> Are you talking about the intrinsics without digit suffix like vuzp_s8? They > looks like legacy for me. If yes, I think this is already covered. I have > test like the followings,
Ah, I see what's happening. A dummy ARM intrinsic that gets expanded to a proper shufflevector in CGBuiltins.cpp. That's fairly disgusting, but obviously has precedent. I may take a poke at it tomorrow if I get the chance. In the meantime, I think this patch is OK for now. Cheers. Tim. _______________________________________________ cfe-commits mailing list [email protected] http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
