avoid inferring when compilation signature differs from call site sig #46581
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
Added in #43415, this was too aggressive for many cases. Unlike the comment suggested, it is unneeded in many cases, so only do it when it is expected to be maximally profitable. Fixes #46492 julia> @time norm(C_212) before 45.959497 seconds (81.85 M allocations: 6.976 GiB, 6.31% gc time, 100.00% compilation time) after 15.781804 seconds (20.81 M allocations: 1.294 GiB, 6.32% gc time, 100.00% compilation time)
aviatesk
approved these changes
Sep 2, 2022
Comment on lines
+216
to
+223
| if infer_compilation_signature(interp) && 1 == seen == napplicable && rettype !== Any && rettype !== Union{} && !is_removable_if_unused(all_effects) | ||
| match = applicable[1]::MethodMatch | ||
| method = match.method | ||
| sig = match.spec_types | ||
| mi = specialize_method(match; preexisting=true) | ||
| if mi !== nothing && !const_prop_methodinstance_heuristic(interp, match, mi::MethodInstance, arginfo, sv) | ||
| csig = get_compileable_sig(method, sig, match.sparams) | ||
| if csig !== nothing && csig !== sig |
There was a problem hiding this comment.
I may want to pack these condition checks into a separate function, e.g.
if infer_compilation_signature(interp)
compilation_sig = maybe_compilation_signature(applicable, seen, rettype, all_effects, arginfo, sv)
if compilation_sig !== nothing
(; method, csig, sparams) = compilation_sig
...
end
endBut maybe_compilation_signature's argument list is so long depending on many local variables of abstract_call_gf_by_type, so I'm fine with the current code if you prefer it.
|
Can this be backported to 1.8 ? |
aviatesk
pushed a commit
that referenced
this pull request
Sep 9, 2022
…#46581) Added in #43415, this was too aggressive for many cases. Unlike the comment suggested, it is unneeded in many cases, so only do it when it is expected to be maximally profitable. Fixes #46492 ``` julia> @time norm(C_212) before 45.959497 seconds (81.85 M allocations: 6.976 GiB, 6.31% gc time, 100.00% compilation time) after 15.781804 seconds (20.81 M allocations: 1.294 GiB, 6.32% gc time, 100.00% compilation time) ```
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.
Added in #43415, this was too aggressive for many cases. Unlike the
comment suggested, it is unneeded in many cases, so only do it when it
is expected to be maximally profitable.
Fixes #46492