Make BoundsError lazy and move Memory boundscheck to Julia
#56167
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
vchuravy
reviewed
Oct 15, 2024
| @@ -347,12 +347,39 @@ struct ErrorException <: Exception | |||
| msg::AbstractString | |||
| end | |||
|
|
|||
| import .Intrinsics: ult_int | |||
| struct Summarized | |||
aviatesk
reviewed
Oct 15, 2024
| struct Summarized | ||
| desc::Union{Nothing,String} | ||
| size::Union{Nothing, Tuple{Vararg{Int}}} | ||
| atype::Type |
There was a problem hiding this comment.
Suggested change
| atype::Type | |
| atype::DataType |
giordano
changed the title
BoundsError lazy and move Memory boundscheck to JuliaBoundsError lazy and move Memory boundscheck to Julia
oscardssmith
force-pushed
the
os/lazy-bounds-error
branch
from
0acb6cd to
bf20618
Compare
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Outdated suggestions cannot be applied.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.
This is a revival of #43738 with a lot of changes. The biggest by far is that with this version of the PR, I'm trying to get the BoundsError to semantically not escape the input argument so that LLVM escape analysis doesn't see bounds error construction as a potential escape (motivated by #55913). That said, this will likely significantly increase IR size, so I'm not sure if it is worth it or if this is a good tradeoff...