alex-t a236991319 [AMDGPU] PHI node cost should not be counted for the size and latency. ... Details: https://reviews.llvm.org/D96805 changed the GCNTTIImpl::getCFInstrCost to return 1 for the PHI nodes for the TTI::TCK_CodeSize and TTI::TCK_SizeAndLatency. This is incorrect because the value moves that are the result of the PHI lowering are inserted into the basic block predecessors - not into the block itself. As a result of this change LoopRotate and LoopUnroll were broken because of the incorrect Loop header and loop body size/cost estimation. Reviewed By: rampitec Differential Revision: https://reviews.llvm.org/D105104		2021-06-30 16:11:17 +03:00
..
add-sub.ll
addrspacecast.ll
bit-ops.ll
control-flow.ll	[AMDGPU] PHI node cost should not be counted for the size and latency.	2021-06-30 16:11:17 +03:00
extractelement.ll
fabs.ll
fadd.ll
fdiv.ll
fma.ll
fmul.ll
fneg.ll
fsub.ll
fused_costs.ll
insertelement.ll
lit.local.cfg
logicalop.ll	[TTI] Consider select form of and/or i1 as having arithmetic cost	2021-03-02 02:18:19 +09:00
mul.ll
reduce-and.ll	[Cost]Canonicalize the cost for logical or/and reductions.	2021-03-19 11:01:58 -07:00
reduce-or.ll	[Cost]Canonicalize the cost for logical or/and reductions.	2021-03-19 11:01:58 -07:00
shifts.ll
shufflevector.ll	[InstructionCost] Don't conflate Invalid costs with Unknown costs.	2021-03-30 09:29:42 +01:00