Add targeted_ops to core ExportPass passes for fast-copy optimization (#21702) - #21702
Open
apullin wants to merge 1 commit into
Open
Add targeted_ops to core ExportPass passes for fast-copy optimization (#21702)#21702apullin wants to merge 1 commit into
apullin wants to merge 1 commit into
Conversation
apullin
requested review from
JacobSzwejbka,
digantdesai and
larryliu0820
as code owners
August 9, 2026 22:51
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/21702
Note: Links to docs will display an error until the docs builds have been completed. ⏳ No Failures, 1 PendingAs of commit 7382466 with merge base 4ce2ec2 ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
meta-cla
Bot
added
the
CLA Signed
Contributor
This PR needs a
|
meta-codesync
Bot
changed the title
apullin
added a commit
to apullin/executorch
that referenced
this pull request
Aug 10, 2026
…pytorch#21702) Summary: Annotate four core `ExportPass` subclasses across three files with `targeted_ops` so they participate in the cold-node fast-copy path introduced by parent D97528110. When fast-copy remains safe, D975 copies eligible non-targeted `call_function` nodes with existing `val` metadata instead of replaying them through FakeTensor dispatch. This diff does not add generic pass skipping: every pass still enters replay, and hot nodes continue through each existing `call_operator` implementation. Each set exactly mirrors the rewrite guard already present in its pass: - `MemoryFormatOpsPass` targets `DimOrderOpsMap` keys. - `DimOrderOpsRevertPass` targets `MemoryFormatOpsMap` keys. - `NormalizeTransposePass` targets `aten.t.default`. - `RemoveMixedTypeOperators` targets `add.Tensor`, `mul.Tensor`, `sub.Tensor`, both tensor `div` overloads, and `minimum.default`. The diff is now stacked directly on D97528110, which owns the consumer infrastructure. Its previous D109215317 parent left these attributes inert. Direct raw-graph regressions prove that transpose normalization and mixed-type rewriting still dispatch hot operators while eligible unrelated operators bypass `call_operator`; the mixed-type test also verifies inserted promotion casts and output dtypes. A/B against D97528110 on the same host, with each revision run twice: - Synthetic add/conv/fc/mul U55 suite: 79.000s / 79.516s here versus 81.238s / 82.371s in D975. Mean speedup 3.1%; warmed speedup 3.5%. - CombinedControl U55 LOWERING: 145.801s / 145.888s here versus 148.465s / 149.726s in D975. Mean speedup 2.2%; warmed speedup 2.6%. Differential Revision: D96247716
apullin
force-pushed
the
export-D96247716
branch
from
August 10, 2026 00:05
059fc38 to
08dd81c
Compare
apullin
added a commit
to apullin/executorch
that referenced
this pull request
Aug 10, 2026
…pytorch#21702) Summary: Annotate four core `ExportPass` subclasses across three files with `targeted_ops` so they participate in the cold-node fast-copy path introduced by parent D97528110. When fast-copy remains safe, D975 copies eligible non-targeted `call_function` nodes with existing `val` metadata instead of replaying them through FakeTensor dispatch. This diff does not add generic pass skipping: every pass still enters replay, and hot nodes continue through each existing `call_operator` implementation. Each set exactly mirrors the rewrite guard already present in its pass: - `MemoryFormatOpsPass` targets `DimOrderOpsMap` keys. - `DimOrderOpsRevertPass` targets `MemoryFormatOpsMap` keys. - `NormalizeTransposePass` targets `aten.t.default`. - `RemoveMixedTypeOperators` targets `add.Tensor`, `mul.Tensor`, `sub.Tensor`, both tensor `div` overloads, and `minimum.default`. The diff is now stacked directly on D97528110, which owns the consumer infrastructure. Its previous D109215317 parent left these attributes inert. Direct raw-graph regressions prove that transpose normalization and mixed-type rewriting still dispatch hot operators while eligible unrelated operators bypass `call_operator`; the mixed-type test also verifies inserted promotion casts and output dtypes. A/B against D97528110 on the same host, with each revision run twice: - Synthetic add/conv/fc/mul U55 suite: 79.000s / 79.516s here versus 81.238s / 82.371s in D975. Mean speedup 3.1%; warmed speedup 3.5%. - CombinedControl U55 LOWERING: 145.801s / 145.888s here versus 148.465s / 149.726s in D975. Mean speedup 2.2%; warmed speedup 2.6%. Differential Revision: D96247716
apullin
force-pushed
the
export-D96247716
branch
from
August 10, 2026 00:28
08dd81c to
6caf061
Compare
apullin
force-pushed
the
export-D96247716
branch
from
September 2, 2026 14:53
6caf061 to
fed7327
Compare
apullin
added a commit
to apullin/executorch
that referenced
this pull request
Sep 2, 2026
…pytorch#21702) Summary: Pull Request resolved: pytorch#21702 Annotate four core `ExportPass` subclasses with `targeted_ops` matching their existing rewrite guards: - `MemoryFormatOpsPass` uses the `DimOrderOpsMap` keys. - `DimOrderOpsRevertPass` uses the `MemoryFormatOpsMap` keys. - `NormalizeTransposePass` targets `aten.t.default`. - `RemoveMixedTypeOperators` targets the operators handled by its promotion path. These declarations enable eligible non-targeted nodes to use the `ExportPass` fast-copy path, improving CombinedControl U55 lowering by ~3%. Differential Revision: D96247716
apullin
added a commit
to apullin/executorch
that referenced
this pull request
Sep 2, 2026
…pytorch#21702) Summary: Pull Request resolved: pytorch#21702 Annotate four core `ExportPass` subclasses with `targeted_ops` matching their existing rewrite guards: - `MemoryFormatOpsPass` uses the `DimOrderOpsMap` keys. - `DimOrderOpsRevertPass` uses the `MemoryFormatOpsMap` keys. - `NormalizeTransposePass` targets `aten.t.default`. - `RemoveMixedTypeOperators` targets the operators handled by its promotion path. These declarations enable eligible non-targeted nodes to use the `ExportPass` fast-copy path, improving CombinedControl U55 lowering by ~3%. Differential Revision: D96247716
apullin
force-pushed
the
export-D96247716
branch
from
September 2, 2026 14:59
fed7327 to
0ab6d1d
Compare
JakeStevens
removed their request for review
September 2, 2026 15:50
apullin
added a commit
to apullin/executorch
that referenced
this pull request
Sep 2, 2026
…pytorch#21702) Summary: Annotate four core `ExportPass` subclasses with `targeted_ops` matching their existing rewrite guards: - `MemoryFormatOpsPass` uses the `DimOrderOpsMap` keys. - `DimOrderOpsRevertPass` uses the `MemoryFormatOpsMap` keys. - `NormalizeTransposePass` targets `aten.t.default`. - `RemoveMixedTypeOperators` targets the operators handled by its promotion path. These declarations enable eligible non-targeted nodes to use the `ExportPass` fast-copy path, improving CombinedControl U55 lowering by ~3%. Differential Revision: D96247716
apullin
force-pushed
the
export-D96247716
branch
from
September 2, 2026 16:16
0ab6d1d to
8f27d21
Compare
apullin
added a commit
to apullin/executorch
that referenced
this pull request
Sep 2, 2026
…pytorch#21702) Summary: Annotate four core `ExportPass` subclasses with `targeted_ops` matching their existing rewrite guards: - `MemoryFormatOpsPass` uses the `DimOrderOpsMap` keys. - `DimOrderOpsRevertPass` uses the `MemoryFormatOpsMap` keys. - `NormalizeTransposePass` targets `aten.t.default`. - `RemoveMixedTypeOperators` targets the operators handled by its promotion path. These declarations enable eligible non-targeted nodes to use the `ExportPass` fast-copy path, improving CombinedControl U55 lowering by ~3%. Differential Revision: D96247716
apullin
force-pushed
the
export-D96247716
branch
from
September 2, 2026 16:18
8f27d21 to
48c0fdc
Compare
apullin
added a commit
to apullin/executorch
that referenced
this pull request
Sep 3, 2026
…pytorch#21702) Summary: Annotate four core `ExportPass` subclasses with `targeted_ops` matching their existing rewrite guards: - `MemoryFormatOpsPass` uses the `DimOrderOpsMap` keys. - `DimOrderOpsRevertPass` uses the `MemoryFormatOpsMap` keys. - `NormalizeTransposePass` targets `aten.t.default`. - `RemoveMixedTypeOperators` targets the operators handled by its promotion path. These declarations enable eligible non-targeted nodes to use the `ExportPass` fast-copy path, improving CombinedControl U55 lowering by ~3%. Differential Revision: D96247716
apullin
force-pushed
the
export-D96247716
branch
from
September 3, 2026 01:47
48c0fdc to
19286d6
Compare
apullin
force-pushed
the
export-D96247716
branch
from
September 3, 2026 02:07
19286d6 to
bf4b1c9
Compare
apullin
added a commit
to apullin/executorch
that referenced
this pull request
Sep 3, 2026
…pytorch#21702) Summary: Annotate four core `ExportPass` subclasses with `targeted_ops` matching their existing rewrite guards: - `MemoryFormatOpsPass` uses the `DimOrderOpsMap` keys. - `DimOrderOpsRevertPass` uses the `MemoryFormatOpsMap` keys. - `NormalizeTransposePass` targets `aten.t.default`. - `RemoveMixedTypeOperators` targets the operators handled by its promotion path. These declarations enable eligible non-targeted nodes to use the `ExportPass` fast-copy path, improving CombinedControl U55 lowering by ~3%. Differential Revision: D96247716
apullin
added a commit
to apullin/executorch
that referenced
this pull request
Sep 3, 2026
…pytorch#21702) Summary: Annotate four core `ExportPass` subclasses with `targeted_ops` matching their existing rewrite guards: - `MemoryFormatOpsPass` uses the `DimOrderOpsMap` keys. - `DimOrderOpsRevertPass` uses the `MemoryFormatOpsMap` keys. - `NormalizeTransposePass` targets `aten.t.default`. - `RemoveMixedTypeOperators` targets the operators handled by its promotion path. These declarations enable eligible non-targeted nodes to use the `ExportPass` fast-copy path, improving CombinedControl U55 lowering by ~3%. Differential Revision: D96247716
apullin
force-pushed
the
export-D96247716
branch
from
September 3, 2026 14:51
bf4b1c9 to
d0b4a09
Compare
apullin
added a commit
to apullin/executorch
that referenced
this pull request
Sep 6, 2026
…pytorch#21702) Summary: Annotate four core `ExportPass` subclasses with `targeted_ops` matching their existing rewrite guards: - `MemoryFormatOpsPass` uses the `DimOrderOpsMap` keys. - `DimOrderOpsRevertPass` uses the `MemoryFormatOpsMap` keys. - `NormalizeTransposePass` targets `aten.t.default`. - `RemoveMixedTypeOperators` targets the operators handled by its promotion path. These declarations enable eligible non-targeted nodes to use the `ExportPass` fast-copy path, improving CombinedControl U55 lowering by ~3%. Differential Revision: D96247716
apullin
force-pushed
the
export-D96247716
branch
from
September 6, 2026 15:19
d0b4a09 to
b0caa01
Compare
apullin
added a commit
to apullin/executorch
that referenced
this pull request
Sep 6, 2026
…pytorch#21702) Summary: Annotate four core `ExportPass` subclasses with `targeted_ops` matching their existing rewrite guards: - `MemoryFormatOpsPass` uses the `DimOrderOpsMap` keys. - `DimOrderOpsRevertPass` uses the `MemoryFormatOpsMap` keys. - `NormalizeTransposePass` targets `aten.t.default`. - `RemoveMixedTypeOperators` targets the operators handled by its promotion path. These declarations enable eligible non-targeted nodes to use the `ExportPass` fast-copy path, improving CombinedControl U55 lowering by ~3%. Differential Revision: D96247716
apullin
force-pushed
the
export-D96247716
branch
from
September 6, 2026 17:07
b0caa01 to
42b032b
Compare
apullin
added a commit
to apullin/executorch
that referenced
this pull request
Sep 6, 2026
…pytorch#21702) Summary: Annotate four core `ExportPass` subclasses with `targeted_ops` matching their existing rewrite guards: - `MemoryFormatOpsPass` uses the `DimOrderOpsMap` keys. - `DimOrderOpsRevertPass` uses the `MemoryFormatOpsMap` keys. - `NormalizeTransposePass` targets `aten.t.default`. - `RemoveMixedTypeOperators` targets the operators handled by its promotion path. These declarations enable eligible non-targeted nodes to use the `ExportPass` fast-copy path, improving CombinedControl U55 lowering by ~3%. Differential Revision: D96247716
apullin
force-pushed
the
export-D96247716
branch
from
September 6, 2026 18:28
42b032b to
4358b11
Compare
apullin
added a commit
to apullin/executorch
that referenced
this pull request
Sep 6, 2026
…pytorch#21702) Summary: Annotate four core `ExportPass` subclasses with `targeted_ops` matching their existing rewrite guards: - `MemoryFormatOpsPass` uses the `DimOrderOpsMap` keys. - `DimOrderOpsRevertPass` uses the `MemoryFormatOpsMap` keys. - `NormalizeTransposePass` targets `aten.t.default`. - `RemoveMixedTypeOperators` targets the operators handled by its promotion path. These declarations enable eligible non-targeted nodes to use the `ExportPass` fast-copy path, improving CombinedControl U55 lowering by ~3%. Differential Revision: D96247716
apullin
force-pushed
the
export-D96247716
branch
from
September 6, 2026 19:22
4358b11 to
6061c40
Compare
apullin
added a commit
to apullin/executorch
that referenced
this pull request
Sep 6, 2026
…pytorch#21702) Summary: Annotate four core `ExportPass` subclasses with `targeted_ops` matching their existing rewrite guards: - `MemoryFormatOpsPass` uses the `DimOrderOpsMap` keys. - `DimOrderOpsRevertPass` uses the `MemoryFormatOpsMap` keys. - `NormalizeTransposePass` targets `aten.t.default`. - `RemoveMixedTypeOperators` targets the operators handled by its promotion path. These declarations enable eligible non-targeted nodes to use the `ExportPass` fast-copy path, improving CombinedControl U55 lowering by ~3%. Differential Revision: D96247716
apullin
force-pushed
the
export-D96247716
branch
from
September 6, 2026 20:16
6061c40 to
49379af
Compare
apullin
added a commit
to apullin/executorch
that referenced
this pull request
Sep 6, 2026
…pytorch#21702) Summary: Annotate four core `ExportPass` subclasses with `targeted_ops` matching their existing rewrite guards: - `MemoryFormatOpsPass` uses the `DimOrderOpsMap` keys. - `DimOrderOpsRevertPass` uses the `MemoryFormatOpsMap` keys. - `NormalizeTransposePass` targets `aten.t.default`. - `RemoveMixedTypeOperators` targets the operators handled by its promotion path. These declarations enable eligible non-targeted nodes to use the `ExportPass` fast-copy path, improving CombinedControl U55 lowering by ~3%. Differential Revision: D96247716
apullin
force-pushed
the
export-D96247716
branch
from
September 6, 2026 20:46
49379af to
c248bf2
Compare
…pytorch#21702) Summary: Annotate four core `ExportPass` subclasses with `targeted_ops` matching their existing rewrite guards: - `MemoryFormatOpsPass` uses the `DimOrderOpsMap` keys. - `DimOrderOpsRevertPass` uses the `MemoryFormatOpsMap` keys. - `NormalizeTransposePass` targets `aten.t.default`. - `RemoveMixedTypeOperators` targets the operators handled by its promotion path. These declarations enable eligible non-targeted nodes to use the `ExportPass` fast-copy path, improving CombinedControl U55 lowering by ~3%. Differential Revision: D96247716
apullin
force-pushed
the
export-D96247716
branch
from
September 6, 2026 21:41
c248bf2 to
7382466
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary:
Annotate four core
ExportPasssubclasses withtargeted_opsmatching their existing rewrite guards:MemoryFormatOpsPassuses theDimOrderOpsMapkeys.DimOrderOpsRevertPassuses theMemoryFormatOpsMapkeys.NormalizeTransposePasstargetsaten.t.default.RemoveMixedTypeOperatorstargets the operators handled by its promotion path.These declarations enable eligible non-targeted nodes to use the
ExportPassfast-copy path, improving CombinedControl U55 lowering by ~3%.Differential Revision: D96247716