dangernoodle-io / breadboard / 27800364247

100%

push

github

jae/bb display active race (#462)

* fix: snapshot s_active to eliminate LoadProhibited race in bb_display

concurrent bb_display_off() on an OTA pause hook can null s_active
between the s_ready guard and the per-glyph blit deref in
bb_display_draw_text, causing a LoadProhibited panic (confirmed via
decoded backtrace at bb_display.c:216, TaipanMiner v0.30.0).

apply the local-snapshot pattern to all public functions that deref
s_active: clear, blit, flush, draw_text, set_rotation. the snapshot is
sufficient because s_active always points to a static-const backend
struct that is never freed; only the pointer variable itself races.

adds four host regression tests confirming the draw/blit/clear/flush
functions are safe no-ops after bb_display_off().

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix(bb_display): re-enable ssd1306/st77xx panels on bb_display_on (missing .on hook)

bb_display_on() calls s_panel->on() to physically re-enable the panel after
an OTA blanks it via bb_display_off(). SSD1306 and ST77xx backends defined
.off (disp_on_off false) but not .on, so panels stayed dark post-OTA. Added
matching .on handlers (disp_on_off true, backlight on for st77xx) wired in
all three backend structs: ssd1306, st7735, st7789. ILI9341 already had .on.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

* test(bb_display): cover set_rotation null-guard after off (B1-297)

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

* test(bb_display): cover s_active null-guard race branches via seam (B1-297)

the 5 !a guards only fire with s_ready=true and s_active=NULL (the concurrent-off race window); bb_display_off() clears both, so added a BB_DISPLAY_TESTING seam to reproduce the window and one test that drives all five draw paths. branch coverage back to 100%.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

---------

Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>

1628 of 1628 branches covered (100.0%)

Branch coverage included in aggregate %.

22 of 22 new or added lines in 1 file covered. (100.0%)

2722 of 2722 relevant lines covered (100.0%)

1207.82 hits per line