| Message ID | 1406654463-15475-1-git-send-email-soren.brinkmann@xilinx.com |
|---|---|
| State | New |
| Headers | show |
On Tue, Jul 29, 2014 at 06:21:03PM +0100, Soren Brinkmann wrote: > Enable the SCU standby mode. > From documentation: > When set, SCU CLK is turned off when all processors are in WFI mode, > there is no pending request on the ACP (if implemented), and there is > no remaining activity in the SCU. > > When SCU CLK is off, ARREADYS, AWREADYS and WREADYS on the ACP are forced > LOW. The clock is turned on when any processor leaves WFI mode, or if > there is a new request on the ACP. > > Signed-off-by: Soren Brinkmann <soren.brinkmann@xilinx.com> > --- > Hi, > > I am working on some PM related patches for Zynq. One of them is > enabling the SCU standby mode. I was wondering whether this is probably > something that I should not do in platform specific code, but rather in > the common code for everybody? Take a look at Shawn's patches for this: http://lists.infradead.org/pipermail/linux-arm-kernel/2014-July/275326.html Will
On Wed, 2014-07-30 at 10:43AM +0100, Will Deacon wrote: > On Tue, Jul 29, 2014 at 06:21:03PM +0100, Soren Brinkmann wrote: > > Enable the SCU standby mode. > > From documentation: > > When set, SCU CLK is turned off when all processors are in WFI mode, > > there is no pending request on the ACP (if implemented), and there is > > no remaining activity in the SCU. > > > > When SCU CLK is off, ARREADYS, AWREADYS and WREADYS on the ACP are forced > > LOW. The clock is turned on when any processor leaves WFI mode, or if > > there is a new request on the ACP. > > > > Signed-off-by: Soren Brinkmann <soren.brinkmann@xilinx.com> > > --- > > Hi, > > > > I am working on some PM related patches for Zynq. One of them is > > enabling the SCU standby mode. I was wondering whether this is probably > > something that I should not do in platform specific code, but rather in > > the common code for everybody? > > Take a look at Shawn's patches for this: > > http://lists.infradead.org/pipermail/linux-arm-kernel/2014-July/275326.html Great, so that is already coming. Is that patch already applied somewhere? Thanks, Sören
diff --git a/arch/arm/kernel/smp_scu.c b/arch/arm/kernel/smp_scu.c index 1aafa0d785eb..82548a76740d 100644 --- a/arch/arm/kernel/smp_scu.c +++ b/arch/arm/kernel/smp_scu.c @@ -22,6 +22,9 @@ #define SCU_INVALIDATE 0x0c #define SCU_FPGA_REVISION 0x10 +#define SCU_CTRL_ENABLE BIT(0) +#define SCU_CTRL_STANDBY_ENABLE BIT(5) + #ifdef CONFIG_SMP /* * Get the number of CPU cores from the SCU configuration @@ -53,7 +56,7 @@ void scu_enable(void __iomem *scu_base) if (scu_ctrl & 1) return; - scu_ctrl |= 1; + scu_ctrl |= SCU_CTRL_ENABLE | SCU_CTRL_STANDBY_ENABLE; writel_relaxed(scu_ctrl, scu_base + SCU_CTRL); /*
Enable the SCU standby mode. From documentation: When set, SCU CLK is turned off when all processors are in WFI mode, there is no pending request on the ACP (if implemented), and there is no remaining activity in the SCU. When SCU CLK is off, ARREADYS, AWREADYS and WREADYS on the ACP are forced LOW. The clock is turned on when any processor leaves WFI mode, or if there is a new request on the ACP. Signed-off-by: Soren Brinkmann <soren.brinkmann@xilinx.com> --- Hi, I am working on some PM related patches for Zynq. One of them is enabling the SCU standby mode. I was wondering whether this is probably something that I should not do in platform specific code, but rather in the common code for everybody? Thanks, Sören --- arch/arm/kernel/smp_scu.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)