[Openswan dev] FC3/FC4 builds and openswan 2.x

[Henrik Nordstrom]

On Mon, 15 Aug 2005, Paul Wouters wrote:

> So it seems one major problem for running KLIPS on Fedora/RedHat kernels has
> been solved, but that in the latest kernel, RedHat changed something else.
> I've attached a diff of config-2.6.11-1.35_FC3 and config-2.6.12-1.1372_FC3,
> but I do not think it is a kernel option that is responsible for the changed
> behaviour.

Only speculating here but perhaps there is some obscure "stack buffer 
under/overflow/misuse" bug in klips not usually tripping over anything 
meaningful if frame pointers are used?

The stack traces even when not using frame-pointers is actually relatively 
accurate in terms of return pointers except that optimization may skip a 
few returns more than without it, similar to how inlining may skip calls. 
The problem is filtering out all the garbage addresses which happened to 
be on unused areas of the stack (unused locals) and does not belong to 
this trace.. Doing this analysis manually from the oops trace is somewhat 
time consuming, but if gdb is given the proper starting point on the stack 
then it does know how to unwind this with the help of the debug info which 
allows gdb to track the stack frames without the help of the frame 
pointer.

Regards
Henrik