[PATCH] [PowerPC] Fix unwind info with dynamic stack realignment

[Jay Foad]

> Yes, we should have accurate unwind info at any position in the function, so that debugging / stack backtraces etc. work correctly.  This means we need to define CFA in terms of SP first, and then switch to a definition in terms of FP/BP as soon as the original definition becomes invalid due to code changing SP. 

OK, thanks. I think the current patch gets this right. If we're using BP then the definition CFA=BP kicks in immediately after SP is updated. I haven't changed the FP case, but it looks correct too.

http://reviews.llvm.org/D6410