(opt): remove cancel_ops phase, subsumed by variable forwarding - #10015

@orizi reviewed 32 files and all commit messages, and made 3 comments.
Reviewable status: 32 of 46 files reviewed, 3 unresolved discussions (waiting on eytan-starkware, ilyalesokhin-starkware, and TomerStarkware).

crates/cairo-lang-lowering/src/optimizations/test_data/cancel_ops at r1 (raw file):
wire to forwarding.
move or delete in following PR.

crates/cairo-lang-semantic/src/expr/expansion_test_data/inline_macros line 272 at r1 (raw file):

 --> lib.cairo:6:1
take_ident!($)
^^^^^^^^^^^^^^

revert

tests/test_data/fib_array.sierra line 114 at r1 (raw file):

store_temp<Box<felt252>>([20]) -> ([20]);
unbox<felt252>([20]) -> ([22]);
snapshot_take<Array<felt252>>([8]) -> ([23], [24]);

regression.

@orizi reviewed 12 files.
Reviewable status: 44 of 46 files reviewed, 3 unresolved discussions (waiting on eytan-starkware, ilyalesokhin-starkware, and TomerStarkware).

@eytan-starkware made 3 comments.
Reviewable status: 44 of 47 files reviewed, 3 unresolved discussions (waiting on ilyalesokhin-starkware, orizi, and TomerStarkware).

crates/cairo-lang-lowering/src/optimizations/test_data/cancel_ops at r1 (raw file):

Done.

crates/cairo-lang-semantic/src/expr/expansion_test_data/inline_macros line 272 at r1 (raw file):

Done.

tests/test_data/fib_array.sierra line 114 at r1 (raw file):

Yes. Acceptable no?

@orizi reviewed 9 files and all commit messages, made 1 comment, and resolved 2 discussions.
Reviewable status: all files reviewed, 1 unresolved discussion (waiting on eytan-starkware, ilyalesokhin-starkware, and TomerStarkware).

tests/test_data/fib_array.sierra line 114 at r1 (raw file):

mostly depends on why this happens.

@eytan-starkware made 1 comment.
Reviewable status: all files reviewed, 1 unresolved discussion (waiting on ilyalesokhin-starkware, orizi, and TomerStarkware).

tests/test_data/fib_array.sierra line 114 at r1 (raw file):

I had claude check this, it is around how the removal cascading works (or doesn't in this case): For the fib_array case the producer is a lowering snapshot statement, which has two outputs (orig, snap). VF forwards the desnap consumers onto orig, which makes snap dead — but it can't DCE the snapshot statement because the other output (orig) is still live. A two-output op with one live output survives ordinary dead-code elimination. Sierra-gen then lowers that surviving statement to exactly what Ori flagged:

@orizi made 1 comment.
Reviewable status: all files reviewed, 1 unresolved discussion (waiting on eytan-starkware, ilyalesokhin-starkware, and TomerStarkware).

tests/test_data/fib_array.sierra line 114 at r1 (raw file):

it previously was removed by cancel ops?

@orizi reviewed 3 files and all commit messages.
Reviewable status: all files reviewed, 1 unresolved discussion (waiting on eytan-starkware, ilyalesokhin-starkware, and TomerStarkware).

@eytan-starkware made 1 comment.
Reviewable status: all files reviewed, 1 unresolved discussion (waiting on ilyalesokhin-starkware, orizi, and TomerStarkware).

tests/test_data/fib_array.sierra line 114 at r1 (raw file):

Yes

@eytan-starkware made 1 comment.
Reviewable status: all files reviewed, 1 unresolved discussion (waiting on ilyalesokhin-starkware, orizi, and TomerStarkware).

tests/test_data/fib_array.sierra line 114 at r1 (raw file):

Though only in combination with VF. Apperantly before VF it didn't

@orizi made 1 comment and resolved 1 discussion.
Reviewable status: complete! all files reviewed, all discussions resolved (waiting on ilyalesokhin-starkware and TomerStarkware).