effects: relax recursion detection for effects analysis #45993
Merged
Conversation
This file contains 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
|
@nanosoldier |
|
Your benchmark job has completed - possible performance regressions were detected. A full report can be found here. |
vtjnash
reviewed
Jul 11, 2022
vtjnash
reviewed
Jul 11, 2022
| effects = Effects(effects; terminates=ALWAYS_FALSE) | ||
| # Some sort of recursion was detected. | ||
| if edge !== nothing && !edgelimited && !is_edge_recursed(edge, sv) | ||
| # no `MethodInstance` cycles -- don't taint :terminate |
There was a problem hiding this comment.
This seems like it may only consider whether there was a cycle that will impact the computation of the result type. If we note that the result is discarded, there won't be an edge or edgecycle here at all (note line 512 above).
IIUC, the correct fix for that may be to poison the effects at line 512 however?
There was a problem hiding this comment.
If we discard the result, we propagate the unknown effects Effects(), whose :terminate effect is also tainted, and it will eventually poison the rest of the call graph.
aviatesk
force-pushed
the
avi/recursion-effects
branch
from
7f70f91
to
703b12c
Compare
aviatesk
added a commit
that referenced
this pull request
Jul 13, 2022
* ~~Add (the equivalent of) `@assume_effects :terminates_globally` to `unwrap_composed`. Although it could be inferred as `Const`, without the annotation, it was not elided for too complex inputs, resulting in unnecessary runtime overhead.~~ EDIT: now #45993 is merged and this part isn't included. * Reverse recursion order in `call_composed`. This prevents potentially changing argument types being piped through the recursion, making inference bail out. With the reversed order, only the tuple of remaining functions is changing during recursion and is becoming strictly simpler, letting inference succeed. Co-authored-by: Shuhei Kadowaki <[email protected]>
aviatesk
added a commit
that referenced
this pull request
Jul 20, 2022
effects: relax recursion detection for effects analysis
32 tasks
Merged
aviatesk
added a commit
that referenced
this pull request
Jul 23, 2022
#45993 fixed this bug, let's make the math effect-tests actually fail
aviatesk
added a commit
that referenced
this pull request
Jul 26, 2022
effects: relax recursion detection for effects analysis
ffucci
pushed a commit
to ffucci/julia
that referenced
this pull request
Aug 11, 2022
* ~~Add (the equivalent of) `@assume_effects :terminates_globally` to `unwrap_composed`. Although it could be inferred as `Const`, without the annotation, it was not elided for too complex inputs, resulting in unnecessary runtime overhead.~~ EDIT: now JuliaLang#45993 is merged and this part isn't included. * Reverse recursion order in `call_composed`. This prevents potentially changing argument types being piped through the recursion, making inference bail out. With the reversed order, only the tuple of remaining functions is changing during recursion and is becoming strictly simpler, letting inference succeed. Co-authored-by: Shuhei Kadowaki <[email protected]>
ffucci
pushed a commit
to ffucci/julia
that referenced
this pull request
Aug 11, 2022
JuliaLang#45993 fixed this bug, let's make the math effect-tests actually fail
pcjentsch
pushed a commit
to pcjentsch/julia
that referenced
this pull request
Aug 18, 2022
* ~~Add (the equivalent of) `@assume_effects :terminates_globally` to `unwrap_composed`. Although it could be inferred as `Const`, without the annotation, it was not elided for too complex inputs, resulting in unnecessary runtime overhead.~~ EDIT: now JuliaLang#45993 is merged and this part isn't included. * Reverse recursion order in `call_composed`. This prevents potentially changing argument types being piped through the recursion, making inference bail out. With the reversed order, only the tuple of remaining functions is changing during recursion and is becoming strictly simpler, letting inference succeed. Co-authored-by: Shuhei Kadowaki <[email protected]>
pcjentsch
pushed a commit
to pcjentsch/julia
that referenced
this pull request
Aug 18, 2022
JuliaLang#45993 fixed this bug, let's make the math effect-tests actually fail
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 a similar spirit to #40561, we can relax the recursion detection to
guarantee
:terminateseffect and allow the effects analysis to nottaint
:terminateseffect when there are no cycles inMethodInstancesin a call graph.
fix #45781