On 03/30/2012 12:43 PM, Richard Henderson wrote: > On 03/30/2012 12:37 PM, Michael Matz wrote: >> Not yet regstrapped, so no rfa, but does this seem sane? > > It definitely seems plausible. I can't immediately think of > anything you might have forgotten.
Stating the obvious, but you'd do well to add at least an x86 test case that does a call to a local function which does throw. r~