[GENERAL] Merged v1.0alpha into master. Added features are:

- A100 support via mma.16816
- Thread swizzling for conflict-free shared memory accesses without
padding
- Complete overhaul of the LLVM code generation in
codegen/selection/generator.cc to remove overengineering
- Added debugging capabilities in the Python binding
- Compilation error for kernels that spill

lib/codegen/transform/coalesce.cc

@@ -66,7 +66,7 @@ void coalesce::run(ir::module &mod) {
 
 
   for(size_t id = 0; id < num_groups; id++) {
-    if(!layout_->get(id)->to_mma884())
+    if(!layout_->get(id)->to_mma())
       continue;
     // extract memory stores
     const auto& values = layout_->values_of(id);