Replace syntax's SmallVector with AccumulateVec #37551
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
|
Also cc #37371, which has some discussion on expanding/generalizing SmallVector. |
Mark-Simulacrum
commented
Nov 3, 2016
| _ => unreachable!() | ||
| } | ||
| } | ||
| } |
There was a problem hiding this comment.
This code would look better with @ticki's replace_with, but it's not altogether that bad without either.
This was referenced Nov 3, 2016
|
LLVM gets confused by |
|
There should probably be two types still, at least a |
Mark-Simulacrum
force-pushed
the
upgrade-accvec
branch
from
c2ea97d
to
2d8ee6f
Compare
|
Alright, the implementation now consists of This currently doesn't compile, but I don't know why--from what I can tell, I didn't change the method signatures shown here. Edit: Also, as a sidenote, this error message could really be improved: I can't tell what is being moved out of borrowed content. cc @jonathandturner. |
|
That's an odd error - are you sure you didn't change a signature by accident? |
Mark-Simulacrum
force-pushed
the
upgrade-accvec
branch
2 times, most recently
from
743fb75
to
1894f71
Compare
Mark-Simulacrum
changed the title
|
The current code segfaults with this backtrace. I don't know why this is, and would appreciate any suggestions. |
|
Hmm, after removing the derives for |
eddyb
suggested changes
Nov 6, 2016
| pub fn pop(&mut self) -> Option<A::Element> { | ||
| if self.count > 0 { | ||
| let arr = &mut self.values as &mut [ManuallyDrop<_>]; | ||
| let value = mem::replace(&mut arr[self.count - 1], ManuallyDrop::new()); |
There was a problem hiding this comment.
You don't need mem::replace, you can just do ptr::read(&arr[self.count - 1].value).
Also, you can decrement first, and use self.count then.
| } | ||
| } | ||
| } | ||
|
|
||
| impl<A: Array> Iterator for ArrayVec<A> { |
There was a problem hiding this comment.
Why? You could make a forward by-value iterator if you wanted to (by having an ArrayVec and a start index), and then implement IntoIterator on ArrayVec.
|
|
||
| fn next(&mut self) -> Option<A::Element> { | ||
| match self.repr { | ||
| IntoIterRepr::Array(ref mut arr) => arr.pop(), |
There was a problem hiding this comment.
This is not a good idea at all, it goes in the reverse order!
Mark-Simulacrum
force-pushed
the
upgrade-accvec
branch
from
86b96e4
to
128d756
Compare
eddyb
reviewed
Nov 6, 2016
| fn into_iter(self) -> Self::IntoIter { | ||
| match *self { | ||
| AccumulateVec::Array(ref arr) => arr.iter(), | ||
| AccumulateVec::Heap(ref vec) => vec.iter(), |
There was a problem hiding this comment.
You can just self.iter() - same for the mutable version below.
Mark-Simulacrum
force-pushed
the
upgrade-accvec
branch
2 times, most recently
from
cb581ec
to
c5b951a
Compare
Mark-Simulacrum
changed the title
eddyb
reviewed
Nov 7, 2016
| fn encode<S: Encoder>(&self, s: &mut S) -> Result<(), S::Error> { | ||
| s.emit_seq(self.len(), |s| { | ||
| for (i, e) in self.iter().enumerate() { | ||
| try!(s.emit_seq_elt(i, |s| e.encode(s))); |
There was a problem hiding this comment.
You can use ? instead of try! here and below.
There was a problem hiding this comment.
I was going to, but then we need to add ? to the feature list and I recalled needing to mess with cfg(stage0) or something like that in order for that to work properly.
I can change it if you'd like, though--let me know.
eddyb
approved these changes
Nov 7, 2016
|
f? @erickt |
Mark-Simulacrum
force-pushed
the
upgrade-accvec
branch
from
77acf35
to
8380d32
Compare
alexcrichton
added
the
T-compiler
|
Eh, it looks fine enough for me. @bors r+ |
|
📌 Commit 8380d32 has been approved by |
|
☔ The latest upstream changes (presumably #37246) made this pull request unmergeable. Please resolve the merge conflicts. |
Mark-Simulacrum
force-pushed
the
upgrade-accvec
branch
from
8380d32
to
7bbebb1
Compare
|
Rebased. |
|
@bors r+ |
|
📌 Commit 7bbebb1 has been approved by |
eddyb
added a commit
to eddyb/rust
that referenced
this pull request
Nov 11, 2016
…ddyb Replace syntax's SmallVector with AccumulateVec This adds a new type to data_structures, `SmallVec`, which wraps `AccumulateVec` with support for re-allocating onto the heap (`SmallVec::reserve`). `SmallVec` is then used to replace the implementation of `SmallVector` in libsyntax. r? @eddyb Fixes rust-lang#37371. Using `SmallVec` instead of libsyntax's `SmallVector` will provide the `N = 2/4` case easily (just needs a few more `Array` impls). cc @nnethercote, probably interested in this area
|
This is nice, thank you. I particularly like that you found a way to let N be choosable. E.g. I was looking at |
eddyb
added a commit
to eddyb/rust
that referenced
this pull request
Nov 12, 2016
…ddyb Replace syntax's SmallVector with AccumulateVec This adds a new type to data_structures, `SmallVec`, which wraps `AccumulateVec` with support for re-allocating onto the heap (`SmallVec::reserve`). `SmallVec` is then used to replace the implementation of `SmallVector` in libsyntax. r? @eddyb Fixes rust-lang#37371. Using `SmallVec` instead of libsyntax's `SmallVector` will provide the `N = 2/4` case easily (just needs a few more `Array` impls). cc @nnethercote, probably interested in this area
bors
added a commit
that referenced
this pull request
Nov 12, 2016
Rollup of 30 pull requests - Successful merges: #37190, #37368, #37481, #37503, #37527, #37535, #37551, #37584, #37600, #37613, #37615, #37659, #37662, #37669, #37682, #37688, #37690, #37692, #37693, #37694, #37695, #37696, #37698, #37699, #37705, #37708, #37709, #37716, #37724, #37727 - Failed merges: #37640, #37689, #37717
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 adds a new type to data_structures,
SmallVec, which wrapsAccumulateVecwith support for re-allocating onto the heap (SmallVec::reserve).SmallVecis then used to replace the implementation ofSmallVectorin libsyntax.r? @eddyb
Fixes #37371. Using
SmallVecinstead of libsyntax'sSmallVectorwill provide theN = 2/4case easily (just needs a few moreArrayimpls).cc @nnethercote, probably interested in this area