SPU LLVM: Rearange FM instruction for better performance #9896
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
- Doesn't eliminate any instructions, but allows for better out of order execution.
|
Got a nice boost here. |
elad335
reviewed
Mar 5, 2021
| const auto cb = eval(bitcast<f32[4]>(bitcast<s32[4]>(b) & ma)); | ||
| set_vr(op.rt, fm(ca, cb)); | ||
| const auto cx = eval(ma & mb); | ||
| const auto x = fm(a, b); |
There was a problem hiding this comment.
Shouldn't you modify fm function as well? llvm expressions detection relies on this.
There was a problem hiding this comment.
fm is currently just an unnecessary alias for multiplication operator
There was a problem hiding this comment.
Shouldn't you modify fm function as well? llvm expressions detection relies on this.
The only time we look for the fm pattern is in is_input_positive, which should still be working since it only looks for the case when a = b
Nekotekina
reviewed
Mar 8, 2021
| const auto ca = eval(bitcast<f32[4]>(bitcast<s32[4]>(a) & mb)); | ||
| const auto cb = eval(bitcast<f32[4]>(bitcast<s32[4]>(b) & ma)); | ||
| set_vr(op.rt, fm(ca, cb)); | ||
| const auto cx = eval(ma & mb); |
There was a problem hiding this comment.
I wonder if it could be more correct to & first, then use sext, but I won't bother trying it for now.
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Outdated suggestions cannot be applied.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.
Rearranges the FM instruction to allow the comparisons and the multiplication to be processed in parallel. This doesn't save any instructions, but still results in a speedup.
In the mandelbrot homebrew performance increased from 120 --> 122fps on my 7700K at 5ghz.
On a cpu with more out of order execution resources available, such as my i5-1135G7 at 2.6ghz performance was increased from 69 --> 74fps.