Optimize MCP2515Class::endPacket() by up to 2x #49
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
1) Use LOAD TX BUFFER instruction to set TXBn* registers sequentially, as well as to write into the TX data registers. This is done with a single CS pull, and requires sending N+6 bytes over SPI. The old code was doing N+5 writeRegister() calls, each doing 1 CS pull and sending 3 bytes. 2) Use the 1-byte RTS SPI instruction instead of WRITEing to TXBnCTRL. This reduces the number of bytes sent over SPI by 2. For N = 8, - we now send 15 bytes over SPI before the while(...TXBnCTRL...) loop vs 42 (2.8x reduction) - we do just 2 CS pulls before the while(...TXBnCTRL...) loop vs 13 (6.5x reduction) We still do at least 3 CS pulls and send at least 10 bytes over SPI for the condition of the while(...TXBnCTRL...) loop, to clear TXnIF and to check for errors. That means for N = 8, in the case when we do zero iterations of the while loop, we end up sending 25 bytes over SPI instead of 52, which is >2x reduction.
# for free
to join this conversation on GitHub.
Already have an account?
# 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Outdated suggestions cannot be applied.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.
Optimize MCP2515Class::endPacket()
Use LOAD TX BUFFER instruction to set TXBn* registers sequentially, as well
as to write into the TX data registers.
This is done with a single CS pull, and requires sending N+6 bytes over SPI.
The old code was doing N+5 writeRegister() calls, each doing 1 CS pull and
sending 3 bytes.
Use the 1-byte RTS SPI instruction instead of WRITEing to TXBnCTRL.
This reduces the number of bytes sent over SPI by 2.
For N = 8,
(2.8x reduction)
(6.5x reduction)
We still do at least 3 CS pulls and send at least 10 bytes over SPI for the
condition of the while(...TXBnCTRL...) loop, to clear TXnIF and to check for
errors.
That means for N = 8, in the case when we do zero iterations of the while loop,
we end up sending 25 bytes over SPI instead of 52, which is >2x reduction.