Christophe Rhodes <csr21@...> writes:
> So, if a mips or alpha expert would stump up the few lines of code
> needed (and test them :-), that would be great; also, if anyone would
> like to test their threaded applications using the new code, that
> would be good too -- particularly if there were previously odd
> failures in calling generic functions.
I ended up writing mips and alpha versions myself. I tested the alpha
version, but the mips version I wrote is completely untested: I don't
even know if it compiles. (I'm pretty sure the assembly code I wrote
is fine; I'm more worried about the directives surrounding it...)
This was committed as sbcl-0.9.17.8: I hope that this improves the
stability of threaded CLOS-using applications.