-
Notifications
You must be signed in to change notification settings - Fork 52
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Add F_POINTED_AT; seems to work after some minimal testing #298
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
lgtm
@@ -329,7 +329,8 @@ class ReachabilityNormalizer(config: NormalizerConfig, ra: ReachabilityAnalyzer) | |||
// add normalized field(s) to the vector | |||
var tn = rf.typeNorm; | |||
var facts = if(tn.size > 1, Fact.F_NORM, Facts.NONE); | |||
if (!rf.raFacts.RF_WRITTEN) facts |= (Fact.F_VALUE | Fact.O_FOLDABLE); | |||
if (!rf.raFacts.RF_WRITTEN && !rf.orig.facts.F_POINTED_AT) facts |= (Fact.F_VALUE | Fact.O_FOLDABLE); |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Note I added this line while resolving a conflict here.
I've confirmed the CI fails because the additional fact brought the size of a |
On 11/19/2024 4:33 AM, Ben L. Titzer wrote:
I've confirmed the CI fails because the additional fact brought the size of a |Fact.set| over 32-bit, which seems to
have exposed some bugs in unboxing (that's why this only fails on the CI configurations that try to unbox all variants).
I'll try to get a repro and fix those other bugs. In the mean time, can you remove the unused |O_ASSOCIATIVE| case?
Yes - but do you mean in the PR or only locally? Eliot
|
In the PR. Since it only exposed a pre-existing bug (which I have a fix for, but not yet a smaller repro), no need to hold up this PR. |
On 11/19/2024 12:02 PM, Ben L. Titzer wrote:
Yes - but do you mean in the PR or only locally? Eliot
In the PR. Since it only exposed a pre-existing bug (which I have a fix for, but not yet a smaller repro), no need to
hold up this PR.
Change developed and added ... EM
|
No description provided.