-
Notifications
You must be signed in to change notification settings - Fork 1
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[CRITICAL] Continuous error spamming to the log file: "[ERROR] [../src/output-state.c:45] Failed to commit frame", сonsumes too much disk space and damages the SD card's write resource #345
Comments
You have not provided any reproduction information, as such I'll close this until it's provided. Note, step 1 of any reproduction information must be "install a new Raspberry Pi OS (lite / desktop) (32bit / 64 bit) |
I'm talking about Raspberry Pi OS Bookworm Desktop 64 bit with all latest updates, device Raspberry Pi 4 with 4GB:
Closing a ticket about critical bug that causes rapid and irreversible depletion of the SD card's write cycles without any fix raises the question of whether this might be an attempt to conceal malicious actions, isn't it? |
Just tested it on a clean install of Raspberry Pi OS Desktop 64-bit and can reproduce this critical bug. Steps to reproduce
Expected result: no log spamming with thousands of errors Actual result: the log is spammed with Note: the same issue happens when you staying on desktop screen and display is going to sleep (display blank). |
Does it happen on a lite image? Or only on a desktop image when switched to the VT? |
I didn't tested it on lite version, but since it is related to UI, I think it is relevant for desktop image only. But if you install desktop on lite version, I think you will get the same issue. |
The logging message is coming from labwc which is triggering wlr_output_commit (with no display to commit to). But the upstream code no longer has this line (this code has been changed recently). Will have a look at it when we're all back in the office, it's not going to affect most people since they'll be running the desktop. It only affects you because you've switched to the VC. Not sure why you're getting 30 per second though, I only get 1 per second... |
The issue affects anyone who using labwc desktop with display sleep or with switching to a virtual terminal.
1 second looks like result of the taskbar clock update every second. I found that the issue occurs when you add the "CPU," "GPU," and "CPU Temp" widgets to the taskbar. Without these widgets, this error is raised rarely. However, when these widgets are added to the desktop taskbar, the error is logged about 30 times per second. My display refresh rate is 75 Hz, not sure but it may also affect the error rate. I discovered this issue because I noticed a strange decrease in disk space when the system was idle (during desktop sleep or when switched to a virtual terminal). The root cause turned out to be the rapid growth of the |
Steps to reproduce
tail -f .xsession-errors
Expected result: no continuous error spamming
Actual result: the following error appears in the log about 33 times per second: [ERROR] [../src/output-state.c:45] Failed to commit frame
As result, the .xsession-errors is spammed with megabytes of the same error text, сonsumes disk space and damages the SD card's write resource.
It seems that the same issue happens when display going into sleep mode, which makes this issue even more critical.
here is a part of the log:
After 24 hours of uptime, I have a .xsession-errors file of 19 megabytes, filled with hundreds of thousands of entries with this error.
Update: As a temporary workaround until a fix is released, you can remove the
~/.xsession-errors
file each time after boot, or add this line to~/.config/labwc/autostart
to do it automatically:/usr/bin/rm /home/pi/.xsession-errors
When this file is removed, it does not seem to be recreated until the next reboot. This approach can help protect your SD card from unnecessary wear and save disk space by removing the file on each OS boot.
The text was updated successfully, but these errors were encountered: