Add notasklocal effect
#45422
Merged
Add notasklocal effect
#45422
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
aviatesk
reviewed
May 23, 2022
base/compiler/types.jl
Outdated
| @@ -76,41 +82,46 @@ function Effects( | |||
| effect_free::TriState, | |||
| nothrow::TriState, | |||
| terminates::TriState, | |||
| nonoverlayed::Bool) | |||
| nonoverlayed::Bool, | |||
| notls::TriState) | |||
There was a problem hiding this comment.
Apologies (and please ignore) if this is unappreciated bikeshedding, but notls is a less immediately obvious name than the other effects. notasklocal would be more similar in clarity and roughly the same length as the other effects names.
There was a problem hiding this comment.
fair enough. The other effects do tend to be more verbose.
There was a problem hiding this comment.
especially since TLS=Transport Layer Security.
There was a problem hiding this comment.
That was how I initially read it.
Split out from #45272. This effect models the legality of moving code between tasks. It is somewhat related to effect-free/consistent, but only with respect to task-local state. As an example consider something like: ``` global glob function bar() @async (global glob = 1; some_other_code()) end ``` The newly created task is not effect-free, but it would be legal to inline the assignment of `glob` into `bar` (as long it is inlined before the creation of the task of `some_other_code` does not access `glob`). For comparison, the following is neither `notls`, nor `effect_free`: ``` function bar() @async (task_local_storage()[:var] = 1; some_other_code()) end ``` The same implies to implicit task-local state such as the RNG state. Implementation wise, there isn't a lot here, because the implicit tainting by ccall is the correct conservative default. In the future, we may want to annotate various ccalls as being permissible for notls, but let's worry about that when we have a case that needs it.
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.
Split out from #45272. This effect models the legality of moving code
between tasks. It is somewhat related to effect-free/consistent, but
only with respect to task-local state. As an example consider something
like:
The newly created task is not effect-free, but it would be legal to inline
the assignment of
globintobar(as long it is inlined before the creationof the task of
some_other_codedoes not accessglob). For comparison,the following is neither
notls, noreffect_free:The same implies to implicit task-local state such as the RNG state.
Implementation wise, there isn't a lot here, because the implicit
tainting by ccall is the correct conservative default. In the future,
we may want to annotate various ccalls as being permissible for
notls, but let's worry about that when we have a case that needs it.