mlir

In ODS we currently use the following all over the place: let printer = "return ::print(p, *this);"; Since this relies on C++'s function overloading, when it can't find a print for a particular operation, all of the visible print methods get dumped into the error message.

We should switch over to let printer = "return ::print$cppClass(p, *this);"; to disambiguate by name and avoid the lon

@honglh added optimized kernels for ARM32 in #432. It would be great if we could add support for bitpacked activations to them to match the AArch64 optimized bgemm kernels.