Messages in this thread | | | Date | Thu, 13 Jan 2005 17:03:52 -0800 | From | George Anzinger <> | Subject | Re: VST patches ported to 2.6.11-rc1 |
| |
Pavel Machek wrote: > Hi! > > >>I really hate sf download system... Here are those patches (only >>common+i386) ported to 2.6.11-rc1. > > > Good news is it booted. But I could not measure any powersavings by > turning it on. (I could measure difference between HZ=100 and > HZ=1000). > > Hmm, it does not want to do anything. threshold used to be 1000, does > it mean that it would not use vst unless there was one second of quiet > state? I tried to lower it to 10 ("get me HZ=100 power consumption") > but it does not seem to be used, anyway: > > root@amd:/proc/sys/kernel/vst# cat successful_vst_exit > 0 > root@amd:/proc/sys/kernel/vst# cat external_intr_exit > 0 > root@amd:/proc/sys/kernel/vst#
Hm.. and this after you lowered the threshold? The skipped interrupts should also be of interest... > > >>+config HIGH_RES_RESOLUTION >>+ int "High Resolution Timer resolution (nanoseconds)" >>+ depends on HIGH_RES_TIMERS >>+ default 1000 >>+ help >>+ This sets the resolution in nanoseconds of the CLOCK_REALTIME_HR and >>+ CLOCK_MONOTONIC_HR timers. Too fine a resolution (small a number) >>+ will usually not be observable due to normal system latencies. For an >>+ 800 MHz processor about 10,000 (10 microseconds) is recommended as a >>+ finest resolution. If you don't need that sort of resolution, >>+ larger values may generate less overhead. > > > Ugh, if minimum recomended value is 10K, why does it have 1K as a > default?
I think I have this wrong. I usually set it to 10 for my testing. I need to look over the conversion code and such to see just what the units are...
> > >>+ The system boots with VST enabled and it can be disabled by: >>+ "echo 0 > /proc/sys/kernel/vst/enable". > > > It definitely booted with vst disabled here... echo 1 did the trick > through.
I missed that. Yes we want to boot with it disabled. It can upset some of the boot up code otherwise. > > >>short_timer_fns This is an array of 5 entries of the form >>... >> 0xc110ea80 when the timer expires. >>Both of these arrays are kept as circular lists and read back such >>that >>the latest entry is presented to the reader first. The entries are >>cleared when read. > > > ...it is bad idea to have them world-readable, then.
This is diagnostic code for the kernel dude who wants to know why he is not getting the VST sleeps he wants. Seems a bit harsh to make him be root to see it....
-- George Anzinger george@mvista.com High-res-timers: http://sourceforge.net/projects/high-res-timers/
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |