Expand description
Original PTX specification:
tcgen05.commit.cta_group.completion_mechanism{.shared::cluster}{.multicast}.b64 [mbar] {, ctaMask}; .completion_mechanism = { .mbarrier::arrive::one }; .cta_group = { .cta_group::1, .cta_group::2 }; .multicast = { .multicast::cluster };