[PATCH v2] [POWERPC] mpic: deal with bogus NIRQ in Feature Reporting Register

[PATCH v2] [POWERPC] mpic: deal with bogus NIRQ in Feature Reporting Register

[Kumar Gala]

Some chips (like the SoCs from Freescale) report the wrong value in NIRQ
and this causes issues if its doesn't match or exceed the value of
irq_count.

Add a flag that board code can set to just use irq_count instead of
FRR[NIRQ].  Eventually we'll add a device tree property with the number
of sources.

Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
---

Screw up with git on the first version.

- k

 arch/powerpc/sysdev/mpic.c |    9 ++++++---
 include/asm-powerpc/mpic.h |    2 ++
 2 files changed, 8 insertions(+), 3 deletions(-)

diff --git a/arch/powerpc/sysdev/mpic.c b/arch/powerpc/sysdev/mpic.c
index 8619f2a..12e285d 100644
--- a/arch/powerpc/sysdev/mpic.c
+++ b/arch/powerpc/sysdev/mpic.c
@@ -1144,9 +1144,12 @@ struct mpic * __init mpic_alloc(struct device_node *node,
 	mpic->num_cpus = ((greg_feature & MPIC_GREG_FEATURE_LAST_CPU_MASK)
 			  >> MPIC_GREG_FEATURE_LAST_CPU_SHIFT) + 1;
 	if (isu_size == 0)
-		mpic->num_sources =
-			((greg_feature & MPIC_GREG_FEATURE_LAST_SRC_MASK)
-			 >> MPIC_GREG_FEATURE_LAST_SRC_SHIFT) + 1;
+		if (flags & MPIC_BROKEN_FRR_NIRQS)
+			mpic->num_sources = mpic->irq_count;
+		else
+			mpic->num_sources =
+				((greg_feature & MPIC_GREG_FEATURE_LAST_SRC_MASK)
+				 >> MPIC_GREG_FEATURE_LAST_SRC_SHIFT) + 1;

 	/* Map the per-CPU registers */
 	for (i = 0; i < mpic->num_cpus; i++) {
diff --git a/include/asm-powerpc/mpic.h b/include/asm-powerpc/mpic.h
index 943c5a3..6802570 100644
--- a/include/asm-powerpc/mpic.h
+++ b/include/asm-powerpc/mpic.h
@@ -353,6 +353,8 @@ struct mpic
 #define MPIC_ENABLE_MCK			0x00000200
 /* Disable bias among target selection, spread interrupts evenly */
 #define MPIC_NO_BIAS			0x00000400
+/* Ignore NIRQS as reported by FRR */
+#define MPIC_BROKEN_FRR_NIRQS		0x00000800

 /* MPIC HW modification ID */
 #define MPIC_REGSET_MASK		0xf0000000
-- 
1.5.4.5

Re: [PATCH v2] [POWERPC] mpic: deal with bogus NIRQ in Feature Reporting Register

[Benjamin Herrenschmidt]

On Wed, 2008-05-21 at 15:59 -0500, Kumar Gala wrote:
> Some chips (like the SoCs from Freescale) report the wrong value in NIRQ
> and this causes issues if its doesn't match or exceed the value of
> irq_count.
> 
> Add a flag that board code can set to just use irq_count instead of
> FRR[NIRQ].  Eventually we'll add a device tree property with the number
> of sources.
> 
> Signed-off-by: Kumar Gala <galak@kernel.crashing.org>

Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

> ---
> 
> Screw up with git on the first version.
> 
> - k
> 
>  arch/powerpc/sysdev/mpic.c |    9 ++++++---
>  include/asm-powerpc/mpic.h |    2 ++
>  2 files changed, 8 insertions(+), 3 deletions(-)
> 
> diff --git a/arch/powerpc/sysdev/mpic.c b/arch/powerpc/sysdev/mpic.c
> index 8619f2a..12e285d 100644
> --- a/arch/powerpc/sysdev/mpic.c
> +++ b/arch/powerpc/sysdev/mpic.c
> @@ -1144,9 +1144,12 @@ struct mpic * __init mpic_alloc(struct device_node *node,
>  	mpic->num_cpus = ((greg_feature & MPIC_GREG_FEATURE_LAST_CPU_MASK)
>  			  >> MPIC_GREG_FEATURE_LAST_CPU_SHIFT) + 1;
>  	if (isu_size == 0)
> -		mpic->num_sources =
> -			((greg_feature & MPIC_GREG_FEATURE_LAST_SRC_MASK)
> -			 >> MPIC_GREG_FEATURE_LAST_SRC_SHIFT) + 1;
> +		if (flags & MPIC_BROKEN_FRR_NIRQS)
> +			mpic->num_sources = mpic->irq_count;
> +		else
> +			mpic->num_sources =
> +				((greg_feature & MPIC_GREG_FEATURE_LAST_SRC_MASK)
> +				 >> MPIC_GREG_FEATURE_LAST_SRC_SHIFT) + 1;
> 
>  	/* Map the per-CPU registers */
>  	for (i = 0; i < mpic->num_cpus; i++) {
> diff --git a/include/asm-powerpc/mpic.h b/include/asm-powerpc/mpic.h
> index 943c5a3..6802570 100644
> --- a/include/asm-powerpc/mpic.h
> +++ b/include/asm-powerpc/mpic.h
> @@ -353,6 +353,8 @@ struct mpic
>  #define MPIC_ENABLE_MCK			0x00000200
>  /* Disable bias among target selection, spread interrupts evenly */
>  #define MPIC_NO_BIAS			0x00000400
> +/* Ignore NIRQS as reported by FRR */
> +#define MPIC_BROKEN_FRR_NIRQS		0x00000800
> 
>  /* MPIC HW modification ID */
>  #define MPIC_REGSET_MASK		0xf0000000

RE: [PATCH v2] [POWERPC] mpic: deal with bogus NIRQ in FeatureReporting Register

[Jin Zhengxiong]

Hi, Kumar,

So I should  change the MSI hwirq num according to this change, The
patches
will be send after this changes updated to your tree. How about applying
the MSI=20
driver firstly?

Jason

> -----Original Message-----
> From: linuxppc-dev-bounces+jinzhengxiong=3Dgmail.com@ozlabs.org=20
> [mailto:linuxppc-dev-bounces+jinzhengxiong=3Dgmail.com@ozlabs.or
> g] On Behalf Of Benjamin Herrenschmidt
> Sent: Thursday, May 22, 2008 5:02 AM
> To: Kumar Gala
> Cc: linuxppc-dev@ozlabs.org
> Subject: Re: [PATCH v2] [POWERPC] mpic: deal with bogus NIRQ=20
> in FeatureReporting Register
>=20
>=20
> On Wed, 2008-05-21 at 15:59 -0500, Kumar Gala wrote:
> > Some chips (like the SoCs from Freescale) report the wrong value in=20
> > NIRQ and this causes issues if its doesn't match or exceed=20
> the value=20
> > of irq_count.
> >=20
> > Add a flag that board code can set to just use irq_count instead of=20
> > FRR[NIRQ].  Eventually we'll add a device tree property with the=20
> > number of sources.
> >=20
> > Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
>=20
> Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
>=20

Re: [PATCH v2] [POWERPC] mpic: deal with bogus NIRQ in FeatureReporting Register

[Kumar Gala]

On May 22, 2008, at 9:15 PM, Jin Zhengxiong wrote:

> Hi, Kumar,
>
> So I should  change the MSI hwirq num according to this change, The
> patches
> will be send after this changes updated to your tree. How about  
> applying
> the MSI
> driver firstly?

I've asked Paul to pick up this patch for powerpc-next.  If you can  
rework the MSI patches with this patch applied and just change the  
board support for mpc85xx_ds and 8610 to set the flag.

- k

RE: [PATCH v2] [POWERPC] mpic: deal with bogus NIRQ in FeatureReporting Register

[Jin Zhengxiong]

=20
> On May 22, 2008, at 9:15 PM, Jin Zhengxiong wrote:
>=20
> > Hi, Kumar,
> >
> > So I should  change the MSI hwirq num according to this change, The=20
> > patches will be send after this changes updated to your tree. How=20
> > about applying the MSI driver firstly?
>=20
> I've asked Paul to pick up this patch for powerpc-next.  If=20
> you can rework the MSI patches with this patch applied and=20
> just change the board support for mpc85xx_ds and 8610 to set the flag.
>=20

Ok, Please find the patchset in the following mail. I'll resend the MSI
driver in this send.

Jason