The commit c590056b6a ("phy: rockchip-inno-usb2: usb remote
wakeup support") power on otg phy in linestate irq handler,
this will cause usb peripheral fail to connect to PC in the
following case:
1. enable otg linestate irq
2. set system enter deep sleep
3. wakeup system by power key
4. connect usb peripheral to PC, pull up D+ to ~3V, trigger
linestate irq and power on otg phy.
5. usb peripheral do BC1.2 detect, but PC try to enumerate
the usb peripheral at the same time and fail at last.
Actually the usb controller drivers (e.g. dwc3 driver)
and otg_sm_work can manage the otg phy power consumption, so
it doesn't need to power on otg phy in linestate irq handler.
Change-Id: Ifd78e4d44ab96f07f75f063ed20af153b4027028
Signed-off-by: William Wu <wulf@rock-chips.com>
When test usb gadget uvc function, we find a isoc in
ep transfer bug that will cause uvc data transfer fail.
The error case is:
1. The current EP request is done, call complete_ep()
to completes the request, and then call start_next_request()
to check the EP request queue, in this error case, the
queue is empty, so it doesn't start next request, just
set ep frame_num to 0xFFFFFFFF.
2. NAK Interrutp is triggered, check isoc ep frame_num
is 0xFFFFFFFF, then reset the frame_num to 0, and then
call start_next_request() to check the EP request queue,
in this error case, the queue is still empty, so set ep
frame_num to 0xFFFFFFFF again.
But afer the above operation, the current code will
modify the ep frame_num in NAK Interrutp handler by
add ep bInterval to frame_num, this cause frame_num
change again, but not keep in 0xFFFFFFFF, so the next
NAK Interrutp handler doesn't start next request any
more.
This patch reset the frame_num to the current frame
number got from DSTS SOFFN register if detect the
frame_num is 0xFFFFFFFF in NAK Interrutp handler.
And modify the frame_num in NAK Interrutp handler
only when the frame_num is not 0xFFFFFFFF.
TEST=Set usb gadget as webcam, use Ubuntu Guvcview
to preview the webcam, observe the preview screen
and the error log "There are no more ISOC requests".
Change-Id: I4403a67b1d5d257d092a2a71d5666c5d6fd5af3c
Signed-off-by: William Wu <wulf@rock-chips.com>
To ensure that the work is executed before system being suspended.
Change-Id: Iec1bd114dfff53e2464540f09ced66cf6be81d1a
Signed-off-by: chenzhen <chenzhen@rock-chips.com>
Until fix this compile error:
ERROR: "Array_MP_8822B_FW_NIC" [drivers/net/wireless/rockchip_wlan/rtl8822be/8822be.ko] undefined!
Change-Id: I78afda57761bb0bd7737864e5d04deeaa1b7df8a
Signed-off-by: Huang, Tao <huangtao@rock-chips.com>
rework for rk3368-tb.dtsi and rk3368-tb-sheep.dts, intergrate them
to rk3368-sheep.dts
Change-Id: Ieb9198be7c80a5c8c31b0a1990bac22079548eea
Signed-off-by: Jianqun Xu <jay.xu@rock-chips.com>
Add cpu-avs node in the device tree for the ARM64 rk3368 SoC.
Change-Id: Ie7eee09c20b06bd755b9277e0acd8eaf810c5331
Signed-off-by: Finley Xiao <finley.xiao@rock-chips.com>
Add a efuse node in the device tree for the ARM64 rk3368 SoC.
Change-Id: I7d13febf73e336d5b9d1046ffe6ed69d99ea9dd6
Signed-off-by: Finley Xiao <finley.xiao@rock-chips.com>
This adds the necessary data for handling efuse on the rk3368.
As efuse of rk3368 is secure, use secure interface to access efuse.
Change-Id: I72c29348b7744b232d75ab51c56dc7de0988c24e
Signed-off-by: Finley Xiao <finley.xiao@rock-chips.com>
Reference the newly added efuse clock-ids in the clock-tree.
Change-Id: Ibbef52bcc44d006ab48e6f1f874e3bc88c681bd6
Signed-off-by: Finley Xiao <finley.xiao@rock-chips.com>
Adds new ids for the pclk supplying the efuse on rk3368 socs.
Change-Id: I69f0daf402d62079e47d8df8f6e9bef0b274239f
Signed-off-by: Finley Xiao <finley.xiao@rock-chips.com>
RGA are used for scale copy buffer form vpu,isp,
no need to flush CPU cache
Change-Id: Id3cfa0560a14cea122321dc1ea814aa8ec5dd4b6
Signed-off-by: Jacob Chen <jacob2.chen@rock-chips.com>
This adds enable mbox and scpi function for rk3368-tb board.
Change-Id: I0b4866b72ad59892bc1c051c60bafd1a15ddce14
Signed-off-by: Frank Wang <frank.wang@rock-chips.com>
This adds support rk3368-mbox and scpi nodes for rk3368 SoC.
Change-Id: Ifdb90f8101c67ec579b5bba28f163527ed096c4c
Signed-off-by: Frank Wang <frank.wang@rock-chips.com>
Add mailbox and scpi protocol function support for rk3368 SoC.
Change-Id: I201c916865eb2729ed135c3f5a77a9dd97007952
Signed-off-by: Frank Wang <frank.wang@rock-chips.com>
The JPEG decoder is also able to output the raw image in
NV12 format without the help of Post Processor of decoder
in VDPU1.
Change-Id: I1b29a5d93d2dd8ae2f21a7537cc3493adf57d022
Signed-off-by: Randy Li <randy.li@rock-chips.com>
The rk3328 saradc is the same as rk3399, so change the compitiable,
they are both 6 channels.
Change-Id: Ia6104e8c5c3590cc745792b8cd3a457a15bb53d2
Signed-off-by: David Wu <david.wu@rock-chips.com>
The RK3328 datasheet is not define the pins of gmac m0,
so remove them.
Change-Id: Ic537586c76bd2f3a937c1e15b5877744598cc702
Signed-off-by: David Wu <david.wu@rock-chips.com>
Add gpio pin index definition to make it easier to describe
GPIO in dts.
Signed-off-by: Andy Yan <andy.yan@rock-chips.com>
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Change-Id: I4aa6c01f96d91fd8ddcb371fd01ca9629f1a3013
bcmdhd driver support ko module or compile into kernel
Change-Id: I260c9d3c3da0a9e249a4ee3883ed3efe03964ddf
Signed-off-by: Xu Xuehui <xxh@rock-chips.com>
If there's a panel connected to the analogix_dp bridge, rely on
the panel driver for modes, rather than reading EDID *and* calling
get_modes() on the panel.
This allows panels with a valid EDID to read it in the panel driver
(e.g. simple_panel), and panels with invalid EDID to homebrew modes
in their get_modes implementation.
BUG=chrome-os-partner:53565
TEST=Boot device and confirm the modes returned are from panel driver
instead of EDID
[from https://patchwork.freedesktop.org/patch/107115/]
Signed-off-by: Sean Paul <seanpaul@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/374959
Commit-Ready: Chris Zhong <zyw@rock-chips.com>
Tested-by: Chris Zhong <zyw@rock-chips.com>
Reviewed-by: Stéphane Marchesin <marcheu@chromium.org>
Change-Id: Ie07dd33e3c121215bf24394cfcb3fff8c7c746a5
Signed-off-by: Mark Yao <mark.yao@rock-chips.com>
It's required by gpu driver, which will load firmware through
/system/vendor/firmware.
If android finds another way to fix it, then can be reverted.
Change-Id: I2fede28f022f10c2e16f68b21159a638a10a53ec
Signed-off-by: Jianqun Xu <jay.xu@rock-chips.com>
