Optimizer: add some floating point optimizations #80404
Merged
Conversation
This file contains hidden or 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
eeckstein
requested review from
rjmccall,
a team,
jckarter,
DougGregor,
hborla,
slavapestov and
xedin
as code owners
|
@swift-ci test |
|
@swift-ci apple silicon benchmark |
eeckstein
force-pushed
the
optimize-floating-point
branch
from
c89ffa6 to
072ed8a
Compare
|
@swift-ci test |
eeckstein
force-pushed
the
optimize-floating-point
branch
from
072ed8a to
fc2d86f
Compare
|
@swift-ci test |
eeckstein
force-pushed
the
optimize-floating-point
branch
from
fc2d86f to
ae2607c
Compare
|
@swift-ci test |
eeckstein
force-pushed
the
optimize-floating-point
branch
from
ae2607c to
08f01de
Compare
|
@swift-ci test |
…erals if it has an AST node This avoids false alarms for floating point literals which are imported from the `var infinity` getter of floating point types.
This forces inlining the annotated function if its arguments are constant.
…ine.constant.arguments"
* `sitofp` signed integer to floating point * `rint` round floating point to integral * `bitcast` between integer and floating point Constant folding `bitcast`s also made it necessary to rewrite constant folding for Nan and inf values, because the old code explicitly checked for `bitcast` intrinsics. Relying on constant folded `bitcast`s makes the new version much simpler. It is important to constant fold these intrinsics already in SIL because it enables other optimizations.
…c floating point literal construction.
eeckstein
force-pushed
the
optimize-floating-point
branch
from
08f01de to
3ec0570
Compare
|
@swift-ci test |
|
@swift-ci test windows |
|
@swift-ci smoke test macos |
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.
The goal is to completely constant fold literals of custom floating point types which use the generic initializers of the stdlib's
BinaryFloatingPointprotocol.Two changes were necessary to make this happen:
It turned out that constant folding the
bitcastintrinsic broke diagnosing floating point literals which end up as "infinity".So, I had to fix this as well. For details see the commit messages.
rdar://147278006