Clean up variables before sroa

In some shaders there are a lot of very large and deeply nested
structures.  This creates a lot of work for scalar replacement.  Also,
since commit ca4457b we have been very aggressive as rewriting
variables.  This has causes a large increase in compile time in creating
and then deleting the instructions.

To help low the costs, I want to run a cleanup of some of the easy loads
and stores to remove.  This reduces the number of symbols sroa has to
work on.  It also reduces the amount of code the simplifier has to
simplify because it was not generated by sroa.

To confirm the improvement, I ran numbers on three different sets of
shaders:

Time to run --legalize-hlsl:

Set #1: 55.89s -> 12.0s
Set #2: 1m44s -> 1m40.5s
Set #3: 6.8s -> 5.7s

Time to run -O

Set #1: 18.8s -> 10.9s
Set #2: 5m44s -> 4m17s
Set #3: 7.8s -> 7.8s

Contributes to #1328.

diff --git a/source/opt/optimizer.cpp b/source/opt/optimizer.cpp
index fa2c767..0478d3a 100644 (file)
--- a/source/opt/optimizer.cpp
+++ b/source/opt/optimizer.cpp
@@ -95,6 +95,9 @@ Optimizer& Optimizer::RegisterLegalizationPasses() {
           // Make private variable function scope
           .RegisterPass(CreateEliminateDeadFunctionsPass())
           .RegisterPass(CreatePrivateToLocalPass())
+          // Propagate the value stored to the loads in very simple cases.
+          .RegisterPass(CreateLocalSingleBlockLoadStoreElimPass())
+          .RegisterPass(CreateLocalSingleStoreElimPass())
           // Split up aggragates so they are easier to deal with.
           .RegisterPass(CreateScalarReplacementPass())
           // Remove loads and stores so everything is in intermediate values.
@@ -123,6 +126,8 @@ Optimizer& Optimizer::RegisterPerformancePasses() {
       .RegisterPass(CreateMergeReturnPass())
       .RegisterPass(CreateInlineExhaustivePass())
       .RegisterPass(CreateAggressiveDCEPass())
+      .RegisterPass(CreateLocalSingleBlockLoadStoreElimPass())
+      .RegisterPass(CreateLocalSingleStoreElimPass())
       .RegisterPass(CreateScalarReplacementPass())
       .RegisterPass(CreateLocalAccessChainConvertPass())
       .RegisterPass(CreateLocalSingleBlockLoadStoreElimPass())