04-25-2021, 04:01 PM
(This post was last modified: 04-25-2021, 04:03 PM by TheOldPresbyope.
Edit Reason: add "and remotely"
)
You may not be alone but I have been unable to repro a failure using a DSI touchscreen display.
I'm not sure I understand your second test. Are you saying moOde works fine with the local display turned off? That was never in doubt.
Last night I finally partially disassembled one of my systems [1] so I could add a HiFiBerry I2S DAC+ Pro to the mix with the touchscreen display. I did this because, while I hadn't seen the issue using a USB DAC, other user reports show they are using an I2S DAC.
I still do not see a failure. I ran my system overnight with MPD idle and no problem [2]. I've been running it this morning with MPD in pause partway through a track and still no problem after three hours.
Sad to say, I don't see us making any progress on this issue until one of the devs can reliably repro it so we can search for a root cause or some user independently discovers a root cause.
Regards,
Kent
[1] RPi3B+, "Official" Raspberry Pi 7-in Touchscreen Display, WiFi client mode, moOde 7.1.0 booting and running from USB memory stick. No other physical device or software additions and not configured for a NAS.
[2] I run a script which logs the output from the top command every 5 minutes. By "no problem", I mean
- the "mem free" and "mem used" numbers don't change appreciably over the test period. It's been shown for the HDMI-related issue, OTOH, that these numbers change dramatically. Notable, available memory drops to a level so low that critical kernel processes like the scheduler cease to function properly. We know this is due to a runaway chromium-browser renderer but we don't know why it does this.
- at the end of test period I can still access the WebUI through the local display and remotely and my SSH connection remains functional.
I'm not sure I understand your second test. Are you saying moOde works fine with the local display turned off? That was never in doubt.
Last night I finally partially disassembled one of my systems [1] so I could add a HiFiBerry I2S DAC+ Pro to the mix with the touchscreen display. I did this because, while I hadn't seen the issue using a USB DAC, other user reports show they are using an I2S DAC.
I still do not see a failure. I ran my system overnight with MPD idle and no problem [2]. I've been running it this morning with MPD in pause partway through a track and still no problem after three hours.
Sad to say, I don't see us making any progress on this issue until one of the devs can reliably repro it so we can search for a root cause or some user independently discovers a root cause.
Regards,
Kent
[1] RPi3B+, "Official" Raspberry Pi 7-in Touchscreen Display, WiFi client mode, moOde 7.1.0 booting and running from USB memory stick. No other physical device or software additions and not configured for a NAS.
[2] I run a script which logs the output from the top command every 5 minutes. By "no problem", I mean
- the "mem free" and "mem used" numbers don't change appreciably over the test period. It's been shown for the HDMI-related issue, OTOH, that these numbers change dramatically. Notable, available memory drops to a level so low that critical kernel processes like the scheduler cease to function properly. We know this is due to a runaway chromium-browser renderer but we don't know why it does this.
- at the end of test period I can still access the WebUI through the local display and remotely and my SSH connection remains functional.
