all sorts of questions ...

Jan Gyselinck (JAN.GYSELINCK@student.kuleuven.ac.be)
Mon, 15 Jun 1998 23:29:23 +0200 (W. Europe Daylight Time)


Hi,

I would like to get some explanation on PCI-IDE interfaces. I have a
National Semiconductor PC87410 in my portable. I do have the impression
that my harddisk is not running at the speed it can run. It supports PIO
3 and DMA. Now, the NS87410 doesn't support that (the NS87415 does
support that, and there's a driver for that in the kernel). I got the
docs for the NS87410, and now I wonna know how to read the PIO/set the
PIO. Now, I guess I just have to adjust the timings, but to which value?
And second, how do I change the values (can setpci do the job? ...)
Anyway, if some IDE-guru can give me a little bit of help or explanation,
that would be great.

Second thing: I got a Chips and Technologies 65545 video card, with a 10"
TFT screen. I got APM enabled, but there's something wierd about that
(maybe 'cauze I messed too much with my hardware, who knows ...), because
if I set the panel-timeout to 2 minutes, it never blanks, but it does
blank when the timeout is 1 minute (I think it did that too under DOS, but
it has been a long time that I ran DOS on this thing). Now, I can live
with that, I guess, but I would be great if I could blank it from linux.
If I set the timeout to 10 minutes, and I wonna blank it, it doesn't do a
thing. Maybe I did some things wrong (and there has been sort of an
explanation, a while ago, about blanking, and what should be done to
really blank). Just repeat in a few words what I have to do. I think I
did tried it the way that was then proposed, but that it didn't blank
either. Now, how can I get it to blank then? Search the code in the
APM-BIOS? Can I get it to blank when APM is disabled? (can I go to
lower-power-consumption when APM is disabled, can I get it to blank when I
chose to never blank my display (in BIOS)?) Tell me what ya think ...

Something about the current kmod-implementation: It does not remove
modules, and somebody said to put 'rmmod -a' in root's crontab. I
objected that, and I still object that, because it causes my HD to wake
up, every time it gets run. And I don't think that's normal, because
everything it needs to read should be cached, and access-time updates
should not be written by my implementation of updated (updated-spindown
patch). (or does the disk-line in /proc/stat change when there is
cache-access?????) So, is it normal that my HD gets waked for rmmod?

(anyway, I'll write a kerneld-thingy that only removes mods, I don't like
to run things frequently with crond ...)

Jan Gyselinck
lector@dds.nl

I want to believe
Linux is to great to be human-made

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu