struct Rav1dIntraPredDspContext: Fix defaults
#583
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
kkysen
force-pushed
the
kkysen/mod-ipred-neon
branch
from
fa222c0 to
f3feffc
Compare
kkysen
force-pushed
the
kkysen/struct-Rav1dIntraPredDSPContext-defaults
branch
from
662f110 to
1d210ad
Compare
rinon
approved these changes
Nov 29, 2023
rinon
reviewed
Nov 29, 2023
rinon
requested changes
Nov 29, 2023
kkysen
force-pushed
the
kkysen/mod-ipred-neon
branch
from
f3feffc to
3aa1661
Compare
kkysen
force-pushed
the
kkysen/struct-Rav1dIntraPredDSPContext-defaults
branch
2 times, most recently
from
ca53d47 to
f2c5e8f
Compare
kkysen
force-pushed
the
kkysen/mod-ipred-neon
branch
from
3aa1661 to
348b7db
Compare
kkysen
force-pushed
the
kkysen/struct-Rav1dIntraPredDSPContext-defaults
branch
from
f2c5e8f to
7d0c2ce
Compare
rinon
reviewed
Dec 1, 2023
… a `initialized: bool` field (on `struct Rav1dDSPContext`) as they are always `Some` even if `.is_none()` is checked. `.is_none()` appears to be checked so that it can lazily initialize DSPs, relying on things being zero-initialized to begin with. Ideally we should move initialization to DSP allocation so its not lazy, but for now (and the way closest to the C), we can use a single initialized flag instead of `Option`s on everything.
…:cfl_pred` to defaults to avoid unitialized memory, as not all (4/6) elements are initialized.
kkysen
force-pushed
the
kkysen/struct-Rav1dIntraPredDSPContext-defaults
branch
from
7d0c2ce to
4609590
Compare
randomPoison
approved these changes
Dec 6, 2023
There was a problem hiding this comment.
The changes here look fine to me, but make sure you've addressed @rinon's requests before merging.
|
I just hadn't gotten a chance to look at it, LGTM thanks 👍 |
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 replaces the
Options on.intra_predwith aninitialized: boolfield as they are alwaysSomeeven if.is_none()is checked..is_none()appears to be checked so that it can lazily initialize DSPs, relying on things being zero-initialized to begin with.Ideally we should move initialization to DSP allocation so its not lazy, but for now (and the way closest to the C), we can use a single initialized flag instead of
Options on everything.This also default initializes
.cfl_predas it otherwise isn't fully initialized and is thus leaving uninitialized memory, which is UB if it is made a reference at any point (which it is). This usesDefaultValue::DEFAULTfor initialization instead ofNone, though, which moves the check from before the call (.unwrap()/.expect()) to inside a call that should never happen (unimplemented!()). Normally thoseunimplemented!()s will be dead code eliminated, but here they won't. Thus, it shouldn't add any perf cost and only a small binary size cost.