You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
It would be nice if there were a canonicalization pass that attempted to improve the likelihood of deterministic output across runs. For example:
The idea of a canonicalizing pass in this instance would be to determine if the conditions tested in an short-circuiting and/or can be evaluated in any order, and if so, impose a specific order. The specific order imposed could be based on sorting the hashes of the sub-nodes, where the hashes are implemented with a clang::StmtProfilerWithoutPointers, which is accessible via
It would be nice if there were a canonicalization pass that attempted to improve the likelihood of deterministic output across runs. For example:
The idea of a canonicalizing pass in this instance would be to determine if the conditions tested in an short-circuiting and/or can be evaluated in any order, and if so, impose a specific order. The specific order imposed could be based on sorting the hashes of the sub-nodes, where the hashes are implemented with a
clang::StmtProfilerWithoutPointers
, which is accessible viaThe text was updated successfully, but these errors were encountered: