In the bandwidth tension environment when close win2, vop will access
the freed memory lead to iommu pagefault. so we add this reset to workaround.
Change-Id: I22b0c0f145d042e3aaf98fb45ffff6304c93963c
Signed-off-by: Sandy Huang <hjc@rock-chips.com>
Because when enable uboot logo display, vop_crtc_enable() will not be
called when power on, this will lead to some vop initial like
axi channel and some debug irq will not be enabled. so we move some
config to vop_initial() and call from vop_crtc_loader_protect().
Change-Id: I86f02e2e7d12b78cce17e278baaf6dff93137167
Signed-off-by: Sandy Huang <hjc@rock-chips.com>
some version vop unsupport pixel alpha add scale, this case
will lead to display error and post empty.
Change-Id: I32820a14292b46ce61fd30bfccdaa5e4f635de49
Signed-off-by: Sandy Huang <hjc@rock-chips.com>
The most pins of rk3308 are 2bits iomux, but the banks's register
width is 0x8.
Change-Id: I3305810b3f75febd6ec7a933b65e3c9d50f003dd
Signed-off-by: David Wu <david.wu@rock-chips.com>
This adds the necessary data for handling io voltage domains on the rk3308.
As interesting tidbit, the rk3308 contains one iodomain area at grf,
Change-Id: Ife72a284a8926d02ef5df7a422d41924494d0300
Signed-off-by: David Wu <david.wu@rock-chips.com>
Otherwise, clk_gpu won't be disabled actually in the runtime.
Change-Id: I92787a5e23bfb92f5a79efda92c130832751cc3b
Signed-off-by: Zhen Chen <chenzhen@rock-chips.com>
The commit 9d9491a7da ("mmc: dw_mmc: Fix the DTO timeout calculation")
and commit 4c2357f57d ("mmc: dw_mmc: Fix the CTO timeout calculation")
made changes, which cause multiply overflow for 32-bit systems. The
broken
timeout calculations leads to unexpected ETIMEDOUT errors and causes
stacktrace splat (such as below) during normal data exchange with
SD-card.
| Running : 4M-check-reassembly-tcp-cmykw2-rotatew2.out -v0 -w1
| - Info: Finished target initialization.
| mmcblk0: error -110 transferring data, sector 320544, nr 2048, cmd
| response 0x900, card status 0x0
DIV_ROUND_UP_ULL helps to escape usage of __udivdi3() from libgcc and so
code gets compiled on all 32-bit platforms as opposed to usage of
DIV_ROUND_UP when we may only compile stuff on a very few arches.
Lets cast this multiply to u64 type to prevent the overflow.
Change-Id: I45462bac22f946c5129eab0e0d5b22b3ed7ca19d
Fixes: 9d9491a7da ("mmc: dw_mmc: Fix the DTO timeout calculation")
Fixes: 4c2357f57d ("mmc: dw_mmc: Fix the CTO timeout calculation")
Tested-by: Vineet Gupta <Vineet.Gupta1@synopsys.com>
Reported-by: Vineet Gupta <Vineet.Gupta1@synopsys.com> # ARC STAR
9001306872 HSDK, sdio: board crashes when copying big files
Signed-off-by: Evgeniy Didin <Evgeniy.Didin@synopsys.com>
Cc: <stable@vger.kernel.org> # 4.14
Reviewed-by: Andy Shevchenko <andy.shevchenko@gmail.com>
Reviewed-by: Douglas Anderson <dianders@chromium.org>
Reviewed-by: Shawn Lin <shawn.lin@rock-chips.com>
Reviewed-by: Jisheng Zhang <Jisheng.Zhang@synaptics.com>
Acked-by: Jaehoon Chung <jh80.chung@samsung.com>
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Shawn Lin <shawn.lin@rock-chips.com>
(cherry picked from c715160225)
Add the clock tree definition for the new RK3308 SoC.
Change-Id: I1c0794b290207d28102e8d30cca13bbbf54ccfb8
Signed-off-by: Finley Xiao <finley.xiao@rock-chips.com>
After a cpufreq transition, update the clockevent's frequency
by fetching the new clock rate from the clock framework and
reprogram the next clock event.
The clock supplying the arm-global-timer on the rk3188 is coming
from the the cpu clock itself and thus changes its rate everytime
cpufreq adjusts the cpu frequency.
Found by code review, real impact not known. Assume what actual
HZ value will be different from expected on platforms using
arm-global-timer as clockevent.
The patch is port of commit 4fd7f9b128 ("ARM: 7212/1: smp_twd:
reconfigure clockevents after cpufreq change") and
commit 2b25d9f64b ("ARM: 7535/1: Reprogram smp_twd based on
new common clk framework notifiers").
Change-Id: I82552f621e30254b9c48f22fb3ebd2866d4476c8
Signed-off-by: Alexander Kochetkov <al.kochet@gmail.com>
Signed-off-by: Ziyuan Xu <xzy.xu@rock-chips.com>
We found that some Sonix usb cameras(e.g. idVendor=0c45,
idProduct=64ab or idProduct=64ac) can't support auto-suspend
well on rockchip platforms(e.g. rk3399).With auto-suspend,
these usb cameras MJPEG will display abnormally on all usb
controllers(DWC2/DWC3/EHCI). So we need to disable auto-
suspend for these special usb cameras.
Change-Id: I08c87cf5c9fa5ebe076b5dd3e873b74c5ec2cb83
Signed-off-by: William Wu <william.wu@rock-chips.com>
for rk322x power down post-PLL must to both set rege0[5]=0 and set
pre pll pre-PLL unlock. So power down pre-PLL before post-PLL power
down
Change-Id: If0eb325b10bb6eb117b0a61d5852e9aae9d92ba6
Signed-off-by: Huicong Xu <xhc@rock-chips.com>
Turn off differential receiver in suspend mode for RK3328 and
PX30 to save power.
The effect of turn off differential receiver on electricity:
USB20_AVDD_1V8: 0.73mA (turn on)
USB20_AVDD_1V8: 0.03mA (tunn off)
Change-Id: I0650d6d4b712a3692eed2564dda36d41b7956bb9
Signed-off-by: Meng Dongyang <daniel.meng@rock-chips.com>
we can power off the da_pwrok to reduce power consumption.
Change-Id: Ie08af149e74408e57750a186cf16d5adf4b3cfb7
Signed-off-by: Wyon Bi <bivvy.bi@rock-chips.com>
If hdmi is enabled in uboot and pluged out when booting kernel,
the hdmi phy is still enabled. It's better to disable it to
match the real status.
Change-Id: Ia1c5ede6499ee277d08c35a85c50e3257305f90f
Signed-off-by: Zheng Yang <zhengyang@rock-chips.com>
Under following processes, rxsense will be not match the real
signal status.
1. HDMI plug in, irq is triggered.
2. HDMI irq is mute in dw_hdmi_hardirq, bring up dw_hdmi_irq.
3. For HDMI connection is not stable, phy_stat read in
dw_hdmi_irq may be zero, then hdmi->rxsense will be false.
4. Connection fallback to stable, but dw_hdmi_irq had not
unmute the irq, irq is not triggered again, and hdmi->rxsense
keep false.
5. repo_hpd_event inform HDMI is pluggned in, dw_hdmi_bridge_enable
is called to enable HDMI. For rxsense is flase, bridge is not
powered up.
When repo_hpd_event is called, we think HDMI connection is stable,
updating rxsense is reliable.
Change-Id: Ie1f52f65b15e9a603dad9200529202053528a390
Signed-off-by: Zheng Yang <zhengyang@rock-chips.com>
most of rk3288 vop lit is same with vop big, but some feature is
different, just like max output resolution.
vop big max output: 3840*2160
vop lit max ougput: 2560*1600
Change-Id: Ie926ed29c9d23159ccf3dd5c3e885e48a21731cb
Signed-off-by: Sandy Huang <hjc@rock-chips.com>
The RK3328 HDMI PHY adds a method to detect the TMDS Data sync
enable status. The counter length is defined as:
Fref * 495 * 4 / Ftmdsclk
If sync enable counter match the counter length, the sync status
defined in reg0xdd will be set to true.
Change-Id: Ib6a58cb50127e84399011cb398e7fa36f72b3a45
Signed-off-by: Zheng Yang <zhengyang@rock-chips.com>
Because vop is unstable when update output timing,
so we close all win make sure stable.
Change-Id: Ifd4e60a41c667426d40e0d57e5180034bab9d5e7
Signed-off-by: Sandy Huang <hjc@rock-chips.com>
Before this update, drm_fb_helper_is_bound will return false
when hdmi plug.
Change-Id: I28c07b1c1c3405af2627bf423721c757f1e84258
Signed-off-by: Sandy Huang <hjc@rock-chips.com>
for some unknown reason, iommu will enable fail, so we set is_iommu_enabled
to false insure the iommu will be enable at next frame.
Change-Id: Icd8779fff816f7498385aff9aa4712e6ad18f93c
Signed-off-by: Sandy Huang <hjc@rock-chips.com>
Replace the specification of a data structure by a pointer dereference
as the parameter for the operator "sizeof" to make the corresponding size
determination a bit safer according to the Linux coding style convention.
This issue was detected by using the Coccinelle software.
Change-Id: Ic9e944a36fb8ed01a71c60922c949c00e6ecda48
Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Tao Huang <huangtao@rock-chips.com>
(cherry picked from commit 283b7ac92b)
Omit extra messages for a memory allocation failure in these functions.
This issue was detected by using the Coccinelle software.
Change-Id: If4c563f14388b60a9c49bcf4c050198a53c89a6c
Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Tao Huang <huangtao@rock-chips.com>
(cherry picked from commit 98c8ee73fe)
We generally leave the GPIO clock disabled, unless an interrupt is
requested or we're accessing IO registers. We forgot to do this for the
->get_direction() callback, which means we can sometimes [1] get
incorrect results [2] from, e.g., /sys/kernel/debug/gpio.
Enable the clock, so we get the right results!
[1] Sometimes, because many systems have 1 or mor interrupt requested on
each GPIO bank, so they always leave their clock on.
[2] Incorrect, meaning the register returns 0, and so we interpret that
as "input".
Change-Id: I8c98cb41719d0666845d0ff893a7de05ce04128f
Signed-off-by: Brian Norris <briannorris@chromium.org>
Reviewed-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Tao Huang <huangtao@rock-chips.com>
(cherry picked from commit 5c9d8c4f6b)
This patch adds the get_direction to support the gpio
interface.
The gpio direction is not used on rockchip platform when use the gpio
debugfs.
Tested on kylin board. (RK3036 SoCs)
The repro steps:
$/sys/class/gpio/
echo 53 > export
$/sys/class/gpio/gpio53# cat direction
in
In general, the gpio53 should be out value, but the direction is the
default value 'in', since the get_direction didn't supported in rockchip
pinctrl.
So, we should add this patch to support it.
Change-Id: I5510f3e60fdcae16b01a8dff1efbe31960dfcc1d
Cc: linux-gpio@vger.kernel.org
Cc: linux-rockchip@lists.infradead.org
Reported-by: Jeffy Chen <jeffy.chen@rock-chips.com>
Signed-off-by: Caesar Wang <wxt@rock-chips.com>
Reviewed-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
(cherry picked from commit 6ba20a00a3)
Rename pinctrl_utils_dt_free_map to pinctrl_utils_free_map, since
it does not depend on device tree despite the current name. This
will enforce a consistent naming in pinctr-utils.c and will make
it clear it can be called from outside device tree (e.g. from
ACPI handling code).
Change-Id: I442cea04967997ed29d6e7a3cfe35f2ec2e9c95f
Signed-off-by: Irina Tirdea <irina.tirdea@intel.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Tao Huang <huangtao@rock-chips.com>
(cherry picked from commit d32f7fd3bb)
for_each_child_of_node performs an of_node_get on each iteration, so a
return from the loop requires an of_node_put.
A simplified version of the semantic patch that fixes this problem is as
follows (http://coccinelle.lip6.fr):
// <smpl>
@@
local idexpression n;
expression e,e1;
@@
for_each_child_of_node(e1,n) {
...
(
of_node_put(n);
|
e = n
|
return n;
|
+ of_node_put(n);
? return ...;
)
...
}
// </smpl>
Change-Id: Ifa715e7d1c475d3f33cfff3ce8b7f32636447a7c
Signed-off-by: Julia Lawall <Julia.Lawall@lip6.fr>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Tao Huang <huangtao@rock-chips.com>
(cherry picked from commit 4fc8a4b2a4)
Add device managed APIs devm_pinctrl_register() and
devm_pinctrl_unregister() for the APIs pinctrl_register()
and pinctrl_unregister().
This helps in reducing code in error path and sometimes
removal of .remove callback for driver unbind.
Change-Id: I6e5f092ed0d830f8fdf357d51ae7d65fe9fb2020
Signed-off-by: Laxman Dewangan <ldewangan@nvidia.com>
Reviewed-by: Philipp Zabel <p.zabel@pengutronix.de>
Acked-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Tao Huang <huangtao@rock-chips.com>
(cherry picked from commit 80e0f8d94d)
Fix devm_pinctrl_dev_match with commit 3024f920eb
("pinctrl: zynq: Use devm_pinctrl_register() for pinctrl registration")
Add device managed APIs devm_gpiochip_add_data() and
devm_gpiochip_remove() for the APIs gpiochip_add_data()
and gpiochip_remove().
This helps in reducing code in error path and sometimes
removal of .remove callback for driver unbind.
Change-Id: I84e3336b05d0c0e82170d16cdae8977c9f962942
Signed-off-by: Laxman Dewangan <ldewangan@nvidia.com>
Signed-off-by: Tao Huang <huangtao@rock-chips.com>
(cherry picked from commit 0cf3292cde)
This makes the driver use the data pointer added to the gpio_chip
to store a pointer to the state container instead of relying on
container_of().
Change-Id: I9a9c0b6558e09efa680dfee0777469db62669e6c
Acked-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Tao Huang <huangtao@rock-chips.com>
(cherry picked from commit 03bf81f1cb)
This adds a void * pointer to gpio_chip so that driver can
assign and retrieve some states. This is done to get rid of
container_of() calls for gpio_chips embedded inside state
containers, so we can remove the need to have the gpio_chip
or later (planned) struct gpio_device be dynamically allocated
at registration time, so that its struct device can be properly
reference counted and not bound to its parent device (e.g.
a platform_device) but instead live on after unregistration
if it is opened by e.g. a char device or sysfs.
The data is added with the new function gpiochip_add_data()
and for compatibility we add static inline wrapper function
gpiochip_add() that will call gpiochip_add_data() with
NULL as argument. The latter will be removed once we have
exorcised gpiochip_add() from the kernel.
gpiochip_get_data() is added as a static inline accessor
for drivers to quickly get their data out.
Change-Id: I00f974c6008c7bd36f9bfef30bf1ef162330a6f6
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Tao Huang <huangtao@rock-chips.com>
(cherry picked from commit b08ea35a32)
