Sorry this took very long. The reason for this type of kernel panic is any driver, that disables interrupts and then allows the CPU to go into idle. One example of this would be a function called as an interrupt handler (irqs off) which uses a mutex instead of a spinlock. So, as MyXelf's case only happens doing PPP networking, this could be any driver in that stack (I would ignore the comments from rodrifra as that case is not in that area and also seems to be fixed by now. Actually there have been stable patches to cifs that address at least the serverino problem. So it might be really fixed).
Looking at the ftdi_sio driver (which is used looking at the dmesg), I found only one mutex and that seems to be in a place that should be ok. Still this would leave usb-serial and the lot of ppp. Before trying to find that needle, I got Karmic packages that at least include 2.6.31.6 on https://launchpad.net/~stefan-bader-canonical/+archive/karmic. MyXelf, could you please try those to make sure the problem is still in that kernel? Thanks -- BUG: scheduling while atomic: swapper - using ppp [kernel 2.6.31-10.35-generic] https://bugs.launchpad.net/bugs/435917 You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. -- ubuntu-bugs mailing list [email protected] https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
