-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Hi,
after some tiny fixes to reiserfs and the makefile for prep bootfile
(using ../lib/lib.a vs. ../lib/libz.a) I managed to succesfully compile
a kernel. It even boots to the point where it frees unused kernel memory
and then stops... this includes succesfully mounting the root
filesystem...
Just to let the world know what does and doesn't work on my powerstack
mobo.
- --
Met vriendelijke groeten,
Remco Post
SARA - Stichting Academisch Rekencentrum Amsterdam http://www.sara.nl
High Performance Computing Tel. +31 20 592 8008 Fax. +31 20 668 3167
PGP keys at http://home.sara.nl/~remco/keys.asc
"I really didn't foresee the Internet. But then, neither did the computer
industry. Not that that tells us very much of course - the computer
industry
didn't even foresee that the century was going to end." -- Douglas Adams
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.7 (Darwin)
iD8DBQE9jwL/BIoCv9yTlOwRAkGMAKC1mv8cF2kfVIWOWruxjLQMGXSgIgCghGQi
tkJONCA76Lz9dWmJPzOc9pg=
=yOO+
-----END PGP SIGNATURE-----
>after some tiny fixes to reiserfs and the makefile for prep bootfile
>(using ../lib/lib.a vs. ../lib/libz.a) I managed to succesfully compile
>a kernel. It even boots to the point where it frees unused kernel memory
>and then stops... this includes succesfully mounting the root
>filesystem...
>
>Just to let the world know what does and doesn't work on my powerstack
>mobo.
Might be related to the mprotect bug signaled by Paul Mackerras
earlier on this list. Try using the linuxppc-2.5 bk tree
Ben.
On maandag, september 23, 2002, at 03:14 , Benjamin Herrenschmidt wrote:
>
> Might be related to the mprotect bug signaled by Paul Mackerras
> earlier on this list. Try using the linuxppc-2.5 bk tree
>
> Ben.
>
>
Hi,
very well possible. If Linus keeps up the pace on kernel releases I'll
just d/l 2.5.39 tomorrow and try again ;)
--
Met vriendelijke groeten,
Remco Post
SARA - Stichting Academisch Rekencentrum Amsterdam http://www.sara.nl
High Performance Computing Tel. +31 20 592 8008 Fax. +31 20 668 3167
PGP keys at http://home.sara.nl/~remco/keys.asc
"I really didn't foresee the Internet. But then, neither did the computer
industry. Not that that tells us very much of course - the computer
industry
didn't even foresee that the century was going to end." -- Douglas Adams
>very well possible. If Linus keeps up the pace on kernel releases I'll
>just d/l 2.5.39 tomorrow and try again ;)
Well, I don't know if a fix went in yet, Paulus did a quick hack but
didn't submit a real patch as we are waiting from someone knowing
sys_mprotect() better to show up
Ben.
On maandag, september 23, 2002, at 03:41 , Benjamin Herrenschmidt wrote:
>> very well possible. If Linus keeps up the pace on kernel releases I'll
>> just d/l 2.5.39 tomorrow and try again ;)
>
> Well, I don't know if a fix went in yet, Paulus did a quick hack but
> didn't submit a real patch as we are waiting from someone knowing
> sys_mprotect() better to show up
>
> Ben.
>
>
I saw Paulus' post a few days ago. I'm fully aware that 2.5 is not ready
for anything serious yet on ppc, I just became curious to know if I
could get it to compile and run. I'll just keep on trying and posting
small patches that I need to compile things. I don't have the time to do
any real hacking, but a small patch now and then is doable ;-)
--
Met vriendelijke groeten,
Remco Post
SARA - Stichting Academisch Rekencentrum Amsterdam http://www.sara.nl
High Performance Computing Tel. +31 20 592 8008 Fax. +31 20 668 3167
PGP keys at http://home.sara.nl/~remco/keys.asc
"I really didn't foresee the Internet. But then, neither did the computer
industry. Not that that tells us very much of course - the computer
industry
didn't even foresee that the century was going to end." -- Douglas Adams
>I saw Paulus' post a few days ago. I'm fully aware that 2.5 is not ready
>for anything serious yet on ppc, I just became curious to know if I
>could get it to compile and run. I'll just keep on trying and posting
>small patches that I need to compile things. I don't have the time to do
>any real hacking, but a small patch now and then is doable ;-)
Well, if you plan to post PPC related patches, please do so on
the linuxppc-dev list (or CC it at least) so we can know what you
are doing. We do maintain a linuxppc-2.5 bk tree which is regulary
merged with Linux but contains our up-to-date fixes for PPC.
It compiles and runs most of the time (thanks Paulus for maintaining
it)
Ben.
On Mon, Sep 23, 2002 at 02:03:02PM +0200, Remco Post wrote:
> after some tiny fixes to reiserfs and the makefile for prep bootfile
> (using ../lib/lib.a vs. ../lib/libz.a) I managed to succesfully compile
> a kernel. It even boots to the point where it frees unused kernel memory
> and then stops... this includes succesfully mounting the root
> filesystem...
What typo exactly? The only 'lib' in the Makefile
(arch/ppc/boot/prep/Makefile) is:
LIBS = ../lib/zlib.a
--
Tom Rini (TR1265)
http://gate.crashing.org/~trini/
On maandag, september 23, 2002, at 04:29 , Tom Rini wrote:
> On Mon, Sep 23, 2002 at 02:03:02PM +0200, Remco Post wrote:
>
>> after some tiny fixes to reiserfs and the makefile for prep bootfile
>> (using ../lib/lib.a vs. ../lib/libz.a) I managed to succesfully compile
>> a kernel. It even boots to the point where it frees unused kernel
>> memory
>> and then stops... this includes succesfully mounting the root
>> filesystem...
>
> What typo exactly? The only 'lib' in the Makefile
> (arch/ppc/boot/prep/Makefile) is:
> LIBS = ../lib/zlib.a
>
That one exactly... I don't recall calling it a typo, though ;-) I guess
that is more a relic from when the only lib routines were libz ones and
we called the lib to be linked libz.a....
There is a simular entry in arch/ppc/boot/openfirmware/Makefile...
removing the z helped over there as well (don't recall exactly, I'm at
work now... ) (not that I dare booting Linus's 2.5 tree on my build
machine, it's falling apart even with stable software.... ;-(
> --
> Tom Rini (TR1265)
> http://gate.crashing.org/~trini/
>
--
Met vriendelijke groeten,
Remco Post
SARA - Stichting Academisch Rekencentrum Amsterdam http://www.sara.nl
High Performance Computing Tel. +31 20 592 8008 Fax. +31 20 668 3167
PGP keys at http://home.sara.nl/~remco/keys.asc
"I really didn't foresee the Internet. But then, neither did the computer
industry. Not that that tells us very much of course - the computer
industry
didn't even foresee that the century was going to end." -- Douglas Adams
On Mon, 23 Sep 2002, Remco Post wrote:
> > What typo exactly? The only 'lib' in the Makefile
> > (arch/ppc/boot/prep/Makefile) is:
> > LIBS = ../lib/zlib.a
> >
>
> That one exactly... I don't recall calling it a typo, though ;-) I guess
> that is more a relic from when the only lib routines were libz ones and
> we called the lib to be linked libz.a....
That's my bad, I renamed the L_TARGET but obviously missed some references
to it. I'll include a fix in my next update.
--Kai
On Mon, Sep 23, 2002 at 04:39:49PM +0200, Remco Post wrote:
>
> On maandag, september 23, 2002, at 04:29 , Tom Rini wrote:
>
> >On Mon, Sep 23, 2002 at 02:03:02PM +0200, Remco Post wrote:
> >
> >>after some tiny fixes to reiserfs and the makefile for prep bootfile
> >>(using ../lib/lib.a vs. ../lib/libz.a) I managed to succesfully compile
> >>a kernel. It even boots to the point where it frees unused kernel
> >>memory
> >>and then stops... this includes succesfully mounting the root
> >>filesystem...
> >
> >What typo exactly? The only 'lib' in the Makefile
> >(arch/ppc/boot/prep/Makefile) is:
> >LIBS = ../lib/zlib.a
>
> That one exactly... I don't recall calling it a typo, though ;-) I guess
> that is more a relic from when the only lib routines were libz ones and
> we called the lib to be linked libz.a....
It's not a relic, it's design. We don't want the kernel's zlib routines
right now, we want our own. Did it not link the way it was? It's been
a few releases since I last compiled for my PReP box, but it was well
after the zlib changes had gone in.
> There is a simular entry in arch/ppc/boot/openfirmware/Makefile...
> removing the z helped over there as well (don't recall exactly, I'm at
> work now... ) (not that I dare booting Linus's 2.5 tree on my build
> machine, it's falling apart even with stable software.... ;-(
Again, we don't want the 'normal' zlib there either. Was it not
linking? If so, what was the error?
--
Tom Rini (TR1265)
http://gate.crashing.org/~trini/
On maandag, september 23, 2002, at 04:55 , Tom Rini wrote:
> On Mon, Sep 23, 2002 at 04:39:49PM +0200, Remco Post wrote:
>>
>> On maandag, september 23, 2002, at 04:29 , Tom Rini wrote:
>>
>>> On Mon, Sep 23, 2002 at 02:03:02PM +0200, Remco Post wrote:
>>>
>>>> after some tiny fixes to reiserfs and the makefile for prep bootfile
>>>> (using ../lib/lib.a vs. ../lib/libz.a) I managed to succesfully
>>>> compile
>>>> a kernel. It even boots to the point where it frees unused kernel
>>>> memory
>>>> and then stops... this includes succesfully mounting the root
>>>> filesystem...
>>>
>>> What typo exactly? The only 'lib' in the Makefile
>>> (arch/ppc/boot/prep/Makefile) is:
>>> LIBS = ../lib/zlib.a
>>
>> That one exactly... I don't recall calling it a typo, though ;-) I
>> guess
>> that is more a relic from when the only lib routines were libz ones and
>> we called the lib to be linked libz.a....
>
> It's not a relic, it's design. We don't want the kernel's zlib routines
> right now, we want our own. Did it not link the way it was? It's been
> a few releases since I last compiled for my PReP box, but it was well
> after the zlib changes had gone in.
>
>> There is a simular entry in arch/ppc/boot/openfirmware/Makefile...
>> removing the z helped over there as well (don't recall exactly, I'm at
>> work now... ) (not that I dare booting Linus's 2.5 tree on my build
>> machine, it's falling apart even with stable software.... ;-(
>
> Again, we don't want the 'normal' zlib there either. Was it not
> linking? If so, what was the error?
>
no rule to make target ../lib/libz.a needed for target
../images/zImage.prep (and simulair for openfirmware...), or something
like that (I have to this from what I remember) I found a file
../lib/lib.a, so I assumed it was renamed, not updating the makefile for
the prep and openfirmware image
This is a linking error in the final stage of the build.... I think
2.5.33 als build ok, after a few patches. Booted as far a to tell me
there was an attempt to kill pid 0 and the do its 180 sec. countdown...
> --
> Tom Rini (TR1265)
> http://gate.crashing.org/~trini/
>
--
Met vriendelijke groeten,
Remco Post
SARA - Stichting Academisch Rekencentrum Amsterdam http://www.sara.nl
High Performance Computing Tel. +31 20 592 8008 Fax. +31 20 668 3167
PGP keys at http://home.sara.nl/~remco/keys.asc
"I really didn't foresee the Internet. But then, neither did the computer
industry. Not that that tells us very much of course - the computer
industry
didn't even foresee that the century was going to end." -- Douglas Adams