Skip to content

Drop inlining on shardy. #1204

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
copybara-service merged 1 commit into from
Apr 28, 2026
Merged

Drop inlining on shardy. #1204

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
11 changes: 6 additions & 5 deletions shardy/dialect/sdy/transforms/import/add_data_flow_edges.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -38,9 +38,10 @@ using func::FuncOp;

// Adds func input and output data flow edges. Adds func input data flow edges
// only for non-main funcs.
void addFuncDataFlowEdges(ModuleOp moduleOp, const SymbolTable& symbolTable,
IRRewriter& rewriter) {
FuncOp mainFuncOp = getMainFuncOrDie(moduleOp, symbolTable);
void addFuncDataFlowEdgeOps(ModuleOp moduleOp, const SymbolTable& symbolTable,
IRRewriter& rewriter) {
FuncOp mainFuncOp =
getMainFuncOrDie(moduleOp, symbolTable, /*useSingleFunc=*/true);
moduleOp.walk([&](FuncOp funcOp) {
if (funcOp == mainFuncOp) {
return;
Expand Down Expand Up @@ -84,8 +85,8 @@ struct AddDataFlowEdgesPass
addDataFlowEdges(op.getBlockArgumentEdgeOwners(), rewriter);
addDataFlowEdges(op.getOpResultEdgeOwners(), rewriter);
});
if (enableNativeNonFlatSupport) {
addFuncDataFlowEdges(moduleOp, symbolTable, rewriter);
if (addFuncDataFlowEdges) {
addFuncDataFlowEdgeOps(moduleOp, symbolTable, rewriter);
}
}
};
Expand Down
2 changes: 1 addition & 1 deletion shardy/dialect/sdy/transforms/import/import_pipeline.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -43,7 +43,7 @@ void addImportPipeline(OpPassManager& pm, int& dumpIndex,
options.dumpDirectory, "before_propagation", dumpIndex++));

pm.addPass(createAddDataFlowEdgesPass(
AddDataFlowEdgesPassOptions{options.enableNativeNonFlatSupport}));
AddDataFlowEdgesPassOptions{/*addFuncDataFlowEdges=*/true}));
pm.addPass(
createApplyShardingConstraintsPass(ApplyShardingConstraintsPassOptions{
options.debugShardingOrigins, options.debugPropagationEdgeSharding}));
Expand Down
6 changes: 2 additions & 4 deletions shardy/dialect/sdy/transforms/import/passes.td
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -92,11 +92,9 @@ def AddDataFlowEdgesPass : Pass<"sdy-add-data-flow-edges", "ModuleOp"> {
let dependentDialects = ["mlir::sdy::SdyDialect"];

let options = [
Option<"enableNativeNonFlatSupport", "enable-native-non-flat-support", "bool",
Option<"addFuncDataFlowEdges", "add-func-data-flow-edges", "bool",
/*default=*/"false",
"Whether to propagate shardings directly on a non-flat graph without "
"flattening it. The default is false, meaning it will flatten the "
"graph and then propagate.">
"Whether to add func data flow edges.">
];
}

Expand Down
2 changes: 1 addition & 1 deletion ...edges_enable_native_non_flat_support.mlir → ..._edges_add_func_data_flow_edges_true.mlir
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,4 +1,4 @@
// RUN: sdy_opt %s -split-input-file -sdy-add-data-flow-edges='enable-native-non-flat-support=true' | FileCheck %s
// RUN: sdy_opt %s -split-input-file -sdy-add-data-flow-edges='add-func-data-flow-edges=true' | FileCheck %s

// CHECK-LABEL: @bar(%arg0: tensor<8xf32>)
func.func @bar(%arg0: tensor<8xf32>) -> tensor<8xf32> {
Expand Down
45 changes: 34 additions & 11 deletions shardy/dialect/sdy/transforms/import/test/import_pipeline.mlir
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -32,12 +32,17 @@ func.func @main(%arg0: tensor<8x8xf32>, %arg1: tensor<8x8xf32>) {

sdy.mesh @mesh = <["c"=2, "a"=2, "b"=2]>

// CHECK-LABEL: @add_manual_axes_to_replicated
func.func @add_manual_axes_to_replicated(%arg0: tensor<8xf32>) -> tensor<8xf32> {
// CHECK-NEXT: sdy.manual_computation(%arg0)
// CHECK-LABEL: func @main
func.func @main(%arg0: tensor<8xf32>) -> tensor<8xf32> {
// CHECK-NEXT: %0 = sdy.manual_computation(%arg0)
// CHECK-SAME{LITERAL}: in_shardings=[<@mesh, [{"c", ?}], replicated={"a"}>]
// CHECK-SAME{LITERAL}: out_shardings=[<@mesh, [{"c", ?}], replicated={"a"}>]
// CHECK-SAME{LITERAL}: manual_axes={"c", "a"} (%arg1: tensor<4xf32>) {
// CHECK-NEXT: %2 = sdy.data_flow_edge %arg1 sharding=<@mesh, [{?}]>
// CHECK-NEXT: sdy.return %2
// CHECK-NEXT: }
// CHECK-NEXT: %1 = sdy.data_flow_edge %0 sharding=<@mesh, [{"c", ?}], replicated={"a"}>
// CHECK-NEXT: return %1
%0 = sdy.manual_computation(%arg0) in_shardings=[<@mesh, [{"c", ?}]>] out_shardings=[<@mesh, [{"c", ?}]>] manual_axes={"c", "a"} (%arg1: tensor<4xf32>) {
sdy.return %arg1 : tensor<4xf32>
} : (tensor<8xf32>) -> tensor<8xf32>
Expand All @@ -50,10 +55,17 @@ sdy.mesh @mesh = <["c"=2, "a"=2, "b"=2]>

// Due to the in_sharding being fully closed, the in_sharding is added to the
// func arg but with the manual axis added as replicated.
// CHECK-LABEL: @add_manual_axes_to_replicated_applied_constraint
// CHECK-LABEL: func @main
// CHECK-SAME %arg0: tensor<16x16xf32> {sdy.sharding = #sdy.sharding<@mesh, [{"c"}, {}], replicated={"a"}>}
// CHECK-SAME -> tensor<16x16xf32> {
func.func @add_manual_axes_to_replicated_applied_constraint(%arg0: tensor<8xf32>) -> tensor<8xf32> {
func.func @main(%arg0: tensor<8xf32>) -> tensor<8xf32> {
// CHECK-NEXT: %0 = sdy.manual_computation(%arg0) in_shardings=[<@mesh, [{"c"}], replicated={"a"}>] out_shardings=[<@mesh, [{"c"}], replicated={"a"}>] manual_axes={"c", "a"} (%arg1: tensor<4xf32>) {
// CHECK-NEXT: %2 = sdy.data_flow_edge %arg1 sharding=<@mesh, [{}]>
// CHECK-NEXT: %3 = stablehlo.add %2, %2
// CHECK-NEXT: sdy.return %3
// CHECK-NEXT: }
// CHECK-NEXT: %1 = sdy.data_flow_edge %0 sharding=<@mesh, [{"c"}], replicated={"a"}>
// CHECK-NEXT: return %1
%0 = sdy.manual_computation(%arg0) in_shardings=[<@mesh, [{"c"}]>] out_shardings=[<@mesh, [{"c"}]>] manual_axes={"c", "a"} (%arg1: tensor<4xf32>) {
%1 = stablehlo.add %arg1, %arg1 : tensor<4xf32>
sdy.return %1 : tensor<4xf32>
Expand All @@ -67,28 +79,39 @@ sdy.mesh @mesh = <["a"=2]>

// This test verifies that the manual axes are cleaned up before adding data
// flow edges.
func.func @manual_axes_cleanup_before_adding_data_flow_edges(%arg0: tensor<8xf32>) -> tensor<8xf32> {
// CHECK-LABEL: func @main
func.func @main(%arg0: tensor<8xf32>) -> tensor<8xf32> {
// CHECK-NEXT: %0 = sdy.manual_computation(%arg0) in_shardings=[<@mesh, [{?}], replicated={"a"}>] out_shardings=[<@mesh, [{?}], replicated={"a"}>] manual_axes={"a"} (%arg1: tensor<8xf32>) {
// CHECK-NEXT: %2 = sdy.data_flow_edge %arg1 sharding=<@mesh, [{?}]>
// CHECK-NEXT: %3 = stablehlo.add %2, %2
// CHECK-NEXT: sdy.return %3
// CHECK-NEXT: }
// CHECK-NEXT: %1 = sdy.data_flow_edge %0 sharding=<@mesh, [{?}], replicated={"a"}>
// CHECK-NEXT: return %1
%0 = sdy.manual_computation(%arg0) in_shardings=[<@mesh, [{?}]>] out_shardings=[<@mesh, [{?}]>] manual_axes={"a"} (%arg1: tensor<8xf32>) {
%1 = stablehlo.add %arg1, %arg1 : tensor<8xf32>
sdy.return %1 : tensor<8xf32>
} : (tensor<8xf32>) -> tensor<8xf32>
// CHECK: sdy.data_flow_edge %0 sharding=<@mesh, [{?}], replicated={"a"}> : tensor<8xf32>
return %0 : tensor<8xf32>
}

// -----

sdy.mesh @mesh = <["a"=2]>

// CHECK-LABEL: func @single_call
func.func @single_call(%arg0: tensor<8xf32>) -> tensor<8xf32> {
// CHECK-NEXT: %0 = call @foo(%arg0) : (tensor<8xf32>) -> tensor<8xf32>
// CHECK-NEXT: return %0 : tensor<8xf32>
// test: single_call
// CHECK-LABEL: func @main
func.func @main(%arg0: tensor<8xf32>) -> tensor<8xf32> {
// CHECK-NEXT: %[[CALL:.*]] = call @foo(%arg0) : (tensor<8xf32>) -> tensor<8xf32>
// CHECK-NEXT: %[[EDGE:.*]] = sdy.func_data_flow_edge %[[CALL]] : tensor<8xf32>
// CHECK-NEXT: return %[[EDGE]] : tensor<8xf32>
%0 = call @foo(%arg0) : (tensor<8xf32>) -> tensor<8xf32>
return %0 : tensor<8xf32>
}

// CHECK-LABEL: func private @foo
func.func private @foo(%arg0: tensor<8xf32> {sdy.sharding = #sdy.sharding<@mesh, [{"a"}]>}) -> tensor<8xf32> {
// CHECK-NEXT: %0 = sdy.func_data_flow_edge %arg0 {sdy.sharding = #sdy.sharding_per_value<[<@mesh, [{"a"}]>]>}
// CHECK-NEXT: return %0
return %arg0 : tensor<8xf32>
}
1 change: 1 addition & 0 deletions shardy/dialect/sdy/transforms/propagation/BUILD
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -57,6 +57,7 @@ cc_library(
":sharding_projection",
":utils",
"//shardy/common:file_utils",
"//shardy/common:logging",
"//shardy/dialect/sdy/ir:dialect",
"//shardy/dialect/sdy/transforms/common:op_properties",
"//shardy/dialect/sdy/transforms/common:propagation_options",
Expand Down
Loading
Loading