4.80 #193
Open
4.80 #193
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
|
Awesome update, looking forward for a new firmware. Thank you for your great work! |
Collaborator
|
@copilot summarize the changes please! |
Use select_bank_fast() instead of select_bank(BANK0) in all UART ISRs. This optimization conditionally skips bank port writes when already in BANK0. Performance improvement: - Old: Always writes to bank port twice per ISR (~27 cycles overhead) - New: Skips both port writes when in BANK0 (~8 cycles overhead) - Savings: ~19 cycles per ISR when in BANK0 (common case) - At 16MHz: ~1.2 microseconds saved per ISR - For 500-byte sysex: ~600 microseconds saved total This reduces UART1 starvation of UART2 during large sysex transfers (e.g., MD kit dumps), helping prevent sequencer drift when receiving external MIDI clock on port 2. Flash cost: +38 bytes (244,478 vs 244,440 bytes)
Replace Bresenham's line algorithm with DDA (Digital Differential Analyzer) using 8.8 fixed-point arithmetic for parameter slides. Performance improvements for ATmega2560: - Code size: 204 bytes (was 336 bytes, -39%) - RAM usage: 10 bytes per lock (was 15 bytes, -33%) - Simpler algorithm: no branching between steep/non-steep paths - Better cache locality: fewer memory accesses per iteration The DDA algorithm provides smooth parameter interpolation that is indistinguishable from Bresenham for MIDI parameter slides while being significantly more efficient on AVR architecture. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude <[email protected]>
Eliminates pointer indirection in rx_isr hot path, reducing overhead from ~7 cycles to ~3 cycles per MIDI byte (~4 cycle savings). Also adds put_byte_bank1_isr for RP2040 platform compatibility. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude <[email protected]>
Mark stopRecord() as noinline to prevent unnecessary register saves in the hot path of MIDI receive ISR. - Change stopRecord() from ALWAYS_INLINE to __attribute__((noinline)) - Reduces register saves from 6 to 3 in RX ISR - 27% faster hot path (44 → 32 cycles on AVR) - Saves 24 bytes of flash Move live_state to MidiUartParent for better memory layout: - Improved offset access in ISR (Y+57 → Y+14 on AVR) - Better architecture - state belongs in parent class 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude <[email protected]>
Improves slide recalculation performance with two optimizations: 1. Optimize find_mask construction (~60% faster) - Build mask from locks_params first, then AND with step locks - Eliminates cur_mask variable and per-iteration shift - Reduces operations from ~44 to ~17 per recalc 2. Add early skip for empty steps in find_next_locks - Skip steps with no locks and no trig entirely - Saves 8 inner loop iterations per empty step - Reduces overhead for sparse lock patterns Binary impact: +12 bytes Runtime impact: Faster for typical slide patterns 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude <[email protected]>
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.
No description provided.