Speed up AC3 demodulator #805
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
By default, reading from cin forces cout to be flushed first (in case you've printed a prompt to cout that needs to be visible). Since these programs both read a lot of data from cin one character at a time, they were spending significant amounts of time flushing cout.
This works the same way as the existing version, but it reworks the buffer code to avoid using % on the indexes in the loop. This means that GCC can use vector instructions, and schedule the unrolled outer loop a bit more effectively. I've also replaced 320 etc. with the derivations of the constants.
This means the comparison can be done with an XOR and popcount, which is quite a bit faster. Since ld-decode is currently C++17, we can't use std::popcount yet; use the GCC/Clang intrinsic if available, otherwise use a from-scratch implementation (based on the version GCC's optimiser recognises -- it's the HAKMEM-inspired algorithm from p69 of "Hacker's Delight").
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Outdated suggestions cannot be applied.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.
CC @LeightonSmallshire.
These changes to
Demodulatormake the program as a whole about five times faster with-O3 -march=native- here's the time taken for the original and the three commits on a 10s sample:The bottleneck now is reading the contents of the buffer for each output bit - since the buffer's only 3 words there's not much that can be done to improve this within this code, but you could get around it by reworking
Reclockerto accept multiple input bits at once...