Hi Segher, > On Thu, Jan 16, 2020 at 12:50:14PM +0000, Wilco Dijkstra wrote: >> The separate shrinkwrapping pass may insert stores in the middle >> of atomics loops which can cause issues on some implementations. >> Avoid this by delaying splitting of atomic patterns until after >> prolog/epilog generation. > > Note that this isn't specific to sws at all: there isn't anything > stopping later passes from doing this either. Is there anything that > protects us from sched2 doing similar here, for example?
The expansions create extra basic blocks and insert various barriers that would stop any reasonable scheduler from doing it. And the current scheduler is basic block based. Wilco
