On 11/14/14 08:12, Maxim Kuvyrkov wrote:
...
Can this be built on top of Bin's work for insn fusion? There's a
lot of commonality in the structure of the insns you care about.
He's already got a nice little priority function that I think you
could utilize to to ensure the insns with smaller offsets fire
first.
I would argue that macro-fusion should have been implemented the way
autopref_model is -- via
targetm.sched.first_cycle_multipass_dfa_lookahead_guard hook. To
implement the autopref model I cleaned up and generalized existing
infrastructure (max_issue and dfa_lookahead_guard hook) instead of
adding yet another decision-making primitive to the scheduler.
Fair enough. We can put unifying the two implementations on the queue
for next stage 1.
My biggest concern would be sched2 coming along and undoing that
work since you're not going to fuse those into move-multiple types
of instructions.
The autoprefetcher will be active only during sched2. It is disabled
during sched1 by the fact that max_issue is not used when scheduling
for register pressure.
Sorry, should have been clearer, that comment was in reference to using
the fusion stuff to handle prefetching. It was not in reference to your
change.
Vlad's approval and my questions crossed last night, I don't have any
objections to this going in, though I do want us to look at unifying the
memory pairing and prefetching stuff during the next stage1.
Jeff
