commit 5012284700 upstream.
Commit 8844618d8a: "ext4: only look at the bg_flags field if it is
valid" will complain if block group zero does not have the
EXT4_BG_INODE_ZEROED flag set. Unfortunately, this is not correct,
since a freshly created file system has this flag cleared. It gets
almost immediately after the file system is mounted read-write --- but
the following somewhat unlikely sequence will end up triggering a
false positive report of a corrupted file system:
mkfs.ext4 /dev/vdc
mount -o ro /dev/vdc /vdc
mount -o remount,rw /dev/vdc
Instead, when initializing the inode table for block group zero, test
to make sure that itable_unused count is not too large, since that is
the case that will result in some or all of the reserved inodes
getting cleared.
This fixes the failures reported by Eric Whiteney when running
generic/230 and generic/231 in the the nojournal test case.
Change-Id: I799bfafa4712e2dd29268c7481b539b23e2cf475
Fixes: 8844618d8a ("ext4: only look at the bg_flags field if it is valid")
Reported-by: Eric Whitney <enwlinux@gmail.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Tao Huang <huangtao@rock-chips.com>
(cherry picked from commit 7736fcede7)
when two dvp sensor connect to rkisp1, only one sensor can connect to sink
pad, another sensor connect to sink params pad.
Change-Id: I6cc9e6db3ad074c82c15a9cc642330ece6997d09
Signed-off-by: Hu Kejun <william.hu@rock-chips.com>
The fiq_tty_proc_show can be called even when fiq is not enabled in dts,
which would cause crash.
Add sanity check to avoid that.
Change-Id: I69d34718f91813aed14c542901060e4fd68b818b
Signed-off-by: Jeffy Chen <jeffy.chen@rock-chips.com>
Android P CDD document chapter "9.7. Security Features"
Kernel integrity and self-protection features are integral to Android
security. Device implementations:
[C-0-8] MUST implement strict kernel memory protections where executable
code is readonly, read-only data is non-executable and non-writable, and
writable data is nonexecutable (e.g. CONFIG_DEBUG_RODATA or CONFIG_STRICT_KERNEL_RWX ).
Change-Id: Id4deb57242b55c8b5de2ad7de8a48f49ffa3c3e3
Signed-off-by: Zhangbin Tong <zebulun.tong@rock-chips.com>
When I tested DP function on excavator, I found that commit 1bf7191033de
("BACKPORT: FROMLIST: phy: rockchip-typec: support variable phy config value"),
which change pll settings, but it does not work as expected on excavator board.
With this patch series, DP works well on excavator and roc.
Change-Id: I48fe0c51e322369d1aff352f4ebaf1096f264834
Signed-off-by: Wyon Bi <bivvy.bi@rock-chips.com>
It looks that switching BIST mode to normal mode, we need
to increase the ADC_STABLE_MS delay (> ~150ms) to avoid
glitches.
Change-Id: Ied7b283cd4a2fa8c2702a655dab464a7f90f63c4
Signed-off-by: Xing Zheng <zhengxing@rock-chips.com>
There is some broken mapped grps when we use the property
'rockchip,en-always-grps', let's fix it.
Change-Id: I1f668138be4133ee06988756917220621758d2de
Signed-off-by: Xing Zheng <zhengxing@rock-chips.com>
- Split DT source files to separate out android fireware specific DT
bindings.
- Add an alias for firmware_android && chosen in px30-android.dtsi.
Change-Id: I63da786d8f70c72497b3def47309efcd8399c171
Signed-off-by: Bian Jin chen <kenjc.bian@rock-chips.com>
The innoslicon GF22FDX MIPI D-PHY integrates a MIPI 1.2
compatible PHY that supports up to 2.5Gbps high speed data
transmitter, plus a MIPI low-power low speed transceiver
that support data transfer in the bi-directional mode.
The IP supports the full specifications described in the
D-PHY spec 1.2. The D-PHY is built in with a standard
digital interface to talk to any third party host controller.
Change-Id: Iad7b9dc8fedaa1a5741830e9c02a593f544c2423
Signed-off-by: Wyon Bi <bivvy.bi@rock-chips.com>
rk1808 have vop lite and vop raw:
1. vop lite: support win1 for display, vop->dsi tx->dphy->lcd.
2. vop raw: transfer data from ddr to csi tx.
Change-Id: I11229b5e61e66e72e4228e7e0ac966f1f85cb49f
Signed-off-by: Sandy Huang <hjc@rock-chips.com>
This algorithm defaults to choosing higher pin config over
lower ones in order to prefer multi-function if desired.
NAME | SIGNALING | OUTPUT TYPE | MULTI-FUNCTION | PIN CONFIG
-------------------------------------------------------------
A | USB G2 | ? | no | 00_0001
B | USB G2 | ? | yes | 00_0010
C | DP | CONVERTED | no | 00_0100
D | PD | CONVERTED | yes | 00_1000
E | DP | DP | no | 01_0000
F | PD | DP | yes | 10_0000
if UFP has NOT asserted multi-function preferred code masks away
B/D/F leaving only A/C/E. For single-output dongles that should
leave only one possible pin config depending on whether its a
converter DP->(VGA|HDMI) or DP output. If UFP is a USB-C receptacle
it may assert C/D/E/F. The DFP USB-C receptacle must always choose
C/D in those cases.
Change-Id: I5594d39a302f27e2d72259f6a18308488d4fa47c
Signed-off-by: Wyon Bi <bivvy.bi@rock-chips.com>
DP firmware uses fixed phy config values to do training, but some
boards need to adjust these values to fit for their unique hardware
design. So get phy config values from dts and use software link training
instead of relying on firmware, if software training fail, keep firmware
training as a fallback if sw training fails.
Signed-off-by: Chris Zhong <zyw@rock-chips.com>
Signed-off-by: Lin Huang <hl@rock-chips.com>
(am from https://patchwork.kernel.org/patch/10420469/)
Conflicts:
drivers/gpu/drm/rockchip/cdn-dp-reg.h
[Context - non-upstream HDCP stuff]
BUG=b:72006974
TEST=DP can display on Dru
Signed-off-by: Brian Norris <briannorris@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/985573
Tested-by: Lin Huang <hl@rock-chips.com>
Reviewed-by: Sean Paul <seanpaul@google.com>
Change-Id: I402c5fd2c83979cee67856e66311d2b1b9c6f774
Signed-off-by: Wyon Bi <bivvy.bi@rock-chips.com>
This patch transitions our rockchip driver to using the upstream
content_protection_property (located in drm_connector) as opposed to the
downstream version that's in mode_config.
In addition to the new location of the property, this patch also makes
HDCP enablement work via atomic commits in addition to the legacy
setproperty ioctl.
The trickiest part of making this work is ensuring we keep the
connector->state->content_protection value in sync with what the
hardware is doing. We're only allowed to change this value in
atomic_check (which we do for certain transitions), however we have to
be careful when updating it outside of atomic_check, this requires
locks. In order to update the property without races, we need a new property
worker whose job is to acquire the connection and dp locks and update the state.
It's not possible to do this without the dedicated worker since the hdcp event
worker doesn't hold the connection mutex, and can't acquire it since we
synchronously cancel it while holding the connection mutex elsewhere. Sigh.
Signed-off-by: Sean Paul <seanpaul@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/864343
Commit-Ready: Sean Paul <seanpaul@google.com>
Tested-by: Sean Paul <seanpaul@google.com>
Reviewed-by: Sean Paul <seanpaul@google.com>
Change-Id: I22d2592096866d2346bc9b48f48e66e845a173f8
Signed-off-by: Wyon Bi <bivvy.bi@rock-chips.com>
This patch adds a new optional connector property to allow userspace to enable
protection over the content it is displaying. This will typically be implemented
by the driver using HDCP.
The property is a tri-state with the following values:
- OFF: Self explanatory, no content protection
- DESIRED: Userspace requests that the driver enable protection
- ENABLED: Once the driver has authenticated the link, it sets this value
The driver is responsible for downgrading ENABLED to DESIRED if the link becomes
unprotected. The driver should also maintain the desiredness of protection
across hotplug/dpms/suspend.
If this looks familiar, I posted [1] this 3 years ago. We have been using this
in ChromeOS across exynos, mediatek, and rockchip over that time.
Changes in v2:
- Pimp kerneldoc for content_protection_property (Daniel)
- Drop sysfs attribute
Changes in v3:
- None
Changes in v4:
- Changed kerneldoc to recommend userspace polling (Daniel)
- Changed kerneldoc to briefly describe how to attach the property (Daniel)
Changes in v5:
- checkpatch whitespace noise
- Change DRM_MODE_CONTENT_PROTECTION_OFF to DRM_MODE_CONTENT_PROTECTION_UNDESIRED
Changes in v6:
- None
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Sean Paul <seanpaul@chromium.org>
[1] https://lists.freedesktop.org/archives/dri-devel/2014-December/073336.html
Link: https://patchwork.freedesktop.org/patch/msgid/20180108195545.218615-4-seanpaul@chromium.org
(cherry picked from commit 24557865c8)
Signed-off-by: Sean Paul <seanpaul@chromium.org>
[downstream changes]
- Fixed some conflicts in comments
- Remove duplicate definition for drm_get_content_protection_name
Change-Id: I825b4863bea715434cb8f76f99fdf6e3fca74a60
Reviewed-on: https://chromium-review.googlesource.com/849079
Commit-Ready: ChromeOS CL Exonerator Bot <chromiumos-cl-exonerator@appspot.gserviceaccount.com>
Tested-by: Sean Paul <seanpaul@google.com>
Reviewed-by: Sean Paul <seanpaul@google.com>
Signed-off-by: Wyon Bi <bivvy.bi@rock-chips.com>
Enabling HDCP on plug/re-plug when already desired was flakey. However,
toggling the property always worked when already plugged in. It seems
like the hardware wants to be explicitly disabled before being enabled.
This patch adds the disable with a short wait before kicking off HDCP.
BUG=b:63816472
TEST=Two cases:
1)
- Boot device with dongle unplugged
- Mark HDCP as desired
- Plug HDMI + dongle
- Ensure HDCP is enabled
2)
- Plug HDMI + dongle
- Mark HDCP as desired, wait for it to enable
- Unplug HDMI from dongle
- Replug HDMI to dongle
- Ensure HDCP is enabled
Signed-off-by: Sean Paul <seanpaul@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/687802
Commit-Ready: Sean Paul <seanpaul@google.com>
Tested-by: Sean Paul <seanpaul@google.com>
Reviewed-by: Kristian H. Kristensen <hoegsberg@chromium.org>
Change-Id: I7c1dfd93bad60483d04525b79c6f75b728096ed4
Signed-off-by: Wyon Bi <bivvy.bi@rock-chips.com>
This patch adds new calls for HDCP disable/enable such that HDCP is
properly disabled/restored across hotplugs (both dongle and cable) and
power on/off.
BUG=b:63816472
TEST=Use hdcp test script below:
attr=/sys/class/drm/card1-DP-1/content_protection
printf "Testing HDCP...\n"
while [ 1 ]; do
printf "\rSetting state to desired... "
echo "Desired" > $attr
sleep $(perl -e 'printf("%.1f\n", rand() * 3)')
printf "\rSetting state to undesired..."
echo "Undesired" > $attr
sleep $(perl -e 'printf("%.1f\n", rand() * 3)')
done
Signed-off-by: Sean Paul <seanpaul@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/657941
Commit-Ready: Sean Paul <seanpaul@google.com>
Tested-by: Sean Paul <seanpaul@google.com>
Reviewed-by: Kristian H. Kristensen <hoegsberg@chromium.org>
Change-Id: I4bcc7decc43f7b648054d841efbade7315e785fe
Signed-off-by: Wyon Bi <bivvy.bi@rock-chips.com>
The rest of the driver uses DRM_DEV_*, so too should the hdcp functions.
This patch also prints error messages when enable/disable fails, as well
as info messages when hdcp is successfully enabled/disabled.
BUG=b:63816472
TEST=Enable hdcp, log messages are prefixed with drm goo
Signed-off-by: Sean Paul <seanpaul@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/657940
Commit-Ready: Sean Paul <seanpaul@google.com>
Tested-by: Sean Paul <seanpaul@google.com>
Reviewed-by: Kristian H. Kristensen <hoegsberg@chromium.org>
Change-Id: I2503ba4ce6839bb8d9db77ae88446da4888732d5
Signed-off-by: Wyon Bi <bivvy.bi@rock-chips.com>
A couple of things to reduce the races during hdcp enable/disable:
- Cancel any active workers currently transition
- Hold lock while disabling hdcp
BUG=b:64438996
TEST=Run the following script, ensure sequencing is correct:
attr=/sys/class/drm/card1-DP-1/content_protection
printf "Testing HDCP...\n"
while [ 1 ]; do
printf "\rSetting state to desired... "
echo "Desired" > $attr
sleep $(perl -e 'printf("%.1f\n", rand() * 3)')
printf "\rSetting state to undesired..."
echo "Undesired" > $attr
sleep $(perl -e 'printf("%.1f\n", rand() * 3)')
done
Signed-off-by: Sean Paul <seanpaul@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/657938
Commit-Ready: Sean Paul <seanpaul@google.com>
Tested-by: Sean Paul <seanpaul@google.com>
Reviewed-by: Kristian H. Kristensen <hoegsberg@chromium.org>
Change-Id: I535b23ffb22eba251a595dcdc4c204de80765414
Signed-off-by: Wyon Bi <bivvy.bi@rock-chips.com>
The Content Protection properties in 4.4 used v2 of the upstream
patches, but chrome (and upstream) prefer v1. This patch removes the ksv
property and reinstates the ENABLED enum value to the content protection
property.
BUG=b:63816472
TEST=Watch protected content on external display, ensure CP is
enabled/disabled properly by chrome
Signed-off-by: Sean Paul <seanpaul@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/687800
Commit-Ready: Sean Paul <seanpaul@google.com>
Tested-by: Sean Paul <seanpaul@google.com>
Reviewed-by: Kristian H. Kristensen <hoegsberg@chromium.org>
Change-Id: I38cecce2d15b4d4b1ce95ef0e572a08f1bc97131
Signed-off-by: Wyon Bi <bivvy.bi@rock-chips.com>
I've seen this:
sysfs: cannot create duplicate filename '/devices/platform/fec00000.dp/hdcp_key'
Presumably because component_add() can -EPROBE_DEFER.
At any rate, we shouldn't leave the sysfs file hanging around. Clean it
up on probe failure, and on removal.
BUG=b:36566733
TEST=boot
Signed-off-by: Brian Norris <briannorris@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/458836
Reviewed-by: Guenter Roeck <groeck@chromium.org>
Change-Id: Ic10f45c0d67cfcc41ab89af108a78f641f424872
Signed-off-by: Wyon Bi <bivvy.bi@rock-chips.com>
Implement HDCP 1.4 authentication function for DP driver.
First, userspace write HDCP cipher to driver via "hdcp_key" sysfs node,
it contains ksv and device private key. DP connector object has a
"Content Proection" property which it's 'Undesired' default. Userspace
set it to 'Desired' as a HDCP authentication request, and it will be set
to 'Enabled' once HDCP authenticated.
BUG=chrome-os-partner:56883
TEST=WITH HDCP KEY SITUATION:
1.flash encrypt key to vpd area, and reboot chromebook.
$ vpd -O; vpd -s hdcp_key_v1_4=something...;dump_vpd_log
--force;
2.write key to DP driver.
$ grep hdcp_key /var/cache/vpd/full-v2.txt | cut -s -d \" -f 4 |
tr -d "\n" > /sys/devices/platform/fec00000.dp/hdcp_key
3.set "Content Protection" property to 'Desired'
$ proptest 36 connector 23 1
4.if the HDCP 1.4 authenticated, the 'Content Proection' will be
changed to 'Enabled'
Signed-off-by: Ziyuan Xu <xzy.xu@rock-chips.com>
Reviewed-on: https://chromium-review.googlesource.com/403975
Reviewed-by: Sean Paul <seanpaul@chromium.org>
Change-Id: I58f11535dd68d9517a3c7b0b9a61ed9fa6d9a1c0
Signed-off-by: Wyon Bi <bivvy.bi@rock-chips.com>
Although this property value have been deleted from the latest
FROMLIST version (CL:266854), but the chromeos would want driver
to report the property to "Enabled" when hardware HDCP have been
enabled successfully, so let's add this back.
BUG=chrome-os-partner:56883
TEST=None
Signed-off-by: Yakir Yang <ykk@rock-chips.com>
Signed-off-by: Ziyuan Xu <xzy.xu@rock-chips.com>
Reviewed-on: https://chromium-review.googlesource.com/403974
Reviewed-by: Sean Paul <seanpaul@google.com>
Change-Id: Icc52d4a83ac434e898be1190cf934ed8333e78bf
Signed-off-by: Wyon Bi <bivvy.bi@rock-chips.com>
Add new standard connector properties to track whether content protection
(ex: hdcp) is desired by userspace. There are two properties involved,
"Content Protection" and "Content Protection KSV".
The "Content Protection" property allows userspace to request protection
on a connector. Set "Desired" to enable, "Undesired" to disable.
The "Content Protection KSV" property reflects the current state of
protection. If the KSV is 0, the connection is not protected. Once the
driver has enabled protection, it will update the the value with the KSV
(or similarly unique identifier, if not using HDCP) of the first-hop
device (sink or repeater).
(am from https://patchwork.kernel.org/patch/5439871/)
BUG=chrome-os-partner:56883
TEST=Tested on kevin, ensured the sysfs file showed up, and
reflected the correct
Signed-off-by: Sean Paul <seanpaul@chromium.org>
Signed-off-by: Yakir Yang <ykk@rock-chips.com>
Signed-off-by: Ziyuan Xu <xzy.xu@rock-chips.com>
Reviewed-on: https://chromium-review.googlesource.com/403973
Reviewed-by: Sean Paul <seanpaul@google.com>
Change-Id: I6bef13729f77de6e37d2da5e12fc69f810a2e286
Signed-off-by: Wyon Bi <bivvy.bi@rock-chips.com>
Provide the "hdcp_key" debug node in sysfs, so that userspace can pass
the legitimate HDCP key to driver.
For ATF security, we should transmit HDCP key data via smc argument one
by one.
Moreover, driver request one page SRAM space as share memory for HDCP
key transmission.
BUG=chrome-os-partner:56883
TEST=WITH HDCP KEY SITUATION:
1.flash encrypt key to vpd area, and reboot chromebook.
$ vpd -O; vpd -s hdcp_key_v1_4=something...; dump_vpd_log
--force
2.write key to DP driver.
$ grep hdcp_key /var/cache/vpd/full-v2.txt | cut -s -d \" -f 4 |
tr -d "\n" > /sys/devices/platform/fec00000.dp/hdcp_key
Signed-off-by: Ziyuan Xu <xzy.xu@rock-chips.com>
Reviewed-on: https://chromium-review.googlesource.com/403972
Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-by: Sean Paul <seanpaul@google.com>
Change-Id: I0b97dcf938623685b6938d8a40285fac3b1d5045
Signed-off-by: Wyon Bi <bivvy.bi@rock-chips.com>
