Displaying 2 results from an estimated 2 matches for "shouldreducelatency".
2019 Sep 10
2
MachineScheduler not scheduling for latency
...<< OtherCritCount /
SchedModel->getMicroOpFactor() << '\n');
+ }
for (unsigned PIdx = 1, PEnd = SchedModel->getNumProcResourceKinds();
PIdx != PEnd; ++PIdx) {
unsigned OtherCount = getResourceCount(PIdx) + Rem->RemainingCounts[PIdx];
As for "shouldReduceLatency should not be relevant at
MicroOpBufferSize = 1": are you suggesting that shouldReduceLatency
should effectively be changed to always return true on in-order CPUs?
Even with that change, latency comes pretty far down the list of
criteria tested in GenericScheduler::tryCandidate.
Thanks,
Jay....
2019 Sep 09
2
Fwd: MachineScheduler not scheduling for latency
Hi,
I'm trying to understand why MachineScheduler does a poor job in
straight line code in cases like the one in the attached debug dump.
This is on AMDGPU, an in-order target, and the problem is that the
IMAGE_SAMPLE instructions have very high (80 cycle) latency, but in
the resulting schedule they are often placed right next to their uses
like this:
1784B %140:vgpr_32 =