Now vpu driver support both drm and ion allocator. And have
test on rk3399-mid(drm android 7.1), rk3399-mid(ion android
6.0), rk3288(linux mini arm) platform.
Change-Id: I0096c3928849b1f11a62378675f4559b4f101445
Signed-off-by: Jung Zhao <jung.zhao@rock-chips.com>
Signed-off-by: Mark Yao <mark.yao@rock-chips.com>
Signed-off-by: Randy Li <randy.li@rock-chips.com>
The return value of iommu_map_sg is size_t, it's unsigned
Change-Id: Ib06f61c020510673bc513e1a8fde6fd3980a7ca3
Signed-off-by: Mark Yao <mark.yao@rock-chips.com>
Since (cacb6f5 FROMLIST: drm/rockchip: Use common IOMMU API to attach
devices), rockchip drm use common IOMMU API, the boot logo buffer
mapping need change to new api.
Change-Id: Ifa2c886e05d2de65de53a868458c56859519a0f2
Signed-off-by: Mark Yao <mark.yao@rock-chips.com>
Some bootloader will check the reboot mode to take different action, so
we treat unrecognized reboot mode as normal mode to prevent the system
run into abnormal case.
Change-Id: I88063a5b41e4e645443229fa490b2b55db5ccf27
Signed-off-by: Andy Yan <andy.yan@rock-chips.com>
in drm mode ion is not support dma fd, in order
to support drm fd rga driver need fix, and user
also need set the buf type by the macro
Change-Id: I3d0eaf7fbefda5aaf715772649d972b9a9d5c06d
Signed-off-by: Zikim,Wei <wzq@rock-chips.com>
In current code, the connect status will be set to connected state when
reset interrupt occur and change to disconnected state in disconnect
interrupt.
But the usb charger may bring about reset signal in accident if we
connect and disconnect quickly. In this case, the dwc3 controller will
change link state and set to connect status, yet not change to
disconnected state when disconnect. So the dwc3 controller suspend
fail and result in a mistake when quick reconnect.
This patch set connect status to connected state when transfer complete
to make sure that usb is connect to PC exactly.
Change-Id: I8e5894d2e08b88bb5434222100d8f5c91c9f1a9d
Signed-off-by: Meng Dongyang <daniel.meng@rock-chips.com>
DWC3 controller still need to read or write memory after dwc3
rockchip receive disconnect notify, the dwc3 controller will not
suspend immediately and the clock is still enabled even if we put
dwc_rockchip sync. So if we reconnect to PC quickly, the transer
will start before dwc3 rockchip receive connect notify and the dwc3
controller will reset during normal transer which result in the gadget
function break.
In order to suspend dwc3 controller and disable clock when transfer
finish immediately, this patch wait for usage_count of dwc3 controller
change to 1 and disconnect interrupt occur then suspend dwc3 controller
and disable clock, make the dwc3 continue transfer after dwc3 rockchip
receive connect notify, otherwise do not reset and get or put dwc3
controller if timeout.
Change-Id: If65344557d77370e9b6cf4bfea84175c37f00057
Signed-off-by: Meng Dongyang <daniel.meng@rock-chips.com>
Floating charger is a kind of special USB charger that
dp/dm not shorted, which is incompatible with BC1.2 spec.
However, we need to support this floating charger, and
consider that the max charge current of floating charger
is the same as standard USB charger (dp/dm is shorted),
so we set the charger type as DCP for floating charger.
Change-Id: Ifaca7269a3d4660ac095c59776d7e935fe6126df
Signed-off-by: William Wu <wulf@rock-chips.com>
gslxxxx wrong irq number with the same vopb, may cause vopb abnormal.
ts->irq is gpio number, gslxxxx irq number is ts->client->irq
ts->client->irq = gpio_to_irq(ts->irq)
Change-Id: Ic1f6a46437a2a185f3218d3f0406deeddd9d670a
Signed-off-by: Bin Yang <yangbin@rock-chips.com>
The newly added OTG support has an obvious uninitialized variable
access that gcc warns about:
drivers/phy/phy-rockchip-inno-usb2.c: In function 'rockchip_chg_detect_work':
drivers/phy/phy-rockchip-inno-usb2.c:717:7: error: 'tmout' may be used uninitialized in this function [-Werror=maybe-uninitialized]
This replaces the use of the uninitialized variable with what
the value was in the previous USB_CHG_STATE_WAIT_FOR_DCD
state.
Change-Id: Ibeed4872e1168e135a332a0978d85a4e48083267
Fixes: 0c42fe48fd23 ("phy: rockchip-inno-usb2: support otg-port for rk3399")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
Signed-off-by: William Wu <wulf@rock-chips.com>
(cherry picked from git.kernel.org kishon/linux-phy next
commit dd796e921e)
The linestate change interrupt may occur during suspend if port is
not connected. This patch pull down dp/dm when suspend.
Change-Id: I31e992727ea63efbda4ecec7ad3af02626eceb44
Signed-off-by: Meng Dongyang <daniel.meng@rock-chips.com>
On the rk3288 USB host-only port (the one that's not the OTG-enabled
port) the PHY can get into a bad state when a wakeup is asserted (not
just a wakeup from full system suspend but also a wakeup from
autosuspend).
We can get the PHY out of its bad state by asserting its "port reset",
but unfortunately that seems to assert a reset onto the USB bus so it
could confuse things if we don't actually deenumerate / reenumerate the
device.
We can also get the PHY out of its bad state by fully resetting it using
the reset from the CRU (clock reset unit) in chip, which does a more full
reset. The CRU-based reset appears to actually cause devices on the bus
to be removed and reinserted, which fixes the problem (albeit in a hacky
way).
It's unfortunate that we need to do a full re-enumeration of devices at
wakeup time, but this is better than alternative of letting the bus get
wedged.
Change-Id: I3120a38a7f646a9d244f04bd2dcfef7474a4a6d1
Signed-off-by: Randy Li <ayaka@soulik.info>
Signed-off-by: Jacob Chen <jacob2.chen@rock-chips.com>
(am from https://mail-archive.com/linux-kernel@vger.kernel.org/msg1254059.html)
The "host1" port (AKA the dwc2 port that isn't the OTG port) on rk3288
has a hardware errata that causes everything to get confused when we get
a remote wakeup. We'll use the reset that's in the CRU to reset the
port when it's in a bad state.
Note that we add the reset to both dwc2 controllers even though only one
has the errata in case we find some other use for this reset that's
unrelated to the current hardware errata. Only the host port gets the
quirk property, though.
Change-Id: I472d33fb1db8b1a6b0c4fcea9ab31fd85b61af40
Signed-off-by: Randy Li <ayaka@soulik.info>
Signed-off-by: Jacob Chen <jacob2.chen@rock-chips.com>
(am from https://patchwork.kernel.org/patch/9324169/)
The opp who contains a opp-suspend property will be configured
during suspend or reboot.
Change-Id: I6b2eede43216435f568db6959127a6e84c8cd4c8
Signed-off-by: Finley Xiao <finley.xiao@rock-chips.com>
Sometimes all cpus of big cluster are closed and its frequency
keep a high value, in order to reduce power and reset normally,
set frequency to a specific value after close all the cpus.
Change-Id: I88bce25812d1b0ff3f78a898cb161642a65cc523
Signed-off-by: Finley Xiao <finley.xiao@rock-chips.com>
We need to put the power status of HEVC IP into IDLE unless
we can't reset that IP or the SoC would crash down.
rockchip_pmu_idle_request(dev, true)---> enter idle
rockchip_pmu_idle_request(dev, false)---> exit idle
Change-Id: I76733efd2de4f7ee183c1b6bd1545d60038ee31b
Signed-off-by: Caesar Wang <wxt@rock-chips.com>
Signed-off-by: Jeffy Chen <jeffy.chen@rock-chips.com>
Signed-off-by: Randy Li <randy.li@rock-chips.com>
When mapping buffers through the PRIME DMA-buf mmap path we might be
given an offset which has to be respected. The DRM GEM mmap path already
takes care of zeroing out the fake mmap offset, so we can just make the
IOMMU mmap implementation always respect the offset.
BUG=chrome-os-partner:56615
TEST=graphics_GLBench
Change-Id: Iec83e720b24ddd35a92f3df8312015bc5af798f0
Signed-off-by: rjan Eide <orjan.eide@arm.com>
Signed-off-by: Tomasz Figa <tfiga@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/386477
Reviewed-by: Daniel Kurtz <djkurtz@chromium.org>
Signed-off-by: Randy Li <randy.li@rock-chips.com>
When mapping external DMA-bufs through the PRIME mmap call, we might be
given an offset which has to be respected. However for the internal DRM
GEM mmap path, we have to ignore the fake mmap offset used to identify
the buffer only. Currently the code always zeroes out vma->vm_pgoff,
which breaks the former.
This patch fixes the problem by moving the vm_pgoff assignment to a
function that is used only for GEM mmap path, so that the PRIME path
retains the original offset.
BUG=chrome-os-partner:56615
TEST=graphics_GLBench
Change-Id: Iec6e996707b0fe7e95a019423a944d98c80beaab
Signed-off-by: rjan Eide <orjan.eide@arm.com>
Signed-off-by: Tomasz Figa <tfiga@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/381332
Reviewed-by: Daniel Kurtz <djkurtz@chromium.org>
Signed-off-by: Randy Li <randy.li@rock-chips.com>
When converting the driver to use shmem-backed GEMs for IOMMU-enabled
systems, we forgot to add calls to drm_gem_object_release(), which gave
us a quite nice memory leak. This patch adds the missing calls.
Fixes: f11d5f0 ("FROMLIST: drm/rockchip: Do not use DMA mapping API if
attached to IOMMU domain")
BUG=chrome-os-partner:57158
TEST=while true; do backlight_dbus_tool --set --percent=0 && sleep 8 &&
backlight_dbus_tool --set --percent=100 && sleep 3 ; done
Signed-off-by: Tomasz Figa <tfiga@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/385456
Reviewed-by: Douglas Anderson <dianders@chromium.org>
Reviewed-by: Daniel Kurtz <djkurtz@chromium.org>
Conflicts:
drivers/gpu/drm/rockchip/rockchip_drm_gem.c
Change-Id: I3c7b21ed22cfb38f512150f76fced3b0cc2b296d
Signed-off-by: Randy Li <randy.li@rock-chips.com>
When freeing the buffer we don't have any means of determining if the
buffer was read or written, so we must assume both and pass true for
both arguments of drm_gem_put_pages(). Let's fix the code which
currently passes false.
BUG=chrome-os-partner:56378
TEST=while true; do backlight_dbus_tool --set --percent=0 && sleep 8 &&
backlight_dbus_tool --set --percent=100 && sleep 3 ; done
Change-Id: I7da2fd81e7e728e0ef242837b70819c4a3aee7bf
Signed-off-by: Tomasz Figa <tfiga@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/382934
Reviewed-by: Daniel Kurtz <djkurtz@chromium.org>
Signed-off-by: Randy Li <randy.li@rock-chips.com>
Rockchip DRM used the arm special API, arm_iommu_*(), to attach
iommu for ARM32 SoCs. This patch convert to common iommu API
so it would support ARM64 like RK3399.
Since previous patch added support for direct IOMMU address space
management, there is no need to use DMA API anymore and this patch wires
things to use the new method.
Signed-off-by: Shunqian Zheng <zhengsq@rock-chips.com>
Signed-off-by: Tomasz Figa <tfiga@chromium.org>
(am from https://patchwork.kernel.org/patch/9196389/)
BUG=chrome-os-partner:53565
TEST=boot to ui on Gru
Reviewed-on: https://chromium-review.googlesource.com/349521
Commit-Ready: Tomasz Figa <tfiga@chromium.org>
Tested-by: Tomasz Figa <tfiga@chromium.org>
Reviewed-by: Tomasz Figa <tfiga@chromium.org>
Conflicts:
drivers/gpu/drm/rockchip/rockchip_drm_drv.c
Change-Id: Ide4ce9f74fd431f0b7cd480e38b683f833733b40
Signed-off-by: Randy Li <randy.li@rock-chips.com>
Some rockchip vop not support iommu, need use non-iommu
buffer for it. And if we get iommu issues, we can compare
the issues with non-iommu path, that would help the debug.
Conflicts:
drivers/gpu/drm/rockchip/rockchip_drm_drv.c
Change-Id: I843f087281300359b07faac1de156cd648325b45
Signed-off-by: Mark Yao <mark.yao@rock-chips.com>
(cherry picked from commit 2d90d47743)
arm_iommu_attach_device() takes its own reference to the mapping we give
it. Since we do not keep a reference to the mapping ourselves, we must
release it before returning.
Also fix the error path, which fails to release the mapping if it has
called arm_iommu_detach_device() since that clears archdata.mapping.
Conflicts:
drivers/gpu/drm/rockchip/rockchip_drm_drv.c
Change-Id: Ia20334afbce08ece5a3238a7b0786547ec0cafb2
Signed-off-by: John Keeping <john@metanate.com>
Signed-off-by: Randy Li <randy.li@rock-chips.com>
Signed-off-by: Huang, Tao <huangtao@rock-chips.com>
(cherry picked from commit 69b007968e)
The call to arm_iommu_detach_device() on the previous line sets
dev->archdata.mapping to NULL so this call is always a no-op.
Change-Id: I09b41c284c61885fb4b989a64839e96bcc316aa6
Signed-off-by: John Keeping <john@metanate.com>
Signed-off-by: Huang, Tao <huangtao@rock-chips.com>
(cherry picked from commit aa1ac27f48)
The prime fd to handle ioctl was not used with rockchip before. Support
was added in order to pass graphics_Gbm and to support potential uses
within Chrome OS (e.g. zero-copy video decode, camera).
Difference from kernel 3.14 implementation:
- prime_import_sg_table passes dma-buf as argument instead of size
- need to handle import sg_table for both DMA and IOMMU paths
TEST=test_that graphics_Gbm on kevin
BUG=chrome-os-partner:56526
Reviewed-on: https://chromium-review.googlesource.com/381991
Tested-by: Haixia Shi <hshi@chromium.org>
Commit-Queue: Haixia Shi <hshi@chromium.org>
Trybot-Ready: Haixia Shi <hshi@chromium.org>
Reviewed-by: Haixia Shi <hshi@chromium.org>
Reviewed-by: Stéphane Marchesin <marcheu@chromium.org>
Conflicts:
drivers/gpu/drm/rockchip/rockchip_drm_gem.c
Change-Id: I7aae5b0e227de61ac17adf98996152da8db097d2
Signed-off-by: Randy Li <randy.li@rock-chips.com>
The API is not suitable for subsystems consisting of multiple devices
and requires severe hacks to use it. To mitigate this, this patch
implements allocation and address space management locally by using
helpers provided by DRM framework, like other DRM drivers do, e.g.
Tegra.
This patch should not introduce any functional changes until the driver
is made to attach subdevices into an IOMMU domain with the generic IOMMU
API, which will happen in following patch. Based heavily on GEM
implementation of Tegra DRM driver.
Signed-off-by: Tomasz Figa <tfiga@chromium.org>
Signed-off-by: Shunqian Zheng <zhengsq@rock-chips.com>
(am from https://patchwork.kernel.org/patch/9196387/)
BUG=chrome-os-partner:53565
TEST=boot to ui on Gru
Reviewed-on: https://chromium-review.googlesource.com/353591
Commit-Ready: Tomasz Figa <tfiga@chromium.org>
Tested-by: Tomasz Figa <tfiga@chromium.org>
Reviewed-by: Tomasz Figa <tfiga@chromium.org>
Reviewed-by: Daniel Kurtz <djkurtz@chromium.org>
Conflicts:
drivers/gpu/drm/rockchip/rockchip_drm_drv.h
drivers/gpu/drm/rockchip/rockchip_drm_gem.c
drivers/gpu/drm/rockchip/rockchip_drm_gem.h
Change-Id: I599b2551e8e8000024894426e4e19d7543af3b0a
Signed-off-by: Randy Li <randy.li@rock-chips.com>
This reverts commit b435f1a281.
Conflicts:
drivers/gpu/drm/rockchip/Kconfig
drivers/gpu/drm/rockchip/rockchip_drm_drv.h
Change-Id: I15d2d598514c2878a574219fe2268ee8d530d6e2
Signed-off-by: Randy Li <randy.li@rock-chips.com>
There is not need to force the video driver dependency in parent
tree.
Also there would be too many entries in a menu, I create a submenu
for them.
Change-Id: Id3bd54fa6274b1311ff6b20f95ffeb0d1851d9f9
Signed-off-by: Randy Li <randy.li@rock-chips.com>
We decide to offer the video support in old way.
Change-Id: I0b30f7b2b559596ee124b8558cd54e0ff090ab79
Signed-off-by: Randy Li <randy.li@rock-chips.com>
If you meet a performance issue, I would suggest you to set
500MHZ for ACLK_HEVC at rk3288. It could solve the decoding quite
slow for some video resources.
Change-Id: Iedc2bc90fb13ae89d204dc6b5f5d897acae6812d
Signed-off-by: Randy Li <randy.li@rock-chips.com>
* linux-linaro-lsk-v4.4-android: (61 commits)
Linux 4.4.36
scsi: mpt3sas: Unblock device after controller reset
flow_dissect: call init_default_flow_dissectors() earlier
mei: fix return value on disconnection
mei: me: fix place for kaby point device ids.
mei: me: disable driver on SPT SPS firmware
drm/radeon: Ensure vblank interrupt is enabled on DPMS transition to on
mpi: Fix NULL ptr dereference in mpi_powm() [ver #3]
parisc: Also flush data TLB in flush_icache_page_asm
parisc: Fix race in pci-dma.c
parisc: Fix races in parisc_setup_cache_timing()
NFSv4.x: hide array-bounds warning
apparmor: fix change_hat not finding hat after policy replacement
cfg80211: limit scan results cache size
tile: avoid using clocksource_cyc2ns with absolute cycle count
scsi: mpt3sas: Fix secure erase premature termination
Fix USB CB/CBI storage devices with CONFIG_VMAP_STACK=y
USB: serial: ftdi_sio: add support for TI CC3200 LaunchPad
USB: serial: cp210x: add ID for the Zone DPMX
usb: chipidea: move the lock initialization to core file
...
Stop wasting IOVA space by over-aligning scatterlist segments for a
theoretical worst-case segment boundary mask, and instead take the real
limits into account to merge consecutive segments wherever appropriate,
so our callers can benefit from getting back nicely simplified lists.
This also represents the last piece of functionality wanted by users of
the current arch/arm implementation, thus brings us a small step closer
to converting that over to the common code.
Change-Id: Ie583142efece02dfa6678d5e8607b5bc7266f50c
Signed-off-by: Robin Murphy <robin.murphy@arm.com>
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Signed-off-by: Huang, Tao <huangtao@rock-chips.com>
(cherry picked from commit 809eac54cd)
when pd power on/off, the qos regs need to save and restore.
Change-Id: Idd6854022fb25538e82238f25a650a687e918a56
Signed-off-by: Elaine Zhang <zhangqing@rock-chips.com>
