[FIRRTL] Replace remove-unused-ports with port-only IMDCE mode#10575
Merged
Conversation
uenoku
force-pushed
the
dev/hidetou/remove-unused-ports
branch
from
df39e83 to
85c55f9
Compare
uenoku
commented
Jun 2, 2026
| @@ -10,11 +10,11 @@ firrtl.circuit "Foo" { | |||
| // CHECK-NOT: %bar_e | |||
| // CHECK: %bar_b, %bar_d = firrtl.instance bar @Bar | |||
| %bar_a, %bar_b, %bar_c, %bar_d, %bar_e = firrtl.instance bar @Bar (in a: !firrtl.uint<1>, in b: !firrtl.uint<1>, out c: !firrtl.uint<1>, out d: !firrtl.uint<1>, out e: !firrtl.uint<1>) | |||
| firrtl.connect %bar_a, %x : !firrtl.uint<1>, !firrtl.uint<1> | |||
Member
Author
There was a problem hiding this comment.
connect was replaced with matching connect since IMDCE doesn't handle normal connect for the simplicity (IMDCE relies on getSingleConnectUserOf. OTOH RemoveUnusedPorts manually inspected use-def connection). I don't expect this restriction reduces capability of reducer.
seldridge
reviewed
Jun 2, 2026
Member
seldridge
left a comment
seldridge
left a comment
There was a problem hiding this comment.
Awesome. Thank you for picking this up! I'm happy to see RemoveUnusedPorts getting dropped.
Member
There was a problem hiding this comment.
Consider a file rename or moving this out of the reduction and into the main test/Dialect/FIRRTL/imdce-* suite of tests.
Member
Author
There was a problem hiding this comment.
I consider this as a test for --include firrtl-imdeadcodeelim-remove-ports so I'd like to keep under reduction. Fair point for file name.
Comment on lines
13
to
+38
| @@ -33,8 +33,8 @@ firrtl.circuit "Foo" { | |||
| ) { | |||
| %invalid_ui1 = firrtl.invalidvalue : !firrtl.uint<1> | |||
| %0 = firrtl.not %b : (!firrtl.uint<1>) -> !firrtl.uint<1> | |||
| firrtl.connect %c, %invalid_ui1 : !firrtl.uint<1>, !firrtl.uint<1> | |||
| firrtl.connect %d, %0 : !firrtl.uint<1>, !firrtl.uint<1> | |||
| firrtl.connect %e, %invalid_ui1 : !firrtl.uint<1>, !firrtl.uint<1> | |||
| firrtl.matchingconnect %c, %invalid_ui1 : !firrtl.uint<1> | |||
| firrtl.matchingconnect %d, %0 : !firrtl.uint<1> | |||
| firrtl.matchingconnect %e, %invalid_ui1 : !firrtl.uint<1> | |||
Member
There was a problem hiding this comment.
Same question about connect vs. matchingconnect.
uenoku
force-pushed
the
dev/hidetou/remove-unused-ports
branch
from
7466173 to
f95cda8
Compare
uenoku
force-pushed
the
dev/hidetou/remove-unused-ports
branch
from
f95cda8 to
4988f5a
Compare
seldridge
approved these changes
Jun 5, 2026
Co-authored-by: Schuyler Eldridge <schuyler.eldridge@gmail.com>
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
2 participants
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.
This PR replaces RemoveUnusedPorts with IMDCE by adding a new mode to IMDCE as mentioned in #10501.
To mimic the old RemoveUnusedPorts behavior, this PR adds two options (1) dropping donttouch (2) remove only ports. Maybe it might be better to move (1) to reduction pattern instead of implementing in IMDCE pass, though.
Invalid value propagation is added as well to pass the old test.
This PR fixes crashes in #10501 and #10504. IMDCE doesn't eliminate property/instance choices at this point though.
Close #3377.
Assisted-by: claude code: sonnet 4.6