Closed eliasbakken closed 1 year ago
It seems having the USB drive with Refactor installed causes some instability and seems to also cause a freeze during booting. It might be related to the double mounting in /etc/fstab
Hmm. In that case we should really be doing a uuid based fstab, and use a command street dd to change the uuid of the emmc partitions. Or the other way around, changing the USB drive's.
On Fri, Apr 21, 2023, 01:42 Elias Bakken @.***> wrote:
It seems having the USB drive with Refactor installed causes some instability and seems to also cause a freeze during booting. It might be related to the double mounting in /etc/fstab
— Reply to this email directly, view it on GitHub https://github.com/intelligent-agent/Refactor/issues/364#issuecomment-1517064064, or unsubscribe https://github.com/notifications/unsubscribe-auth/AEEI46P2RAHFTHMHV47ZPDDXCHCVVANCNFSM6AAAAAAW7ALR4M . You are receiving this because you are subscribed to this thread.Message ID: @.***>
I'm getting an error message musb-sunxi 1c19000.usb: Error unknown readw offset 0
and repeated several times before the freeze.
Upgrading to kernel 6.1 and running cat /sys/kernel/debug/usb/musb-hdrc.4.auto/regdump
produces the Error unknown readw offset
issue.
It seems that doing anything with /sys/kernel/debug/psci
can also lead to a crash.
Uh... The values in the debug may need to be read by special utils or when the kernel is booted with debug arguments?
Otherwise, yeah, i can see you could cause crashes with that, especially if you're trying to access values that are otherwise restricted because of security features that are enabled by default?
On Mon, Apr 24, 2023 at 9:43 PM Elias Bakken @.***> wrote:
Upgrading to kernel 6.1 and running cat /sys/kernel/debug/usb/musb-hdrc.4.auto/regdump produces the Error unknown readw offset issue. It seems that doing anything with /sys/kernel/debug/psci can also lead to a crash.
— Reply to this email directly, view it on GitHub https://github.com/intelligent-agent/Refactor/issues/364#issuecomment-1520729630, or unsubscribe https://github.com/notifications/unsubscribe-auth/AEEI46LKH4QTP2B4EMH74MLXC3JXLANCNFSM6AAAAAAW7ALR4M . You are receiving this because you commented.Message ID: @.***>
I think it should be possible to run a grep on the entire filesystem without causing a crash, but it's not a showstopper really. I will close this for now as "wontfix". It's an embedded system, most people will never access via ssh anyway.
Running
grep -r "test-test" /
for instance causes a complete freeze.