Re: [PATCH V2 3/4] soc: c3: Add support for power domains controller

Xianwei Zhao <xianwei.zhao@xxxxxxxxxxx> · Thu, 27 Jul 2023 10:24:03 +0800

Hi Kevin,
     Thanks for your reply.

On 2023/7/27 06:40, Kevin Hilman wrote:
[ EXTERNAL EMAIL ]

Xianwei Zhao <xianwei.zhao@xxxxxxxxxxx> writes:

Add support for C3 Power controller. C3 power control
registers are in secure domain, and should be accessed by SMC.

Signed-off-by: Xianwei Zhao <xianwei.zhao@xxxxxxxxxxx>
---
V1 -> V2: Fixed some formatting.
---
  drivers/soc/amlogic/meson-secure-pwrc.c | 26 +++++++++++++++++++++++++
  1 file changed, 26 insertions(+)

diff --git a/drivers/soc/amlogic/meson-secure-pwrc.c b/drivers/soc/amlogic/meson-secure-pwrc.c
index c11d65a3e3d9..a1ffebf70de3 100644
--- a/drivers/soc/amlogic/meson-secure-pwrc.c
+++ b/drivers/soc/amlogic/meson-secure-pwrc.c
@@ -11,6 +11,7 @@
  #include <linux/platform_device.h>
  #include <linux/pm_domain.h>
  #include <dt-bindings/power/meson-a1-power.h>
+#include <dt-bindings/power/amlogic,c3-pwrc.h>
  #include <dt-bindings/power/meson-s4-power.h>
  #include <linux/arm-smccc.h>
  #include <linux/firmware/meson/meson_sm.h>
@@ -120,6 +121,22 @@ static struct meson_secure_pwrc_domain_desc a1_pwrc_domains[] = {
       SEC_PD(RSA,     0),
  };

+static struct meson_secure_pwrc_domain_desc c3_pwrc_domains[] = {
+     SEC_PD(C3_NNA,  0),
+     SEC_PD(C3_AUDIO,        GENPD_FLAG_ALWAYS_ON),
+     SEC_PD(C3_SDIOA,        GENPD_FLAG_ALWAYS_ON),
+     SEC_PD(C3_EMMC, GENPD_FLAG_ALWAYS_ON),
+     SEC_PD(C3_USB_COMB, GENPD_FLAG_ALWAYS_ON),
+     SEC_PD(C3_SDCARD,       GENPD_FLAG_ALWAYS_ON),
+     SEC_PD(C3_ETH,  GENPD_FLAG_ALWAYS_ON),
+     SEC_PD(C3_GE2D, GENPD_FLAG_ALWAYS_ON),
+     SEC_PD(C3_CVE,  GENPD_FLAG_ALWAYS_ON),
+     SEC_PD(C3_GDC_WRAP,     GENPD_FLAG_ALWAYS_ON),
+     SEC_PD(C3_ISP_TOP,              GENPD_FLAG_ALWAYS_ON),
+     SEC_PD(C3_MIPI_ISP_WRAP, GENPD_FLAG_ALWAYS_ON),
+     SEC_PD(C3_VCODEC,       0),
+};

All of these domains being hard-coded to ALWAYS_ON looks suspicious, and
can also be an indicator that the drivers for these domains are not
(properly) using runtime PM, or not connected to the correct domains the DT.

Similar to the tables for s4 and a1 in this same file, please describe
the reason that each of these domains needs to be hard coded to be
always on.
Will do.

Thanks,

Kevin






