19 Nov, 2020
23 commits
-
Due to commit b3a420c9cf3f (MLK-24998-4 drm/bridge: sec-dsim:
use 12MHz for default PHY REF clock), the dsi PHY reference
clock source need to be assigned to osc_24m clock.Signed-off-by: Fancy Fang
Reviewed-by: Jacky Bai
(cherry picked from commit 8e43cd16c8bbfe5b7e3c0fc1e7c3ddf738d8db01) -
Due to commit b3a420c9cf3f (MLK-24998-4 drm/bridge: sec-dsim:
use 12MHz for default PHY REF clock), the dsi PHY reference
clock source need to be assigned to osc_24m clock.Signed-off-by: Fancy Fang
Reviewed-by: Jacky Bai
(cherry picked from commit 2972241b831ed65f641ccdb80b504cadef0ba591) -
Add mode_valid() implementation for CRTC to filter out any
mode which cannot be supported by LCDIFv3. Only check the
CEA and DMT modes for pixel clock round rate is same with
the value from mode.Signed-off-by: Fancy Fang
Reviewed-by: Liu Ying
(cherry picked from commit f252a44da9f90951614c0bf513df6bd4d145e76e) -
Due to commit 82586f0aa1c2 (arm64: dts: imx8mp: correct
assigned-clock-rates for video_pll1), so remove unused
2079M clock from imx_pll1443x_tbl.Signed-off-by: Fancy Fang
Reviewed-by: Jacky Bai
(cherry picked from commit b96af227c28b1dfdbdf656de2a77bc4de99136e2) -
According to i.MX8MP Architecture Defition Document, the maximum
clock rate comes generated by 'ccm_media_disp2_pix_clk_root' is
160MHz, so 1039.5MHz clock rate is not supported. And besides,
this clock rate will be set to the matched rate with display mode
in lcdif driver, so it is not necessary to set its rate in its
assigned-clock-rates property, and just leave it to be 0.Signed-off-by: Fancy Fang
Reviewed-by: Liu Ying
(cherry picked from commit 0e3556f282466e6b91def024afc815ef77733161) -
After using osc_24m for MIPI PHY reference clock source,
the default PHY reference clock rate should be changed
also accordingly. Here choose 12MHz rate for this since
below usual DSI output DDR clock rates can be derived
from 12MHz reference:891000,
810000,
792000,
648000,
472500,
445500,
390000,
297000,
240000,
189000,Signed-off-by: Fancy Fang
Reviewed-by: Liu Ying
(cherry picked from commit b3a420c9cf3fe40c408d4eb58841a0d047c186a4) -
Due to commit 26ef2488a2ef (MLK-24998-1 arm64: dts: imx8mp: correct
assigned-clock-rates for video_pll1), default 27MHz dsi PHY reference
clock cannot be derived from 'vide_pll1', so change to use osc_24m
for the clock source and use 12MHz for dsi reference clock rate, since
below usual DDR clock rates can be derived through 12MHz clock rate:891000,
810000,
792000,
648000,
472500,
445500,
390000,
297000,
240000,
189000,All these clock rates comes from ADV7535 bridge driver.
Signed-off-by: Fancy Fang
Reviewed-by: Jacky Bai
(cherry picked from commit f3915cb61639821fbdcdc9db3cf3a8e0880cbca3) -
According to i.MX8MP Architecture Defition Document, the maximum
output frequency generated by video_pll1 is 1190MHz, so correct
its assigned-clock-rates to be 1039.5MHz to meet the spec.Signed-off-by: Fancy Fang
Reviewed-by: Jacky Bai
(cherry picked from commit 1dff13053bf83c2d4fb818562a086ad834f2a0bf) -
According to i.MX8MP Architecture Defition Document, the maximum
output frequency generated by video_pll1 is 1190MHz, so correct
its assigned-clock-rates to be 1039.5MHz to meet the spec.Signed-off-by: Fancy Fang
Reviewed-by: Jacky Bai
(cherry picked from commit b935595aa00859887a407dc6900763bfd41dfac2) -
kernel debug file is moved to linux osal part,
Fix: "MGS-6060 gpu-viv: sync 20201116 patch release"Signed-off-by: Xianzhong
-
The config for lcdif should be removed, and need to add vpu
qos setting by default. this patch is just to add back the
vpu qos config that is missed when resolving cherry-pick
conflict for patch:
3a3f54750294: MLK-19380 driver: soc: update the noc QoS setting on imx8mqSigned-off-by: Jacky Bai
Reviewed-by: Anson Huang -
include critical bug-fixings for 6.4.3.p1 driver
Signed-off-by: Xianzhong
-
fix drm header and function related changes for new kernel
Signed-off-by: Xianzhong
-
weston-simple-dmabuf-egl has failure without drm render node,
need enable vivante drm to fix dmabuf failure issue.Signed-off-by: Xianzhong
-
checked return value
Date:30th Oct,2020
Signed-off-by Ya Zhou -
Signed-off-by: Jiyu Yang
-
Signed-off-by: Jiyu Yang
-
Signed-off-by: Jiyu Yang
-
this also fix LF-2472
Signed-off-by: Jiyu Yang -
Signed-off-by: Jiyu Yang
-
Signed-off-by: Jiyu Yang
-
checked return value
Date:30th Oct,2020
Signed-off-by Ya Zhou -
Signed-off-by: Xianzhong
18 Nov, 2020
1 commit
-
For 88w8987/88w8997 wireless chips requires many module parameters
set by user and don't expect to be auto-loaded, remove the module
device table.Fixes: 04b2fa2cd5d3 ("MLK-25013 net: wireless: nxp: mxm_wifiex: upgrade to mxm5x16210 release")
Reviewed-by: yang.tian
Signed-off-by: Fugang Duan
17 Nov, 2020
7 commits
-
Enable CONFIG_UAPI_HEADER_TEST turns out the following build issues
HDRTEST usr/include/linux/mxc_asrc.h
In file included from :32:0:
./usr/include/linux/mxc_asrc.h:125:2: error: unknown type name ‘snd_pcm_format_t’
snd_pcm_format_t input_format;
^~~~~~~~~~~~~~~~
./usr/include/linux/mxc_asrc.h:126:2: error: unknown type name ‘snd_pcm_format_t’
snd_pcm_format_t output_format;
^~~~~~~~~~~~~~~~
./usr/include/linux/mxc_asrc.h:134:2: error: unknown type name ‘uint64_t’
uint64_t supported_in_format;
^~~~~~~~
./usr/include/linux/mxc_asrc.h:135:2: error: unknown type name ‘uint64_t’
uint64_t supported_out_format;But even add the missing/necessary header file, it finally failed too in the end:
HDRTEST usr/include/linux/mxc_asrc.h
In file included from ./usr/include/linux/mxc_asrc.h:23:0,
from :32:
./usr/include/sound/asound.h:459:18: error: field ‘trigger_tstamp’ has incomplete type
struct timespec trigger_tstamp; /* time when stream was started/stopped/paused */
^~~~~~~~~~~~~~
./usr/include/sound/asound.h:460:18: error: field ‘tstamp’ has incomplete type
struct timespec tstamp; /* reference timestamp */
^~~~~~
./usr/include/sound/asound.h:469:18: error: field ‘audio_tstamp’ has incomplete type
struct timespec audio_tstamp; /* sample counter, wall clock, PHC or on-demand sync'ed */
^~~~~~~~~~~~
./usr/include/sound/asound.h:470:18: error: field ‘driver_tstamp’ has incomplete type
struct timespec driver_tstamp; /* useful in case reference system tstamp is reported with delay */
^~~~~~~~~~~~~
./usr/include/sound/asound.h:472:37: error: invalid application of ‘sizeof’ to incomplete type ‘struct timespec’
unsigned char reserved[52-2*sizeof(struct timespec)]; /* must be filled with zero */
^~~~~~
./usr/include/sound/asound.h:479:18: error: field ‘tstamp’ has incomplete type
struct timespec tstamp; /* Timestamp */
^~~~~~
./usr/include/sound/asound.h:481:18: error: field ‘audio_tstamp’ has incomplete type
struct timespec audio_tstamp; /* from sample counter or wall clock */
^~~~~~~~~~~~
./usr/include/sound/asound.h:651:18: error: field ‘tstamp’ has incomplete type
struct timespec tstamp; /* Timestamp */
^~~~~~
./usr/include/sound/asound.h:763:18: error: field ‘tstamp’ has incomplete type
struct timespec tstamp; /* Timestamp - last update */
^~~~~~
./usr/include/sound/asound.h:813:18: error: field ‘tstamp’ has incomplete type
struct timespec tstamp;
^~~~~~
./usr/include/sound/asound.h:956:18: error: field ‘tstamp’ has incomplete type
struct timespec tstamp;
^~~~~~
./usr/include/sound/asound.h:957:36: error: invalid application of ‘sizeof’ to incomplete type ‘struct timespec’
unsigned char reserved[128-sizeof(struct timespec)];This is kernel known issue, just like other fixes as the following, put mxc_asrc.h into the skip-list:
header-test- += sound/asequencer.h
header-test- += sound/asoc.h
header-test- += sound/asound.h
header-test- += sound/compress_offload.h
header-test- += sound/emu10k1.h
header-test- += sound/sfnt_info.hSigned-off-by: Jason Liu
(cherry picked from commit 4018355af34d9b25ff7bd4d27960d72b8d304e30) -
Enable CONFIG_UAPI_HEADER_TEST turns on the following build issue
./usr/include/linux/ipu.h:167:2: error: unknown type name ‘u32’
u32 x;
^~~
./usr/include/linux/ipu.h:168:2: error: unknown type name ‘u32’
u32 y;
^~~
./usr/include/linux/ipu.h:173:2: error: unknown type name ‘u32’
u32 w;
^~~
./usr/include/linux/ipu.h:174:2: error: unknown type name ‘u32’
u32 h;
^~~
./usr/include/linux/ipu.h:179:2: error: unknown type name ‘u8’
u8 motion; /*see ipu_motion_sel*/The fix is to use the __u32/__u8 instead and remove the internal definiton
But it finally will result in another failure as the followings:
HDRTEST usr/include/linux/ipu.h
In file included from ./usr/include/linux/ipu.h:29:0,
from :32:
./usr/include/linux/videodev2.h:2353:20: error: field ‘timestamp’ has incomplete type
struct timespec timestamp;This is kernel known issue. The fix is as others by adding the ipu.h to skip-list:
header-test- += linux/v4l2-mediabus.h
header-test- += linux/v4l2-subdev.h
header-test- += linux/videodev2.hSigned-off-by: Jason Liu
(cherry picked from commit 24c36527086fff43595bde2822f8277138ebcb55) -
Enable CONFIG_UAPI_HEADER_TEST turns out the following build issues:
HDRTEST usr/include/linux/tsn.h
In file included from :32:0:
./usr/include/linux/tsn.h:384:2: error: unknown type name ‘__u8’
__u8 admin_state;
^~~~
./usr/include/linux/tsn.h:390:2: error: unknown type name ‘__u32’
__u32 hold_advance;
^~~~~
./usr/include/linux/tsn.h:397:2: error: unknown type name ‘__u32’
__u32 release_advance;
^~~~~
./usr/include/linux/tsn.h:403:2: error: unknown type name ‘__u8’
__u8 preemption_active;
^~~~
./usr/include/linux/tsn.h:412:2: error: unknown type name ‘__u8’
__u8 hold_request;
^~~~
./usr/include/linux/tsn.h:426:2: error: unknown type name ‘__u8’
__u8 delta_bw; /* percentage, 0~100 */
^~~~
./usr/include/linux/tsn.h:427:2: error: unknown type name ‘__u32’
__u32 idleslope;
^~~~~
./usr/include/linux/tsn.h:428:2: error: unknown type name ‘__s32’
__s32 sendslope;
^~~~~
./usr/include/linux/tsn.h:429:2: error: unknown type name ‘__u32’
__u32 maxframesize;
^~~~~The fix is to add the necessary header file into the tsn.h file
Signed-off-by: Jason Liu
(cherry picked from commit 328bf99d8b42aca38538b6122878c7a9da9234d0) -
Eanble CONFIG_UAPI_HEADER_TEST turns out the following build issues:
HDRTEST usr/include/linux/mxc_v4l2.h
In file included from :32:0:
./usr/include/linux/mxc_v4l2.h:52:2: error: unknown type name ‘uint32_t’
uint32_t u_offset;
^~~~~~~~
./usr/include/linux/mxc_v4l2.h:53:2: error: unknown type name ‘uint32_t’
uint32_t v_offset;
^~~~~~~~
./usr/include/linux/mxc_v4l2.h:57:2: error: unknown type name ‘__u32’
__u32 type; /* enum v4l2_buf_type */
^~~~~The fix is to add the necessary header file into mxc_v4l2.h
Signed-off-by: Jason Liu
(cherry picked from commit ca2be4571361d04aa291e0ed0fcb755feba8783a) -
Enable CONFIG_UAPI_HEADER_TEST turns out the following build issues:
HDRTEST usr/include/linux/mxc_dsp.h
In file included from :32:0:
./usr/include/linux/mxc_dsp.h:82:2: error: C++ style comments are not allowed in ISO C90
//UNIA_CHANNEL_MASK,
^
./usr/include/linux/mxc_dsp.h:82:2: error: (this will be reported only once per input file)The fix is to use C style comment insteading of C++
Signed-off-by: Jason Liu
(cherry picked from commit 4c016b8114b3db2bba7ec579c34e248766feac6e) -
Enable CONFIG_UAPI_HEADER_TEST or make headers_install result in build issues:
include/uapi/linux/fmd/Peripherals/fm_port_ioctls.h: leak CONFIG_COMPAT to user-space
include/uapi/linux/fmd/Peripherals/fm_pcd_ioctls.h: leak CONFIG_COMPAT to user-space
include/uapi/linux/fmd/Peripherals/fm_ioctls.h: leak CONFIG_COMPAT to user-spaceThe comments from DN fix of the fmd related failure is to add to the skip-list due to the
historial reasons from DN(Layerscape) drivers - to avoid break the legacy usersSigned-off-by: Jason Liu
(cherry picked from commit 569d6d3cc2a664de66ee980bd611687f4d785b4e) -
Upgrade to mxm5x16210 verison:
- Fixed WPA3 SAE pre-cert requirement where there is no assoc
request from DUT, if EX-AP sends the Auth confirm frame
immediately after Auth commit frameReviewed-by: yang.tian
Signed-off-by: Fugang Duan
(cherry picked from commit: 0a14a53a273aecb7dfb837cf90f15744e893e6bf)
16 Nov, 2020
7 commits
-
update the noc QoS setting for CPU & VPU on i.MX8MQ.
Signed-off-by: Bai Ping
Reviewed-by: Jian Li
(cherry picked from commit 45d2dcaecce6d83e5c4a7e9488c651a05b0f05ac) -
Config NOC to limit bandwidth to 4GB for both VPU
and CPU to avoid lcdif flickering only when lcdif is enabled.Signed-off-by: Anson Huang
Acked-by: Peng Fan
Reviewed-by: Bai Ping
(cherry picked from commit 8ab89ebeb94a423792bf588bdf2354c5960d8f13) -
Add widget to fix error prompt:
imx-cdnhdmi sound-hdmi: ASoC: DAPM unknown pin HDMI JackSigned-off-by: Shengjiu Wang
Reviewed-by: Peng Zhang -
The A53 900MHz opp is actually 896MHz when reading back from SCFW, so
use the 896MHz instead of 900MHz to make it accurate enough.Signed-off-by: Anson Huang
Reviewed-by: Jacky Bai
(cherry picked from commit 58be8bf3e6e8f51ed3a3b8f992baaa9eeaf3516d) -
Merge contiguous TCML/U regions into one to avoid load elf files which
has large sections failure.Reviewed-by: Ye Li
Signed-off-by: Peng Fan -
Correct LDO3/4OUT_MASK for pca9450 though that's not a real issue on
current i.mx8M family boards with pca9450 pmic since LDO3/4 used as
the default value and kernel never touch it.Signed-off-by: Robin Gong
Reviewed-by: Anson Huang -
When set_params return error, we should avoid to delete the component
in free function again. This may cause kernel crash.
Check cstream state to avoid such issue.Signed-off-by: Zhang Peng
13 Nov, 2020
2 commits
-
Adding "id" parameter in the ISP deivce nodes.
Signed-off-by: Oliver F. Brown
-
Set the V4L2_SUBDEV_FL_HAS_DEVNODE flag in mipi_csis_subdev_init.
Signed-off-by: Oliver F. Brown
