Something's still fishy here. or1k_compare_op[01] are never assigned anywhere, as you said, but or1k_output_bf and or1k_output_cmov refer to them. It seems that there should be segfaults in those functions, too, but there isn't, which is odd.
I want to look at this a bit more before I apply it. -Pete On Fri, May 25, 2012 at 12:24 PM, Stefan Kristiansson <[email protected]> wrote: > On Fri, May 25, 2012 at 12:02:02PM +0200, Olof Kindgren wrote: >> >> Hmm... is this related to good ol' bug bug 27? Would be interesting to see >> if your changes affect that one too. >> > > Yes, this ought to kill good ol' bug 27 > > Stefan _______________________________________________ OpenRISC mailing list [email protected] http://lists.openrisc.net/listinfo/openrisc
