search

cyberman54 / ESP32-Paxcounter

Wifi & BLE driven passenger flow metering with cheap ESP32 boards
https://cyberman54.github.io/ESP32-Paxcounter/
Other
1.73k stars 405 forks source link

Op code 88 (set time) will crash on TGGO T3 V1.6 // halfile = ttgov21new.h #781

Closed hansju-11 closed 3 years ago

hansju-11 commented 3 years ago

first try

[I][reset.cpp:125] enter_deepsleep(): Waiting until LMIC is idle... [I][lorawan.cpp:449] myRxCallback(): Received 5 byte(s) of payload on port 2 [I][rcommand.cpp:379] set_time(): Remote command: set time to 1617306575 [I][timekeeper.cpp:100] setMyTime(): [71.542] UTC time: 1617306575.000 sec Guru Meditation Error: Core 1 panic'ed (LoadProhibited). Exception was unhandled. Core 1 register dump: PC : 0x4009166f PS : 0x00060b30 A0 : 0x800814dc A1 : 0x3ffe6770 A2 : 0x00000000 A3 : 0x0000003f A4 : 0x00000000 A5 : 0x00000000 A6 : 0x3ffc7068 A7 : 0x4051e2b0 A8 : 0x800eae55 A9 : 0x3ffe6710 A10 : 0x3ffbb350 A11 : 0x40088edc A12 : 0x3ffba6b8 A13 : 0x00000000 A14 : 0x00000000 A15 : 0xffffffb3 SAR : 0x00000004 EXCCAUSE: 0x0000001c EXCVADDR: 0x00000000 LBEG : 0x400029ac LEND : 0x400029cb LCOUNT : 0x00000000

ELF file SHA256: 0000000000000000

Backtrace: 0x4009166f:0x3ffe6770 0x400814d9:0x3ffe6790 0x400d6825:0x3ffe67c0 0x400d6b15:0x3ffe67e0 0x400d6bd6:0x3ffe6820 0x40092d22:0x3ffe6850

Rebooting... ets Jun 8 2016 00:22:57

rst:0xc (SW_CPU_RESET),boot:0x13 (SPI_FAST_FLASH_BOOT)

second try

[I][lorawan.cpp:449] myRxCallback(): Received 5 byte(s) of payload on port 2 [I][rcommand.cpp:379] set_time(): Remote command: set time to 1617306650 [I][timekeeper.cpp:100] setMyTime(): [61.234] UTC time: 1617306650.000 sec Guru Meditation Error: Core 1 panic'ed (LoadProhibited). Exception was unhandled. Core 1 register dump: PC : 0x4009166f PS : 0x00060930 A0 : 0x800814dc A1 : 0x3ffe6770 A2 : 0x00000000 A3 : 0x0000003f A4 : 0x00000000 A5 : 0x00000000 A6 : 0x3ffc7068 A7 : 0x404e9df3 A8 : 0x800eae55 A9 : 0x3ffe6710 A10 : 0x3ffbb38c A11 : 0x40088edc A12 : 0x3ffba71c A13 : 0x00000000 A14 : 0x00000000 A15 : 0xffffffb3 SAR : 0x00000004 EXCCAUSE: 0x0000001c EXCVADDR: 0x00000000 LBEG : 0x400029ac LEND : 0x400029cb LCOUNT : 0x00000000

ELF file SHA256: 0000000000000000

Backtrace: 0x4009166f:0x3ffe6770 0x400814d9:0x3ffe6790 0x400d6825:0x3ffe67c0 0x400d6b15:0x3ffe67e0 0x400d6bd6:0x3ffe6820 0x40092d22:0x3ffe6850

Rebooting... ets Jun 8 2016 00:22:57

rst:0xc (SW_CPU_RESET),boot:0x13 (SPI_FAST_FLASH_BOOT) configsip: 188777542, SPIWP:0xee clk_drv:0x00,q_drv:0x00,d_drv:0x00,cs0_drv:0x00,hd_drv:0x00,wp_drv:0x00 mode:DIO, clock div:2 load:0x3fff0018,len:4 load:0x3fff001c,len:1044 load:0x40078000,len:10124 load:0x40080400,len:5828 entry 0x400806a8 [I][configmanager.cpp:102] loadConfig(): Loading device configuration from NVRAM... [I][configmanager.cpp:127] loadConfig(): Runtime configuration v2.4.0 loaded [I][main.cpp:127] setup(): Starting paxcounter_f56af210 v2.4.0 (runmode=4 / restarts=3)

cyberman54 commented 3 years ago

please increase version number in platformio.ini (this triggers clearing NVRAM values during first start). Currently i cannot reproduce this issue.

cyberman54 commented 3 years ago

This may happen if the device is going to sleep, while it receives the rcommand. Need to be checked. I can't reproduce this issue with a device that does not go to sleep.

cyberman54 commented 3 years ago

This should be fixed now in code in development branch: https://github.com/cyberman54/ESP32-Paxcounter/commit/d098d6ada37931ac3625ebc58c634345a1b9d96e @hansju-11 Can you please test and close this issue, if this works for you? Thanks.

cyberman54 commented 3 years ago

PS: time-of-day, once set by rcommand 0x88 or by other timesource (LORA, GPS, RTC) is now preserved during deep sleep.

hansju-11 commented 3 years ago

I change the version to 3.4.0 but still the same problem.

[I][lorawan.cpp:449] myRxCallback(): Received 5 byte(s) of payload on port 2 [D][lorawan.cpp:412] myEventCallback(): TXCOMPLETE [I][rcommand.cpp:379] set_time(): Remote command: set time to 1617813000 [I][timekeeper.cpp:101] setMyTime(): [131.553] UTC time: 1617813000.000 sec Guru Meditation Error: Core 1 panic'ed (LoadProhibited). Exception was unhandled. Core 1 register dump: PC : 0x4009171f PS : 0x00060f30 A0 : 0x8008150d A1 : 0x3ffef7b0 A2 : 0x00000000 A3 : 0x0000003f A4 : 0x00000000 A5 : 0x00000000 A6 : 0x3ffc70d8 A7 : 0x406071b2 A8 : 0x800eb429 A9 : 0x3ffef750 A10 : 0x3ffcd7c8 A11 : 0x40088f8c A12 : 0x3ffba6b8 A13 : 0x00000000 A14 : 0x00000000 A15 : 0xffffffb2 SAR : 0x00000004 EXCCAUSE: 0x0000001c EXCVADDR: 0x00000000 LBEG : 0x400029ac LEND : 0x400029cb LCOUNT : 0x00000000

ELF file SHA256: 0000000000000000

Backtrace: 0x4009171f:0x3ffef7b0 0x4008150a:0x3ffef7d0 0x400d6a89:0x3ffef800 0x400d6d79:0x3ffef820 0x400d6e3a:0x3ffef860 0x40092dd2:0x3ffef890

Rebooting... ets Jun 8 2016 00:22:57

rst:0xc (SW_CPU_RESET),boot:0x13 (SPI_FAST_FLASH_BOOT) configsip: 188777542, SPIWP:0xee clk_drv:0x00,q_drv:0x00,d_drv:0x00,cs0_drv:0x00,hd_drv:0x00,wp_drv:0x00 mode:DIO, clock div:2 load:0x3fff0018,len:4 load:0x3fff001c,len:1044 load:0x40078000,len:10124 load:0x40080400,len:5828 entry 0x400806a8 [I][configmanager.cpp:102] loadConfig(): Loading device configuration from NVRAM... [I][configmanager.cpp:127] loadConfig(): Runtime configuration v3.4.0 loaded [I][main.cpp:127] setup(): Starting paxcounter_f56af210 v3.4.0 (runmode=1 / restarts=1)

I will do some deeper debug tomorow

cyberman54 commented 3 years ago

Did you use the current code of development branch?

Please enable esp32 backtrace in platformio logging, this will show the lines of code which cause the crash.

Please show your paxcounter.conf and the used hal file, thx.

hansju-11 commented 3 years ago

Hello yes I just uploaded it 1 hour ago.

for me the error comes from timekeeper.cpp line 112 (timerWrite ...)

if (!defined GPS_INT && !defined RTC_INT)

timerWrite(ppsIRQ, 0); // reset pps timer
CLOCKIRQ();            // fire clock pps, this advances time 1 sec

endif

if I disable the function in esp32-hal-timer.c line 106

void IRAM_ATTR timerWrite(hw_timer_t *timer, uint64_t val){

ESP_LOGD(TAG, "hjb error 1 esp32-hal-timer.c 108");

// timer->dev->load_high = (uint32_t) (val >> 32); // timer->dev->load_low = (uint32_t) (val); // timer->dev->reload = 1; ESP_LOGD(TAG, "hjb error 1 esp32-hal-timer.c 112"); }

I did not get an error.

I will activate the esp backtrace to see if I get more informations for you.

It looks like ppsIRQ pointer is NULL

hansju-11 commented 3 years ago

here are my paxcounter.conf and the hal file I use ( I put .txt at the end of the file to be able to upload it)

ttgov21new.h.txt

paxcounter.conf.txt

hansju-11 commented 3 years ago

I did a check on NULL. in timekeeper.cpp

if (!defined GPS_INT && !defined RTC_INT)

if (ppsIRQ == NULL) { ESP_LOGD(TAG, "ppsIRQ is NULL"); } else { timerWrite(ppsIRQ, 0); // reset pps timer } CLOCKIRQ(); // fire clock pps, this advances time 1 sec

endif

my output : [I][timekeeper.cpp:101] setMyTime(): [71.560] UTC time: 1617817930.000 sec [D][timekeeper.cpp:115] setMyTime(): ppsIRQ is NULL ....

and than it works .... :-)

cyberman54 commented 3 years ago

Yes, this is the bug. Thanks for spotting it! Now it should be finally solved: https://github.com/cyberman54/ESP32-Paxcounter/commit/c00612ad76906c3fc67151c1f6dc5d83926f01b8