Display dma optimization - #99
Merged
Merged
Conversation
cofob
marked this pull request as draft
August 29, 2026 13:18
|
Development firmware for this pull request: https://rskey.fob.wtf/preview?pr=99 The link always opens the latest successful preview build. |
TheMaxMur
marked this pull request as ready for review
September 2, 2026 16:23
TheMaxMur
approved these changes
Sep 2, 2026
TheMaxMur
added a commit
that referenced
this pull request
Sep 8, 2026
Brings PR #99 (the retained framebuffer-less display compositor over a PIO DMA link) and 27 other commits from main onto develop. Conflicts, and how each was settled: - firmware/src/main.rs — bcdDevice: develop's 0x09C6 wins; main's 0x0989 is 345 builds behind it. - scripts/check.sh — kept develop's FLASH_SIZE=16M display row (it compiles the geometry the shipped package actually is) and main's new stack-floor row. - crates/rsk-display/src/tests.rs — both sides added independent recorder fields; kept both. - crates/rsk-ui/src/render_tests.rs — kept main's new hold-fill damage test and develop's wording of the pin-dots doc comment. - formal/RSKeyAppletSeams.tla — neither side's line span was right after the merge. Re-derived by content: ccid.rs:310-319 did not move, pin.rs is 704-712 now (PR #99 shifted it and turned self.panel into self.frame()). - formal/citations.lock — took develop's side wholesale. main's half is locked against pre-develop ctaphid.rs/main.rs/reset.rs/fs.rs; it is stale, not a competing claim. The lock gets re-derived against the merged tree separately. - CHANGELOG.md — PR #99's entry folded into develop's [Unreleased] Changed. Verified: the display flavor builds (LED_KIND=none FLASH_SIZE=16M, --features display). Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
TheMaxMur
added a commit
that referenced
this pull request
Sep 8, 2026
…tchets and a matrix cell Bookkeeping the `main` merge owes, and none of it is a re-label. The 41 drifted citations were repaired BY CONTENT, not by `--relock`. Running --relock first is what showed why: it re-pointed `PLAT-TIMER-002` from `hp.spawn(led::gpio_task(led).unwrap());` at `18 => gpio_pin!(p.PIN_18),` and `PLAT-TIMER-003` from `usb_attach::mark();` at a bare `};`, silently, because a stale line number still resolves to whatever now sits on it. The tool says so in its own output — "re-locked, not repaired" — and it is right. So the citing pages were renumbered from the gate's own old->new report, the gate re-run until it reported zero drift, and only then re-locked. The lock's content columns are byte-identical across all 1177 rows except one: `RSKeyAppletSeams`'s citation of `pin.rs`, where PR #99 turned `self.panel` into `self.frame()`. `SEC-SEAM-001..003` and `SEC-TRANS-001..003` x waveshare-touch-lcd were re-decided rather than re-labelled. `display.spi_freq_hz` stopped being only a panel knob when the PIO transport started taking `clk_sys / 2` from it, so this column now runs the part at 160 MHz where every other runs 150. That is real and it is recorded, but it does not reach these six: no clause in `RSKeyAppletSeams` or `RSKeyTransport` quantifies over time, a rate or a deadline, so a core that retires instructions faster cannot make one true or false. The cell says that, and says what would have owed `conditional` instead. Regenerated: assurance-matrix, assurance-vector, assurance-bounds and platform-assumptions. Only one status-bearing change is in them — `PLAT-TIMER-001` and `PLAT-UNSAFE-009` go fresh -> stale, TIMER-001 because the clock moved under the evidence. Committed so that shows up in a diff, which is the whole point of the row that demanded it. Two ratchets the merge moved and did not carry: `SUITE_CASES` 2608 -> 2609 (main's `test_spdx_gate.py` gained a case) and `verified-compilation.md`'s row counts 121/19 -> 123/21 (`firmware stack floor (display)` and `preview publisher`). `reproduce.sh` also had no shape for `node --test`, so the preview publisher row was UNCLAIMED and red; it is a gate-phase row like every other `check.sh` command. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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.
What
How it was tested
nix develop -c ./scripts/check.shpasses locallyChecklist
config.device_release(bcdDevice) bumped by one (hex)in
firmware/src/main.rs— host-only (CLI / docs / CI / build) does not (CONTRIBUTING.md)rollback epoch (
docs/production.md, stage 3) — a seal-time--rollbackcall, not a code bumpunsafe→ justified indocs/unsafe.mddocs/updatedAGPL-3.0-only)fido:,piv:,rsk:,docs:, …)