[ELDK] pthread / TQM5200 problem
wd at denx.de
Fri Feb 1 22:31:39 CET 2008
in message <003d01c864e2$d7d8df40$0101a8c0 at tis> you wrote:
> Here's some additional information I forgot when originally posting my
> question today morning: I'm using the DENX ELDK 3.1.1 and a kernel version
Note that the TB5200 is not really supported by DENX. Actually, we
never had such a device in our lab yet. You might want to bother TQ
support about this. At least one of the guys is subscribed to the list
The big question is: which *exact* version of the Linux kernel are you
running, i. e. which git commit ID is this?
Is there any chance that you are using quite obsolete versions of
U-Boot and Linux on your board? Or did you try a current code (top of
tree from git repository), with the same results?
> "spurious wakeup" and "lost signal" problems. It runs 159 additional threads
> (that seems to be some internal limit), each of them continuously increments
This rings some bell to me. The "159 threads limit" probably results
from PCI mappings that overlap with CONFIG_TASKSIZE. This can cause
mysterious reboots and other problems, when the process address space
gets exercised intensively (like big number of simultaneous threads
running, heavier contention for memory regions etc.).
However, that bug was fixed a long, long time ago (September 1006),
so that would mean that you are actually running pretty ancient code:
Author: Marian Balakowicz <m8 at semihalf.com>
Date: Thu Sep 28 15:35:20 2006 +0200
Fix PCI mappings so they don't overlap with CONFIG_TASKSIZE
I think you should just update your Linux kernel source tree from the
top of tree of the git repository and try again.
DENX Software Engineering GmbH, MD: Wolfgang Denk & Detlev Zundel
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd at denx.de
Any technology distinguishable from magic is insufficiently advanced.
More information about the eldk