OM2007.2 continuously freezes!

Andy Green andy at
Thu Jul 31 14:35:32 CEST 2008

Hash: SHA1

Somebody in the thread at some point said:
| Andy Green wrote:
|> | To confirm this, I've just removed from my FSO installation Xglamo and
|> | replaced it with Xfbdev and all worked without freezes. It's all a
|> | little slower, but it works without freezes.
|> |
|> | Do you think that this is an hardware problem? :|
|> No, XGlamo contains a bunch of kernel code in userspace with no attempt
|> at cooperating with the kernel locking.  There's other fun stuff to do
|> with its synchronizing with the Glamo.  So this is "improvements needed
|> in XGlamo" problem I would think, not hardware.
| Thanks for replying Andy I did hope you would have done :P.
| I'm quite happy that it shouldn't be an hw bug, btw I can't understand
| why this happens so often to me. Only, it seems (well, in an Italian
| forum some one had this too, but not so often as I have).
| However, couldn't you suggest me a way to log this (also to help your
| develompent), or are those known bugs and you don't need to know more of
| my crashes?

I run at the moment with patches that force a panic on pressing AUX, and
another patch takes care to spew any pending syslog and give me a
backtrace of where we were spinning when we panic.  So if interrupts
were up when we lockup I have a chance to see through the freeze.  The
second patch even has its own "by hand" serial stuff so it works deep in
suspend before the serial ports resume.

But often Glamo freezes are especially ugly to debug because they are
true hard freezes on the external bus, it is forcing the CPU to wait
forever.  I have a way around that too but I don't recommend anyone else
use my violent techniques.

Someone was saying they will look at XGlamo and I don't think you're the
only one seeing some shakiness.

- -Andy
Version: GnuPG v1.4.9 (GNU/Linux)
Comment: Using GnuPG with Fedora -


More information about the support mailing list