fix: Fix inline_constant_functions pass
#2135
Conversation
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## main #2135 +/- ##
=======================================
Coverage 82.15% 82.16%
=======================================
Files 227 227
Lines 39891 39897 +6
Branches 35990 35996 +6
=======================================
+ Hits 32774 32781 +7
+ Misses 5287 5286 -1
Partials 1830 1830
Flags with carried forward coverage won't be shown. Click here to find out more. ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
| hugr.insert_hugr(func_node, func_hugr); | ||
| func_hugr.replace_op(func_hugr.root(), func_defn).unwrap(); | ||
| let func_node = hugr.insert_hugr(hugr.root(), func_hugr).new_root; | ||
| hugr.set_num_ports(func_node, 0, 1); |
There was a problem hiding this comment.
use add_ports here, and use the port it returns
There was a problem hiding this comment.
This is about removing ports: If the inlined function value is a DFG, then it already has some inputs/outputs that we need to remove because it's now a FuncDefn
|
|
||
| for lcn in load_constant_ns { | ||
| hugr.replace_op(lcn, LoadFunction::try_new(polysignature.clone(), [])?)?; | ||
|
|
||
| let src_port = hugr.node_outputs(func_node).next().unwrap(); | ||
| let tgt_port = hugr.node_inputs(lcn).next().unwrap(); |
There was a problem hiding this comment.
It would be better to delegate to LoadConstant what the input is. You don't have one at hand, so maybe not worth it.
mark-koch
force-pushed
the
fix/inline_constant_functions
branch
from
15988dd to
527ad1c
Compare
Fixes #2132