[PATCH v2 1/1] ACPI / LPSS: enable hard LLP for DMA

[PATCH v2 1/1] ACPI / LPSS: enable hard LLP for DMA

[Andy Shevchenko]

Right now the DMA support of hard LLP (*) is fused. Enable it via specific
message sent to SoC at run time.

(*) Hard LLP stands for the multi-block transfer feature of DMA controller
supported by hardware.

Tested-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
---
In v2:
- massage commit message
- add Mika's tag
 drivers/acpi/acpi_lpss.c | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)

diff --git a/drivers/acpi/acpi_lpss.c b/drivers/acpi/acpi_lpss.c
index 373657f..8ea836c 100644
--- a/drivers/acpi/acpi_lpss.c
+++ b/drivers/acpi/acpi_lpss.c
@@ -718,13 +718,14 @@ static int acpi_lpss_resume_early(struct device *dev)
 #define LPSS_GPIODEF0_DMA1_D3		BIT(2)
 #define LPSS_GPIODEF0_DMA2_D3		BIT(3)
 #define LPSS_GPIODEF0_DMA_D3_MASK	GENMASK(3, 2)
+#define LPSS_GPIODEF0_DMA_LLP		BIT(13)
 
 static DEFINE_MUTEX(lpss_iosf_mutex);
 
 static void lpss_iosf_enter_d3_state(void)
 {
 	u32 value1 = 0;
-	u32 mask1 = LPSS_GPIODEF0_DMA_D3_MASK;
+	u32 mask1 = LPSS_GPIODEF0_DMA_D3_MASK | LPSS_GPIODEF0_DMA_LLP;
 	u32 value2 = LPSS_PMCSR_D3hot;
 	u32 mask2 = LPSS_PMCSR_Dx_MASK;
 	/*
@@ -768,8 +769,9 @@ static void lpss_iosf_enter_d3_state(void)
 
 static void lpss_iosf_exit_d3_state(void)
 {
-	u32 value1 = LPSS_GPIODEF0_DMA1_D3 | LPSS_GPIODEF0_DMA2_D3;
-	u32 mask1 = LPSS_GPIODEF0_DMA_D3_MASK;
+	u32 value1 = LPSS_GPIODEF0_DMA1_D3 | LPSS_GPIODEF0_DMA2_D3 |
+		     LPSS_GPIODEF0_DMA_LLP;
+	u32 mask1 = LPSS_GPIODEF0_DMA_D3_MASK | LPSS_GPIODEF0_DMA_LLP;
 	u32 value2 = LPSS_PMCSR_D0;
 	u32 mask2 = LPSS_PMCSR_Dx_MASK;
 
-- 
2.10.2

Re: [PATCH v2 1/1] ACPI / LPSS: enable hard LLP for DMA

[Rafael J. Wysocki]

On Thursday, November 17, 2016 04:30:06 PM Andy Shevchenko wrote:
> Right now the DMA support of hard LLP (*) is fused. Enable it via specific
> message sent to SoC at run time.
> 
> (*) Hard LLP stands for the multi-block transfer feature of DMA controller
> supported by hardware.
> 
> Tested-by: Mika Westerberg <mika.westerberg@linux.intel.com>
> Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>

Applied.

Thanks,
Rafael