[PATCH 2/3] pmdomain: amlogic: Add support for A9 power domains controller
Neil Armstrong
neil.armstrong at linaro.org
Tue Jul 28 01:34:47 PDT 2026
On 7/22/26 10:39, Xianwei Zhao via B4 Relay wrote:
> From: Xianwei Zhao <xianwei.zhao at amlogic.com>
>
> Add support for the A9 power controller, whose registers are
> in the secure domain and should be accessed via SMC.
>
> Signed-off-by: Xianwei Zhao <xianwei.zhao at amlogic.com>
> ---
> drivers/pmdomain/amlogic/meson-secure-pwrc.c | 69 ++++++++++++++++++++++++++++
> 1 file changed, 69 insertions(+)
>
> diff --git a/drivers/pmdomain/amlogic/meson-secure-pwrc.c b/drivers/pmdomain/amlogic/meson-secure-pwrc.c
> index 1d2f371d2d7f..133549956efa 100644
> --- a/drivers/pmdomain/amlogic/meson-secure-pwrc.c
> +++ b/drivers/pmdomain/amlogic/meson-secure-pwrc.c
> @@ -16,6 +16,7 @@
> #include <dt-bindings/power/amlogic,t7-pwrc.h>
> #include <dt-bindings/power/amlogic,a4-pwrc.h>
> #include <dt-bindings/power/amlogic,a5-pwrc.h>
> +#include <dt-bindings/power/amlogic,a9-pwrc.h>
> #include <dt-bindings/power/amlogic,s6-pwrc.h>
> #include <dt-bindings/power/amlogic,s7-pwrc.h>
> #include <dt-bindings/power/amlogic,s7d-pwrc.h>
> @@ -175,6 +176,65 @@ static const struct meson_secure_pwrc_domain_desc a5_pwrc_domains[] = {
> SEC_PD(A5_DSPA, 0),
> };
>
> +static const struct meson_secure_pwrc_domain_desc a9_pwrc_domains[] = {
> + SEC_PD(A9_DSPA, 0),
> + SEC_PD(A9_U3HSG_U3, 0),
> + SEC_PD(A9_DP, 0),
> + SEC_PD(A9_DOS_HCODEC, 0),
> + SEC_PD(A9_TAHOE, 0),
> + SEC_PD(A9_DOS_HEVC, 0),
> + SEC_PD(A9_U2H, 0),
> + SEC_PD(A9_U3DRD_B, 0),
> + SEC_PD(A9_VPU_HDMI, 0),
> + SEC_PD(A9_U2DRD, 0),
> + SEC_PD(A9_U3DRD_A, 0),
> + SEC_PD(A9_SD_EMMC_C, 0),
> + SEC_PD(A9_GE2D, 0),
> + SEC_PD(A9_AMFC, 0),
> + SEC_PD(A9_EDPTX, 0),
> + SEC_PD(A9_OPP, 0),
> + SEC_PD(A9_VICP, 0),
> + SEC_PD(A9_SD_EMMC_A, 0),
> + SEC_PD(A9_SD_EMMC_B, 0),
> + SEC_PD(A9_ETH, 0),
> + SEC_PD(A9_PCIE_A, 0),
> + SEC_PD(A9_PCIE_B, 0),
> + SEC_PD(A9_NNA_4T, 0),
> + SEC_PD(A9_HDMIRX, 0),
> + SEC_PD(A9_CVE, 0),
> + SEC_PD(A9_ISP, 0),
> + SEC_PD(A9_ETH_1G, 0),
> + SEC_PD(A9_U3HSG_HSG, 0),
> + SEC_PD(A9_U3DPPHY_U3, 0),
> + SEC_PD(A9_U3DPPHY_DP, 0),
> + SEC_PD(A9_PCIE3PHY, 0),
> + SEC_PD(A9_U3HSG_PCIE2, 0),
> + SEC_PD(A9_MALI_TOP, 0),
> + SEC_PD(A9_AO_SED, 0),
> + SEC_PD(A9_AO_IR, 0),
> + /* UART_B is for console log, and should be always on */
> + SEC_PD(A9_AO_UART_B, GENPD_FLAG_ALWAYS_ON),
> + SEC_PD(A9_AO_UART_C, 0),
> + SEC_PD(A9_AO_UART_D, 0),
> + SEC_PD(A9_AO_SPISG, 0),
> + SEC_PD(A9_AO_UART_E, 0),
> + SEC_PD(A9_AO_CEC, 0),
> + /* SRAMA is used as ATF runtime memory, and should be always on */
> + SEC_PD(A9_EE_SRAMA, GENPD_FLAG_ALWAYS_ON),
> + SEC_PD(A9_AUDIO, 0),
> + /* DMC0 is for DDR PHY ana/dig and DMC, and should be always on */
> + SEC_PD(A9_DMC0, GENPD_FLAG_ALWAYS_ON),
> + /*GIC needs to be used consistently, and should be always on */
> + SEC_PD(A9_GIC, GENPD_FLAG_ALWAYS_ON),
> + /* DDRPHY is for DDR Work, and should be always on */
> + SEC_PD(A9_DDRPHY, GENPD_FLAG_ALWAYS_ON),
> + SEC_PD(A9_AUCPU, 0),
> + SEC_PD(A9_DSI0, 0),
> + SEC_PD(A9_DSI1, 0),
> + SEC_PD(A9_CAN0, 0),
> + SEC_PD(A9_CAN1, 0),
> +};
> +
> static const struct meson_secure_pwrc_domain_desc c3_pwrc_domains[] = {
> SEC_PD(C3_NNA, 0),
> SEC_PD(C3_AUDIO, 0),
> @@ -425,6 +485,11 @@ static const struct meson_secure_pwrc_domain_data amlogic_secure_a5_pwrc_data =
> .count = ARRAY_SIZE(a5_pwrc_domains),
> };
>
> +static const struct meson_secure_pwrc_domain_data amlogic_secure_a9_pwrc_data = {
> + .domains = a9_pwrc_domains,
> + .count = ARRAY_SIZE(a9_pwrc_domains),
> +};
> +
> static const struct meson_secure_pwrc_domain_data amlogic_secure_c3_pwrc_data = {
> .domains = c3_pwrc_domains,
> .count = ARRAY_SIZE(c3_pwrc_domains),
> @@ -468,6 +533,10 @@ static const struct of_device_id meson_secure_pwrc_match_table[] = {
> .compatible = "amlogic,a5-pwrc",
> .data = &amlogic_secure_a5_pwrc_data,
> },
> + {
> + .compatible = "amlogic,a9-pwrc",
> + .data = &amlogic_secure_a9_pwrc_data,
> + },
> {
> .compatible = "amlogic,c3-pwrc",
> .data = &amlogic_secure_c3_pwrc_data,
>
Reviewed-by: Neil Armstrong <neil.armstrong at linaro.org>
Thanks,
Neil
More information about the linux-amlogic
mailing list