	[flat|nested] 23+ messages in thread

* Re: Recommended toolchain
  2001-03-29 17:28     ` Houten K.H.C. van (Karel)
@ 2001-03-29 17:28       ` Houten K.H.C. van (Karel)
  2001-03-29 17:39       ` Karel van Houten
  2001-03-29 17:51       ` Maciej W. Rozycki
  2 siblings, 0 replies; 23+ messages in thread
From: Houten K.H.C. van (Karel) @ 2001-03-29 17:28 UTC (permalink / raw)
  To: Maciej W. Rozycki; +Cc: Karel van Houten, Simon Gee, linux-mips


Hi all,

"Maciej W. Rozycki" writes:
> What's wrong with the toolchain wrt the kernel now?  I've been compiling
>2.4 kernels successfully till the end of January -- it's just a lack of
>time and a nasty bug I want to track down that stop me from trying to
>compile a new kernel these days.

This happens when I compile kernel 2.4.0-test9 with
binutils 2.10.1, gcc 2.95.3, glibc 2.2.2 on my 5000/260 (R4k)
(the same source/config compiles fine with 2.8.1/egcs-2.90.27/glibc-2.0.6):

# make boot
....
ld -static -G 0 -T arch/mips/ld.script.little -Ttext 0x80040000 arch/mips/kernel
/head.o arch/mips/kernel/init_task.o init/main.o init/version.o \
        --start-group \
        arch/mips/kernel/kernel.o arch/mips/mm/mm.o kernel/kernel.o mm/mm.o fs/f
s.o ipc/ipc.o arch/mips/dec/dec.o \
        drivers/block/block.o drivers/char/char.o drivers/misc/misc.o drivers/ne
t/net.o drivers/media/media.o drivers/parport/parport.a  drivers/scsi/scsidrv.o 
drivers/cdrom/cdrom.a drivers/tc/tc.a \
        net/network.o \
        arch/mips/lib/lib.a /usr/src/linux/lib/lib.a arch/mips/dec/prom/rexlib.a
 \
        --end-group \
        -o vmlinux
nm vmlinux | grep -v '\(compiled\)\|\(\.o$\)\|\( [aU] \)\|\(\.\.ng$\)\|\(LASH[RL
]DI\)' | sort > System.map
make[1]: Entering directory `/usr/src/linux-2.4.0t9-R4k/arch/mips/boot'
./elf2ecoff /usr/src/linux/vmlinux vmlinux.ecoff -a
wrote 20 byte file header.
wrote 56 byte a.out header.
wrote 240 bytes of section headers.
wrote 4 byte pad.
writing 30560 bytes...
Intersegment gap (-2147256160 bytes) too large.
make[1]: *** [vmlinux.ecoff] Error 1
make[1]: Leaving directory `/usr/src/linux-2.4.0t9-R4k/arch/mips/boot'
make: *** [boot] Error 2

^ permalink raw reply	[flat|nested] 23+ messages in thread

* Re: Recommended toolchain
  2001-03-29 17:28     ` Houten K.H.C. van (Karel)
  2001-03-29 17:28       ` Houten K.H.C. van (Karel)
@ 2001-03-29 17:39       ` Karel van Houten
  2001-03-29 17:39         ` Karel van Houten
  2001-03-29 18:10         ` Maciej W. Rozycki
  2001-03-29 17:51       ` Maciej W. Rozycki
  2 siblings, 2 replies; 23+ messages in thread
From: Karel van Houten @ 2001-03-29 17:39 UTC (permalink / raw)
  To: linux-mips; +Cc: Maciej W. Rozycki

I wrote:
> This happens when I compile kernel 2.4.0-test9 with
> binutils 2.10.1, gcc 2.95.3, glibc 2.2.2 on my 5000/260 (R4k)
> (the same source/config compiles fine with 2.8.1/egcs-2.90.27/glibc-2.0.6):
> ....

Oh, and the resulting (ELF) kernel doesn't boot at all:
>>boot 3/rz0 1/new
delo V0.7 Copyright 2000 Florian Lohoff <flo@rfc822.org>
Loading /etc/delo.conf .. ok
Loading /boot/vmlinux.new ....... ok

KN05 V2.1k    (PC: 0xa002cab8, SP: 0x8043fef0)
>>                            

-- 
Karel van Houten

----------------------------------------------------------
The box said "Requires Windows 95 or better."
I can't understand why it won't work on my Linux computer. 
----------------------------------------------------------

^ permalink raw reply	[flat|nested] 23+ messages in thread

* Re: Recommended toolchain
  2001-03-29 17:39       ` Karel van Houten
@ 2001-03-29 17:39         ` Karel van Houten
  2001-03-29 18:10         ` Maciej W. Rozycki
  1 sibling, 0 replies; 23+ messages in thread
From: Karel van Houten @ 2001-03-29 17:39 UTC (permalink / raw)
  To: linux-mips; +Cc: Maciej W. Rozycki

I wrote:
> This happens when I compile kernel 2.4.0-test9 with
> binutils 2.10.1, gcc 2.95.3, glibc 2.2.2 on my 5000/260 (R4k)
> (the same source/config compiles fine with 2.8.1/egcs-2.90.27/glibc-2.0.6):
> ....

Oh, and the resulting (ELF) kernel doesn't boot at all:
>>boot 3/rz0 1/new
delo V0.7 Copyright 2000 Florian Lohoff <flo@rfc822.org>
Loading /etc/delo.conf .. ok
Loading /boot/vmlinux.new ....... ok

KN05 V2.1k    (PC: 0xa002cab8, SP: 0x8043fef0)
>>                            

-- 
Karel van Houten

----------------------------------------------------------
The box said "Requires Windows 95 or better."
I can't understand why it won't work on my Linux computer. 
----------------------------------------------------------

^ permalink raw reply	[flat|nested] 23+ messages in thread

* Re: Recommended toolchain
  2001-03-29 17:28     ` Houten K.H.C. van (Karel)
  2001-03-29 17:28       ` Houten K.H.C. van (Karel)
  2001-03-29 17:39       ` Karel van Houten
@ 2001-03-29 17:51       ` Maciej W. Rozycki
  2 siblings, 0 replies; 23+ messages in thread
From: Maciej W. Rozycki @ 2001-03-29 17:51 UTC (permalink / raw)
  To: vhouten; +Cc: Karel van Houten, Simon Gee, linux-mips

On Thu, 29 Mar 2001, Houten K.H.C. van (Karel) wrote:

> This happens when I compile kernel 2.4.0-test9 with
> binutils 2.10.1, gcc 2.95.3, glibc 2.2.2 on my 5000/260 (R4k)
> (the same source/config compiles fine with 2.8.1/egcs-2.90.27/glibc-2.0.6):
[...]
> ./elf2ecoff /usr/src/linux/vmlinux vmlinux.ecoff -a
> wrote 20 byte file header.
> wrote 56 byte a.out header.
> wrote 240 bytes of section headers.
> wrote 4 byte pad.
> writing 30560 bytes...
> Intersegment gap (-2147256160 bytes) too large.
> make[1]: *** [vmlinux.ecoff] Error 1
> make[1]: Leaving directory `/usr/src/linux-2.4.0t9-R4k/arch/mips/boot'
> make: *** [boot] Error 2

 I recall I needed to edit the ld script for linux 2.4.0-test12 that I am
currently using.  Just watch out which sections have improper addresses. 
Alternatively grab a newer version -- the script was fixed later (by Ralf,
IIRC).

-- 
+  Maciej W. Rozycki, Technical University of Gdansk, Poland   +
+--------------------------------------------------------------+
+        e-mail: macro@ds2.pg.gda.pl, PGP key available        +

^ permalink raw reply	[flat|nested] 23+ messages in thread

* Re: Recommended toolchain
  2001-03-29 17:39       ` Karel van Houten
  2001-03-29 17:39         ` Karel van Houten
@ 2001-03-29 18:10         ` Maciej W. Rozycki
  2001-04-01 14:39           ` Houten K.H.C. van (Karel)
  1 sibling, 1 reply; 23+ messages in thread
From: Maciej W. Rozycki @ 2001-03-29 18:10 UTC (permalink / raw)
  To: vhouten; +Cc: linux-mips

On Thu, 29 Mar 2001, Karel van Houten wrote:

> Oh, and the resulting (ELF) kernel doesn't boot at all:
> >>boot 3/rz0 1/new
> delo V0.7 Copyright 2000 Florian Lohoff <flo@rfc822.org>
> Loading /etc/delo.conf .. ok
> Loading /boot/vmlinux.new ....... ok
> 
> KN05 V2.1k    (PC: 0xa002cab8, SP: 0x8043fef0)
> >>                            

 Well, this might be related to R4k you are using.  I have been able to
run kernels up to 2.4.0-test12 fine on my /240.  I tried the final 2.4.0
(or was it 2.4.1? -- I don't rememeber) on my machine and it booted but it
was very unstable -- I got a shell prompt but any moderate fs activity
killed it immediately.

 As it was the time I was building binutils 2.10.91 package I checked if
the version of binutils matters.  It didn't.  The 2.4.0-test12 version was
stable and the other one was not regardless of the binutils version used. 

 So I guess there may be something wrong with the R4k code generation in
gcc 2.95.2(3) (or possibly binutils, but the latter is quite unlikely).  I
can't run-time test R4k code but I may see if I can review the generated
binary of startup code up to the first line output for any
inconsistencies.  Don't hold your breath, though... 

-- 
+  Maciej W. Rozycki, Technical University of Gdansk, Poland   +
+--------------------------------------------------------------+
+        e-mail: macro@ds2.pg.gda.pl, PGP key available        +

^ permalink raw reply	[flat|nested] 23+ messages in thread

* Re: Recommended toolchain
  2001-03-29 18:10         ` Maciej W. Rozycki
@ 2001-04-01 14:39           ` Houten K.H.C. van (Karel)
  2001-04-01 18:54             ` Florian Lohoff
  0 siblings, 1 reply; 23+ messages in thread
From: Houten K.H.C. van (Karel) @ 2001-04-01 14:39 UTC (permalink / raw)
  To: Maciej W. Rozycki; +Cc: vhouten, linux-mips


"Maciej W. Rozycki" writes:
> So I guess there may be something wrong with the R4k code generation in
>gcc 2.95.2(3) (or possibly binutils, but the latter is quite unlikely).  I
>can't run-time test R4k code but I may see if I can review the generated
>binary of startup code up to the first line output for any
>inconsistencies.  Don't hold your breath, though... 

I've now done a native compile of a CVS checkout from Friday, and
it does compile. It even boots, up until:

esp0: target 1 [period 248ns offset 15 4.03MHz synchronous SCSI]
sr0: scsi-1 drive
Uniform CD-ROM driver Revision: 3.12
NET4: Linux TCP/IP 1.0 for NET4.0
IP Protocols: ICMP, UDP, TCP
IP: routing cache hash table of 2048 buckets, 16Kbytes
TCP: Hash tables configured (established 16384 bind 16384)
Unhandled kernel unaligned access in unaligned.c:emulate_load_store_insn, line 372:
$0 : 00000000 10010c00 8017fed0 00000000
$4 : 801fe33c 80200000 813de000 802db000
$8 : 10010c01 8019af20 00000000 00000000
$12: 00000000 fffff000 fffffff7 802db060
$16: 00000000 813de000 00000001 00000000
$20: 00000005 801cbe34 801eb070 801eb1ac
$24: 00000000 0000000a
$28: 8022e000 8022fe88 00000000 80066cc0
epc   : 8017ff08
Status: 10010c03
Cause : 00002010
Process swapper (pid: 1, stackpage=8022e000)

etc.
I have not yet found the time to look further into that...

^ permalink raw reply	[flat|nested] 23+ messages in thread

* Re: Recommended toolchain
  2001-04-01 14:39           ` Houten K.H.C. van (Karel)
@ 2001-04-01 18:54             ` Florian Lohoff
  0 siblings, 0 replies; 23+ messages in thread
From: Florian Lohoff @ 2001-04-01 18:54 UTC (permalink / raw)
  To: vhouten; +Cc: Maciej W. Rozycki, linux-mips

On Sun, Apr 01, 2001 at 04:39:34PM +0200, Houten K.H.C. van (Karel) wrote:
> IP Protocols: ICMP, UDP, TCP
> IP: routing cache hash table of 2048 buckets, 16Kbytes
> TCP: Hash tables configured (established 16384 bind 16384)
> Unhandled kernel unaligned access in unaligned.c:emulate_load_store_insn, line 372:
> $0 : 00000000 10010c00 8017fed0 00000000
> $4 : 801fe33c 80200000 813de000 802db000
> $8 : 10010c01 8019af20 00000000 00000000
> $12: 00000000 fffff000 fffffff7 802db060
> $16: 00000000 813de000 00000001 00000000
> $20: 00000005 801cbe34 801eb070 801eb1ac
> $24: 00000000 0000000a
> $28: 8022e000 8022fe88 00000000 80066cc0
> epc   : 8017ff08
> Status: 10010c03
> Cause : 00002010
> Process swapper (pid: 1, stackpage=8022e000)
> 
> etc.
> I have not yet found the time to look further into that...

Guido did that last weekend and discovered that this is a toolchain
problem which disappears when compiling with gcc 3.0

Flo
-- 
Florian Lohoff                  flo@rfc822.org             +49-5201-669912
     Why is it called "common sense" when nobody seems to have any?

^ permalink raw reply	[flat|nested] 23+ messages in thread

* recommended toolchain
@ 2003-04-10 19:23 carbonated beverage
  0 siblings, 0 replies; 23+ messages in thread
From: carbonated beverage @ 2003-04-10 19:23 UTC (permalink / raw)
  To: linux-mips

Hi,

	What's the currently recommended toolchain for building MIPS kernels?
I'm working on a NEC VR4111 (mipsel, MIPS16, I, II, III, no LL/SC :P) and
would like to avoid tracking down toolchain bugs.

	Also, Are partial patches (i.e. platform not fully functional yet)
accepted?  Or should I finish a full port (hits userland) first?

	Vadem Clio 1000, reading the 2.4.0-test9 tree that's the last version
I see in CVS and porting to 2.5 (bk pull as of... a while ago).

	Thanks,

-- DN
Daniel

^ permalink raw reply	[flat|nested] 23+ messages in thread

end of thread, other threads:[~2003-04-10 19:24 UTC | newest]

Thread overview: 23+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2001-03-21 14:00 Recommended toolchain Phil Thompson
2001-03-21 12:11 ` Erik Mullinix
2001-03-21 16:06 ` Keith M Wesolowski
  -- strict thread matches above, loose matches on Subject: below --
2003-04-10 19:23 recommended toolchain carbonated beverage
2001-03-22  9:39 Recommended toolchain Phil Thompson
2001-03-22  9:43 ` Geert Uytterhoeven
2001-03-21 18:46 Phil Thompson
2001-03-21 22:06 ` Keith M Wesolowski
2001-03-22  7:49 ` Geert Uytterhoeven
2001-03-20  3:06 Simon Gee
2001-03-20 20:12 ` Karel van Houten
2001-03-20 20:12   ` Karel van Houten
2001-03-20 20:37   ` Ralf Baechle
2001-03-20 20:37     ` Ralf Baechle
2001-03-27 19:02   ` Maciej W. Rozycki
2001-03-29 17:28     ` Houten K.H.C. van (Karel)
2001-03-29 17:28       ` Houten K.H.C. van (Karel)
2001-03-29 17:39       ` Karel van Houten
2001-03-29 17:39         ` Karel van Houten
2001-03-29 18:10         ` Maciej W. Rozycki
2001-04-01 14:39           ` Houten K.H.C. van (Karel)
2001-04-01 18:54             ` Florian Lohoff
2001-03-29 17:51       ` Maciej W. Rozycki

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