Skip to content

Commit 5e1dcff

Browse files
lmiqtimholy
andauthored
note about performance regressions in <1.8
Co-authored-by: Tim Holy <[email protected]>
1 parent 138e92e commit 5e1dcff

File tree

1 file changed

+1
-1
lines changed

1 file changed

+1
-1
lines changed

docs/src/index.md

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -16,7 +16,7 @@ Finally, `PrecompileTools` includes `@recompile_invalidations` to mitigate the u
1616
These different tools are demonstrated below.
1717

1818
!!! note
19-
There have been cases where precompilation led to performance regressions in packages (see for example [[1](https://github.com/JuliaLang/julia/issues/50749),[2](https://github.com/JuliaLang/julia/pull/50766),[3](https://github.com/JuliaLang/julia/issues/35972)]). While fixes have been released, some problems persist in older Julia versions and new issues might also slow down runtime after precompilation. Hence, it's advised to test packages for performance regressions in all supported Julia versions when introducing precompilation directives.
19+
The latency reductions from PrecompileTools are maximally effective for Julia versions 1.9 and higher, and intermediate for Julia 1.8. Julia versions 1.7 and earlier may see some limited benefit as well, but have also occasionally been found to suffer from [precompilation-induced runtime performance regressions](https://github.com/JuliaLang/julia/issues/35972). If you wish, you can disable precompilation on older Julia versions by wrapping precompilation statements (see below) with `if Base.VERSION >= v"1.8" ... end`. On older Julia versions, you may wish to test packages for performance regressions when introducing precompilation directives.
2020

2121
## Tutorial: forcing precompilation with workloads
2222

0 commit comments

Comments
 (0)