FiPy Crashing on boot.

[smaug]

[More info]
After playing around with the safe boot options I got the error message below. I've gotten the "esp_image: Checksum failed" message a couple of times now with different values for "Calculated" and "read".

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

rst:0xc (SW_CPU_RESET),boot:0x12 (SPI_FAST_FLASH_BOOT)
configsip: 0, SPIWP:0xee
clk_drv:0x00,q_drv:0x00,d_drv:0x00,cs0_drv:0x00,hd_drv:0x00,wp_drv:0x00
mode:DIO, clock div:1
load:0x3fff8028,len:8
load:0x3fff8030,len:1984
load:0x4009fa00,len:0
load:0x4009fa00,len:19112
entry 0x400a05e4
E (534) esp_image: Checksum failed. Calculated 0x58 read 0x4d

[robert-hh]

@smaug I had that when the firmware image was too large for the partition. Even if that's not to be expected when you run 1.18.0.r1, try the option "full erase" at the firmware upload. The re-builds the flash landscape.
Edit: Are you just using the Pycom firmware or an self made image with many frozen modules?

[smaug]

This post is deleted!

[smaug]

@robert-hh Thank you for the reply! Unfortunately that didn't help. I've tried the "full erase" option every time I've reflashed trying to get this resolved. Yes, I am using Pycom's utility to flash.

[robert-hh]

@smaug Try esptool.py for flash erase. Then you're sure to have a clean starting state.

[smaug]

@robert-hh I'll try that. Do you know the correct command arguments for that? I have this command for flashing .bin files to an ESP32 devkitC:

esptool.py    \
        --chip esp32                                \
        --port /dev/tty.SLAB_USBtoUART              \
        --baud 921600                               \
        --after hard_reset write_flash              \
        -z                                          \
        --flash_mode dio                            \
        --flash_freq 40m                            \
        --flash_size detect                         \
        0x1000 bootloader.bin                       \
        0x8000 partitions_espruino.bin              \
        0x10000 espruino_esp32.bin

[robert-hh]

@robert-hh as usual, you have to specify port and baudrate with -p and -b, the chip with -c, and the operation is erase_flash. The device has to be in bootloader mode, GPIO0 at GND. Like
esptool.py -p /dev/.... -b921600 --chip esp32 erase_flash
If in doubt, try esptool.py --help

[smaug]

@robert-hh Thank you for that. I've erased flash with esptool.py and then uploaded v1.18r1 using Pycom's software. The device still cannot boot into micropython :(. UART messages below:

ets Jun  8 2016 00:22:57

rst:0x1 (POWERON_RESET),boot:0x13 (SPI_FAST_FLASH_BOOT)
configsip: 0, SPIWP:0xee
clk_drv:0x00,q_drv:0x00,d_drv:0x00,cs0_drv:0x00,hd_drv:0x00,wp_drv:0x00
mode:DIO, clock div:1
load:0x3fff8028,len:8
load:0x3fff8030,len:1728
load:0x4009fa00,len:0
load:0x4009fa00,len:14584
entry 0x400a059c
user code done

ets Jun  8 2016 00:22:57

rst:0x1 (POWERON_RESET),boot:0x13 (SPI_FAST_FLASH_BOOT)
configsip: 0, SPIWP:0xee
clk_drv:0x00,q_drv:0x00,d_drv:0x00,cs0_drv:0x00,hd_drv:0x00,wp_drv:0x00
mode:DIO, clock div:1
load:0x3fff8028,len:8
load:0x3fff8030,len:1728
load:0x4009fa00,len:0
load:0x4009fa00,len:14584
entry 0x400a059c
E (404) esp_image: Checksum failed. Calculated 0x54 read 0x7c

[robert-hh]

@smaug That's strange. What's the difference between the two dumps?

[smaug]

@robert-hh The top dump happens first, then the bottom dump. The only difference between them is the last line.

user code done

vs

E (404) esp_image: Checksum failed. Calculated 0x54 read 0x7c

Clarification: Every dump I've posted is from a different attempt to boot into micropython, usually after pressing the reset button or powering up.

[smaug]

It seems that the FTDI chip I was using started to brown out. I switched to external power with a DC supply and everything is back to normal. I should've tried that first. I didn't even think of it since for the first couple of weeks the FTDI chip had no issues supplying sufficient power. Regardless, thank you very much for your help @robert-hh