espflash and cargo espflash are flashing different bootloader

[andresv]

espflash 1.7.0 flashes a bootloader that is bundled with espflash, but cargo espflash takes bootloader from: target/xtensa-esp32-espidf/release/build/esp-idf-sys-049b1b8192557540/out/build/bootloader/bootloader.bin

I tried to enable CONFIG_BOOTLOADER_APP_ROLLBACK_ENABLE and it worked well on my dev machine where I always flash with cargo espflash, but this feature did not work for devices that were flashed with espflash.

It took me quite some time to figure out that it is because of bootloader.

I noticed that after cargo espflash logs start with:

I (27) boot: ESP-IDF v4.4.2-dirty 2nd stage bootloader
I (27) boot: compile time 11:54:03
I (27) boot: chip revision: 1
I (30) boot_comm: chip revision: 1, min. bootloader chip revision: 0
I (37) boot.esp32: SPI Speed      : 40MHz
I (42) boot.esp32: SPI Mode       : DIO
I (46) boot.esp32: SPI Flash Size : 4MB
I (51) boot: Enabling RNG early entropy source...
I (56) boot: Partition Table:
I (60) boot: ## Label            Usage          Type ST Offset   Length
I (67) boot:  0 nvs              WiFi data        01 02 00009000 00006000
I (75) boot:  1 otadata          OTA data         01 00 0000f000 00002000
I (82) boot:  2 phy_init         RF data          01 01 00011000 00001000
I (90) boot:  3 app_0            OTA app          00 10 00020000 001db000
I (97) boot:  4 app_1            OTA app          00 11 00200000 001db000

But after espflash they started with:

I (476) cpu_start: Pro cpu up.
I (477) cpu_start: Starting app cpu, entry point is 0x40082f2c
0x40082f2c - call_start_cpu1
    at /Users/firmware/.embuild/espressif/esp-idf/v4.4.2/components/esp_system/port/cpu_start.c:160
I (0) cpu_start: App cpu up.
I (491) cpu_start: Pro cpu start user code
I (491) cpu_start: cpu freq: 160000000
I (491) cpu_start: Application information:
I (495) cpu_start: Project name:     libespidf
I (500) cpu_start: App version:      1
I (505) cpu_start: Compile time:     Nov 10 2022 11:54:00
I (511) cpu_start: ELF file SHA256:  0000000000000000...
I (517) cpu_start: ESP-IDF:          v4.4.2-dirty
I (523) heap_init: Initializing. RAM available for dynamic allocation:

Not sure what is correct way to solve this confusion, maybe:

• print out which particular bootloader is being flashed
• add more info about this to docs

[jessebraham]

This has been the case since 1.6.0, and was added in #186. The behaviour is different between cargo-espflash and espflash because espflash does not build the project and as such has no information regarding the location of the build artifacts. You're able to use the --bootloader option to provide a path to espflash.

I will document this for cargo-espflash for the time being.

[jessebraham]

I've at least mentioned this auto-detection feature in the README with #327.

[jessebraham]

I don't think there's really anything actionable here, so I'm going to close this issue. If you feel this is a mistake please feel free to re-open it, but other than mentioning this behaviour in the README I'm not sure what more to do. I will at least add some logging output to indicate the bootloader being used, that should be simple enough.

[andresv]

Yes, it would be perfect if it is just logged which one is actually used.

[jessebraham]

Will be done in the next release, see #339.