AT91SAM9261EK kit is broken, commit bd60299 has broken it

AT91SAM9261EK kit is broken, commit bd60299 has broken it

[Remy Bohmer]

Hi Jean-Christophe,

Currently the Atmel AT91SAM9261-EK kit is broken within Linux 3.0,
3.01 and latest git.

These are the symptoms:
* Booting stops if LCD driver is enabled in the kernel. It stops
within the probe of the LCD driver. (at a msleep(10))
* schedule_timeout is broken.
* The jiffies based sched_clock does not run properly and stalls
during boot, before the LCD driver is being initialised. (You can see
this with printk-timestamps enabled)
* When LCD driver disabled, boards boots up to a prompt, but every
tool (e.g. 'top') that does a nanosleep syscall stays blocked in the
kernel.

After bisecting the tree searching for the offending commit, I came to
the following patch, after reverting it in mainline, everything works
again as normal. I do not see immediately what is wrong with this
commit, do you have an idea? Otherwise I will dig deeper in this
commit...

bd60299594fb1d28ae66563c9e76a0b89b0412cf is the first bad commit
commit bd60299594fb1d28ae66563c9e76a0b89b0412cf
Author: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
Date:   Wed Feb 2 07:27:07 2011 +0100

    at91: switch to CLKDEV_LOOKUP

    we do not change the clock naming convention so does not need to switch
    the AVR32 yet

    Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
    Acked-by: Nicolas Ferre <nicolas.ferre@atmel.com>
    Cc: Patrice Vilchez <patrice.vilchez@atmel.com>

Kind regards,

Remy

AT91SAM9261EK kit is broken, commit bd60299 has broken it

[Jean-Christophe PLAGNIOL-VILLARD]

Hi.

	I'm taking a look on it

Best Regads,
J.
On 23:27 Sat 06 Aug     , Remy Bohmer wrote:
> Hi Jean-Christophe,
> 
> Currently the Atmel AT91SAM9261-EK kit is broken within Linux 3.0,
> 3.01 and latest git.
> 
> These are the symptoms:
> * Booting stops if LCD driver is enabled in the kernel. It stops
> within the probe of the LCD driver. (at a msleep(10))
> * schedule_timeout is broken.
> * The jiffies based sched_clock does not run properly and stalls
> during boot, before the LCD driver is being initialised. (You can see
> this with printk-timestamps enabled)
> * When LCD driver disabled, boards boots up to a prompt, but every
> tool (e.g. 'top') that does a nanosleep syscall stays blocked in the
> kernel.
> 
> After bisecting the tree searching for the offending commit, I came to
> the following patch, after reverting it in mainline, everything works
> again as normal. I do not see immediately what is wrong with this
> commit, do you have an idea? Otherwise I will dig deeper in this
> commit...
> 
> bd60299594fb1d28ae66563c9e76a0b89b0412cf is the first bad commit
> commit bd60299594fb1d28ae66563c9e76a0b89b0412cf
> Author: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
> Date:   Wed Feb 2 07:27:07 2011 +0100
> 
>     at91: switch to CLKDEV_LOOKUP
> 
>     we do not change the clock naming convention so does not need to switch
>     the AVR32 yet
> 
>     Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
>     Acked-by: Nicolas Ferre <nicolas.ferre@atmel.com>
>     Cc: Patrice Vilchez <patrice.vilchez@atmel.com>
> 
> Kind regards,
> 
> Remy