mirror of
https://github.com/RPCS3/llvm-mirror.git
synced 2024-11-22 10:42:39 +01:00
ef6c8d6574
Noticed while investigating PR50364, the truncation costs for v4i64->v4i16/v4i8 and v8i32->v8i8 were way too optimistic for a shuffle sequence that usually matches the AVX1 codegen (they matched AVX512 numbers which have actual truncation instructions!). |
||
---|---|---|
.. | ||
AliasSet | ||
AssumptionCache | ||
BasicAA | ||
BlockFrequencyInfo | ||
BranchProbabilityInfo | ||
CallGraph | ||
CFLAliasAnalysis | ||
CostModel | ||
DDG | ||
Delinearization | ||
DemandedBits | ||
DependenceAnalysis | ||
DivergenceAnalysis | ||
DominanceFrontier | ||
Dominators | ||
FunctionPropertiesAnalysis | ||
GlobalsModRef | ||
IRSimilarityIdentifier | ||
IVUsers | ||
LazyCallGraph | ||
LazyValueAnalysis | ||
LegacyDivergenceAnalysis | ||
Lint | ||
LoopAccessAnalysis | ||
LoopCacheAnalysis/PowerPC | ||
LoopInfo | ||
LoopNestAnalysis | ||
MemoryDependenceAnalysis | ||
MemorySSA | ||
MustExecute | ||
PhiValues | ||
PostDominators | ||
ProfileSummary | ||
RegionInfo | ||
ScalarEvolution | ||
ScopedNoAliasAA | ||
StackSafetyAnalysis | ||
TypeBasedAliasAnalysis | ||
ValueTracking | ||
alias-analysis-uses.ll |