On Tue, Sep 27, 2011 at 02:15:24AM +0200, Bernd Schmidt wrote: > That looks perfectly reasonable, and even if it is ordered in the way > shown above, dwarf2cfi still should be able to deal with it if the > prologue and epilogue are annotated correctly.
Here I was thinking that it was someone else's problem. :-) I looked at this before and saw some epilogue cfa_restores, but I see now we're missing some for -m64. More work needed in rs6000_emit_epilogue.. Sorry for the noise. -- Alan Modra Australia Development Lab, IBM