2.4.0-test9-pre5, although this has existed since at least 2.4.0-test1.
VT console on vga. printk -> vt_console_print -> hide_cursor ->
vgacons_cursor -> write_vga -> cli -> __global_cli -> get_irqlock ->
wait_on_irq -> show -> printk -> SMP deadlock!
I hit this on an SMP box, one cpu was dead and inside an interrupt.
The other cpu tried to print to the console, ran down that chain, timed
out in wait_on_irq and tried to print. It was big red button time.
Not only is this path *slow* but it introduces a nasty deadlock
condition into printk. We cannot rely on printk to get diagnostics for
SMP hangs.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Sat Sep 23 2000 - 21:00:24 EST