-
Notifications
You must be signed in to change notification settings - Fork 10.5k
Remove the "shrink" part of the solver. #17922
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
Conversation
At one point this phase was providing a lot of value in reducing the number of expressions we ultimately decided were too complex, but at this point it doesn't seem to be helping very often. Based on experimenting with some of our slow test cases, in several cases it actually increases the amount of work we do before we decide an expression is too complex, sometimes making it take 2-3x longer. I will open a bug to track restoring the two expressions that we are now "too complex" on, so that the tests can be restored once we fix some of the underlying problems.
@swift-ci Please smoke test |
@swift-ci Please test source compatibility |
@swift-ci Please test compiler performance |
The source compatibility suite failures are unrelated to my changes and are all happening on the builders. |
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.
Changes looks good to me, but from souce compatibility perspective Kitura
and SPR
now both fail with “too complex” expressions...
@xedin Ah, I missed that in all the noise, and was waiting on swiftlang/swift-source-compat-suite#218 before re-running. I'll take a look. |
@swift-ci Please test source compatibility |
It's easy enough to rip this out when the time comes. There are conflicts with this PR, so I'll close and re-do later. |
At one point this phase was providing a lot of value in reducing the
number of expressions we ultimately decided were too complex, but at
this point it doesn't seem to be helping very often.
Based on experimenting with some of our slow test cases, in several
cases it actually increases the amount of work we do before we decide
an expression is too complex, sometimes making it take 2-3x longer.
I will open a bug to track restoring the two expressions that we are
now "too complex" on, so that the tests can be restored once we fix
some of the underlying problems.