Make Compiler use separate sorting algorithm. #47066
Merged
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
oscardssmith
added
needs tests
vchuravy
reviewed
Oct 5, 2022
|
This is a combination of
The algorithmic changes in #46586 and #46587 have little practical performance impact, but what impact there is is positive. I think what is holding this back right now is that we need review of those changes by someone familiar with the appropriate sections of the compiler. |
|
@nanosoldier |
|
Your benchmark job has completed - no performance regressions were detected. A full report can be found here. |
ranocha
reviewed
Oct 7, 2022
Co-authored-by: Valentin Churavy <vchuravy@users.noreply.github.com>
Co-authored-by: Hendrik Ranocha <ranocha@users.noreply.github.com>
oscardssmith
force-pushed
the
oscardssmith-Compiler-Sort
branch
from
3533c60
to
1ab8885
Compare
aviatesk
reviewed
Oct 9, 2022
Comment on lines
+110
to
+124
| @testset "many basic blocks" begin | ||
| n = 1000 | ||
| ex = :(return 1) | ||
| for _ in 1:n | ||
| ex = :(if rand()<.1 | ||
| $(ex) end) | ||
| end | ||
| @eval begin | ||
| function f_1000() | ||
| $ex | ||
| return 0 | ||
| end | ||
| end | ||
| @test f_1000()===0 | ||
| end |
There was a problem hiding this comment.
I'm confused at this test case. What's being tested here?
There was a problem hiding this comment.
I believe that this is testing compilation of a function with a bunch of basic blocks. Is there a better version of this test?
There was a problem hiding this comment.
Is the compilation performance being tested, or is there any correctness issue? And why f_1000 always returns 0?
There was a problem hiding this comment.
There was a correctness issue (#47065) where the compiler would try to sort a big list and fail. f_1000 doesn't technically always return 0, but it returns 1 1 in 10^1000 times which is close enough to never that it will never fail.
There was a problem hiding this comment.
Okay, I understand. I'd prefer not having @testset there though.
LilithHafner
reviewed
Oct 9, 2022
|
|
||
| # Make sure that the compiler can sort things. | ||
| # https://github.com/JuliaLang/julia/issues/47065 | ||
| @testset "Compiler Sorting" begin |
There was a problem hiding this comment.
I think this is redundant with (or at least should be in the same place as) test/compiler/sort.jl
aviatesk
added a commit
that referenced
this pull request
Oct 10, 2022
aviatesk
added a commit
that referenced
this pull request
Oct 10, 2022
aviatesk
added a commit
that referenced
this pull request
Oct 10, 2022
aviatesk
added a commit
to aviatesk/JET.jl
that referenced
this pull request
Oct 13, 2022
aviatesk
added a commit
that referenced
this pull request
Feb 8, 2024
This commit refines `interpreter_exec.jl` further, on top of #53218. In detail, by using `Base.Experimental.@compiler_options compile=min`, it's now possible to mimic the effect of `julia --compile=[min|yes]`. This commit is leveraging it and includes `interpreter_exec.jl` from `test/compiler/ssair.jl`, allowing us to get better stack traces and test summaries. In addition, I've introduced several tests concerning the `src.inferred` values and have relocated the tests from #47066 within `interpreter_exec.jl` to `test/compiler/irpasses.jl`, since it is related to Julia-level compilation and not to interpreter execution.
aviatesk
added a commit
that referenced
this pull request
Feb 8, 2024
This commit refines `interpreter_exec.jl` further, on top of #53218. In detail, by using `Base.Experimental.@compiler_options compile=min`, it's now possible to mimic the effect of `julia --compile=[min|yes]`. This commit is leveraging it and includes `interpreter_exec.jl` from `test/compiler/ssair.jl`, allowing us to get better stack traces and test summaries. In addition, I've introduced several tests concerning the `src.inferred` values and have relocated the tests from #47066 within `interpreter_exec.jl` to `test/compiler/irpasses.jl`, since it is related to Julia-level compilation and not to interpreter execution.
aviatesk
added a commit
that referenced
this pull request
Feb 9, 2024
This commit refines `interpreter_exec.jl` further, on top of #53218. In detail, by using `Base.Experimental.@compiler_options compile=min`, it's now possible to mimic the effect of `julia --compile=[min|yes]`. This commit is leveraging it and includes `interpreter_exec.jl` from `test/compiler/ssair.jl`, allowing us to get better stack traces and test summaries. In addition, I've introduced several tests concerning the `src.inferred` values and have relocated the tests from #47066 within `interpreter_exec.jl` to `test/compiler/irpasses.jl`, since it is related to Julia-level compilation and not to interpreter execution.
aviatesk
added a commit
that referenced
this pull request
Feb 9, 2024
This commit refines `interpreter_exec.jl` further, on top of #53218. In detail, by using `Base.Experimental.@compiler_options compile=min`, it's now possible to mimic the effect of `julia --compile=[min|yes]`. This commit is leveraging it and includes `interpreter_exec.jl` from `test/compiler/ssair.jl`, allowing us to get better stack traces and test summaries. In addition, I've introduced several tests concerning the `src.inferred` values and have relocated the tests from #47066 within `interpreter_exec.jl` to `test/compiler/irpasses.jl`, since it is related to Julia-level compilation and not to interpreter execution.
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.
Separated from #46679 as the easiest way to fix #47065. Needs tests.
closes #46586
closes #46587