diff mbox

ARM: dts: berlin2q.dtsi: set L2CC tag and data latency as 2 cycles

Message ID 1402565920-5636-1-git-send-email-jszhang@marvell.com (mailing list archive)
State New, archived
Headers show

Commit Message

Jisheng Zhang June 12, 2014, 9:38 a.m. UTC
For all BG2Q SoCs, 2 cycles is the best/correct value

Signed-off-by: Jisheng Zhang <jszhang@marvell.com>
---
 arch/arm/boot/dts/berlin2q.dtsi | 2 ++
 1 file changed, 2 insertions(+)

Comments

Russell King - ARM Linux June 12, 2014, 9:44 a.m. UTC | #1
On Thu, Jun 12, 2014 at 05:38:40PM +0800, Jisheng Zhang wrote:
> For all BG2Q SoCs, 2 cycles is the best/correct value

It would be a good idea to set all these parameters if you need to set
them at all - in other words, setting arm,dirty-latency as well, as
that's all part of the timing specification.
Jisheng Zhang June 12, 2014, 10:15 a.m. UTC | #2
Hi Russell,

On Thu, 12 Jun 2014 02:44:23 -0700
Russell King - ARM Linux <linux@arm.linux.org.uk> wrote:

> On Thu, Jun 12, 2014 at 05:38:40PM +0800, Jisheng Zhang wrote:
> > For all BG2Q SoCs, 2 cycles is the best/correct value
> 
> It would be a good idea to set all these parameters if you need to set
> them at all - in other words, setting arm,dirty-latency as well, as
> that's all part of the timing specification.
> 

Thanks for reviewing this patch. I will check with SoC people to find the correct
dirty-latency value.

Thanks,
Jisheng
Jisheng Zhang June 12, 2014, 10:19 a.m. UTC | #3
Hi Russell,

On Thu, 12 Jun 2014 03:15:03 -0700
Jisheng Zhang <jszhang@marvell.com> wrote:

> Hi Russell,
> 
> On Thu, 12 Jun 2014 02:44:23 -0700
> Russell King - ARM Linux <linux@arm.linux.org.uk> wrote:
> 
> > On Thu, Jun 12, 2014 at 05:38:40PM +0800, Jisheng Zhang wrote:
> > > For all BG2Q SoCs, 2 cycles is the best/correct value
> > 
> > It would be a good idea to set all these parameters if you need to set
> > them at all - in other words, setting arm,dirty-latency as well, as
> > that's all part of the timing specification.
> > 
> 
> Thanks for reviewing this patch. I will check with SoC people to find the
> correct dirty-latency value.

The BG2Q L2 cache controller is PL310, so no "dirty-latency"

Thanks,
Jisheng
Sebastian Hesselbarth June 16, 2014, 11:24 a.m. UTC | #4
On 06/12/2014 11:38 AM, Jisheng Zhang wrote:
> For all BG2Q SoCs, 2 cycles is the best/correct value
>
> Signed-off-by: Jisheng Zhang <jszhang@marvell.com>

Applied to berlin/dt with following fixed patch title:
   "ARM: dts: berlin2q: set L2CC tag and data latency to 2 cycles"

Thanks!

> ---
>   arch/arm/boot/dts/berlin2q.dtsi | 2 ++
>   1 file changed, 2 insertions(+)
>
> diff --git a/arch/arm/boot/dts/berlin2q.dtsi b/arch/arm/boot/dts/berlin2q.dtsi
> index 635a16a..3f95dc5 100644
> --- a/arch/arm/boot/dts/berlin2q.dtsi
> +++ b/arch/arm/boot/dts/berlin2q.dtsi
> @@ -90,6 +90,8 @@
>   			compatible = "arm,pl310-cache";
>   			reg = <0xac0000 0x1000>;
>   			cache-level = <2>;
> +			arm,data-latency = <2 2 2>;
> +			arm,tag-latency = <2 2 2>;
>   		};
>
>   		scu: snoop-control-unit@ad0000 {
>
diff mbox

Patch

diff --git a/arch/arm/boot/dts/berlin2q.dtsi b/arch/arm/boot/dts/berlin2q.dtsi
index 635a16a..3f95dc5 100644
--- a/arch/arm/boot/dts/berlin2q.dtsi
+++ b/arch/arm/boot/dts/berlin2q.dtsi
@@ -90,6 +90,8 @@ 
 			compatible = "arm,pl310-cache";
 			reg = <0xac0000 0x1000>;
 			cache-level = <2>;
+			arm,data-latency = <2 2 2>;
+			arm,tag-latency = <2 2 2>;
 		};
 
 		scu: snoop-control-unit@ad0000 {