tried on feather_nrf52840 - same results -- reboots to safe mode -- FS erased

@jerryneedell Thanks for testing. Your test is not running code.py; I'm thinking I need to do the same filesystem_flush() on the user-initiated soft reload. Will check on that later tonight.

@jerryneedell Added another flush, and I believe I tested your scenario successfully.

I just tried your commit @dhalbert and it already is appearing to function better (i.e. no crashes yet) on the Feather M4 Express.

Would this have affected 3.x too?

works great on the feather_nrf52840!

works on pyportal as well!!
Nice job!

@kevinjwalters it's possible it was a latent issue in 3.x, but it might have been ameliorated by different timing. There is a ram flash cache in 3.x as well.

That didn't do it.
I merged your pull request, compiled and flashed it to a pca10059, then I copied my BLE colourpicker code + libs to the drive.
The code started to run, but stopped by itself after about 20 seconds, without me doing anything.
After unplugging and replugging, the drive was wiped.

I tried this PR on my PCA10059 - It has been awhile since I used it and I don't recall the state I left it in. After loading the new image, I found that my FS was corrupted, but then after erasing and reloading the FS (using storage.erase_filesystem()), it has been worked normally -- with a BLE Uart test and Eddystone beacon test.

Hmmm -- Update -- PCA10059 wiped itself -- not sure when or why but I first found that I could not run a script I had just loaded, and os.listdir() just showed a lot of 00's. I did a hard RESET and it booted OK, but all the files were gone.

added:
I have not been able to reproduce this behavior on the feather_nrf52840 -- running eddystone beacon - copying files to FS and running other scripts.

Is there a difference in how the PCA10059 handles this since it does not have an SPI Flash for the FS?

This PR is a huge improvement over Beta4 in terms of stability for the FS for most boards. While the pca10059 appears to be an exception there may be a separate issue going on there.

In my experience, the PyPortal and feather_nrf5840 are working well and were extremely frustrating with Beta4.
Even if not perfect, I think a release of this update would be helpful.

I tested the pca10059 with a simple main.py, that doesn't use any external libs.
It runs stable. I can even torture the drive with f3write. No errors. The problems seem to be linked to bluetooth here.

With the pca10056, it's completely different.
It runs the bluetooth code without an error.
When I write test data to the drive with f3, while the code is running, the test data shows errors afterwards.
When I write test data with stopped code, I get no errors.

It seems that there are still some timing problems.

The difference here is that the PCA10059 uses internal flash for CIRCUITPY, and the PCA10056 uses external. So there's probably a different bug related to that. There's code to use the SoftDevice internal flash writing routines when the SoftDevice is enabled, but clearly there's still some issue. I will look at that.

I think I'm going to release a beta 5 immediately with just this fix, since beta 4 is too broken for most people to use.

Does this post need an edit or a reply on when beta 4 should or should not be used? https://forums.adafruit.com/viewtopic.php?f=60&t=149034

If the bug affects the pca10056, then it should also affect the feather_nrf52840, as both are nRF52840 with external flash. Unless it's a SPI timing issue, and the flash chips behave differently.

so far, I have found the feather_nrf52840 and a particle_argon to both behave well with this PR. I am just doing "normal" use -- running code and copying files to the FS then issuing soft reboot. I have not had any problems.

One related observation - not new but still occasionally occurs.
I have had seen few "odd" behaviors were a newly copied file does not run on the first attempt, but does run after a reboot or recopying - I don't have a clean example of this. This is not a new thing and I have always attributed it to the delayed file writes - I seem to avoid it by always syncing the FS and theOS after a copy.

@jerryneedell Would you mind trying this code?
It needs the libraries adafruit_ble and adafruit_bluefruit_connect. It's written for the pca10059, so you'll have to adapt the LED port pins.

from adafruit_ble.uart import UARTServer
import board, digitalio, pulseio
from adafruit_bluefruit_connect.packet import Packet
from adafruit_bluefruit_connect.color_packet import ColorPacket

ledr = pulseio.PWMOut(board.LED2_R, frequency=5000, duty_cycle=65535)
ledg = pulseio.PWMOut(board.LED2_G, frequency=5000, duty_cycle=65535)
ledb = pulseio.PWMOut(board.LED2_B, frequency=5000, duty_cycle=65535)

ledy = digitalio.DigitalInOut(board.LED1)
ledy.direction = digitalio.Direction.OUTPUT
ledy.value = True

uart_server = UARTServer()

def rgbled(colors):
    ledr.duty_cycle = (255 - colors[0]) << 8
    ledg.duty_cycle = (255 - colors[1]) << 8
    ledb.duty_cycle = (255 - colors[2]) << 8

while True:
    # Advertise when not connected.
    uart_server.start_advertising()
    ledy.value = False
    while not uart_server.connected:
        pass

    while uart_server.connected:
        ledy.value = True
        packet = Packet.from_stream(uart_server)
        if isinstance(packet, ColorPacket):
            print(packet.color) 
            rgbled(packet.color) 

This code wipes itself on the pca10059, but runs on the pca10056, so I guess it'll also run on the feather_nrf52840.

@uhrheber I made a modified version since the feather_nrf52840 does not have the same LEDs - I just use 1 LED
I have run it on a particle argon - so far without issue.
Does this test it adequately?

from adafruit_ble.uart import UARTServer
import board, digitalio, pulseio
from adafruit_bluefruit_connect.packet import Packet
from adafruit_bluefruit_connect.color_packet import ColorPacket

ledr = pulseio.PWMOut(board.D7, frequency=5000, duty_cycle=65535)


uart_server = UARTServer()

def rgbled(colors):
    ledr.duty_cycle = (255 - colors[0]) << 8

while True:
    # Advertise when not connected.
    uart_server.start_advertising()
    while not uart_server.connected:
        pass

    while uart_server.connected:
        packet = Packet.from_stream(uart_server)
        if isinstance(packet, ColorPacket):
            print(packet.color) 
            rgbled(packet.color) 

it is also running OK on a feather_nrf52840

@uhrheber and @jerryneedell could you open a new issue for the PCA10059 issue, so we can track it? Thanks.