On Sat, Dec 21, 2002 at 02:08:02PM -0500, Preston A. Elder wrote:
> Denis Vlasenko <[email protected]> said I should forward this
> to you, as I got no response from the LKML.
Well, I'm doing keyboards in 2.5 only, that's why I didn't comment on
the LKML posting.
> So, if you have any ideas, please, let me know. I suppose the more important
> issue is the 'time warp' issue (with the hwclock command failing, and
> concequently the time jumping forward and back again). But the keyboard
> timeout is also important, and it is symtomatic of the other problem. If the
> keyboard works, the clock is fine.
>
> Original message Wednesday 18 December 2002 08:05 am:
> > [1.] One line summary of the problem:
> > Keyboard not found, but it exists!
> >
> > [2.] Full description of the problem/report:
> > The PS/2 keyboard fails during the kernel boot (before init), saying:
> > keyboard: Timeout - AT keyboard not present?(ed)
> > keyboard: Timeout - AT keyboard not present?(f4)
> >
> > The keyboard does indeed exist, and I am able to use it in BIOS, in the
> > boot manager, and booting to DOS. The above errors, however disable the
> > keyboard after the kenel boot, and so I am unable to use it in the running
> > system.
That's not correct. The above messages don't disable the keyboard - even
when they appear the kernel still listens to keypresses from the
keyboard - it only doesn't send LED commands to the keyboard anymore
(because it'd have to timeout on each LED command). It'll start sending
LED commands as soon as a keypress from the keyboard arrives.
Which means: The messages are just saying that the keyboard doesn't
work. And it doesn't for some reason. But the reason is not the
messages.
> > Thats not the full extent of the problem though. I also have problems
> > running "hwclock" without the --directisa option, it will just block (and I
> > can't abort it with no keyboard). I've added the --directisa option to
> > startup/shutdown for now.
This may be caused by the same problem.
What BIOS timing setup are you using? Have you by any chance 'tuned' the
board for better performance?
> > I've also experienced some problems with the system time just jumping ahead
> > by 35 minutes, and then back again. The back again part could be because
> > I'm running NTP to keep the system in sync, but I'm getting weird behavior
> > with time on this system, which is causing havoc on applications.
This is a known vt82c686a bug. Workarounds exist in various kernels.
2.4.18 has an incomplete workaround which isn't used if the CPU supports
TSC.
> > [3.] Keywords (i.e., modules, networking, kernel):
> > 2.4.19 keyboard ps2
I hate bugreports done by exactly following the bugreport cookbook. ;)
> > [6.] A small shell script or example program which triggers the
> > problem (if possible)
> > N/A
... because I don't need to know a small shell script isn't applicable,
for example ...
> > 00:07.0 ISA bridge: VIA Technologies, Inc. VT82C686 [Apollo Super South] (rev 22)
> > Subsystem: VIA Technologies, Inc. VT82C686/A PCI to ISA Bridge
Yes, a vt82c686a buggy chip ...
--
Vojtech Pavlik
SuSE Labs
> > > [3.] Keywords (i.e., modules, networking, kernel):
> > > 2.4.19 keyboard ps2
>
> I hate bugreports done by exactly following the bugreport cookbook. ;)
>
> > > [6.] A small shell script or example program which triggers the
> > > problem (if possible)
> > > N/A
>
> ... because I don't need to know a small shell script isn't applicable,
> for example ...
Quoting the FAQ:
"Don't attach huge files to your post. One major culprit is people
attaching their kernel .config file to their post."
These days people get flamed for _not_ attaching their .config to
their post :-)
Also:
"Stick to the subject. This is a Linux kernel list, mainly for
developers."
--- developers."
+++ developers to discuss technical issues, start flame wars, and
+++ generally use at least 50% of the bandwidth on non-kernel issues.
:-)
John.
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Well, I upgraded the kernel (the old one was a 2.4.18, with a crapload of
patches applied, the new one is actually a 2.4.19 kernel). In short, the
problem persists under v2.4.19. And from the changelog, I can't see any
changes done to this fix for 2.4.20.
It still jumps back and forth with time, and the keyboard (and mouse) are both
still unresponsive. I can try 2.4.20 if you think it will change anything
over 2.4.19. Otherwise, are there any other suggestions you can make as to
what I can try? I'm not as worried about the keyboard/mouse since its a
server, and I usually ssh to it, however the 'time warp' situation (it
jumping ahead in time, and then back again) is much more destructive to the
applications I run on the server.
Also, is there any way to know if it activated the workaround or not? I don't
see anything special in my boot dmesg. I can send the boot dmesg file (my
system saves it immediately after boot) and my .config file to anyone who
thinks they might get some useful information out of it. I'd rather not have
to go out and buy a new motherboard for this.
Thanks in advance,
- --
PreZ
Systems Administrator
GOTH.NET
Goth Code '98: tSKeba5qaSabsaaaGbaa75KAASWGuajmsvbieqcL4BaaLb3F4
nId5mefqmDjmmgm#haxthgzpj4GiysNkycSRGHabiabOkauNSW
GOTH.NET - http://www.goth.net
Free online resource for the gothic community.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.7 (GNU/Linux)
iD8DBQE+BRtVuULtzKdGMboRAteLAJ92Wicv05G+PT3GAvHa1LUCSZtFmACgs2Az
+j6zBGa81RboJRituL4Vf0Y=
=xulE
-----END PGP SIGNATURE-----