Teach SCEVExpander to expand arithmetic involving pointers into GEP

instructions. It attempts to create high-level multi-operand GEPs,
though in cases where this isn't possible it falls back to casting
the pointer to i8* and emitting a GEP with that. Using GEP instructions
instead of ptrtoint+arithmetic+inttoptr helps pointer analyses that
don't use ScalarEvolution, such as BasicAliasAnalysis.

Also, make the AddrModeMatcher more aggressive in handling GEPs.
Previously it assumed that operand 0 of a GEP would require a register
in almost all cases. It now does extra checking and can do more
matching if operand 0 of the GEP is foldable. This fixes a problem
that was exposed by SCEVExpander using GEPs.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@72093 91177308-0d34-0410-b5e6-96231b3b80d8

1 files changed, 7 insertions, 0 deletions

diff --git a/lib/Analysis/ScalarEvolution.cpp b/lib/Analysis/ScalarEvolution.cpp
index eaa847aa10..0857014863 100644
--- a/lib/Analysis/ScalarEvolution.cpp
+++ b/lib/Analysis/ScalarEvolution.cpp
@@ -456,6 +456,13 @@ namespace {
       if (const SCEVUnknown *LU = dyn_cast<SCEVUnknown>(LHS)) {
         const SCEVUnknown *RU = cast<SCEVUnknown>(RHS);
 
+        // Order pointer values after integer values. This helps SCEVExpander
+        // form GEPs.
+        if (isa<PointerType>(LU->getType()) && !isa<PointerType>(RU->getType()))
+          return false;
+        if (isa<PointerType>(RU->getType()) && !isa<PointerType>(LU->getType()))
+          return true;
+
         // Compare getValueID values.
         if (LU->getValue()->getValueID() != RU->getValue()->getValueID())
           return LU->getValue()->getValueID() < RU->getValue()->getValueID();