...unfortunately, its kind of hard to debug, as I'm testing kernels
using kexec and only see messages late in boot. 2.6.26 works ok.
Pavel
--
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
Pavel Machek wrote:
> ...unfortunately, its kind of hard to debug, as I'm testing kernels
> using kexec and only see messages late in boot. 2.6.26 works ok.
Lost patch:
http://lists.arm.linux.org.uk/lurker/message/20081007.113509.c17e1e65.en.html
OT: 2.6.26 does not work OK. You need I2C init patch.
--
Best Regards / S pozdravem,
Stanislav Brabec
software developer
---------------------------------------------------------------------
SUSE LINUX, s. r. o. e-mail: [email protected]
Lihovarsk? 1060/12 tel: +420 284 028 966, +49 911 740538747
190 00 Praha 9 fax: +420 284 028 951
Czech Republic http://www.suse.cz/
On Thu, Oct 09, 2008 at 02:00:46PM +0200, Stanislav Brabec wrote:
> Pavel Machek wrote:
> > ...unfortunately, its kind of hard to debug, as I'm testing kernels
> > using kexec and only see messages late in boot. 2.6.26 works ok.
>
> Lost patch:
> http://lists.arm.linux.org.uk/lurker/message/20081007.113509.c17e1e65.en.html
Not really lost. I asked for it, but it didn't arrive in time for the
upcoming merge window.
However, since it is a regression, as identified by Pavel, it is -rc
material, not merge window material which isn't clear from the thread
or the problem description.
So, let's get it in for -final.
--
Russell King
Linux kernel 2.6 ARM Linux - http://www.arm.linux.org.uk/
maintainer of:
On Thu, Oct 09, 2008 at 02:00:46PM +0200, Stanislav Brabec wrote:
> Pavel Machek wrote:
> > ...unfortunately, its kind of hard to debug, as I'm testing kernels
> > using kexec and only see messages late in boot. 2.6.26 works ok.
>
> Lost patch:
> http://lists.arm.linux.org.uk/lurker/message/20081007.113509.c17e1e65.en.html
>
> OT: 2.6.26 does not work OK. You need I2C init patch.
... and now queued for Linus.
--
Russell King
Linux kernel 2.6 ARM Linux - http://www.arm.linux.org.uk/
maintainer of:
On Thu 2008-10-09 14:00:46, Stanislav Brabec wrote:
> Pavel Machek wrote:
> > ...unfortunately, its kind of hard to debug, as I'm testing kernels
> > using kexec and only see messages late in boot. 2.6.26 works ok.
>
> Lost patch:
> http://lists.arm.linux.org.uk/lurker/message/20081007.113509.c17e1e65.en.html
Thanks!
> OT: 2.6.26 does not work OK. You need I2C init patch.
It boots for me :-)
--
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
> On Thu, Oct 09, 2008 at 02:00:46PM +0200, Stanislav Brabec wrote:
> > Pavel Machek wrote:
> > > ...unfortunately, its kind of hard to debug, as I'm testing kernels
> > > using kexec and only see messages late in boot. 2.6.26 works ok.
> >
> > Lost patch:
> > http://lists.arm.linux.org.uk/lurker/message/20081007.113509.c17e1e65.en.html
>
> Not really lost. I asked for it, but it didn't arrive in time for the
> upcoming merge window.
>
> However, since it is a regression, as identified by Pavel, it is -rc
> material, not merge window material which isn't clear from the thread
> or the problem description.
>
> So, let's get it in for -final.
Thanks for the effort, but it looks like it did not make it, right?
Anyway, thanks for the patch, it worked for me. I will scream if
2.6.28-rc1 does not work.
Pavel
--
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
On Wed, Oct 15, 2008 at 12:17:47PM +0200, Pavel Machek wrote:
> Thanks for the effort, but it looks like it did not make it, right?
Correct - Linus was unpredictable and released 2.6.27 that evening,
four days early giving me no time to push the fix to him.
However, it's gone in as part of the first ARM merge, so should be in
-rc1.
--
Russell King
Linux kernel 2.6 ARM Linux - http://www.arm.linux.org.uk/
maintainer of:
On Wednesday, 15 of October 2008, Russell King wrote:
> On Wed, Oct 15, 2008 at 12:17:47PM +0200, Pavel Machek wrote:
> > Thanks for the effort, but it looks like it did not make it, right?
>
> Correct - Linus was unpredictable and released 2.6.27 that evening,
> four days early giving me no time to push the fix to him.
>
> However, it's gone in as part of the first ARM merge, so should be in
> -rc1.
Is it in -stable? if not, is it considered as -stable material?
Rafael
Hi!
> > > Thanks for the effort, but it looks like it did not make it, right?
> >
> > Correct - Linus was unpredictable and released 2.6.27 that evening,
> > four days early giving me no time to push the fix to him.
> >
> > However, it's gone in as part of the first ARM merge, so should be in
> > -rc1.
>
> Is it in -stable? if not, is it considered as -stable material?
Yes, I guess it should go to stable.
commit 69fc7eed5f56bce15b239e5110de2575a6970df4
Author: Dmitry Baryshkov <[email protected]>
Date: Thu Oct 9 16:58:13 2008 +0100
[ARM] 5300/1: fixup spitz reset during boot
Some machines don't have the pullup/down on their reset
pin, so configuring the reset generating pin as input makes
them reset immediately. Fix that by making reset pin direction
configurable.
Signed-off-by: Dmitry Baryshkov <[email protected]>
Signed-off-by: Russell King <[email protected]>
Signed-off-by: Pavel Machek <[email protected]>
diff --git a/arch/arm/mach-pxa/include/mach/reset.h b/arch/arm/mach-pxa/include/mach/reset.h
index 9489a48..7b8842c 100644
--- a/arch/arm/mach-pxa/include/mach/reset.h
+++ b/arch/arm/mach-pxa/include/mach/reset.h
@@ -10,9 +10,12 @@
extern unsigned int reset_status;
extern void clear_reset_status(unsigned int mask);
-/*
- * register GPIO as reset generator
+/**
+ * init_gpio_reset() - register GPIO as reset generator
+ *
+ * @gpio - gpio nr
+ * @output - set gpio as out/low instead of input during normal work
*/
-extern int init_gpio_reset(int gpio);
+extern int init_gpio_reset(int gpio, int output);
#endif /* __ASM_ARCH_RESET_H */
diff --git a/arch/arm/mach-pxa/reset.c b/arch/arm/mach-pxa/reset.c
index 31d7885..1b2af57 100644
--- a/arch/arm/mach-pxa/reset.c
+++ b/arch/arm/mach-pxa/reset.c
@@ -20,7 +20,7 @@ static void do_hw_reset(void);
static int reset_gpio = -1;
-int init_gpio_reset(int gpio)
+int init_gpio_reset(int gpio, int output)
{
int rc;
@@ -30,9 +30,12 @@ int init_gpio_reset(int gpio)
goto out;
}
- rc = gpio_direction_input(gpio);
+ if (output)
+ rc = gpio_direction_output(gpio, 0);
+ else
+ rc = gpio_direction_input(gpio);
if (rc) {
- printk(KERN_ERR "Can't configure reset_gpio for input\n");
+ printk(KERN_ERR "Can't configure reset_gpio\n");
gpio_free(gpio);
goto out;
}
diff --git a/arch/arm/mach-pxa/spitz.c b/arch/arm/mach-pxa/spitz.c
index 9b4f24c..524f656 100644
--- a/arch/arm/mach-pxa/spitz.c
+++ b/arch/arm/mach-pxa/spitz.c
@@ -618,7 +618,7 @@ static void spitz_restart(char mode)
static void __init common_init(void)
{
- init_gpio_reset(SPITZ_GPIO_ON_RESET);
+ init_gpio_reset(SPITZ_GPIO_ON_RESET, 1);
pm_power_off = spitz_poweroff;
arm_pm_restart = spitz_restart;
diff --git a/arch/arm/mach-pxa/tosa.c b/arch/arm/mach-pxa/tosa.c
index 9f3ef9e..130e37e 100644
--- a/arch/arm/mach-pxa/tosa.c
+++ b/arch/arm/mach-pxa/tosa.c
@@ -781,7 +781,7 @@ static void __init tosa_init(void)
gpio_set_wake(MFP_PIN_GPIO1, 1);
/* We can't pass to gpio-keys since it will drop the Reset altfunc */
- init_gpio_reset(TOSA_GPIO_ON_RESET);
+ init_gpio_reset(TOSA_GPIO_ON_RESET, 0);
pm_power_off = tosa_poweroff;
arm_pm_restart = tosa_restart;
--
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
On Sun, 26 Oct 2008 19:41:09 +0100
Pavel Machek <[email protected]> wrote:
> Hi!
>
> > > > Thanks for the effort, but it looks like it did not make it, right?
> > >
> > > Correct - Linus was unpredictable and released 2.6.27 that evening,
> > > four days early giving me no time to push the fix to him.
> > >
> > > However, it's gone in as part of the first ARM merge, so should be in
> > > -rc1.
> >
> > Is it in -stable? if not, is it considered as -stable material?
>
> Yes, I guess it should go to stable.
>
> commit 69fc7eed5f56bce15b239e5110de2575a6970df4
> Author: Dmitry Baryshkov <[email protected]>
> Date: Thu Oct 9 16:58:13 2008 +0100
>
> [ARM] 5300/1: fixup spitz reset during boot
>
> Some machines don't have the pullup/down on their reset
> pin, so configuring the reset generating pin as input makes
> them reset immediately. Fix that by making reset pin direction
> configurable.
>
> Signed-off-by: Dmitry Baryshkov <[email protected]>
> Signed-off-by: Russell King <[email protected]>
> Signed-off-by: Pavel Machek <[email protected]>
>
> diff --git a/arch/arm/mach-pxa/include/mach/reset.h b/arch/arm/mach-pxa/include/mach/reset.h
> index 9489a48..7b8842c 100644
> --- a/arch/arm/mach-pxa/include/mach/reset.h
> +++ b/arch/arm/mach-pxa/include/mach/reset.h
> @@ -10,9 +10,12 @@
> extern unsigned int reset_status;
> extern void clear_reset_status(unsigned int mask);
>
> -/*
> - * register GPIO as reset generator
> +/**
> + * init_gpio_reset() - register GPIO as reset generator
> + *
> + * @gpio - gpio nr
> + * @output - set gpio as out/low instead of input during normal work
Use colon (':') after function parameters, not hyphen/dash.
> */
> -extern int init_gpio_reset(int gpio);
> +extern int init_gpio_reset(int gpio, int output);
>
> #endif /* __ASM_ARCH_RESET_H */
> diff --git a/arch/arm/mach-pxa/reset.c b/arch/arm/mach-pxa/reset.c
> index 31d7885..1b2af57 100644
> --- a/arch/arm/mach-pxa/reset.c
> +++ b/arch/arm/mach-pxa/reset.c
> @@ -20,7 +20,7 @@ static void do_hw_reset(void);
>
> static int reset_gpio = -1;
>
> -int init_gpio_reset(int gpio)
> +int init_gpio_reset(int gpio, int output)
> {
> int rc;
>
> @@ -30,9 +30,12 @@ int init_gpio_reset(int gpio)
> goto out;
> }
>
> - rc = gpio_direction_input(gpio);
> + if (output)
> + rc = gpio_direction_output(gpio, 0);
> + else
> + rc = gpio_direction_input(gpio);
> if (rc) {
> - printk(KERN_ERR "Can't configure reset_gpio for input\n");
> + printk(KERN_ERR "Can't configure reset_gpio\n");
> gpio_free(gpio);
> goto out;
> }
> diff --git a/arch/arm/mach-pxa/spitz.c b/arch/arm/mach-pxa/spitz.c
> index 9b4f24c..524f656 100644
> --- a/arch/arm/mach-pxa/spitz.c
> +++ b/arch/arm/mach-pxa/spitz.c
> @@ -618,7 +618,7 @@ static void spitz_restart(char mode)
>
> static void __init common_init(void)
> {
> - init_gpio_reset(SPITZ_GPIO_ON_RESET);
> + init_gpio_reset(SPITZ_GPIO_ON_RESET, 1);
> pm_power_off = spitz_poweroff;
> arm_pm_restart = spitz_restart;
>
> diff --git a/arch/arm/mach-pxa/tosa.c b/arch/arm/mach-pxa/tosa.c
> index 9f3ef9e..130e37e 100644
> --- a/arch/arm/mach-pxa/tosa.c
> +++ b/arch/arm/mach-pxa/tosa.c
> @@ -781,7 +781,7 @@ static void __init tosa_init(void)
> gpio_set_wake(MFP_PIN_GPIO1, 1);
> /* We can't pass to gpio-keys since it will drop the Reset altfunc */
>
> - init_gpio_reset(TOSA_GPIO_ON_RESET);
> + init_gpio_reset(TOSA_GPIO_ON_RESET, 0);
>
> pm_power_off = tosa_poweroff;
> arm_pm_restart = tosa_restart;
>
>
--
~Randy
On Mon, 27 Oct 2008 09:25:12 -0700 Randy Dunlap <[email protected]> wrote:
> On Sun, 26 Oct 2008 19:41:09 +0100
> Pavel Machek <[email protected]> wrote:
>
> > Hi!
> >
> > > > > Thanks for the effort, but it looks like it did not make it, right?
> > > >
> > > > Correct - Linus was unpredictable and released 2.6.27 that evening,
> > > > four days early giving me no time to push the fix to him.
> > > >
> > > > However, it's gone in as part of the first ARM merge, so should be in
> > > > -rc1.
> > >
> > > Is it in -stable? if not, is it considered as -stable material?
> >
> > Yes, I guess it should go to stable.
> >
> > commit 69fc7eed5f56bce15b239e5110de2575a6970df4
> > Author: Dmitry Baryshkov <[email protected]>
> > Date: Thu Oct 9 16:58:13 2008 +0100
> >
> > [ARM] 5300/1: fixup spitz reset during boot
> >
> > Some machines don't have the pullup/down on their reset
> > pin, so configuring the reset generating pin as input makes
> > them reset immediately. Fix that by making reset pin direction
> > configurable.
> >
> > Signed-off-by: Dmitry Baryshkov <[email protected]>
> > Signed-off-by: Russell King <[email protected]>
> > Signed-off-by: Pavel Machek <[email protected]>
> >
> > diff --git a/arch/arm/mach-pxa/include/mach/reset.h b/arch/arm/mach-pxa/include/mach/reset.h
> > index 9489a48..7b8842c 100644
> > --- a/arch/arm/mach-pxa/include/mach/reset.h
> > +++ b/arch/arm/mach-pxa/include/mach/reset.h
Somebody applied this.
> > @@ -10,9 +10,12 @@
> > extern unsigned int reset_status;
> > extern void clear_reset_status(unsigned int mask);
> >
> > -/*
> > - * register GPIO as reset generator
> > +/**
> > + * init_gpio_reset() - register GPIO as reset generator
> > + *
> > + * @gpio - gpio nr
> > + * @output - set gpio as out/low instead of input during normal work
>
> Use colon (':') after function parameters, not hyphen/dash.
Without doing that.
On Mon, Oct 27, 2008 at 11:42:46PM -0700, Andrew Morton wrote:
> On Mon, 27 Oct 2008 09:25:12 -0700 Randy Dunlap <[email protected]> wrote:
> > On Sun, 26 Oct 2008 19:41:09 +0100
> > Pavel Machek <[email protected]> wrote:
> > > Yes, I guess it should go to stable.
> > >
> > > commit 69fc7eed5f56bce15b239e5110de2575a6970df4
> > > Author: Dmitry Baryshkov <[email protected]>
> > > Date: Thu Oct 9 16:58:13 2008 +0100
> > >
> > > [ARM] 5300/1: fixup spitz reset during boot
>
> Somebody applied this.
Yes, I did, on October 11th.
> > > +/**
> > > + * init_gpio_reset() - register GPIO as reset generator
> > > + *
> > > + * @gpio - gpio nr
> > > + * @output - set gpio as out/low instead of input during normal work
> >
> > Use colon (':') after function parameters, not hyphen/dash.
>
> Without doing that.
Comments received after the patch has been committed and sent to
Linus can't be retrospectively corrected. Another patch is required.
Since it's a trivial documentation matter, don't we have someone
handling such patches? Is Rusty still the Trivial Patch Monkey?
--
Russell King
Linux kernel 2.6 ARM Linux - http://www.arm.linux.org.uk/
maintainer of:
> Since it's a trivial documentation matter, don't we have someone
> handling such patches? Is Rusty still the Trivial Patch Monkey?
Randy is the documentation maintainer (see MAINTAINERS)
Sounds like it should just go in directly not through the ARM tree
On Tue, 28 Oct 2008 09:24:49 +0000 Alan Cox wrote:
> > Since it's a trivial documentation matter, don't we have someone
> > handling such patches? Is Rusty still the Trivial Patch Monkey?
No, Jesper Juhl is (see MAINTAINTERS file).
> Randy is the documentation maintainer (see MAINTAINERS)
Yes, mostly for Documentation/...
> Sounds like it should just go in directly not through the ARM tree
whatever.
---
~Randy
Hi!
> > > > > Thanks for the effort, but it looks like it did not make it, right?
> > > >
> > > > Correct - Linus was unpredictable and released 2.6.27 that evening,
> > > > four days early giving me no time to push the fix to him.
> > > >
> > > > However, it's gone in as part of the first ARM merge, so should be in
> > > > -rc1.
> > >
> > > Is it in -stable? if not, is it considered as -stable material?
> >
> > Yes, I guess it should go to stable.
> >
> > commit 69fc7eed5f56bce15b239e5110de2575a6970df4
> > Author: Dmitry Baryshkov <[email protected]>
> > Date: Thu Oct 9 16:58:13 2008 +0100
> >
> > [ARM] 5300/1: fixup spitz reset during boot
> >
> > Some machines don't have the pullup/down on their reset
> > pin, so configuring the reset generating pin as input makes
> > them reset immediately. Fix that by making reset pin direction
> > configurable.
> >
> > Signed-off-by: Dmitry Baryshkov <[email protected]>
> > Signed-off-by: Russell King <[email protected]>
> > Signed-off-by: Pavel Machek <[email protected]>
> >
> > diff --git a/arch/arm/mach-pxa/include/mach/reset.h b/arch/arm/mach-pxa/include/mach/reset.h
> > index 9489a48..7b8842c 100644
> > --- a/arch/arm/mach-pxa/include/mach/reset.h
> > +++ b/arch/arm/mach-pxa/include/mach/reset.h
> > @@ -10,9 +10,12 @@
> > extern unsigned int reset_status;
> > extern void clear_reset_status(unsigned int mask);
> >
> > -/*
> > - * register GPIO as reset generator
> > +/**
> > + * init_gpio_reset() - register GPIO as reset generator
> > + *
> > + * @gpio - gpio nr
> > + * @output - set gpio as out/low instead of input during normal work
>
> Use colon (':') after function parameters, not hyphen/dash.
Okay, sorry about that. If I promise to fix that in mainline, will you
let me apply the -stable patch?
...I believe this one should go in. Documentation problem is not that
bad, and it fixes booting.
Alternatively, I can submit fixed patch, but it will be in trivial way
different from what is currently in mainline... which is okay for this
I guess.
Greg, will you take the patch as-is, or can you simply replace - with
: or do you want me to submit new one?
Pavel
--
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
On Tue, 28 Oct 2008 19:24:25 +0100 Pavel Machek wrote:
> Hi!
>
> > > > > > Thanks for the effort, but it looks like it did not make it, right?
> > > > >
> > > > > Correct - Linus was unpredictable and released 2.6.27 that evening,
> > > > > four days early giving me no time to push the fix to him.
> > > > >
> > > > > However, it's gone in as part of the first ARM merge, so should be in
> > > > > -rc1.
> > > >
> > > > Is it in -stable? if not, is it considered as -stable material?
> > >
> > > Yes, I guess it should go to stable.
> > >
> > > commit 69fc7eed5f56bce15b239e5110de2575a6970df4
> > > Author: Dmitry Baryshkov <[email protected]>
> > > Date: Thu Oct 9 16:58:13 2008 +0100
> > >
> > > [ARM] 5300/1: fixup spitz reset during boot
> > >
> > > Some machines don't have the pullup/down on their reset
> > > pin, so configuring the reset generating pin as input makes
> > > them reset immediately. Fix that by making reset pin direction
> > > configurable.
> > >
> > > Signed-off-by: Dmitry Baryshkov <[email protected]>
> > > Signed-off-by: Russell King <[email protected]>
> > > Signed-off-by: Pavel Machek <[email protected]>
> > >
> > > diff --git a/arch/arm/mach-pxa/include/mach/reset.h b/arch/arm/mach-pxa/include/mach/reset.h
> > > index 9489a48..7b8842c 100644
> > > --- a/arch/arm/mach-pxa/include/mach/reset.h
> > > +++ b/arch/arm/mach-pxa/include/mach/reset.h
> > > @@ -10,9 +10,12 @@
> > > extern unsigned int reset_status;
> > > extern void clear_reset_status(unsigned int mask);
> > >
> > > -/*
> > > - * register GPIO as reset generator
> > > +/**
> > > + * init_gpio_reset() - register GPIO as reset generator
> > > + *
> > > + * @gpio - gpio nr
> > > + * @output - set gpio as out/low instead of input during normal work
> >
> > Use colon (':') after function parameters, not hyphen/dash.
>
> Okay, sorry about that. If I promise to fix that in mainline, will you
> let me apply the -stable patch?
Sure, but it's really up to GregKH and cdub.
> ...I believe this one should go in. Documentation problem is not that
> bad, and it fixes booting.
>
> Alternatively, I can submit fixed patch, but it will be in trivial way
> different from what is currently in mainline... which is okay for this
> I guess.
>
> Greg, will you take the patch as-is, or can you simply replace - with
> : or do you want me to submit new one?
> Pavel
---
~Randy
From: Randy Dunlap <[email protected]>
Fix kernel-doc notation to use correct syntax.
Even though this should be moved to where the function is
actually implemented...
Signed-off-by: Randy Dunlap <[email protected]>
---
arch/arm/mach-pxa/include/mach/reset.h | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
--- linux-next-20081107.orig/arch/arm/mach-pxa/include/mach/reset.h
+++ linux-next-20081107/arch/arm/mach-pxa/include/mach/reset.h
@@ -12,9 +12,8 @@ extern void clear_reset_status(unsigned
/**
* init_gpio_reset() - register GPIO as reset generator
- *
- * @gpio - gpio nr
- * @output - set gpio as out/low instead of input during normal work
+ * @gpio: gpio nr
+ * @output: set gpio as out/low instead of input during normal work
*/
extern int init_gpio_reset(int gpio, int output);