Add gc use intrinisic #23562
Merged
Add gc use intrinisic #23562
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
yuyichao
approved these changes
Sep 2, 2017
JeffBezanson
reviewed
Sep 2, 2017
base/boot.jl
Outdated
| @@ -438,4 +438,6 @@ show(@nospecialize a) = show(STDOUT, a) | |||
| print(@nospecialize a...) = print(STDOUT, a...) | |||
| println(@nospecialize a...) = println(STDOUT, a...) | |||
|
|
|||
| gcuse(a) = ccall(:jl_gc_use, Void, (Any,), a) | |||
There was a problem hiding this comment.
Only if that doesn't prevent inlining.
Now that our gc root placement pass is significantly more aggressive about dropping roots, we need to be a bit more careful about our use of pointer, etc. This adds a low-level intrinsic for annotating gc uses to keep objects alive even if they would be otherwise unreferenced. As an initial use case, we get rid of a number of uses of `pointer` in string, but creating a new `unsafe_load` that keeps the string alive.
yuyichao
reviewed
Sep 3, 2017
| end | ||
| cnum | ||
| end | ||
|
|
||
| @noinline function slow_utf8_next(p::Ptr{UInt8}, b::UInt8, i::Int, l::Int) |
There was a problem hiding this comment.
The noinline can probably be removed?
There was a problem hiding this comment.
I think that was there for performance reasons to keep the slow path out of the icache?
There was a problem hiding this comment.
OK, sure. (Not that the current heuristic inlines the big function anyway....)
vtjnash
reviewed
Sep 4, 2017
| @@ -59,6 +59,13 @@ String(s::Symbol) = unsafe_string(Cstring(s)) | |||
| pointer(s::String) = unsafe_convert(Ptr{UInt8}, s) | |||
| pointer(s::String, i::Integer) = pointer(s)+(i-1) | |||
|
|
|||
| function unsafe_load(s::String, i::Integer=1) | |||
There was a problem hiding this comment.
This doesn't seem like the right name for this function, as we've been deprecating names of this form. It seems like it should be spelled @inbounds codeunit(...) instead, following the example of the array code.
Keno
added a commit
that referenced
this pull request
Sep 4, 2017
Addresses post-commit review in #23562.
This was referenced Feb 14, 2018
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.
Now that our gc root placement pass is significantly more aggressive about
dropping roots, we need to be a bit more careful about our use of pointer,
etc. This adds a low-level, zero-overhead intrinsic for annotating gc uses to keep objects
alive even if they would be otherwise unreferenced.
As an initial use case, we get rid of a number of uses of
pointerin string,but creating a new
unsafe_loadthat keeps the string alive.