Solve the pd could only ever turn off but never turn them on again,
if the pd registers have the writemask bits.
So far this affects the rk3328 only.
Fixes: 79bb17ce8e ("soc: rockchip: power-domain: Support domain control in hiword-registers")
Cc: stable@vger.kernel.org
Signed-off-by: Finley Xiao <finley.xiao@rock-chips.com>
Signed-off-by: Elaine Zhang <zhangqing@rock-chips.com>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
The of_count_phandle_with_args() can fail and return error(for example,
rk3399 pd_vio doesn't have clocks). That would break the pd probe.
Add a sanity check on pd->num_clks to avoid that.
Fixes: 65084121d59d ("soc: rockchip: power-domain: use clk_bulk APIs")
Reported-by: Shawn Lin <shawn.lin@rock-chips.com>
Signed-off-by: Jeffy Chen <jeffy.chen@rock-chips.com>
Tested-by: Shawn Lin <shawn.lin@rock-chips.com>
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Use clk_bulk APIs, and also add error handling for clk enable.
Signed-off-by: Jeffy Chen <jeffy.chen@rock-chips.com>
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Set the newly introduced GENPD_FLAG_ACTIVE_WAKEUP, which allows to
remove the driver's own flag-based callback.
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Acked-by: Ulf Hansson <ulf.hansson@linaro.org>
Acked-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The rk3328 uses the newly introduced support for power-domain control
in hiword-mask registers.
Signed-off-by: Elaine Zhang <zhangqing@rock-chips.com>
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
New Rockchips SoCs may have their power-domain control in registers
using a writemask-based access scheme (upper 16bit being the write
mask). So add a DOMAIN_M type and handle this case accordingly.
Signed-off-by: Elaine Zhang <zhangqing@rock-chips.com>
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
It was a bit surprising that the device was reported to have probed just
fine, but the provider hadn't been registered.
So handle any errors when registering the provider and fail the probe
accordingly.
Signed-off-by: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Cc: Caesar Wang <wxt@rock-chips.com>
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
The newly introduced pm_genpd_remove reverts the initialization done
by pm_genpd_init and is necessary in the error path of the rockchip
power-domain driver.
Without it the driver will in the error case cleanup the devm-allocated
structures including the elements referenced in the gpd_list thus making
deactivation of unused domains (and probably later genpd accesses as
well) fail by accessing invalid pointers.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
In some cases, we have met the infinite loop in
rockchip_pmu_set_idle_request() or rockchip_do_pmu_set_power_domain().
As the crosbug.com/p/57351 reported, the boot hangs right after this
[1.629163] bootconsole [uart8250] disabled
[1.639286] [drm:drm_core_init] Initialized drm 1.1.0 20060810
[1.645926] [drm:drm_get_platform_dev] Initialized vgem 1.0.0 20120112..
[1.654558] iommu: Adding device ff8f0000.vop to group 0
[1.660569] iommu: Adding device ff900000.vop to group 1
<hang>
This patch adds the error message and timeout to avoid infinite loop if
it fails to get the ack.
Signed-off-by: Caesar Wang <wxt@rock-chips.com>
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
On rk3288 it was important that powerdown and powerup counts for the
CPU/GPU in the kernel because:
* The power on default was crazy long.
* We couldn't rely on the firmware to set this up because really this
wasn't the firmware's job--the kernel was the only one that really
cared about bringing up / down CPUs and the GPU and doing suspend /
resume (which involves bringing up / down CPUs).
On newer ARM systems (like rk3399) ARM Trusted Firmware is in charge of
bringing up and down the CPUs and it really should be in charge of
setting all these counts right. After all ATF is in charge of suspend /
resume and CPU up / down. Let's get out of the way and let ATF do its
job.
A few other motivations for doing this:
* Depending on another configuration (PMU_24M_EN_CFG) these counts can
be either in 24M or 32k cycles. Thus, though ATF isn't really so
involved in bringing up the GPU, ATF should probably manage the counts
for everything so it can also manage the 24M / 32k choice.
* It turns out that (right now) 24M mode is broken on rk3399 and not
being used. That means that the count the kernel was programming
in (24) was not 1 us (which it seems was intended) but was actually
.75 ms
* On rk3399 there are actually 2 separate registers for setting CPU
up/down time plus 1 register for GPU up/down time. The curent kernel
code actually was putting the register for the "little" cores in the
"CPU" slot and the register for the "big" cores in the "GPU" slot. It
was never initting the GPU counts.
Note: this change assumes that ATF will actually set these values at
boot, as I'm proposing in <http://crosreview.com/372381>.
Signed-off-by: Douglas Anderson <dianders@chromium.org>
[ATF change has landed]
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Register gpd_dev_ops.active_wakeup function to support keep power
during suspend state. And add flag to each power domain to
decide whether keep power during suspend or not.
Signed-off-by: Elaine Zhang <zhangqing@rock-chips.com>
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
support qos save and restore when power domain on/off.
Signed-off-by: Elaine Zhang <zhangqing@rock-chips.com>
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Check return value of syscon_node_to_regmap for
rockchip_pm_domain_probe. If err value is returned, probe
procedure should abort.
Signed-off-by: Shawn Lin <shawn.lin@rock-chips.com>
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
This driver is modified to support RK3399 SoC.
Signed-off-by: Elaine Zhang <zhangqing@rock-chips.com>
[small indentation fixups]
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
This patch adds support for making one power domain a sub-domain of
other domain. This is useful for modeling power dependences,
which needs to have more than one power domain enabled to be operational.
Signed-off-by: Elaine Zhang <zhangqing@rock-chips.com>
[restructured error handling in subdomain-addition]
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
On some Rockchip SoC there exist child-domains only handling their
idle state with the actual power-state handled by a (shared) parent-
domain.
So allow such types of domains. For them, we can determine their
state (on/off) by checking the inverse idle-state instead.
There exist one special case if both idle as well power handling
were set as not present, but as the domain-data is defined in the
code itself, we can expect the reasonable developer to define them
in a correct way, without adding more checks.
Signed-off-by: Elaine Zhang <zhangqing@rock-chips.com>
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Not all new socs need to handle idle states on domain state changes,
so add the possibility to make them optional.
Signed-off-by: Elaine Zhang <zhangqing@rock-chips.com>
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
If we fail to probe the driver, we should not directly break
from the for_each_available_child_of_node since it calls of_node_get
while iterating. This patch add of_node_put to fix the unbalanced
call pair.
Fixes: 7c696693a4 ("soc: rockchip: power-domain: Add power domain driver")
Signed-off-by: Shawn Lin <shawn.lin@rock-chips.com>
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
This driver is modified to support RK3368 SoC.
Signed-off-by: zhangqing <zhangqing@rock-chips.com>
Reviewed-by: Caesar Wang <wxt@rock-chips.com>
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
When we never got the the clock-reference, i.e. when IS_ERR(clk) is true,
don't try to print the clock name via %pC as this of course produces a
null-pointer-dereference in __clk_get_name().
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Reviewed-by: Caesar Wang <wxt@rock-chips.com>
This driver is found on RK3288 SoCs.
In order to meet high performance and low power requirements, a power
management unit is designed or saving power when RK3288 in low power
mode.
The RK3288 PMU is dedicated for managing the power of the whole chip.
PMU can work in the Low Power Mode by setting bit[0] of PMU_PWRMODE_CON
register. After setting the register, PMU would enter the Low Power mode.
In the low power mode, pmu will auto power on/off the specified power
domain, send idle req to specified power domain, shut down/up pll and
so on. All of above are configurable by setting corresponding registers.
Signed-off-by: Caesar Wang <wxt@rock-chips.com>
Reviewed-by: Kevin Hilman <khilman@linaro.org>
[replace dsb() with dsb(sy) for arm64 buildability; sy is the default,
so no functional change; adapt to per-user clocks in genpd]
Signed-off-by: Heiko Stuebner <heiko@sntech.de>