Robin Dapp <rdapp....@gmail.com> writes: >> - Change the second mode to vec_extract_optab. This is only a name >> lookup, and it seems more natural to continue using the real element mode. > > Am I understanding correctly that this implies we should provide > a vec_extract<bimode>bi expander? (with the innermode being BImode > here).
Yeah, I think so. That way the interpretation of the mode stays in sync with the interpretation of the bit position. If instead we used QImode with a bitnum of <number of elements - 1>, the top 7 bits of the read would logically be out of bounds. Thanks, Richard
