Hi! With the rather timing critical pwm-demos (see my previsious mail), i have discovered a strange effect, using an Athlon 600 MHz and a PentiumII/333 (within a Notebook) running RTL2.0: When logged in (in graphical mode / X-Server), the screen-saver is activated, although it is set to 10 minutes, and i have just typed something or moved the mouse. Additionaly, some running "xanim" and "xquake" were stopped randomly (experienced that only with the notebook, but who knows ...) The screen-sever problem also appeard in an early RTAI-version (0.2). Since RTL and RTAI have basicaly the same architecture, i assumed the same bug. I met Paolo Montegazza in Vienna at the Linux Real-Time Workshop. He told me, that he had solved that problem some time ago, and he showed me, what he has altered to make the effect going away: in an RTL2.0-patched kernel-2.2.13 in file /usr/src/linux/arch/i386/kernel/time.c in function static void timer_interrupt() you have to comment the following equation (line 456): delay_at_last_interrupt = (count + LATCH/2) / LATCH; I didn't dig deeper, but i assume, this is some sort of optimization, that is corrupted/disturbed by RT-Linux (early RTAI as well). Now, everything seems to work just fine: i didn't discovered the undesired screen-blanking any more, and the apps (xanim/xquake) on the notebook are also working without stopping randomly for several hours (usuals, the effect showed up within 20 minutes). comments? Bernhard Kuhn P.S.: Many thanks to Paolo Montegazza, who showed my solution! --- [rtl] --- To unsubscribe: echo "unsubscribe rtl" | mail [EMAIL PROTECTED] OR echo "unsubscribe rtl <Your_email>" | mail [EMAIL PROTECTED] ---- For more information on Real-Time Linux see: http://www.rtlinux.org/~rtlinux/
