Rewriting the lower-buffer-element-type pass to avoid unnecessary packing/unpacking. (#8526)

Part of the effort to improve the performance of generated SPIRV code.

The existing lower-buffer-element-type pass works by loading the entire
buffer element content from memory, and translate it to logical type
stored in a local variable at the earliest reference of a buffer handle.
This means that is can generate inefficient code that reads more than
necessary.

Consider this example:
```
struct BigStruct { bool values[1024]; }
ConstantBuffer<BigStruct> cb;

void test(BigStruct v)
{
      if (v.values[0]) { printf("ok"); }
}

[numthreads(1,1,1)]
void computeMain()
{
    test(cb);
}
```

In IR, the `computeMain` function before lower-buffer-element-type pass
is something like following:
```
func test:
   %v = param : BigStruct
   %barr = fieldExtract(%v, "values")
   %element = elementExtract(%barr, 0)
    ... // uses %element 

func computeMain:
  %v = load(cb)
  call %test %v
```

The existing lower-buffer-element-type pass will rewrite the bool array
in `BigStruct` into `int` array so it is legal in SPIRV. However, it
does so by inserting the translation on the first `load` of the constant
buffer:

```
struct BigStruct_std430 {
    int values[1024];
}
var cb : ConstantBuffer<BigStruct_std430>;
func computeMain:
   %tmpVar : var<BigStruct>
    call %unpackStorage(%tmpVar, cb)
   %v : BigStruct = load %tmpVar
   call %test %v
```

This means that the entire array will be loaded and translated to int,
before calling `test`, which only uses one element. It turns out that
the downstream compiler isn't always able to optimize out this
inefficient translation/copy.

This PR completely rewrites the way buffer-element-type lowering is
handled to avoid producing this inefficient code. It works in two parts:
first we turn on the `transformParamsToConstRef` pass for SPIRV target
as well, so we will translate the `test` function to take the `v`
parameter as `constref`. The second part is a redesigned
buffer-element-type pass that defers the storage-type to logical-type
translation until a value is actually used by a `load` instruction.

In this example, after `transformParamsToConstRef`, the IR is:

```
func test:
   %v = param : ConstRef<BigStruct>
   %barr = fieldAddr(%v, "values")
   %elementPtr = elementAddr(%barr, 0)
   %element = load(%elementPtr)
    ... // uses %element 

func computeMain:
  call %test %cb
```

The new `buffer-element-type-lowering` pass will take this IR, and
insert translation at latest possible time across the entire call graph,
and translate the IR into:

```
func test:
   %v = param : ConstRef<BigStruct_std430>
   %barr = fieldAddr(%v, "values")
   %elementPtr : ptr<int> = elementAddr(%barr, 0)
   %element_int = load(%elementPtr)
    %element = cast(%element_int) : %bool
    ... // uses %element 

func computeMain:
  call %test %cb
```

In this new IR, there is no longer a load and conversion of the entire
array.

See new comment in `slang-ir-lower-buffer-element-type.cpp` for more
details of how the pass works.

This PR also address many other issues surfaced by turning on
`transformParamsToConstRef` pass on SPIRV backend.

---------

Co-authored-by: slangbot <186143334+slangbot@users.noreply.github.com>

1 files changed, 3 insertions, 35 deletions

diff --git a/source/slang/slang-ir-defer-buffer-load.cpp b/source/slang/slang-ir-defer-buffer-load.cpp
index e71892fe5..51c6a161b 100644
--- a/source/slang/slang-ir-defer-buffer-load.cpp
+++ b/source/slang/slang-ir-defer-buffer-load.cpp
@@ -67,38 +67,6 @@ struct DeferBufferLoadContext
         return result;
     }
 
-    static bool isImmutableLocation(IRInst* loc)
-    {
-        switch (loc->getOp())
-        {
-        case kIROp_GetStructuredBufferPtr:
-        case kIROp_ImageSubscript:
-            return isImmutableLocation(loc->getOperand(0));
-        default:
-            break;
-        }
-
-        auto type = loc->getDataType();
-        if (!type)
-            return false;
-
-        switch (type->getOp())
-        {
-        case kIROp_HLSLStructuredBufferType:
-        case kIROp_HLSLByteAddressBufferType:
-        case kIROp_ConstantBufferType:
-        case kIROp_ParameterBlockType:
-            return true;
-        default:
-            break;
-        }
-
-        if (auto textureType = as<IRTextureType>(type))
-            return textureType->getAccess() == SLANG_RESOURCE_ACCESS_READ;
-
-        return false;
-    }
-
     static bool isImmutableBufferLoad(IRInst* inst)
     {
         // Note: we cannot defer loads from RWStructuredBuffer because there can be other
@@ -111,7 +79,7 @@ struct DeferBufferLoadContext
         case kIROp_Load:
             {
                 auto rootAddr = getRootAddr(inst->getOperand(0));
-                return isImmutableLocation(rootAddr);
+                return isPointerToImmutableLocation(rootAddr);
             }
         default:
             return false;
@@ -132,14 +100,14 @@ struct DeferBufferLoadContext
             align = load->findAttr<IRAlignedAttr>();
         if (!as<IRModuleInst>(ptr->getParent()))
         {
-            builder.setInsertAfter(ptr);
+            setInsertAfterOrdinaryInst(&builder, ptr);
             IRType* valueType = tryGetPointedToType(&builder, ptr->getFullType());
             result = builder.emitLoad(valueType, ptr, align);
             mapPtrToValue[ptr] = result;
         }
         else
         {
-            builder.setInsertBefore(loadInst);
+            setInsertBeforeOrdinaryInst(&builder, loadInst);
             IRType* valueType = tryGetPointedToType(&builder, ptr->getFullType());
             result = builder.emitLoad(valueType, ptr, align);
             // Since we are inserting the load in a local scope, we can't register