Let's root our Max Lumi

[Renate]

Faking VSYNC due to driver stall

I see that a lot in my logcat too and I don't have any problems.

Originally Posted by Grorom

Is there a launcher you suggest being good with black & white screens

I'm boring and don't use a lot of apps on my eReaders (and the ones I do, I wrote). This is my launcher, but it wouldn't suit you as the links are fixed and soon I'll be ready to delete Footsie.

[Grorom]

Okay, thanks !

The developer options "Show touches" shows the points when I use the pen, absolutely nothing with the fingers.
I've tried a last Magisk de-install + factory reset : nothing better.

So I'm resigned to send it to the Onyx support (in Poland) next week.

[metalkode]

I'm having the same issues as Grorom.

Installed the patched boot, and touch stopped working.
Reflashed original one and still not touch.

Something seems to mess up the touch drivers somehow.

[metalkode]

I might have found the problem.
It looks like the boot.img supplied in the OP corresponds to build 3.1_ad716f8 (Build 1834), however my max lumi runs 3.1_0631e26 (Build 2026) according to the update history.

It seems however, that build 2026 is not available for download, at least, I couldn't find it.

I'll try to flash the complete decrypted ROM and report back.

[metalkode]

Welp, managed to get hold of a boot.img from build 1904. (It was not a lot of fun, applying the 1904 patch to boot.img by hand)
Tried flashing it but no luck.
Guess it needs a complete firmware reflash...

Does anyone have a full device partition dump?
Otherwise I'll have to contact customer service

[yokos]

What a mess. I have added a big warning in #1. It all started when onyx didn't release any full firmware update files anymore.

updater-script MaxLumi:10/2021-03-31_19-14_3.1_0631e26

ui_print("Patching boot image...");
show_progress(0.100000, 10);
patch_partition("EMMC:/dev/block/bootdevice/by-name/boot:12522792:436ca212b3eff288a5c71bf72fa5e5be00d0e023",
                "EMMC:/dev/block/bootdevice/by-name/boot:12522792:e1380eeddffc778f949977747b8ac6db63e91821",
                package_extract_file("boot.img.p")) ||
    abort("E3008: Failed to apply patch to EMMC:/dev/block/bootdevice/by-name/boot:12522792:e1380eeddffc778f949977747b8ac6db63e91821");

[metalkode]

I did manage to find some interesting things in dmesg though:

During initialization:

[   31.560119] cyttsp5_i2c_adapter 4-0024: cyttsp5_hid_send_output_and_wait_: HID output cmd execution timed out
[   31.560144] cyttsp5_i2c_adapter 4-0024: cyttsp5_startup_: Error on start bootloader r=-62
[   31.560173] ONYX_TP: cyttsp5_i2c_adapter: cyttsp5_startup_:finish, rc: -62
[   31.560181] cyttsp5_i2c_adapter 4-0024: cyttsp5_startup_work_function: Fail queued startup r=-62
[   32.760596] cyttsp5_i2c_adapter 4-0024: cyttsp5_hid_send_output_and_wait_: HID output cmd execution timed out
[   32.760618] cyttsp5_i2c_adapter 4-0024: cyttsp5_watchdog_work: failed to access device in watchdog timer r=-62
[   32.760626] cyttsp5_i2c_adapter 4-0024: cyttsp5_queue_startup_: cyttsp5_startup queued
[   32.760672] ONYX_TP: cyttsp5_i2c_adapter: cyttsp5_startup_:start
[   32.870132] cyttsp5_i2c_adapter 4-0024: cyttsp5_xres: RESET CYTTSP gpio=66 r=0

And later on, this seems to be repeating:

[ 1349.982818] healthd: battery l=100 v=4323 t=26.0 h=2 st=5 c=1 fc=4255000 cc=0 chg=u
[ 1350.029363] __create_fence: fb0 vsync pending first update
[ 1350.029430] mdss_mdp_layer_handle_buffer_fences: unable to retrieve release fence
[ 1350.073908] mdss_mdp_cmd_epdc_done: invalid ctx
[ 1350.522093] <GLINK_PKT> err: glink_pkt_open: open failed on dev id:0 rc:-110
[ 1353.682212] ONYX: Pending Wakeup Sources: a800000.ssusb onyx_charging_lock
[ 1379.813764] ################ fb_blank():  caller = do_fb_ioctl+0x568/0x5f8, task:composer@2.1-se, blank=0
[ 1379.839180] mdss_mdp_overlay_on: reenabling vsync for fb0
[ 1379.840208] lm3630a_bank_update_status: led a already disabled!
[ 1379.840210] lm3630a_bank_update_status: led a already disabled!
[ 1379.840218] cyttsp5_i2c_adapter 4-0024: fb_notifier_callback: UNBLANK!
[ 1379.840219] ONYX_TP: cyttsp5_i2c_adapter: fb_notifier_callback:FB_ON, start cyttsp5_core_wake!
[ 1379.840237] ONYX_TP: cyttsp5_i2c_adapter: onyx_tp_power:power on success
[ 1379.950463] cyttsp5_i2c_adapter 4-0024: cyttsp5_xres: RESET CYTTSP gpio=66 r=0
[ 1383.761368] ONYX: Pending Wakeup Sources: PowerManager.SuspendLockout a800000.ssusb onyx_charging_lock
[ 1383.950269] cyttsp5_i2c_adapter 4-0024: cyttsp5_get_hid_descriptor_: HID get descriptor timed out
[ 1383.950274] cyttsp5_i2c_adapter 4-0024: _fast_startup: Error on getting HID descriptor r=-62
[ 1383.962107] ################ fb_blank():  END!

I wonder what the -62 return code is...

[metalkode]

Ah it looks like I got the numbers mixed up, shame I can't edit my posts...
These are the correct versions with the corresponding build numbers:

- 3.1_ad716f8 (Build 1834) <- The files of the OP correspond to this version
- 3.1_0631e26 (Build 1914) <- Latest version I was able to find, and patch boot.img from
- 3.1_a207066 (Build 2026) <- Version my e-reader runs, and of which I wasnt able to find any update files.

[yokos]

On 20th October the new Max Lumi 2 will be presented. I have seen it on reddit. It is sure that our Max Lumi will get soon a full 3.2 firmware update. It should fix the finger touch problems.

onyx-zeng on reddit:

Some highlights of 3.2 are

NeoReader, we improve document layout engine v2 including better epub/txt/doc/docx support. We also improve page turning performance. The full data sync for library and books, will be released after 3.2, hopefully in 3.3.

Note, we improved the sync stability greatly and fixed all known issues we found and user reported. we add pen shortcut, so you can switch among different pens easily.

add handwriting support for 3rd party apps like oneNote, the handwriting performance is almost the same as native note app.

New BOOXDrop upload/download content from/to PC.

New look and feel for 3.2

[metalkode]

So I tried binding and unbinding the driver for the touchscreen to see how it reacts, and this is the result:

[15227.410645] ONYX_TP: cyttsp5_i2c_adapter: cyttsp5_i2c_probe:start
[15227.411186] ONYX_TP: cyttsp5_i2c_adapter: onyx_tp_dt_parse:vdd: yes, vdd_ctrl: gpio:62, vcc: yes, vcc_ctrl: gpio61, reset: gpio66, irq: gpio67
[15227.411195] ONYX_TP: cyttsp5_i2c_adapter: cyttsp5_devtree_create_and_get_pdata:irq_gpio:67 rst_gpio:66, hid_desc_register:1 level_irq_udelay:1, vendor_id:0 product_id:0, flags:6 easy_wakeup_gesture:0
[15227.411210] ONYX_TP: cyttsp5_i2c_adapter: onyx_tp_hw_init:start
[15227.411272] ONYX_TP: cyttsp5_i2c_adapter: onyx_tp_power:power on success
[15227.411275] ONYX_TP: cyttsp5_i2c_adapter: onyx_tp_hw_init:success!
[15227.411282] cyttsp5_i2c_adapter 4-0024: cyttsp5_probe: Detect HW
[15227.520391] cyttsp5_i2c_adapter 4-0024: cyttsp5_xres: RESET CYTTSP gpio=66 r=0
[15227.634075] ONYX_TP: cyttsp5_i2c_adapter: cyttsp5_probe:HW detected
[15227.634089] ONYX_TP: cyttsp5_i2c_adapter: cyttsp5_probe:Setup watchdog timer
[15227.634737] ONYX_TP: cyttsp5_i2c_adapter: cyttsp5_probe:setup_irq_gpio success!
[15227.741452] cyttsp5_i2c_adapter 4-0024: cyttsp5_xres: RESET CYTTSP gpio=66 r=0
[15227.744226] ONYX_TP: cyttsp5_i2c_adapter: cyttsp5_startup_:start
[15231.741527] cyttsp5_i2c_adapter 4-0024: cyttsp5_get_hid_descriptor_: HID get descriptor timed out
[15231.741543] cyttsp5_i2c_adapter 4-0024: cyttsp5_startup_: Error on getting HID descriptor r=-62
[15231.852303] cyttsp5_i2c_adapter 4-0024: cyttsp5_xres: RESET CYTTSP gpio=66 r=0
[15233.931247] cyttsp5_i2c_adapter 4-0024: cyttsp5_hid_send_output_and_wait_: HID output cmd execution timed out
[15233.931263] cyttsp5_i2c_adapter 4-0024: cyttsp5_startup_: Error on start bootloader r=-62
[15234.050427] cyttsp5_i2c_adapter 4-0024: cyttsp5_xres: RESET CYTTSP gpio=66 r=0
[15236.121431] cyttsp5_i2c_adapter 4-0024: cyttsp5_hid_send_output_and_wait_: HID output cmd execution timed out
[15236.121448] cyttsp5_i2c_adapter 4-0024: cyttsp5_startup_: Error on start bootloader r=-62
[15236.231483] cyttsp5_i2c_adapter 4-0024: cyttsp5_xres: RESET CYTTSP gpio=66 r=0
[15238.311514] cyttsp5_i2c_adapter 4-0024: cyttsp5_hid_send_output_and_wait_: HID output cmd execution timed out
[15238.311528] cyttsp5_i2c_adapter 4-0024: cyttsp5_startup_: Error on start bootloader r=-62
[15238.311535] ONYX_TP: cyttsp5_i2c_adapter: cyttsp5_startup_:finish, rc: -62
[15238.311545] ONYX_TP: cyttsp5_i2c_adapter: cyttsp5_mt_probe:start
[15238.311562] cyttsp5_i2c_adapter 4-0024: cyttsp5_mt_probe: Fail get sysinfo pointer from core p=          (null)
[15238.311566] ONYX_TP: cyttsp5_i2c_adapter: cyttsp5_mt_probe:finish
[15238.311628] cyttsp5_i2c_adapter 4-0024: cyttsp5_device_access_probe: Fail get sysinfo pointer from core p=          (null)
[15238.311657] ONYX_TP: cyttsp5_i2c_adapter: onyx_cyttsp5_tp_add_update_handle:cyttsp5 fw update handle added!
[15238.311662] cyttsp5_i2c_adapter 4-0024: cyttsp5_loader_probe: Successful probe 4-0024
[15238.311668] ONYX_TP: cyttsp5_i2c_adapter: cyttsp5_probe:cyttsp5 probe successed!
[15238.311673] ONYX_TP: cyttsp5_i2c_adapter: cyttsp5_i2c_probe:finish, rc: 0
[15238.311993] cyttsp5_i2c_adapter 4-0024: cyttsp5_parse_cmcp_threshold_builtin: Fail get sysinfo pointer from core
[15238.312001] cyttsp5_i2c_adapter 4-0024: cyttsp5_fw_and_config_upgrade: Fail get sysinfo pointer from core
[15239.510114] cyttsp5_i2c_adapter 4-0024: cyttsp5_hid_send_output_and_wait_: HID output cmd execution timed out
[15239.510128] cyttsp5_i2c_adapter 4-0024: cyttsp5_watchdog_work: failed to access device in watchdog timer r=-62
[15239.510134] cyttsp5_i2c_adapter 4-0024: cyttsp5_queue_startup_: cyttsp5_startup queued
[15239.510141] ONYX_TP: cyttsp5_i2c_adapter: cyttsp5_startup_:start

Looks like the startup is fine until the HID descriptor needs to be retrieved. I also wonder what the ONYX_TP is, and where it gets its settings from. It must be from system.img since it only stopped working after flashing a different kernel.

[metalkode]

I managed to extract system.img and vendor.img from the 1834 patch.
Flashed them, but still no result...

Version does say build 1834 though, so everything is on the correct version...
I'm starting to lose hope in solving this...

[metalkode]

I'm thinking there was a hardware revision with different touch drivers for 2026.
Which could also explain it not showing up on the update screen for the other devices. Shame there is no way to recover that update file with the correct boot.img

One thing I haven't tried yet is flashing the persist.img. As far as I know it shouldn't contain data about the touchscreen though... I'll try unpacking it and comparing it first.

EDIT: Yeah, its just a file with the number 1 in it.

[Renate]

Originally Posted by metalkode

I wonder what the -62 return code is...

62 is ETIME, timeout.
The Cypress TrueTouch controller is being reset, then the first thing they do is ask for the HID report descriptor and they are not getting it. This is basically a total failure.
Using HID reports on an I²C device is something new to me. I'm very used to it in USB and Bluetooth, just never seen it this way.

This could just be that you've got software for the completely wrong hardware.
If it were me, I might just scan I²C bus #4 to see if there is anything responding at address 24.

Just out of curiosity:

# ls -l /sys/bus/i2c/devices

FWIW, this is my Poke3:

4-001f  i2c_camera
4-0038  fts_ts      FocalTech touch Screen
6-0009  wacom
6-0038  lm3630a_bl  TI LED driver LM3630A
8-000b  tc358762    Toshiba display interface bridge
8-0068  tps6518x    TI power management for Eink TPS65185, TPS65186

Hmm, does the Poke3 support HDMI over USB Type C?

[Renate]

Hmm, looking into this more I see that there are drivers for things that aren't even supposed to be in there. I knocked out a little utility for scanning I²C and correlating that with whether there is stuff in sysfs for it.
My Poke3 again:

Poke3:/data/local/tmp # ./i2cscan
/dev/i2c-2
/dev/i2c-4
  04  exists
  05  exists
  06  exists
  07  exists
  1f          qcom,camera  i2c_camera
  38  exists  focaltech (FocalTech touch screen controller)  fts_ts
/dev/i2c-6
  04  exists
  05  exists
  06  exists
  07  exists
  09          wacom (Wacom digitizer)
  38  exists  lm3630a (TI LED controller)  lm3630a_bl
/dev/i2c-8
  04  exists
  05  exists
  06  exists
  07  exists
  0b  exists  tc358762 (Toshiba display interface bridge)  tc358762
  68          tps6518x (TI Eink power management)  tps6518x

So we can see that neither a camera, a Wacom or a TPS exists although the drivers are there. The stuff on addresses 4-7 are probably bus expanders or something.

If you want to try this (please), just ADB it somewhere (I like /data/local/tmp), chmod 755 it and run it. This is not a zip file (just named that for upload). Just rename it i2cscan, don't unzip it.

Edit: Minor beauty enhancements to i2cscan, same functionality.

[metalkode]

Originally Posted by Renate

Hmm, looking into this more I see that there are drivers for things that aren't even supposed to be in there. I knocked out a little utility for scanning I²C and correlating that with whether there is stuff in sysfs for it.
My Poke3 again:

Poke3:/data/local/tmp # ./i2cscan
/dev/i2c-2
/dev/i2c-4
  04  exists
  05  exists
  06  exists
  07  exists
  1f          qcom,camera  i2c_camera
  38  exists  focaltech (FocalTech touch screen controller)  fts_ts
/dev/i2c-6
  04  exists
  05  exists
  06  exists
  07  exists
  09          wacom (Wacom digitizer)
  38  exists  lm3630a (TI LED controller)  lm3630a_bl
/dev/i2c-8
  04  exists
  05  exists
  06  exists
  07  exists
  0b  exists  tc358762 (Toshiba display interface bridge)  tc358762
  68          tps6518x (TI Eink power management)  tps6518x

So we can see that neither a camera, a Wacom or a TPS exists although the drivers are there. The stuff on addresses 4-7 are probably bus expanders or something.

If you want to try this (please), just ADB it somewhere (I like /data/local/tmp), chmod 755 it and run it. This is not a zip file (just named that for upload). Just rename it i2cscan, don't unzip it.

This is the result:

/dev/i2c-2
  04  exists
  05  exists
  06  exists
  07  exists
  60  exists
/dev/i2c-4
  04  exists
  05  exists
  06  exists
  07  exists
  1f          qcom,camera  i2c_camera
  24  exists  tsc  cyttsp5_i2c_adapter
  34  exists  aw881xx_smartpa  aw881xx_smartpa
  35  exists  aw881xx_smartpa  aw881xx_smartpa
/dev/i2c-6
  00  exists
  04  exists
  05  exists
  06  exists
  07  exists
  09  exists  wacom (Wacom digitizer)  wacom_ts
  38  exists  lm3630a (TI LED controller)  lm3630a_bl
/dev/i2c-8
  04  exists
  05  exists
  06  exists
  07  exists
  0b  exists  tc358762 (Toshiba display interface bridge)  tc358762
  48  exists  max17135  max17135

IIRC the settings for which drivers should load on I²C are set in the devicetree in boot.img right?

Sorry for my late replies, my posts have to be checked by a mod beforehand...