[PATCH V2] arm: pj4: check cpu id for pj4 cp0 access

[PATCH V2] arm: pj4: check cpu id for pj4 cp0 access

[Chao Xie]

Check cpu id in pj4_cp0_init. So for no-PJ4 V7 cpus,
pj4_cpu0_init just return.
This fix will help to make the all the V7 cpus(pj4 and no-PJ4)
use same platform code.

Signed-off-by: Chao Xie <chao.xie@marvell.com>
---
 arch/arm/kernel/pj4-cp0.c |   11 +++++++++++
 1 files changed, 11 insertions(+), 0 deletions(-)

diff --git a/arch/arm/kernel/pj4-cp0.c b/arch/arm/kernel/pj4-cp0.c
index 679cf4d..bf658d3 100644
--- a/arch/arm/kernel/pj4-cp0.c
+++ b/arch/arm/kernel/pj4-cp0.c
@@ -17,6 +17,10 @@
 #include <linux/init.h>
 #include <linux/io.h>
 #include <asm/thread_notify.h>
+#include <asm/cputype.h>
+
+#define CORE_PJ4_MASK	0xfffffff0
+#define CORE_PJ4_ID	0x562f5840
 
 static int iwmmxt_do(struct notifier_block *self, unsigned long cmd, void *t)
 {
@@ -79,6 +83,13 @@ static void __init pj4_cp_access_write(u32 value)
 static int __init pj4_cp0_init(void)
 {
 	u32 cp_access;
+	unsigned int cpuid;
+
+	cpuid = read_cpuid_id();
+
+	/* If it is not a pj4 core, skip it. */
+	if ((cpuid & CORE_PJ4_MASK) != CORE_PJ4_ID)
+		return 0;
 
 	cp_access = pj4_cp_access_read() & ~0xf;
 	pj4_cp_access_write(cp_access);
-- 
1.7.4.1

[PATCH V2] arm: pj4: check cpu id for pj4 cp0 access

[Arnd Bergmann]

On Friday 14 June 2013, Chao Xie wrote:
> Check cpu id in pj4_cp0_init. So for no-PJ4 V7 cpus,
> pj4_cpu0_init just return.
> This fix will help to make the all the V7 cpus(pj4 and no-PJ4)
> use same platform code.
> 
> Signed-off-by: Chao Xie <chao.xie@marvell.com>

Acked-by: Arnd Bergmann <arnd@arndb.de>

[V2] arm: pj4: check cpu id for pj4 cp0 access

[Matt Porter]

On Fri, Jun 14, 2013 at 05:03:57AM -0400, Chao Xie wrote:
> Check cpu id in pj4_cp0_init. So for no-PJ4 V7 cpus,
> pj4_cpu0_init just return.
> This fix will help to make the all the V7 cpus(pj4 and no-PJ4)
> use same platform code.
> 
> Signed-off-by: Chao Xie <chao.xie@marvell.com>
> Acked-by: Arnd Bergmann <arnd@arndb.de>

This patch fixes a boot failure in next now that ARCH_DOVE (and thus
CPU_PJ4) are enabled in multi_v7_defconfig. I verified that boot works
again in next with the BCM28155-AP board.

Tested-by: Matt Porter <mporter@linaro.org>

> arch/arm/kernel/pj4-cp0.c |   11 +++++++++++
>  1 files changed, 11 insertions(+), 0 deletions(-)
> 
> diff --git a/arch/arm/kernel/pj4-cp0.c b/arch/arm/kernel/pj4-cp0.c
> index 679cf4d..bf658d3 100644
> --- a/arch/arm/kernel/pj4-cp0.c
> +++ b/arch/arm/kernel/pj4-cp0.c
> @@ -17,6 +17,10 @@
>  #include <linux/init.h>
>  #include <linux/io.h>
>  #include <asm/thread_notify.h>
> +#include <asm/cputype.h>
> +
> +#define CORE_PJ4_MASK	0xfffffff0
> +#define CORE_PJ4_ID	0x562f5840
>  
>  static int iwmmxt_do(struct notifier_block *self, unsigned long cmd, void *t)
>  {
> @@ -79,6 +83,13 @@ static void __init pj4_cp_access_write(u32 value)
>  static int __init pj4_cp0_init(void)
>  {
>  	u32 cp_access;
> +	unsigned int cpuid;
> +
> +	cpuid = read_cpuid_id();
> +
> +	/* If it is not a pj4 core, skip it. */
> +	if ((cpuid & CORE_PJ4_MASK) != CORE_PJ4_ID)
> +		return 0;
>  
>  	cp_access = pj4_cp_access_read() & ~0xf;
>  	pj4_cp_access_write(cp_access);

[V2] arm: pj4: check cpu id for pj4 cp0 access

[Kevin Hilman]

On Tue, Mar 18, 2014 at 1:15 PM, Matt Porter <mporter@linaro.org> wrote:
> On Fri, Jun 14, 2013 at 05:03:57AM -0400, Chao Xie wrote:
>> Check cpu id in pj4_cp0_init. So for no-PJ4 V7 cpus,
>> pj4_cpu0_init just return.
>> This fix will help to make the all the V7 cpus(pj4 and no-PJ4)
>> use same platform code.
>>
>> Signed-off-by: Chao Xie <chao.xie@marvell.com>
>> Acked-by: Arnd Bergmann <arnd@arndb.de>
>
> This patch fixes a boot failure in next now that ARCH_DOVE (and thus
> CPU_PJ4) are enabled in multi_v7_defconfig. I verified that boot works
> again in next with the BCM28155-AP board.
>
> Tested-by: Matt Porter <mporter@linaro.org>

I've also verified that this fixes the boot regression on the BCM28155-AP board.

Tested-by: Kevin Hilman <khilman@linaro.org>

Chao, can you collect the acks/tested-bys and submit this to Russell's
patch tracker[1] so he can apply it please?.  Thanks.

Kevin

[1] http://www.arm.linux.org.uk/developer/patches/

[V2] arm: pj4: check cpu id for pj4 cp0 access

[Russell King - ARM Linux]

On Tue, Mar 18, 2014 at 01:19:35PM -0700, Kevin Hilman wrote:
> On Tue, Mar 18, 2014 at 1:15 PM, Matt Porter <mporter@linaro.org> wrote:
> > On Fri, Jun 14, 2013 at 05:03:57AM -0400, Chao Xie wrote:
> >> Check cpu id in pj4_cp0_init. So for no-PJ4 V7 cpus,
> >> pj4_cpu0_init just return.
> >> This fix will help to make the all the V7 cpus(pj4 and no-PJ4)
> >> use same platform code.
> >>
> >> Signed-off-by: Chao Xie <chao.xie@marvell.com>
> >> Acked-by: Arnd Bergmann <arnd@arndb.de>
> >
> > This patch fixes a boot failure in next now that ARCH_DOVE (and thus
> > CPU_PJ4) are enabled in multi_v7_defconfig. I verified that boot works
> > again in next with the BCM28155-AP board.
> >
> > Tested-by: Matt Porter <mporter@linaro.org>
> 
> I've also verified that this fixes the boot regression on the BCM28155-AP board.
> 
> Tested-by: Kevin Hilman <khilman@linaro.org>
> 
> Chao, can you collect the acks/tested-bys and submit this to Russell's
> patch tracker[1] so he can apply it please?.  Thanks.

Any reason not put the IDs in asm/cputype.h, where all the others are -
and maybe also a cpu_is_pj4() inline function too?

-- 
FTTC broadband for 0.8mile line: now at 9.7Mbps down 460kbps up... slowly
improving, and getting towards what was expected from it.

[V2] arm: pj4: check cpu id for pj4 cp0 access

[Chao Xie]

On Wed, Mar 19, 2014 at 4:32 AM, Russell King - ARM Linux
<linux@arm.linux.org.uk> wrote:
> On Tue, Mar 18, 2014 at 01:19:35PM -0700, Kevin Hilman wrote:
>> On Tue, Mar 18, 2014 at 1:15 PM, Matt Porter <mporter@linaro.org> wrote:
>> > On Fri, Jun 14, 2013 at 05:03:57AM -0400, Chao Xie wrote:
>> >> Check cpu id in pj4_cp0_init. So for no-PJ4 V7 cpus,
>> >> pj4_cpu0_init just return.
>> >> This fix will help to make the all the V7 cpus(pj4 and no-PJ4)
>> >> use same platform code.
>> >>
>> >> Signed-off-by: Chao Xie <chao.xie@marvell.com>
>> >> Acked-by: Arnd Bergmann <arnd@arndb.de>
>> >
>> > This patch fixes a boot failure in next now that ARCH_DOVE (and thus
>> > CPU_PJ4) are enabled in multi_v7_defconfig. I verified that boot works
>> > again in next with the BCM28155-AP board.
>> >
>> > Tested-by: Matt Porter <mporter@linaro.org>
>>
>> I've also verified that this fixes the boot regression on the BCM28155-AP board.
>>
>> Tested-by: Kevin Hilman <khilman@linaro.org>
>>
>> Chao, can you collect the acks/tested-bys and submit this to Russell's
>> patch tracker[1] so he can apply it please?.  Thanks.
>
> Any reason not put the IDs in asm/cputype.h, where all the others are -
> and maybe also a cpu_is_pj4() inline function too?
>

Sure, i can put it at arch/arm/asm/include/cputype.h

I will send the patch later.

Thanks.

> --
> FTTC broadband for 0.8mile line: now at 9.7Mbps down 460kbps up... slowly
> improving, and getting towards what was expected from it.

[V2] arm: pj4: check cpu id for pj4 cp0 access

[Stephen Warren]

On 03/18/2014 02:15 PM, Matt Porter wrote:
> On Fri, Jun 14, 2013 at 05:03:57AM -0400, Chao Xie wrote:
>> Check cpu id in pj4_cp0_init. So for no-PJ4 V7 cpus,
>> pj4_cpu0_init just return.
>> This fix will help to make the all the V7 cpus(pj4 and no-PJ4)
>> use same platform code.
>>
>> Signed-off-by: Chao Xie <chao.xie@marvell.com>
>> Acked-by: Arnd Bergmann <arnd@arndb.de>
> 
> This patch fixes a boot failure in next now that ARCH_DOVE (and thus
> CPU_PJ4) are enabled in multi_v7_defconfig. I verified that boot works
> again in next with the BCM28155-AP board.
> 
> Tested-by: Matt Porter <mporter@linaro.org>

This also solves the Tegra SoC boot failures with next-20130320.
Tested-by: Stephen Warren <swarren@nvidia.com>