* Re: 2.6.16-rc3-mm1
2006-02-14 9:41 2.6.16-rc3-mm1 Andrew Morton
@ 2006-02-14 11:07 ` Marc Koschewski
2006-02-14 11:16 ` 2.6.16-rc3-mm1 Michal Piotrowski
` (5 subsequent siblings)
6 siblings, 0 replies; 24+ messages in thread
From: Marc Koschewski @ 2006-02-14 11:07 UTC (permalink / raw)
To: Andrew Morton; +Cc: linux-kernel
* Andrew Morton <akpm@osdl.org> [2006-02-14 01:41:57 -0800]:
> - To fetch this -mm trees using git, use (for example)
>
> git fetch git://git.kernel.org/pub/scm/linux/kernel/git/smurf/linux-trees.git v2.6.16-rc2-mm1
>
Andrew,
the git link has the wrong release. It should be v2.6.16-rc3-mm1,
shouldn't it?
Marc
^ permalink raw reply [flat|nested] 24+ messages in thread* Re: 2.6.16-rc3-mm1
2006-02-14 9:41 2.6.16-rc3-mm1 Andrew Morton
2006-02-14 11:07 ` 2.6.16-rc3-mm1 Marc Koschewski
@ 2006-02-14 11:16 ` Michal Piotrowski
2006-02-14 12:16 ` 2.6.16-rc3-mm1 Sander
2006-02-14 13:17 ` 2.6.16-rc3-mm1: i386 compilation broken Adrian Bunk
` (4 subsequent siblings)
6 siblings, 1 reply; 24+ messages in thread
From: Michal Piotrowski @ 2006-02-14 11:16 UTC (permalink / raw)
To: Andrew Morton; +Cc: linux-kernel, sam
Hi,
On 14/02/06, Andrew Morton <akpm@osdl.org> wrote:
>
> ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.16-rc3/2.6.16-rc3-mm1/
>
> - Various fixes, updates and cleanups. Nothing very exciting, unless you
> spend a lot of your time waiting for msync() to complete.
>
> - Again, please cast an eye across this patch series for things which should
> go into 2.6.16.
It's strange... rc3-mm1 vs. rc2-mm1
:/usr/src/linux-mm$ uname -a
Linux ltg01-sid 2.6.16-rc2-mm1 #15 SMP PREEMPT Thu Feb 9 18:12:08 CET
2006 i686 GNU/Linux
:/usr/src/linux-mm$ head Makefile
VERSION = 2
PATCHLEVEL = 6
SUBLEVEL = 16
EXTRAVERSION =-rc3-mm1
there is something wrong with build system.
I had a lot of "modprobe: FATAL: Could not load
/lib/modules/2.6.16-rc2-mm1/ modules.dep: No such file or directory"
messages while boot.
Workaround: copy files from /lib/modules/2.6.16-rc3-mm1 to
/lib/modules/2.6.16-rc2-mm1
Regards,
Michal Piotrowski
^ permalink raw reply [flat|nested] 24+ messages in thread* Re: 2.6.16-rc3-mm1
2006-02-14 11:16 ` 2.6.16-rc3-mm1 Michal Piotrowski
@ 2006-02-14 12:16 ` Sander
2006-02-14 12:46 ` 2.6.16-rc3-mm1 Michal Piotrowski
0 siblings, 1 reply; 24+ messages in thread
From: Sander @ 2006-02-14 12:16 UTC (permalink / raw)
To: Michal Piotrowski; +Cc: Andrew Morton, linux-kernel, sam
Michal Piotrowski wrote (ao):
> It's strange... rc3-mm1 vs. rc2-mm1
>
> :/usr/src/linux-mm$ uname -a
> Linux ltg01-sid 2.6.16-rc2-mm1 #15 SMP PREEMPT Thu Feb 9 18:12:08 CET
> 2006 i686 GNU/Linux
>
>
> :/usr/src/linux-mm$ head Makefile
> VERSION = 2
> PATCHLEVEL = 6
> SUBLEVEL = 16
> EXTRAVERSION =-rc3-mm1
>
> there is something wrong with build system.
You just booted an old kernel (see the date in your uname output).
Kind regards, Sander
--
Humilis IT Services and Solutions
http://www.humilis.net
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: 2.6.16-rc3-mm1
2006-02-14 12:16 ` 2.6.16-rc3-mm1 Sander
@ 2006-02-14 12:46 ` Michal Piotrowski
0 siblings, 0 replies; 24+ messages in thread
From: Michal Piotrowski @ 2006-02-14 12:46 UTC (permalink / raw)
To: sander; +Cc: Andrew Morton, linux-kernel, sam
Hi,
On 14/02/06, Sander <sander@humilis.net> wrote:
> Michal Piotrowski wrote (ao):
> > It's strange... rc3-mm1 vs. rc2-mm1
> >
> > :/usr/src/linux-mm$ uname -a
> > Linux ltg01-sid 2.6.16-rc2-mm1 #15 SMP PREEMPT Thu Feb 9 18:12:08 CET
> > 2006 i686 GNU/Linux
> >
> >
> > :/usr/src/linux-mm$ head Makefile
> > VERSION = 2
> > PATCHLEVEL = 6
> > SUBLEVEL = 16
> > EXTRAVERSION =-rc3-mm1
> >
> > there is something wrong with build system.
>
> You just booted an old kernel (see the date in your uname output).
>
> Kind regards, Sander
>
> --
> Humilis IT Services and Solutions
> http://www.humilis.net
>
Ups...
But wait! It's my first -mm kernel on that box after system
reinstallation. uname contain my new hostname "ltg01-sid" - it's very
strange.
I will check my build script.
Regards,
Michal Piotrowski
^ permalink raw reply [flat|nested] 24+ messages in thread
* 2.6.16-rc3-mm1: i386 compilation broken
2006-02-14 9:41 2.6.16-rc3-mm1 Andrew Morton
2006-02-14 11:07 ` 2.6.16-rc3-mm1 Marc Koschewski
2006-02-14 11:16 ` 2.6.16-rc3-mm1 Michal Piotrowski
@ 2006-02-14 13:17 ` Adrian Bunk
2006-02-14 13:27 ` Andi Kleen
2006-02-14 14:00 ` 2.6.16-rc3-mm1: ISDN_DRV_GIGASET driver Adrian Bunk
` (3 subsequent siblings)
6 siblings, 1 reply; 24+ messages in thread
From: Adrian Bunk @ 2006-02-14 13:17 UTC (permalink / raw)
To: Andrew Morton, Andi Kleen; +Cc: linux-kernel
On Tue, Feb 14, 2006 at 01:41:57AM -0800, Andrew Morton wrote:
>...
> Changes since 2.6.16-rc2-mm1:
>...
> +x86_64-fix-string.patch
>...
> x86_64 tree updates.
>...
This patch breaks the compilation on i386:
<-- snip -->
...
LD .tmp_vmlinux1
arch/i386/kernel/built-in.o: In function
`show_type':intel_cacheinfo.c:(.text+0x768b): undefined reference to `strcpy'
:intel_cacheinfo.c:(.text+0x769d): undefined reference to `strcpy'
:intel_cacheinfo.c:(.text+0x76af): undefined reference to `strcpy'
:intel_cacheinfo.c:(.text+0x76c1): undefined reference to `strcpy'
kernel/built-in.o: In function `prof_cpu_mask_read_proc':profile.c:(.text+0x4a84): undefined reference to `strcpy'
kernel/built-in.o:clocksource.c:(.text+0x17bc1): more undefined references to `strcpy' follow
drivers/built-in.o: In function `zoran_write':zoran_procfs.c:(.text+0x41edd6): undefined reference to `strchr'
drivers/built-in.o: In function `cpia_read_proc':cpia.c:(.text+0x42d175): undefined reference to `strcpy'
:cpia.c:(.text+0x42d34c): undefined reference to `strcpy'
:cpia.c:(.text+0x42d35c): undefined reference to `strcpy'
:cpia.c:(.text+0x42d4d7): undefined reference to `strcpy'
:cpia.c:(.text+0x42d5df): undefined reference to `strcpy'
drivers/built-in.o:cpia.c:(.text+0x42d8b9): more undefined references to `strcpy' follow
make: *** [.tmp_vmlinux1] Error 1
<-- snip -->
Andi, you should have known that your patch could breaks i386 now that
we are no longer using no-unit-at-a-time:
http://lkml.org/lkml/2004/11/8/284
cu
Adrian
--
"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon Seed
^ permalink raw reply [flat|nested] 24+ messages in thread* Re: 2.6.16-rc3-mm1: i386 compilation broken
2006-02-14 13:17 ` 2.6.16-rc3-mm1: i386 compilation broken Adrian Bunk
@ 2006-02-14 13:27 ` Andi Kleen
2006-02-14 13:54 ` Michal Piotrowski
` (2 more replies)
0 siblings, 3 replies; 24+ messages in thread
From: Andi Kleen @ 2006-02-14 13:27 UTC (permalink / raw)
To: Adrian Bunk; +Cc: Andrew Morton, linux-kernel
On Tuesday 14 February 2006 14:17, Adrian Bunk wrote:
> On Tue, Feb 14, 2006 at 01:41:57AM -0800, Andrew Morton wrote:
> >...
> > Changes since 2.6.16-rc2-mm1:
> >...
> > +x86_64-fix-string.patch
> >...
> > x86_64 tree updates.
> >...
>
> This patch breaks the compilation on i386:
Ok then the -ffreestanding was apparently still needed on other architectures too.
I guess that part of the patch can be just dropped.
Andrew can you drop that please?
-Andi
Index: linux/Makefile
===================================================================
--- linux.orig/Makefile
+++ linux/Makefile
@@ -339,8 +339,7 @@ LINUXINCLUDE := -Iinclude \
CPPFLAGS := -D__KERNEL__ $(LINUXINCLUDE)
CFLAGS := -Wall -Wundef -Wstrict-prototypes -Wno-trigraphs \
- -fno-strict-aliasing -fno-common \
- -ffreestanding
+ -fno-strict-aliasing -fno-common
AFLAGS := -D__ASSEMBLY__
# Read KERNELRELEASE from .kernelrelease (if it exists)
^ permalink raw reply [flat|nested] 24+ messages in thread* Re: 2.6.16-rc3-mm1: i386 compilation broken
2006-02-14 13:27 ` Andi Kleen
@ 2006-02-14 13:54 ` Michal Piotrowski
2006-02-14 20:23 ` Andrew Morton
2006-02-14 21:08 ` Adrian Bunk
2006-02-15 13:19 ` Roman Zippel
2 siblings, 1 reply; 24+ messages in thread
From: Michal Piotrowski @ 2006-02-14 13:54 UTC (permalink / raw)
To: Andi Kleen; +Cc: Adrian Bunk, Andrew Morton, linux-kernel
Hi,
On 14/02/06, Andi Kleen <ak@suse.de> wrote:
> On Tuesday 14 February 2006 14:17, Adrian Bunk wrote:
> > On Tue, Feb 14, 2006 at 01:41:57AM -0800, Andrew Morton wrote:
> > >...
> > > Changes since 2.6.16-rc2-mm1:
> > >...
> > > +x86_64-fix-string.patch
> > >...
> > > x86_64 tree updates.
> > >...
> >
> > This patch breaks the compilation on i386:
>
> Ok then the -ffreestanding was apparently still needed on other architectures too.
> I guess that part of the patch can be just dropped.
>
> Andrew can you drop that please?
>
> -Andi
Thanks, problem solved!
Andrew can you add this to hot-fixes?
Regards,
Michal Piotrowski
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: 2.6.16-rc3-mm1: i386 compilation broken
2006-02-14 13:54 ` Michal Piotrowski
@ 2006-02-14 20:23 ` Andrew Morton
2006-02-14 21:05 ` Andi Kleen
0 siblings, 1 reply; 24+ messages in thread
From: Andrew Morton @ 2006-02-14 20:23 UTC (permalink / raw)
To: Michal Piotrowski; +Cc: ak, bunk, linux-kernel
Michal Piotrowski <michal.k.k.piotrowski@gmail.com> wrote:
>
> Hi,
>
> On 14/02/06, Andi Kleen <ak@suse.de> wrote:
> > On Tuesday 14 February 2006 14:17, Adrian Bunk wrote:
> > > On Tue, Feb 14, 2006 at 01:41:57AM -0800, Andrew Morton wrote:
> > > >...
> > > > Changes since 2.6.16-rc2-mm1:
> > > >...
> > > > +x86_64-fix-string.patch
> > > >...
> > > > x86_64 tree updates.
> > > >...
> > >
> > > This patch breaks the compilation on i386:
> >
> > Ok then the -ffreestanding was apparently still needed on other architectures too.
> > I guess that part of the patch can be just dropped.
> >
> > Andrew can you drop that please?
> >
> > -Andi
>
> Thanks, problem solved!
Andi like to break x86 - I think it's a market-share thing.
I wonder why I didn't hit that problem.
> Andrew can you add this to hot-fixes?
Done. I backed out the whole patch.
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: 2.6.16-rc3-mm1: i386 compilation broken
2006-02-14 20:23 ` Andrew Morton
@ 2006-02-14 21:05 ` Andi Kleen
0 siblings, 0 replies; 24+ messages in thread
From: Andi Kleen @ 2006-02-14 21:05 UTC (permalink / raw)
To: Andrew Morton; +Cc: Michal Piotrowski, bunk, linux-kernel
On Tuesday 14 February 2006 21:23, Andrew Morton wrote:
>
> Andi like to break x86 - I think it's a market-share thing.
:) I'll try to do 32bit builds more often.
> I wonder why I didn't hit that problem.
It only hits with allyesconfig or obscure configurations I think
>
> > Andrew can you add this to hot-fixes?
>
> Done. I backed out the whole patch.
It would have been enough to just back out that hunk. Or if you do a resync
you'll get a fixed patch.
-Andi
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: 2.6.16-rc3-mm1: i386 compilation broken
2006-02-14 13:27 ` Andi Kleen
2006-02-14 13:54 ` Michal Piotrowski
@ 2006-02-14 21:08 ` Adrian Bunk
2006-02-15 13:19 ` Roman Zippel
2 siblings, 0 replies; 24+ messages in thread
From: Adrian Bunk @ 2006-02-14 21:08 UTC (permalink / raw)
To: Andi Kleen; +Cc: Andrew Morton, linux-kernel
On Tue, Feb 14, 2006 at 02:27:49PM +0100, Andi Kleen wrote:
> On Tuesday 14 February 2006 14:17, Adrian Bunk wrote:
> > On Tue, Feb 14, 2006 at 01:41:57AM -0800, Andrew Morton wrote:
> > >...
> > > Changes since 2.6.16-rc2-mm1:
> > >...
> > > +x86_64-fix-string.patch
> > >...
> > > x86_64 tree updates.
> > >...
> >
> > This patch breaks the compilation on i386:
>
> Ok then the -ffreestanding was apparently still needed on other architectures too.
> I guess that part of the patch can be just dropped.
>...
This means you'd no longer use the builtins on x86_64 ???
> -Andi
cu
Adrian
--
"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon Seed
^ permalink raw reply [flat|nested] 24+ messages in thread* Re: 2.6.16-rc3-mm1: i386 compilation broken
2006-02-14 13:27 ` Andi Kleen
2006-02-14 13:54 ` Michal Piotrowski
2006-02-14 21:08 ` Adrian Bunk
@ 2006-02-15 13:19 ` Roman Zippel
2006-02-15 13:32 ` Andi Kleen
2 siblings, 1 reply; 24+ messages in thread
From: Roman Zippel @ 2006-02-15 13:19 UTC (permalink / raw)
To: Andi Kleen; +Cc: Adrian Bunk, Andrew Morton, linux-kernel
Hi,
On Tue, 14 Feb 2006, Andi Kleen wrote:
> Ok then the -ffreestanding was apparently still needed on other architectures too.
> I guess that part of the patch can be just dropped.
The main problem is still the sprintf optimization, so
--fno-builtin-sprintf should fix it too.
That leaves only the single strchr, which is caused by an strpbrk
optimization in zoran_procfs.c, where we could use --fno-builtin-strpbrk
or simply directly replace that strpbrk with strchr.
If we really want to keep -ffreestanding, we have to rework how string.h
is organized to allow enabling builtin functions, but still provide fall
back functions. For example we had to add a lot of "#define foo
__builtin_foo" to linux/string.h and "#undef foo" to lib/string.c.
bye, Roman
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: 2.6.16-rc3-mm1: i386 compilation broken
2006-02-15 13:19 ` Roman Zippel
@ 2006-02-15 13:32 ` Andi Kleen
0 siblings, 0 replies; 24+ messages in thread
From: Andi Kleen @ 2006-02-15 13:32 UTC (permalink / raw)
To: Roman Zippel; +Cc: Adrian Bunk, Andrew Morton, linux-kernel
On Wednesday 15 February 2006 14:19, Roman Zippel wrote:
> Hi,
>
> On Tue, 14 Feb 2006, Andi Kleen wrote:
>
> > Ok then the -ffreestanding was apparently still needed on other architectures too.
> > I guess that part of the patch can be just dropped.
>
> The main problem is still the sprintf optimization, so
> --fno-builtin-sprintf should fix it too.
Currently it will just use out of line strcpy etc. on x86-64.
Not quite optimal - probably need to go back to fix this
> That leaves only the single strchr, which is caused by an strpbrk
> optimization in zoran_procfs.c, where we could use --fno-builtin-strpbrk
> or simply directly replace that strpbrk with strchr.
>
> If we really want to keep -ffreestanding,
I think we should drop it, just i386 has to be fixed first.
> we have to rework how string.h
> is organized to allow enabling builtin functions, but still provide fall
> back functions. For example we had to add a lot of "#define foo
> __builtin_foo" to linux/string.h and "#undef foo" to lib/string.c.
Yes it would be ugly.
-Andi
^ permalink raw reply [flat|nested] 24+ messages in thread
* 2.6.16-rc3-mm1: ISDN_DRV_GIGASET driver
2006-02-14 9:41 2.6.16-rc3-mm1 Andrew Morton
` (2 preceding siblings ...)
2006-02-14 13:17 ` 2.6.16-rc3-mm1: i386 compilation broken Adrian Bunk
@ 2006-02-14 14:00 ` Adrian Bunk
2006-02-15 1:57 ` Tilman Schmidt
2006-02-14 14:33 ` 2.6.16-rc3-mm1 Brice Goglin
` (2 subsequent siblings)
6 siblings, 1 reply; 24+ messages in thread
From: Adrian Bunk @ 2006-02-14 14:00 UTC (permalink / raw)
To: Andrew Morton, Hansjoerg Lipp, Tilman Schmidt
Cc: linux-kernel, kkeil, kai.germaschewski, isdn4linux
On Tue, Feb 14, 2006 at 01:41:57AM -0800, Andrew Morton wrote:
>...
> Changes since 2.6.16-rc2-mm1:
>...
> +isdn4linux-siemens-gigaset-drivers-kconfigs-and-makefiles.patch
> +isdn4linux-siemens-gigaset-drivers-common-module.patch
> +isdn4linux-siemens-gigaset-drivers-event-layer.patch
> +isdn4linux-siemens-gigaset-drivers-isdn4linux-interface.patch
> +isdn4linux-siemens-gigaset-drivers-tty-interface.patch
> +isdn4linux-siemens-gigaset-drivers-procfs-interface.patch
> +isdn4linux-siemens-gigaset-drivers-direct-usb-connection.patch
> +isdn4linux-siemens-gigaset-drivers-isochronous-data-handler.patch
> +isdn4linux-siemens-gigaset-drivers-m105-usb-dect-adapter.patch
>
> New ISDN driver.
>...
I see two big problems with this driver:
- Do we really want to add new non-CAPI drivers to the kernel?
- A new driver that can only be built modular is not acceptable.
cu
Adrian
--
"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon Seed
^ permalink raw reply [flat|nested] 24+ messages in thread* Re: 2.6.16-rc3-mm1: ISDN_DRV_GIGASET driver
2006-02-14 14:00 ` 2.6.16-rc3-mm1: ISDN_DRV_GIGASET driver Adrian Bunk
@ 2006-02-15 1:57 ` Tilman Schmidt
2006-02-15 22:10 ` Adrian Bunk
0 siblings, 1 reply; 24+ messages in thread
From: Tilman Schmidt @ 2006-02-15 1:57 UTC (permalink / raw)
To: Adrian Bunk
Cc: Andrew Morton, Hansjoerg Lipp, linux-kernel, kkeil,
kai.germaschewski, isdn4linux
[-- Attachment #1: Type: text/plain, Size: 1826 bytes --]
Adrian,
thank you very much for taking the time to comment.
On 14.02.2006 15:00, you wrote:
> - Do we really want to add new non-CAPI drivers to the kernel?
I have been in contact with the isdn4linux maintainer Karsten Keil on
that topic for quite some time and he didn't voice any objections to
submitting the driver in its current state.
Personally I am a great fan of CAPI, and of course we'll be happy to
port the driver to CAPI as soon as the capi4linux / mISDN framework is
ready for such an endeavour. This may however take some time yet, if I
understand Karsten correctly. In particular, we are talking here about
the mISDN L3 interface which seems to be the most appropriate for this
purpose, but has not been documented so far.
In the meantime, I take it from the discussions on lkml that it is
strongly discouraged to maintain working drivers outside the kernel
tree, which is what prompted us to submit ours in the first place.
Therefore I think it's in the best interest of everybody to merge its
current isnd4linux gestalt now, and convert it to CAPI at the earliest
convenience.
> - A new driver that can only be built modular is not acceptable.
No problem. In fact, the submitted drivers work fine if linked directly
into the kernel, too. The dependency on modular build in gigaset/Kconfig
only exists for the benefit of the ser_gigaset driver which we didn't
submit anyway. (See part 0 for the reasons.) We left it in because tests
have been done almost exclusively with modular builds; but if that turns
out to be a problem we'll just remove it.
Regards
Tilman
--
Tilman Schmidt E-Mail: tilman@imap.cc
Bonn, Germany
Diese Nachricht besteht zu 100% aus wiederverwerteten Bits.
Ungeöffnet mindestens haltbar bis: (siehe Rückseite)
[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 250 bytes --]
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: 2.6.16-rc3-mm1: ISDN_DRV_GIGASET driver
2006-02-15 1:57 ` Tilman Schmidt
@ 2006-02-15 22:10 ` Adrian Bunk
0 siblings, 0 replies; 24+ messages in thread
From: Adrian Bunk @ 2006-02-15 22:10 UTC (permalink / raw)
To: Tilman Schmidt
Cc: Andrew Morton, Hansjoerg Lipp, linux-kernel, kkeil,
kai.germaschewski, isdn4linux
On Wed, Feb 15, 2006 at 02:57:01AM +0100, Tilman Schmidt wrote:
> Adrian,
Hi Tilman,
> thank you very much for taking the time to comment.
>
> On 14.02.2006 15:00, you wrote:
> > - Do we really want to add new non-CAPI drivers to the kernel?
>
> I have been in contact with the isdn4linux maintainer Karsten Keil on
> that topic for quite some time and he didn't voice any objections to
> submitting the driver in its current state.
>
> Personally I am a great fan of CAPI, and of course we'll be happy to
> port the driver to CAPI as soon as the capi4linux / mISDN framework is
> ready for such an endeavour. This may however take some time yet, if I
> understand Karsten correctly. In particular, we are talking here about
> the mISDN L3 interface which seems to be the most appropriate for this
> purpose, but has not been documented so far.
>
> In the meantime, I take it from the discussions on lkml that it is
> strongly discouraged to maintain working drivers outside the kernel
> tree, which is what prompted us to submit ours in the first place.
> Therefore I think it's in the best interest of everybody to merge its
> current isnd4linux gestalt now, and convert it to CAPI at the earliest
> convenience.
OK.
> > - A new driver that can only be built modular is not acceptable.
>
> No problem. In fact, the submitted drivers work fine if linked directly
> into the kernel, too. The dependency on modular build in gigaset/Kconfig
> only exists for the benefit of the ser_gigaset driver which we didn't
> submit anyway. (See part 0 for the reasons.) We left it in because tests
> have been done almost exclusively with modular builds; but if that turns
> out to be a problem we'll just remove it.
All drivers in the kernel should work built statically into the kernel,
and I don't see an obvious reason why your driver should work worse when
being compiled statically into the kernel.
> Regards
> Tilman
cu
Adrian
--
"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon Seed
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: 2.6.16-rc3-mm1
2006-02-14 9:41 2.6.16-rc3-mm1 Andrew Morton
` (3 preceding siblings ...)
2006-02-14 14:00 ` 2.6.16-rc3-mm1: ISDN_DRV_GIGASET driver Adrian Bunk
@ 2006-02-14 14:33 ` Brice Goglin
2006-02-14 20:28 ` 2.6.16-rc3-mm1 Andrew Morton
2006-02-14 15:22 ` [-mm patch] block/blktrace.c: make blk_trace_cleanup() static Adrian Bunk
2006-02-16 23:28 ` 2.6.16-rc3-mm1 Pavel Machek
6 siblings, 1 reply; 24+ messages in thread
From: Brice Goglin @ 2006-02-14 14:33 UTC (permalink / raw)
To: Andrew Morton; +Cc: linux-kernel
Andrew Morton wrote:
>ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.16-rc3/2.6.16-rc3-mm1/
>
>
>
Hi Andrew,
WARNING: speedstep-centrino.ko needs unknown symbol cpu_online_map
This symbol is in include/linux/cpumask.h but actually only defined and
exported in smpboot.c which is not compiled on UP.
Regards,
Brice
^ permalink raw reply [flat|nested] 24+ messages in thread* Re: 2.6.16-rc3-mm1
2006-02-14 14:33 ` 2.6.16-rc3-mm1 Brice Goglin
@ 2006-02-14 20:28 ` Andrew Morton
2006-02-14 23:26 ` 2.6.16-rc3-mm1 Con Kolivas
0 siblings, 1 reply; 24+ messages in thread
From: Andrew Morton @ 2006-02-14 20:28 UTC (permalink / raw)
To: Brice Goglin; +Cc: linux-kernel
Brice Goglin <Brice.Goglin@ens-lyon.org> wrote:
>
> Andrew Morton wrote:
>
> >ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.16-rc3/2.6.16-rc3-mm1/
> >
> >
> >
>
> Hi Andrew,
>
> WARNING: speedstep-centrino.ko needs unknown symbol cpu_online_map
>
> This symbol is in include/linux/cpumask.h but actually only defined and
> exported in smpboot.c which is not compiled on UP.
>
diff -puN arch/i386/kernel/cpu/cpufreq/speedstep-centrino.c~git-acpi-up-fix-2 arch/i386/kernel/cpu/cpufreq/speedstep-centrino.c
--- devel/arch/i386/kernel/cpu/cpufreq/speedstep-centrino.c~git-acpi-up-fix-2 2006-02-14 12:27:41.000000000 -0800
+++ devel-akpm/arch/i386/kernel/cpu/cpufreq/speedstep-centrino.c 2006-02-14 12:27:41.000000000 -0800
@@ -654,8 +654,10 @@ static int centrino_target (struct cpufr
return -EINVAL;
}
+#ifdef CONFIG_SMP
/* cpufreq holds the hotplug lock, so we are safe from here on */
cpus_and(online_policy_cpus, cpu_online_map, policy->cpus);
+#endif
saved_mask = current->cpus_allowed;
first_cpu = 1;
_
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: 2.6.16-rc3-mm1
2006-02-14 20:28 ` 2.6.16-rc3-mm1 Andrew Morton
@ 2006-02-14 23:26 ` Con Kolivas
2006-02-14 23:38 ` 2.6.16-rc3-mm1 Brice Goglin
2006-02-15 2:58 ` 2.6.16-rc3-mm1 Andrew Morton
0 siblings, 2 replies; 24+ messages in thread
From: Con Kolivas @ 2006-02-14 23:26 UTC (permalink / raw)
To: Andrew Morton; +Cc: Brice Goglin, linux-kernel
On Wednesday 15 February 2006 07:28, Andrew Morton wrote:
> Brice Goglin <Brice.Goglin@ens-lyon.org> wrote:
> > Andrew Morton wrote:
> > >ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.16-rc3
> > >/2.6.16-rc3-mm1/
> >
> > Hi Andrew,
> >
> > WARNING: speedstep-centrino.ko needs unknown symbol cpu_online_map
> >
> > This symbol is in include/linux/cpumask.h but actually only defined and
> > exported in smpboot.c which is not compiled on UP.
>
> diff -puN
> arch/i386/kernel/cpu/cpufreq/speedstep-centrino.c~git-acpi-up-fix-2
> arch/i386/kernel/cpu/cpufreq/speedstep-centrino.c ---
> devel/arch/i386/kernel/cpu/cpufreq/speedstep-centrino.c~git-acpi-up-fix-2 2
>006-02-14 12:27:41.000000000 -0800 +++
> devel-akpm/arch/i386/kernel/cpu/cpufreq/speedstep-centrino.c 2006-02-14
> 12:27:41.000000000 -0800 @@ -654,8 +654,10 @@ static int centrino_target
> (struct cpufr
> return -EINVAL;
> }
>
> +#ifdef CONFIG_SMP
> /* cpufreq holds the hotplug lock, so we are safe from here on */
> cpus_and(online_policy_cpus, cpu_online_map, policy->cpus);
> +#endif
Shouldn't the cpu_online_map be hardcoded to the first/only cpu on UP instead?
Cheers,
Con
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: 2.6.16-rc3-mm1
2006-02-14 23:26 ` 2.6.16-rc3-mm1 Con Kolivas
@ 2006-02-14 23:38 ` Brice Goglin
2006-02-15 2:58 ` 2.6.16-rc3-mm1 Andrew Morton
1 sibling, 0 replies; 24+ messages in thread
From: Brice Goglin @ 2006-02-14 23:38 UTC (permalink / raw)
To: Con Kolivas; +Cc: Andrew Morton, linux-kernel
Con Kolivas wrote:
>On Wednesday 15 February 2006 07:28, Andrew Morton wrote:
>
>
>>Brice Goglin <Brice.Goglin@ens-lyon.org> wrote:
>>
>>
>>>WARNING: speedstep-centrino.ko needs unknown symbol cpu_online_map
>>>
>>>This symbol is in include/linux/cpumask.h but actually only defined and
>>>exported in smpboot.c which is not compiled on UP.
>>>
>>>
>>diff -puN
>>arch/i386/kernel/cpu/cpufreq/speedstep-centrino.c~git-acpi-up-fix-2
>>arch/i386/kernel/cpu/cpufreq/speedstep-centrino.c ---
>>devel/arch/i386/kernel/cpu/cpufreq/speedstep-centrino.c~git-acpi-up-fix-2 2
>>006-02-14 12:27:41.000000000 -0800 +++
>>devel-akpm/arch/i386/kernel/cpu/cpufreq/speedstep-centrino.c 2006-02-14
>>12:27:41.000000000 -0800 @@ -654,8 +654,10 @@ static int centrino_target
>>(struct cpufr
>> return -EINVAL;
>> }
>>
>>+#ifdef CONFIG_SMP
>> /* cpufreq holds the hotplug lock, so we are safe from here on */
>> cpus_and(online_policy_cpus, cpu_online_map, policy->cpus);
>>+#endif
>>
>>
>
>Shouldn't the cpu_online_map be hardcoded to the first/only cpu on UP instead?
>
>Cheers,
>Con
>
>
It's actually defined on UP in kernel/sched.c:
#ifndef CONFIG_SMP
cpumask_t cpu_online_map __read_mostly = CPU_MASK_ALL;
cpumask_t cpu_possible_map __read_mostly = CPU_MASK_ALL;
#endif
What about adding an EXPORT_SYMBOL_GPL there ?
Brice
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: 2.6.16-rc3-mm1
2006-02-14 23:26 ` 2.6.16-rc3-mm1 Con Kolivas
2006-02-14 23:38 ` 2.6.16-rc3-mm1 Brice Goglin
@ 2006-02-15 2:58 ` Andrew Morton
1 sibling, 0 replies; 24+ messages in thread
From: Andrew Morton @ 2006-02-15 2:58 UTC (permalink / raw)
To: Con Kolivas; +Cc: Brice.Goglin, linux-kernel
Con Kolivas <kernel@kolivas.org> wrote:
>
> On Wednesday 15 February 2006 07:28, Andrew Morton wrote:
> > Brice Goglin <Brice.Goglin@ens-lyon.org> wrote:
> > > Andrew Morton wrote:
> > > >ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.16-rc3
> > > >/2.6.16-rc3-mm1/
> > >
> > > Hi Andrew,
> > >
> > > WARNING: speedstep-centrino.ko needs unknown symbol cpu_online_map
> > >
> > > This symbol is in include/linux/cpumask.h but actually only defined and
> > > exported in smpboot.c which is not compiled on UP.
> >
> > diff -puN
> > arch/i386/kernel/cpu/cpufreq/speedstep-centrino.c~git-acpi-up-fix-2
> > arch/i386/kernel/cpu/cpufreq/speedstep-centrino.c ---
> > devel/arch/i386/kernel/cpu/cpufreq/speedstep-centrino.c~git-acpi-up-fix-2 2
> >006-02-14 12:27:41.000000000 -0800 +++
> > devel-akpm/arch/i386/kernel/cpu/cpufreq/speedstep-centrino.c 2006-02-14
> > 12:27:41.000000000 -0800 @@ -654,8 +654,10 @@ static int centrino_target
> > (struct cpufr
> > return -EINVAL;
> > }
> >
> > +#ifdef CONFIG_SMP
> > /* cpufreq holds the hotplug lock, so we are safe from here on */
> > cpus_and(online_policy_cpus, cpu_online_map, policy->cpus);
> > +#endif
>
> Shouldn't the cpu_online_map be hardcoded to the first/only cpu on UP instead?
>
Probably. That's just a make-it-build patch. The ACPI guys will need to
come up with a suitable permanent fix.
^ permalink raw reply [flat|nested] 24+ messages in thread
* [-mm patch] block/blktrace.c: make blk_trace_cleanup() static
2006-02-14 9:41 2.6.16-rc3-mm1 Andrew Morton
` (4 preceding siblings ...)
2006-02-14 14:33 ` 2.6.16-rc3-mm1 Brice Goglin
@ 2006-02-14 15:22 ` Adrian Bunk
2006-02-14 15:23 ` Jens Axboe
2006-02-16 23:28 ` 2.6.16-rc3-mm1 Pavel Machek
6 siblings, 1 reply; 24+ messages in thread
From: Adrian Bunk @ 2006-02-14 15:22 UTC (permalink / raw)
To: Andrew Morton, axboe; +Cc: linux-kernel
On Tue, Feb 14, 2006 at 01:41:57AM -0800, Andrew Morton wrote:
>...
> Changes since 2.6.16-rc2-mm1:
>...
> git-blktrace.patch
>...
blk_trace_cleanup() is needlessly global.
Signed-off-by: Adrian Bunk <bunk@stusta.de>
--- linux-2.6.16-rc3-mm1-full/block/blktrace.c.old 2006-02-14 15:42:12.000000000 +0100
+++ linux-2.6.16-rc3-mm1-full/block/blktrace.c 2006-02-14 15:42:23.000000000 +0100
@@ -171,7 +171,7 @@
return dir;
}
-void blk_trace_cleanup(struct blk_trace *bt)
+static void blk_trace_cleanup(struct blk_trace *bt)
{
relay_close(bt->rchan);
relayfs_remove_file(bt->dropped_file);
^ permalink raw reply [flat|nested] 24+ messages in thread* Re: 2.6.16-rc3-mm1
2006-02-14 9:41 2.6.16-rc3-mm1 Andrew Morton
` (5 preceding siblings ...)
2006-02-14 15:22 ` [-mm patch] block/blktrace.c: make blk_trace_cleanup() static Adrian Bunk
@ 2006-02-16 23:28 ` Pavel Machek
6 siblings, 0 replies; 24+ messages in thread
From: Pavel Machek @ 2006-02-16 23:28 UTC (permalink / raw)
To: Andrew Morton; +Cc: linux-kernel
Hi!
> - Again, please cast an eye across this patch series for things which should
> go into 2.6.16.
...
> -suspend-to-ram-allow-video-options-to-be-set-at-runtime-fix.patch
>
> Folded into suspend-to-ram-allow-video-options-to-be-set-at-runtime.patch
I'd quite like this one to go in. It is not critical and it is not bugfix,
but it is fairly simple...
--
64 bytes from 195.113.31.123: icmp_seq=28 ttl=51 time=448769.1 ms
^ permalink raw reply [flat|nested] 24+ messages in thread