On Sun, Aug 3, 2025 at 7:31 AM Andi Kleen <a...@linux.intel.com> wrote: > > "H.J. Lu" <hjl.to...@gmail.com> writes: > > > Don't hoist non all 0s/1s vector set outside of the loop to avoid extra > > spills. > > It seems this could be a loss if there are actually enough registers. > So you need to make it depend on the register pressure?
IMO further code motion is job of sched2 or invariant motion. Note even the transform as it remains can in odd circumstances cause spilling due to register pressure. That's a general issue with many transforms. As proposed I think the heuristics are sane. Richard. > > -Andi
