Dataflow: Refactor access paths to split TypedContent into an explicit pair #12930
Conversation
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
aschackmull
force-pushed
the
dataflow/split-typedcontent
branch
from
fd8e96d to
6f21781
Compare
aschackmull
force-pushed
the
dataflow/split-typedcontent
branch
from
6f21781 to
71ae090
Compare
|
DCA is looking good. The few additional results for Java appear to be caused by loss of precise type tracking stored in |
hvitved
approved these changes
May 1, 2023
| @@ -2138,6 +2163,8 @@ module Impl<FullStateConfigSig Config> { | |||
| PrevStage::revFlow(node, _) and result = TApproxFrontNil(node.getDataFlowType()) | |||
| } | |||
|
|
|||
| Typ getTyp(DataFlowType t) { result = t } | |||
There was a problem hiding this comment.
I expect it to be extremely likely that the optimiser will in fact inline these, so no need to be explicit about that, I think.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
In data flow we're tracking both a type and an access path. These two things are somewhat intertwined as we want to be able to reestablish the tracked type after a store-to-read step sequence.
We used to achieve this by merging the type into the front of the access paths resulting in a list of the form
(typ1, content1) :: (typ2, content2) :: ... :: nil(typn). This refactor pulls the front type out of the access path, such that it becomes an explicit pair rather than being merged withContent. We still need the types nested inside the access path tails, so now, instead of being a list ofTypedContents, access paths are a list ofContents where each nested tail is paired with aDataFlowType, i.e.(typ1, content1 :: (typ2, content2 :: ... :: (typn, nil)) ... )).In the automaton view of access paths, we're switching from a graph with access-path nodes and type-content-pair-labelled edges to a graph where the nodes are type-access-path pairs and the edges are merely content-labelled.
This refactor is intended to pave the way for a future enhancement where we'll be able to update the tracked type independently from the access path to improve precision.
This PR is intended to be reviewed commit-by-commit. The commits progress by first duplicating the type information from the access paths where needed before replacing
TypedContentwithContent.