diff options
| author | Yong He <yonghe@outlook.com> | 2025-09-29 17:45:08 -0700 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2025-09-30 00:45:08 +0000 |
| commit | a6deb5ed82cb8fc6b4f4c5c5fee264e09f97ff89 (patch) | |
| tree | 1c374bd52498cad2e142e3c7f5482fd42dca966f /source/slang/slang-ir-specialize-address-space.cpp | |
| parent | 2827c94de5901cac42a67f73a78ab2548771b28c (diff) | |
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>
Diffstat (limited to 'source/slang/slang-ir-specialize-address-space.cpp')
| -rw-r--r-- | source/slang/slang-ir-specialize-address-space.cpp | 39 |
1 files changed, 21 insertions, 18 deletions
diff --git a/source/slang/slang-ir-specialize-address-space.cpp b/source/slang/slang-ir-specialize-address-space.cpp index 29f1ec516..c4a155eec 100644 --- a/source/slang/slang-ir-specialize-address-space.cpp +++ b/source/slang/slang-ir-specialize-address-space.cpp @@ -168,6 +168,7 @@ struct AddressSpaceContext : public AddressSpaceSpecializationContext { case kIROp_Var: case kIROp_RWStructuredBufferGetElementPtr: + case kIROp_Load: { // The address space of these insts should be assigned by the initial // address space assigner. @@ -204,16 +205,6 @@ struct AddressSpaceContext : public AddressSpaceSpecializationContext } } break; - case kIROp_Load: - { - if (auto addrSpace = - mapVarValueToAddrSpace.tryGetValue(inst->getOperand(0))) - { - mapInstToAddrSpace[inst] = *addrSpace; - changed = true; - } - } - break; case kIROp_Param: if (!isFirstBlock) { @@ -248,22 +239,24 @@ struct AddressSpaceContext : public AddressSpaceSpecializationContext if (callee) { List<AddressSpace> argAddrSpaces; - bool fullySpecialized = true; + bool hasSpecializableArg = false; for (UInt i = 0; i < callInst->getArgCount(); i++) { auto arg = callInst->getArg(i); - auto argAddrSpace = getAddrSpace(arg); argAddrSpaces.add(getAddrSpace(arg)); - if (argAddrSpace == AddressSpace::Generic && - as<IRPtrTypeBase>(arg->getDataType())) + if (as<IRPtrTypeBase>(arg->getDataType())) { - fullySpecialized = false; - break; + hasSpecializableArg = true; } } - if (!fullySpecialized) + if (!hasSpecializableArg) + { + workList.add(callee); + break; + } + // If callee doesn't have a body, don't specialize. + if (!callee->getFirstBlock()) break; - FuncSpecializationKey key(callee, argAddrSpaces); IRFunc* specializedCallee = nullptr; if (IRFunc** specializedFunc = @@ -484,4 +477,14 @@ void propagateAddressSpaceFromInsts(List<IRInst*>&& workList) } } +AddressSpace NoOpInitialAddressSpaceAssigner::getAddressSpaceFromVarType(IRInst*) +{ + return AddressSpace::Generic; +} + +AddressSpace NoOpInitialAddressSpaceAssigner::getLeafInstAddressSpace(IRInst*) +{ + return AddressSpace::Generic; +} + } // namespace Slang |
