On Fri, Feb 23, 2007 at 11:34:47AM +0000, Raymond Meyer wrote:
> Does the new threading subsystem eliminate those bugs or are they still
> present?

We can not tell really. Since most of the suspect code paths have been
eliminated, it is likely that they are gone. On the other hand I have
been unable to reproduce SA threading related bugs lately anyway.

Btw., the newlock2 branch was not meant as a new threading system, but
a reorganization of kernel internal locking infrastructure. The threading
changes were done late in the process when it became clear that it
would be hard to do the right new style locking for the kernel SA support
code.

Martin