[arch-general] Computer blocking due to kernel <-> userspace incompatibility?
Hello list Is it possible a computer gets stuck (not responding to keyboard, the only way to bring it back alive is a hard reset) because of a incompatibility between the kernel and userspace? The cursor is still flickering though, that is what leads me to believe it's software crash, not a hardware crash. This is the software in question: glibc 2.9-4 kernel-headers 2.6.27.6-2 kernel26 2.6.25.10-1 Thanks, Best regards, Glenn Matthys
RedShift schrieb:
Is it possible a computer gets stuck (not responding to keyboard, the only way to bring it back alive is a hard reset) because of a incompatibility between the kernel and userspace? The cursor is still flickering though, that is what leads me to believe it's software crash, not a hardware crash.
This is the software in question:
glibc 2.9-4 kernel-headers 2.6.27.6-2 kernel26 2.6.25.10-1
If there was a problem, you should get "kernel too old" as soon as init is started. Our userspace should work with 2.6.16 and newer (some kernel-related things like udev might fail, but the binaries themselves should be able to execute). However, you can never rule out a bug that causes this, I am only stating what it _should_ support.
look like an not-recognized-by-X keyboard. i've get some weird problem like that when I've install X. But maybe your cursor have already work. In this case, sorry for the useless post. 2009/3/30, Thomas Bächler <thomas@archlinux.org>:
RedShift schrieb:
Is it possible a computer gets stuck (not responding to keyboard, the only way to bring it back alive is a hard reset) because of a incompatibility between the kernel and userspace? The cursor is still flickering though, that is what leads me to believe it's software crash, not a hardware crash.
This is the software in question:
glibc 2.9-4 kernel-headers 2.6.27.6-2 kernel26 2.6.25.10-1
If there was a problem, you should get "kernel too old" as soon as init is started. Our userspace should work with 2.6.16 and newer (some kernel-related things like udev might fail, but the binaries themselves should be able to execute).
However, you can never rule out a bug that causes this, I am only stating what it _should_ support.
participants (3)
-
ludovic coues
-
RedShift
-
Thomas Bächler