Hi Richard,
On Mon, Feb 11, 2019 at 08:27:32PM -0800, Richard Henderson wrote: > On 2/11/19 10:19 AM, Sven Schnelle wrote: > > - f = (orig & 4) / 4; > > - > > - return do_log_cond(c * 2 + f, res); > > Given that this used to reference do_log_cond, and you've fixed do_log_cond, > why is there any reason for a change here? You're right, this patch can be dropped. Regards Sven