Age | Commit message (Collapse) | Author |
|
am: 600cbf2580
Change-Id: I2c7bb90457e9e15e20002522917d41731a97c30b
|
|
Change bus interface from SPI to I2C will build break.
Bug: 142166340
Change-Id: I1be6c183005cc163bb131dac82460b271b8261a5
Signed-off-by: davidycchen <davidycchen@google.com>
|
|
data new API to know the status of the Info block
am: eb9dce5460
Change-Id: I359f0a06fe76091d0c76883a9480ff7dc45137a8
|
|
Option to read Filter raw data
new API to know the status of the Info block
Signed-off-by: Sivasankari <Sivasankari.muthusamy@st.com>
Change-Id: Ibc098031740716c2b5314a5faea6139ca5f7344f
|
|
am: 6873e1f341
Change-Id: I7999e2fe086778d52dfa801bb23cf02d4fe21144
|
|
Print the failed sensor routing node in the FATP ITO test
Bug: 141721123
Change-Id: I19f9d5908029bc01a1cafe3e367e567457a70922
Signed-off-by: davidycchen <davidycchen@google.com>
|
|
am: 7f16a8546f
Change-Id: Iaa124362bf9514fa607d31d8a75b8e7310c3f938
|
|
Bug: 141372918
Change-Id: I8826607f1aaf5b54c05f9fecfa96b93acd1b5847
Signed-off-by: Chenglu Lin <chenglulin@google.com>
|
|
Add heatmap config for no heatmap driver situation.
Bug: 138816997
Change-Id: I6f018aefd6751cd57f271c782f165e947bef91f9
Signed-off-by: davidycchen <davidycchen@google.com>
|
|
dsi_panel_read_vendor_extinfo is used in c2f2 dsi_panel driver.
Skip this function first for bring up.
Will revert this commit after dsi_panel driver ready.
Bug: 138816997
Change-Id: Ib5dec6e58622fcfac23ab036bbb99a1362766e29
Signed-off-by: davidycchen <davidycchen@google.com>
|
|
Add CONFIG_DRM flag in touch driver.
Undefine CONFIG_DRM first for bring up stage.
Bug: 138816997
Change-Id: Icc3b35317018a59fdae2c9701c03306335f28ccb
Signed-off-by: davidycchen <davidycchen@google.com>
|
|
Bug: 138816997
Bug: 138817231
Change-Id: I29ea5750e4cd063c339730f847fded560a97685c
Signed-off-by: davidycchen <davidycchen@google.com>
|
|
Bug: 138816997
Change-Id: I043714e646112b7e71c657755496a69e71b55065
|
|
Added new production test item for ITO data evaluation
Signed-off-by: Sivasankari <Sivasankari.muthusamy@st.com>
Change-Id: If3b1400c2a1bdf1349849ab5ff486acf9bfa4b59
|
|
update in production test items
Modification in FPI/PI initialization
SPI4 setting change
Signed-off-by: Sivasankari <Sivasankari.muthusamy@st.com>
Change-Id: Id0646c576cefd1c3e444bdce27bcb2ee0ebf120a
|
|
Added Support for B5/S5 projects, update in the Dchip id, Change in the DMA chunk size
Added Support for Flash organization change
FW API update for Echo Status event, firmware Event truncated to 4 bytes,
Signed-off-by: Sivasankari <Sivasankari.muthusamy@st.com>
Change-Id: I50909b4966b6190f6fd2549564276902fe3bf131
|
|
Context: With the potential new API change to timestamp, it isn't
necessary anymore to use the input_event call for MSC_TIMESTAMP.
Bug: 139785163
Bug: 119840121
Change-Id: I3bf5f0e6deb7be032155544b30d2994de124ae1e
Signed-off-by: atifniyaz@google.com
|
|
Bug: 119840121
Bug: 139785163
Change-Id: I98d073c72b2bd52563da9447679afa807fdd0a3e
Signed-off-by: atifniyaz@google.com
|
|
There is an unconditional flush of the suspend/resume work queue in the
screen state callback from the display driver. There is no need to
immediately flush the queue since any pending suspend/resume work will
be canceled or completed before the next suspend/resume is queued.
Bug: 139785163
Bug: 137754842
Change-Id: Ibe26f057c5ff942b975710af43f40bf104f97917
Signed-off-by: Steve Pfetsch <spfetsch@google.com>
|
|
- The mananer in which handled nonzero rows and zero columns were
incompatible, resulting field crash.
- There was no handling of malloc failures.
Bug: 138535981
Test: Run Dumpstate commands.
Change-Id: Iaf3fbf9377ab9da519f759015b4493f0e35e6836
Signed-off-by: Harpreet "Eli" Sangha <eliptus@google.com>
|
|
Bug: 136781155
Change-Id: Ia1c25bc8a24dd4f94b9925ed6f25f37af3838b69
Signed-off-by: Super Liu <supercjliu@google.com>
|
|
The mutex behavior made it possible for the one client with access to
the procfs node (/proc/fts/driver_test) to fail read/write/lseek/release
if another client is attempting to open the node. Even worse, if the
client cannot acquire the lock during release, it will fail to clean-up
its ownership of the node and the node will become inaccessible.
Instead, rely on the initial lock in fts_driver_test_open to guarantee
there will only be one client using the other entry points and therefore
they do not need locking.
Bug: 136024296
Change-Id: I63a2c2addf54a9570a5677e64b0002d40e623d40
Signed-off-by: Steve Pfetsch <spfetsch@google.com>
|
|
The line that computed the length of the golden mutual raw data
formatted for output over procfs was longer than necessary. Shorten to
the correct length.
Bug: 136021057
Change-Id: I3bbc015607db1adbd86561159312f4aa479dde60
Signed-off-by: Steve Pfetsch <spfetsch@google.com>
|
|
Demoted noise messages to pr_debug. In order to catch future undecoded
status events, explicitly indicate "unknown" in default status handling.
Bug: 133451980
Change-Id: Ib396446b2278300378768aad730fa1ad8e4aaa5f
Signed-off-by: Steve Pfetsch <spfetsch@google.com>
|
|
Merge List:
* 6183607 input: touchscreen: stm: Driver version 5.2.16.8
Bug: 133451980
Change-Id: Ia7d228e0db008b76fd1f701b95f89ef42746e032
Signed-off-by: Steve Pfetsch <spfetsch@google.com>
|
|
In some cases, the flushFIFO command waits for 5s for completion,
times out, then repeats the command and wait two more times. This can
result in a total of 15s delay at a time when the touch IC appears to be
unresponsive during suspend, so drastically shorten the echo delay to
50ms per attempt. This does not solve the underlying problem but should
improve the user experience when it does occur.
Bug: 134638807
Change-Id: I4d58020170009dd1a241f73d19047ac7e1786400
Signed-off-by: Steve Pfetsch <spfetsch@google.com>
|
|
Somehow there is the corner case will fill fifo with coordinate events
by palm or intentional trick. Therefore, add fts_system_reset() to back
touch IC original state before flushFIFO() will eliminate this error.
Error logs:
[17304.432940] [ FTS ] fts_mode_handler: Sense OFF!
[17304.433165] [ FTS ] writeSysCmd: Command = A4 00 01
[17304.433172] [ FTS ] writeSysCmd: Writing Sys command...
[17304.442835] [ FTS ] READ EVENT = 23 11 00 D3 68 2A A4 00
[17304.462823] [ FTS ] READ EVENT = 27 11 FB D2 67 35 49 01
[17304.463047] [ FTS ] READ EVENT = 43 09 00 01 0B 00 00 00
(repeated...)
[17309.422950] [ FTS ] READ EVENT = 27 11 CA 82 5F 31 79 00
[17309.432665] [ FTS ] pollForEvent: ERROR 80000007
[17309.432691] [ FTS ] checkEcho: Echo Event not found! ERROR 80000007
[17309.435183] [ FTS ] READ EVENT = 27 11 CE E2 63 36 79 00
[17309.442940] [ FTS ] READ EVENT = 27 11 D7 02 68 36 7A 00
[17309.452998] [ FTS ] READ EVENT = 27 11 E1 92 6B 36 7A 00
(repeated...)
[17314.403044] [ FTS ] READ EVENT = 27 01 18 73 5C 43 8F 00
[17314.413038] [ FTS ] READ EVENT = 27 01 08 03 5A 45 8F 00
[17314.423336] [ FTS ] READ EVENT = 27 01 E0 F2 56 3F 8E 00
[17314.433094] [ FTS ] pollForEvent: ERROR 80000007
[17314.433120] [ FTS ] checkEcho: Echo Event not found! ERROR 80000007
[17314.435619] [ FTS ] READ EVENT = 27 01 9B 12 53 38 8E 01
[17314.436020] [ FTS ] READ EVENT = 27 01 45 D2 4E 31 8D 00
[17314.446245] [ FTS ] READ EVENT = 27 01 EA A1 4A 27 2B 00
(repeated...)
[17319.532793] [ FTS ] READ EVENT = 27 01 E9 41 4A 40 9C 00
[17319.546992] [ FTS ] READ EVENT = 2B 01 20 A2 4D 5D 02 01
[17319.547173] [ FTS ] READ EVENT = 2B 01 44 E2 4F 6C 02 00
[17319.555906] [ FTS ] pollForEvent: ERROR 80000007
[17319.555917] [ FTS ] checkEcho: Echo Event not found! ERROR 80000007
[17319.557926] [ FTS ] fts_writeFwCmd: check echo ERROR 800E0007
[17319.557934] [ FTS ] writeSysCmd: ERROR 800E0007
[17319.557941] [ FTS ] flushFIFO: ERROR 800E0007
Bug: 134185999
Change-Id: Ide67ce1f5f360d72437ba07f983ba71513f309bd
Signed-off-by: Super Liu <supercjliu@google.com>
|
|
flushFIFO() in suspend work use pollForEvent() with msleep to
complete. Therefore, keep stay awake to avoid device suspend and
additional check senser_sleep for pm suspend.
Bug: 134046469
Change-Id: Ie87db22530063e193ee9094ae7876338f2a1689f
Signed-off-by: Super Liu <supercjliu@google.com>
|
|
Kernel log info updated to decode the events
Signed-off-by: Sivasankari <Sivasankari.muthusamy@st.com>
Change-Id: I75242527a6affa44f07e838b790c45ae3a090b9d
|
|
Change-Id: I76e7561137cad7df9169d9912eb9ec5ed025f942
Signed-off-by: Steve Pfetsch <spfetsch@google.com>
|
|
fts_suspend_work runs while holding the bus_mutex, which prevents the
interrupt handler from running and using the bus as the touch driver
suspends. Once fts_suspend_work disables touch sensing, the touch IC
generates a status event and pulls down its INT line, but the driver
disables interrupt handling and does not have an opportunity to handle
the event before suspending. Since the event is expected and harmless,
flush the touch IC's FIFO after interrupts are disabled.
Bug: 128865249
Change-Id: Ib2a14cfc407793d0989bbef0068f389945493386
Signed-off-by: Steve Pfetsch <spfetsch@google.com>
|
|
Normally, touch driver use display notifier(MSM_DRM_EVENT_BLANK
and MSM_DRM_EARLY_EVENT_BLANK) with bus reference mask(*) to trigger
suspend and resume.
Somehow system will enter deep sleep without MSM_DRM_BLANK_POWERDOWN
and MSM_DRM_BLANK_LP to suspend touch. Therefore, register pm
suspend/resume to correct report current touch driver state.
* Bits for the bus reference mask
enum {
FTS_BUS_REF_SCREEN_ON = 0x01,
FTS_BUS_REF_IRQ = 0x02,
FTS_BUS_REF_FW_UPDATE = 0x04,
FTS_BUS_REF_SYSFS = 0x08,
FTS_BUS_REF_FORCE_ACTIVE = 0x10
}
Bug: 131944527
Change-Id: I3b673536a787a71976d8f8a13e10720c18b1908f
Signed-off-by: Super Liu <supercjliu@google.com>
|
|
Bug: 132609975
Change-Id: Id5ba3fc5daa0547a28264c38dce17e1b480dfbfc
Signed-off-by: Super Liu <supercjliu@google.com>
|
|
into android-msm-floral-4.14
|
|
pollForEvent is designed to consume events from the touch IC until a
desired event is received, but it incorrectly waits between all read
operations regardless of how many events are queued in the FIFO.
Instead, read continuosly while the touch IC is responsive and it is
not reporting that the FIFO is empty.
Bug: 131922554
Change-Id: Id1cfc83ab2bec9dafa6460d0d142b52fb353c838
Signed-off-by: Steve Pfetsch <spfetsch@google.com>
|
|
Display refresh rate is now switching dynamically, causing the touch
driver to print these three lines frequently:
[ FTS ] Refresh rate changed to 60 Hz.
[ FTS ] fts_screen_state_chg_callback: UNBLANK
[ FTS ] fts_set_bus_ref: reference is unexpectedly set: mask=0x0001,
ref=0x0001, enable=1.
Bug: 129716641
Bug: 131714407
Change-Id: I52b8a8d2f4d83bfa174b0bdf14b0bac316b675b0
Signed-off-by: Steve Pfetsch <spfetsch@google.com>
|
|
Merge List
* 538f0ded input: touchscreen: stm: Driver version 5.2.16.7
* c4619736 input: touchscreen: stm: Driver version 5.2.16.6
* d1f0ddf9 Merge android-msm-floral-4.14 into partner-drops-c2f2
Change-Id: I5097dcc2e379958ca81f55989178dc48d3e17ed6
Signed-off-by: Super Liu <supercjliu@google.com>
|
|
fixed missing AFE version check in case of Panel Cfg area CRC error
Signed-off-by: Sivasankari Muthusami <sivasankari.muthusamy@st.com>
Change-Id: I138a30217bfde4e2f75ee45694bd502f089304ca
|
|
Kernel log for status events are updated in file fts.c
Signed-off-by: Sivasankari Muthusamy <sivasankari.muthusamy@st.com>
Change-Id: I9a55b23e1e63823297aa7ce5acc7cd301b666692
|
|
Bug: 130922273
Test:
Check below test items with kasan built.
BUILD_CONFIG=private/msm-google/build.config.kasan ./build/build.sh
- adb bugreport
- echo -n 23 00 > /proc/fts/driver_test ; cat /proc/fts/driver_test
- echo 23 00 > /proc/fts/driver_test ; cat /proc/fts/driver_test
- echo -n 23 > /proc/fts/driver_test ; cat /proc/fts/driver_test
- echo 23 > /proc/fts/driver_test ; cat /proc/fts/driver_test
- echo -n 08 stm_fts_production_limits_c2.csv >
/proc/fts/driver_test ; cat /proc/fts/driver_test
- echo 08 stm_fts_production_limits_c2.csv > /proc/fts/driver_test
; cat /proc/fts/driver_test
- echo -n 08 > /proc/fts/driver_test ; cat /proc/fts/driver_test
- echo 08 > /proc/fts/driver_test ; cat /proc/fts/driver_test
- echo -n 08 > /proc/fts/driver_test ; cat /proc/fts/driver_test
- echo 09 > /proc/fts/driver_test ; cat /proc/fts/driver_test
Change-Id: I1d688cadbe04768a9c04e07ef92ca776d0159834
Signed-off-by: Super Liu <supercjliu@google.com>
|
|
Change-Id: Ie5f3572141e0cc76c5bba645e4af8bb9f5463efb
Signed-off-by: Steve Pfetsch <spfetsch@google.com>
|
|
Stimpad status events were hitting the default case in the
fts_status_event_handler() and printed using error verbose
as an unhandled status event. This change adds a new case to
handle the stimpad event id and print them using info verbose.
Bug: 129855432
Change-Id: I70c0e5434398384fca39caf48ef6d00a3266b0e3
Signed-off-by: Biswajit Dash <bisdash@google.com>
|
|
Bug: 129723288
Test:
C2 FW change between v9 and v11.
F2 FW change between v13.31 and v14.3
adb shell "echo 16 > /sys/class/spi_master/spi1/spi1.0/stm_fts_cmd ;
cat /sys/class/spi_master/spi1/spi1.0/stm_fts_cmd"
adb shell "echo 33 12 > /proc/fts/driver_test ; cat /proc/fts/driver_test"
adb shell "echo 33 13 > /proc/fts/driver_test ; cat /proc/fts/driver_test"
Change-Id: I759a315b614de9795e3d47091d0e897c356f4f54
Signed-off-by: Harpreet "Eli" Sangha <eliptus@google.com>
Signed-off-by: Super Liu <supercjliu@google.com>
|
|
This is only used for dogfood debugging and factory validation.
Bug: 130283593
Change-Id: I3da83f80216ebca47195fc12f157b587e700d673
Signed-off-by: Super Liu <supercjliu@google.com>
|
|
Enabling the readout for production timestamp value and
Self compensation Ix0 values
Checkecho() function is updated to choose different
timeouts for FPI and other events and those timeout
values are updated accordingly in fts_time.h
Change-Id: I4201fdb8e0dee84e222ca392443c661ef6e24254
Signed-off-by: Sivasankari Muthusamy <sivasankari.muthusamy@st.com>
|
|
android-msm-floral-4.14
|
|
Bug: 120141999
Bug: 120141031
Bug: 120279144
Change-Id: Ic73b97bd66b9edc76c1624ed3c43e56fbceed537
Signed-off-by: Super Liu <supercjliu@google.com>
|
|
Added support in fts proc fs to dump Golden Mutual Raw data.
'echo 34 > /proc/fts/driver_test' will prepare the data in driver
'cat /proc/fts/driver_test' will dump the data on terminal.
Bug:128437605
Change-Id: Id95cb3c374cf65aa9d9043c9421f64b655eba857
Signed-off-by: Biswajit Dash <bisdash@google.com>
|
|
For factory touch run-in test(ITO), driver needs I/O read and write
by SPI bus after screen-off. Therefore, when force touch active
is triggered by sysfs, keep wake-up to avoid into deep suspension for
later SPI communication.
Bug: 123489319
Change-Id: Iac6afea5b89c262de2280f7abbd26f007cb9a964
Signed-off-by: Super Liu <supercjliu@google.com>
|
|
In current design, the motion filter will be changed dynamically
for heatmap feature. But, for robot test validation, it needs to
fall back to FW default motion filter without change.
Bug: 121266276
Change-Id: I5c57a3aa806c60f2b5057bed0c6a97bd34fe3b47
Signed-off-by: Super Liu <supercjliu@google.com>
|