[ltp] T20 locking up

Cliff Vasicek linux-thinkpad@www.bm-soft.com
Thu, 26 Oct 2000 15:08:33 -0500


I was wondering if anybody has been able to resolve this problem.  If I
leave my T20(RedHat 6.2, kernel 2.2.16-3) on at work overnight, at least
75% of the time it will be locked up hard when I come in the next
morning.  If anybody can help with this, I would appreciate it.

Cliff


As I mentioned recently, my T20 (2647-44: 700MHz, 128MB, 12 GB, DVD,
Intel PRO/100+) occasionally locks up hard.

Tim Roberts, the author of the Savage IX/MV XFree Server, told me that
he had received two similar reports, that he had not yet beeen able to
reproduce (AFAIK, he uses Toshibas himself).  However, he could
provide me with the crucial information that the `Matrix' mode of
xlock was efficient in triggering the lock (no pun intended).

Indeed with `xlock -mode matrix', I can lock up my T20 hard _reliably_
inside a couple of minutes.

Armed with this information, I could look for a workaround, even if I
don't know enough about hardware in general and the Savage in
particular to really fix the problem.

  Workaround 1: add 

      Option "noaccel"

    to the "Device" section of your XF86Config file.  This will
    _really_ slow down the server, but it appears work around the
    lockups without touching the X server.

  Workaround 2: recompile the server and disable only the accelerated
    *ScreenToScreenCopy functions.  Scrolling will still be annoyingly
    slow, but at least screen fills are fast.

Of course, these both are just workarounds, because I want the
acceleration active.  More importantly: it's possible that they just
remove the symptom by relaxing timing constraints and not the cause.

Question:

    Which X server is IBM shipping with the preinstalled T20s?
    If it is not Tim Roberts' server: does it have the same problem?
    If not: where can we get it?

Otherwise: how can IBM help with resolving the problem?  I assume that
it is more efficient to ask the IBM people reading this list before
going through official support channels.

It could alse be useful to collect more data by running `xlock -mode
matrix' for a few minutes on more T20s out there.  This way one could
find out if the problem is common to all T20s or a bug in some Savage
chips.

   WARNING: the probably ensuing hard lockups can do serious damage to
   your filesystems: back up and unmout (or remount readonly) as many
   partitions as possible before making an experiment!  I don't want
   to be held responsible for any filesystem damage!
    
Cheers,
-Thorsten
-- 
Thorsten Ohl, Physics Department, TU Darmstadt --
ohl@hep.tu-darmstadt.de
http://heplix.ikp.physik.tu-darmstadt.de/~ohl/ [<=== PGP public key
here]
----- The Linux ThinkPad mailing list -----
The linux-thinkpad mailing list home page is at:
http://www.bm-soft.com/~bm/tp_mailing.html


-- 
Cliff Vasicek
Net Quotient Consulting Group
(512) 651-5050 
cvasicek@netquotient.com
----- The Linux ThinkPad mailing list -----
The linux-thinkpad mailing list home page is at:
http://www.bm-soft.com/~bm/tp_mailing.html