[MLIR][TORCH] Only unroll prim loop-like ops within a `torch.shape.calculate` region (#3812)

Reports a match failure for the pattern `FullyUnrollPrimLoop` when the
loop op is not in a region defined by a `torch.shape.calculate` op.

This is needed to avoid unrolling prim loops generated by ONNX IR, since
we are applying shape refinement in the
`torch-onnx-to-torch-backend-pipeline` introduced in fa4794d .

See also the discussion in
<https://github.com/iree-org/iree/pull/18867#discussion_r1811101655>
pull/3790/head
zjgarvey 2024-10-23 03:08:55 -05:00 committed by GitHub
parent aca33f1742
commit 55ff110dc2
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
2 changed files with 23 additions and 3 deletions

View File

@ -32,9 +32,6 @@ public:
} // namespace } // namespace
namespace { namespace {
// TODO: Only unroll inside the shape calculation region.
// Maybe do this by only applying patterns and folding greedily on the ops
// inside the region + the shape.calculate op itself?
class FullyUnrollPrimLoopOp : public OpRewritePattern<PrimLoopOp> { class FullyUnrollPrimLoopOp : public OpRewritePattern<PrimLoopOp> {
public: public:
using OpRewritePattern::OpRewritePattern; using OpRewritePattern::OpRewritePattern;
@ -42,6 +39,12 @@ public:
PatternRewriter &rewriter) const override { PatternRewriter &rewriter) const override {
Location loc = op->getLoc(); Location loc = op->getLoc();
MLIRContext *context = op->getContext(); MLIRContext *context = op->getContext();
// Only unroll loops if they are contained in a shape calculate region.
Region *region = op->getParentRegion();
Operation *parentOp = region->getParentOp();
if (!parentOp || !isa<Torch::ShapeCalculateOp>(parentOp))
return rewriter.notifyMatchFailure(
op, "Loop is not contained in a shape calculation region.");
if (!op.isForLike()) if (!op.isForLike())
return rewriter.notifyMatchFailure(op, "Loop is not for-like"); return rewriter.notifyMatchFailure(op, "Loop is not for-like");
int64_t maxTripCount; int64_t maxTripCount;

View File

@ -152,6 +152,23 @@ func.func @fully_unroll_prim_loop$no_unroll(%arg0: !torch.vtensor, %arg1: !torch
return %0 : !torch.vtensor return %0 : !torch.vtensor
} }
// CHECK-LABEL: func.func @fully_unroll_prim_loop$outside_region(
// CHECK: %[[LOOP:.*]] = torch.prim.Loop
func.func @fully_unroll_prim_loop$outside_region(%arg0: !torch.vtensor, %arg1: !torch.list<int>, %arg2: !torch.int) -> !torch.vtensor {
%true = torch.constant.bool true
%0 = torch.prim.Loop %arg2, %true, init(%arg0) {
^bb0(%arg3: !torch.int, %arg4: !torch.vtensor):
%1 = torch.shape.calculate {
torch.shape.calculate.yield %arg4 : !torch.vtensor
} shapes {
torch.prim.Print(%arg3) : !torch.int
torch.shape.calculate.yield.shapes %arg1 : !torch.list<int>
} : !torch.vtensor
torch.prim.Loop.condition %true, iter(%1 : !torch.vtensor)
} : (!torch.int, !torch.bool, !torch.vtensor) -> !torch.vtensor
return %0 : !torch.vtensor
}
// CHECK-LABEL: func.func @abstractly_interpret_list_ops$basic( // CHECK-LABEL: func.func @abstractly_interpret_list_ops$basic(
// CHECK-SAME: %[[ARG0:.*]]: !torch.vtensor, // CHECK-SAME: %[[ARG0:.*]]: !torch.vtensor,
// CHECK-SAME: %[[ARG1:.*]]: !torch.int, // CHECK-SAME: %[[ARG1:.*]]: !torch.int,