gcc-3.3.2 vs 2.6.0-test11

gcc-3.3.2 vs 2.6.0-test11

[J.A. Magallon]

hi all..

Are there any known issues wrt gcc-3.3.2 ?
I built test11 with gcc-3.3.1 and worked fine, the same config built with
3.3.2 does not pass init launch:

INIT version 2.85 booting

and nothing more....

I have to check, but I think it also miscompiles 2.4. I rebuilt a kernel on a
remote box (2.4.23 + assorted patches), that worked fine under 3.3.1, and
after reboot the box didn't came to life, no ping.

TIA

-- 
J.A. Magallon <jamagallon()able!es>     \                 Software is like sex:
werewolf!able!es                         \           It's better when it's free
Mandrake Linux release 10.0 (Cooker) for i586
Linux 2.6.0-test11-jam2 (gcc 3.3.1 (Mandrake Linux 9.2 3.3.1-4mdk))

Re: gcc-3.3.2 vs 2.6.0-test11

[Andrew Walrond]

On Wednesday 17 Dec 2003 11:37 am, J.A. Magallon wrote:
> hi all..
>
> Are there any known issues wrt gcc-3.3.2 ?
> I built test11 with gcc-3.3.1 and worked fine, the same config built with
> 3.3.2 does not pass init launch:

I'm running test11 built with 3.3.2 without any (obviously gcc related) 
problems

Andrew Walrond

Re: gcc-3.3.2 vs 2.6.0-test11

[venom]

using gcc 3.3.2 with 2.6.0-test11 and 2.4.23. boots cleanly on Ayhlon XP,
Athlon-trbind, PentiumIII.

which compilation options are you using?

Luigi

On Wed, 17 Dec 2003, J.A. Magallon wrote:

> Date: Wed, 17 Dec 2003 12:37:42 +0100
> From: J.A. Magallon <jamagallon@able.es>
> To: Lista Linux-Kernel <linux-kernel@vger.kernel.org>
> Subject: gcc-3.3.2 vs 2.6.0-test11
>
> hi all..
>
> Are there any known issues wrt gcc-3.3.2 ?
> I built test11 with gcc-3.3.1 and worked fine, the same config built with
> 3.3.2 does not pass init launch:
>
> INIT version 2.85 booting
>
> and nothing more....
>
> I have to check, but I think it also miscompiles 2.4. I rebuilt a kernel on a
> remote box (2.4.23 + assorted patches), that worked fine under 3.3.1, and
> after reboot the box didn't came to life, no ping.
>
> TIA
>
> --
> J.A. Magallon <jamagallon()able!es>     \                 Software is like sex:
> werewolf!able!es                         \           It's better when it's free
> Mandrake Linux release 10.0 (Cooker) for i586
> Linux 2.6.0-test11-jam2 (gcc 3.3.1 (Mandrake Linux 9.2 3.3.1-4mdk))
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
>

Re: gcc-3.3.2 vs 2.6.0-test11

[J.A. Magallon]

[-- Attachment #1: Type: text/plain, Size: 485 bytes --]


On 12.17, venom@sns.it wrote:
> 
> using gcc 3.3.2 with 2.6.0-test11 and 2.4.23. boots cleanly on Ayhlon XP,
> Athlon-trbind, PentiumIII.
> 
> which compilation options are you using?
> 

.config attached

-- 
J.A. Magallon <jamagallon()able!es>     \                 Software is like sex:
werewolf!able!es                         \           It's better when it's free
Mandrake Linux release 10.0 (Cooker) for i586
Linux 2.6.0-test11-jam2 (gcc 3.3.1 (Mandrake Linux 9.2 3.3.1-4mdk))

[-- Attachment #2: config.bz2 --]
[-- Type: application/x-bzip, Size: 5697 bytes --]

Re: gcc-3.3.2 vs 2.6.0-test11

[J.A. Magallon]

On 12.17, J.A. Magallon wrote:
> hi all..
> 
> Are there any known issues wrt gcc-3.3.2 ?
> I built test11 with gcc-3.3.1 and worked fine, the same config built with
> 3.3.2 does not pass init launch:
> 
> INIT version 2.85 booting
> 
> and nothing more....
> 
> I have to check, but I think it also miscompiles 2.4. I rebuilt a kernel on a
> remote box (2.4.23 + assorted patches), that worked fine under 3.3.1, and
> after reboot the box didn't came to life, no ping.
> 

Well, it fails to compile also 2.4. A 2.4 kernel that worked fine hangs when
launching init if built with this gcc. It boots again if is built at -O instead
of -O2.

I think it is a Mandrake specific problem, as changelog reads:

* Tue Dec 16 2003 Gwenole Beauchesne <gbeauchesne@mandrakesoft.com> 3.3.2-2mdk

- Add gcj(1), aka make gc happy
- Add [ep]mmintrin.h for even for SSE/PNI intrinsics
- Move cc1 to gcc-cpp, thusly nuking gcc dep for XFree86
- Remove -funit-at-a-time from -O2 since it is memory hungry

* Mon Dec 15 2003 Gwenole Beauchesne <gbeauchesne@mandrakesoft.com> 3.3.2-1mdk

- 3.3.2 + 3.3-hammer branch (2003/12/08)

I think I don't like that hammer branch...

Thanks.

-- 
J.A. Magallon <jamagallon()able!es>     \                 Software is like sex:
werewolf!able!es                         \           It's better when it's free
Mandrake Linux release 10.0 (Cooker) for i586
Linux 2.6.0-test11-jam2 (gcc 3.3.1 (Mandrake Linux 9.2 3.3.1-4mdk))

Re: gcc-3.3.2 vs 2.6.0-test11

[Martin Schlemmer]

[-- Attachment #1: Type: text/plain, Size: 1538 bytes --]

On Thu, 2003-12-18 at 02:04, J.A. Magallon wrote:
> On 12.17, J.A. Magallon wrote:
> > hi all..
> > 
> > Are there any known issues wrt gcc-3.3.2 ?
> > I built test11 with gcc-3.3.1 and worked fine, the same config built with
> > 3.3.2 does not pass init launch:
> > 
> > INIT version 2.85 booting
> > 
> > and nothing more....
> > 
> > I have to check, but I think it also miscompiles 2.4. I rebuilt a kernel on a
> > remote box (2.4.23 + assorted patches), that worked fine under 3.3.1, and
> > after reboot the box didn't came to life, no ping.
> > 
> 
> Well, it fails to compile also 2.4. A 2.4 kernel that worked fine hangs when
> launching init if built with this gcc. It boots again if is built at -O instead
> of -O2.
> 
> I think it is a Mandrake specific problem, as changelog reads:
> 
> * Tue Dec 16 2003 Gwenole Beauchesne <gbeauchesne@mandrakesoft.com> 3.3.2-2mdk
> 
> - Add gcj(1), aka make gc happy
> - Add [ep]mmintrin.h for even for SSE/PNI intrinsics
> - Move cc1 to gcc-cpp, thusly nuking gcc dep for XFree86
> - Remove -funit-at-a-time from -O2 since it is memory hungry
> 
> * Mon Dec 15 2003 Gwenole Beauchesne <gbeauchesne@mandrakesoft.com> 3.3.2-1mdk
> 
> - 3.3.2 + 3.3-hammer branch (2003/12/08)
> 
> I think I don't like that hammer branch...

I wont add my opinion here about the hammer branch, but try
the gcc-3_3-rhl-branch which is gcc-3.3.relative_current maintained
by Jakub, and used by us with a patch or two added.


Regards,

-- 

Martin Schlemmer




[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]