track(platform): validate STM32 / STM32duino FastLED build support

[zackees]

Context

Several open FastLED issues are STM32 platform-support or STM32duino build failures. fbuild already has STM32 board metadata and smoke fixtures for F1/F4/H7/U5-class targets, so this tracker should reconcile current fbuild coverage with the old FastLED issues and identify any remaining board/backend gaps.

Source FastLED issues

• Fastled STM32 support FastLED#590 - Fastled STM32 support
• Support STM32F030C8T6 FastLED#750 - Support STM32F030C8T6
• Does not compile for STM32F103 blue-pill FastLED#805 - Does not compile for STM32F103 blue-pill
• Problem adapting FastLED library for stm32l476rg board FastLED#975 - Problem adapting FastLED library for stm32l476rg board
• STM32duino support? FastLED#1134 - STM32duino support?
• STM32 based boards supported, but fail to compile. FastLED#1312 - STM32 based boards supported, but fail to compile.
• Support for Portenta H7 FastLED#1304 - Support for Portenta H7

Acceptance criteria

• Map each source issue to an existing fbuild board/env where possible, or add a missing smoke fixture when the board is not represented.
• Run current FastLED minimal builds for representative STM32F0/F1/F4/L4/H7 targets, including Portenta H7 / STM32H747 if available in fbuild.
• Record which failures are fbuild backend/toolchain issues versus FastLED pin/platform code issues.
• Add or update fbuild regression coverage for any fbuild-owned gap.
• Comment back on each source FastLED issue with the fbuild tracker, result, and any remaining FastLED-side blocker.

[zackees]

Mapping table: source FastLED issues -> fbuild coverage

FastLED issue	Target MCU/board	fbuild board	fbuild smoke fixture	fbuild family	Build result	Side of remaining work
FastLED/FastLED#590	STM32F103RB Nucleo / maple mini	bluepill_f103c8, maple_mini_b20	tests/platform/stm32f103c8, tests/platform/stm32f103cb	STM32F1 (Cortex-M3)	PASS (17.7 KB flash)	FastLED platform headers
FastLED/FastLED#750	STM32F030C8T6	nucleo_f030r8 (same F030x8 die family)	tests/platform/nucleo_f030r8 (new, marker)	STM32F0 (Cortex-M0)	FAIL - fbuild gap: no stm32f0.json mcu_config	fbuild backend (then FastLED)
FastLED/FastLED#805	STM32F103 blue-pill	bluepill_f103c8	tests/platform/stm32f103c8	STM32F1 (Cortex-M3)	PASS (17.7 KB flash)	FastLED platform headers
FastLED/FastLED#975	STM32L476RG	nucleo_l476rg	tests/platform/nucleo_l476rg (new, marker)	STM32L4 (Cortex-M4F, no DSP)	FAIL - fbuild gap: no stm32l4.json mcu_config	fbuild backend (then FastLED)
FastLED/FastLED#1134	STM32duino broadly	F1/F4/H7/U5 covered today	existing F1/F4/H7 fixtures	mixed	PASS for F1/F4/H7	FastLED platform headers
FastLED/FastLED#1304	Portenta H7 / STM32H747	portenta_h7_m7 (same MCU as giga_r1_m7)	tests/platform/stm32h747xi (giga_r1_m7)	STM32H7 (Cortex-M7)	PASS (94.1 KB flash)	FastLED platform headers
FastLED/FastLED#1312	bluepill F103C8 redefines	bluepill_f103c8	tests/platform/stm32f103c8	STM32F1 (Cortex-M3)	PASS (17.7 KB flash)	FastLED platform headers (HAL/pin collisions)

fbuild backend results (Arduino framework, ststm32 platform, arm-none-eabi-gcc 15.2.1)

stm32f103c8 (bluepill F103C8)         OK   17712 bytes flash / 2816 bytes ram
stm32f103cb (maple_mini_b20)          OK   17712 bytes flash / 2816 bytes ram
stm32f103tb (hy_tinystm103tb)         OK   17472 bytes flash / 2816 bytes ram
stm32f411ce (blackpill_f411ce)        OK   18852 bytes flash / 2840 bytes ram
nucleo_f429zi                         OK   21096 bytes flash / 2920 bytes ram
stm32h747xi (giga_r1_m7 / Portenta H7) OK  96304 bytes flash / 324224 bytes ram (mbed)
nucleo_f030r8                         FAIL  unsupported STM32 MCU 'stm32f030r8t6'
nucleo_l476rg                         FAIL  unsupported STM32 MCU 'stm32l476rgt6'

fbuild-owned gaps surfaced

1. crates/fbuild-build/src/stm32/mcu_config.rs does not map stm32f0* -> Cortex-M0 config. Blocks Support STM32F030C8T6 FastLED#750.
2. crates/fbuild-build/src/stm32/mcu_config.rs does not map stm32l4* -> Cortex-M4F (no DSP) config. Blocks Problem adapting FastLED library for stm32l476rg board FastLED#975.

Both require a new JSON config under crates/fbuild-build/src/stm32/configs/ plus a prefix branch in get_stm32_config_for_mcu. Out of scope for this tracker per the issue's "documented status, not a full STM32 test suite expansion" guidance - the smoke fixtures are added as red regression markers so the gap is self-evident and CI will turn green automatically once the configs land.

Changes landed for this tracker

PR #396:

• tests/platform/nucleo_f030r8/ - STM32F0 gap-tracking smoke fixture
• tests/platform/nucleo_l476rg/ - STM32L4 gap-tracking smoke fixture
• crates/fbuild-build/src/stm32/mcu_config.rs - unsupported-MCU error now names the gap and cross-links track(platform): validate STM32 / STM32duino FastLED build support #385
• .gitignore - ignore transient tests/platform/*/build_info*.json

FastLED-side comment links

• Fastled STM32 support FastLED#590 -> Fastled STM32 support FastLED#590 (comment)
• Support STM32F030C8T6 FastLED#750 -> Support STM32F030C8T6 FastLED#750 (comment)
• Does not compile for STM32F103 blue-pill FastLED#805 -> Does not compile for STM32F103 blue-pill FastLED#805 (comment)
• Problem adapting FastLED library for stm32l476rg board FastLED#975 -> Problem adapting FastLED library for stm32l476rg board FastLED#975 (comment)
• STM32duino support? FastLED#1134 -> STM32duino support? FastLED#1134 (comment)
• STM32 based boards supported, but fail to compile. FastLED#1312 -> STM32 based boards supported, but fail to compile. FastLED#1312 (comment)
• Support for Portenta H7 FastLED#1304 -> Support for Portenta H7 FastLED#1304 (comment)

[zackees]

Closing as tracked + documented.

Outcome of the reconciliation:

• F1, F4, H7, U5 STM32 families: green on fbuild backend today. Smoke builds verified for bluepill_f103c8, maple_mini_b20, hy_tinystm103tb, blackpill_f411ce, nucleo_f429zi, and giga_r1_m7 (same MCU as Portenta H7 M7). Source FastLED issues chore(release): bump Cargo.toml workspace version 2.2.27 -> 2.2.28 #590, audit: blocking operations with no timeout in fbuild-packages (sub-issue of #802) #805, fix(ch32v): select D8 register map for V303 #1134, fix(rp): recover application USB without replug #1304, and refactor(platform): phase 6/9 migrate daemon IPC and lifecycle mechanics #1312 now hinge on FastLED's own src/platforms/arm/stm32/ platform-header work; fbuild is no longer the blocker for any of them.
• F0 and L4 families: gap surfaced and now tracked with red smoke fixtures (tests/platform/nucleo_f030r8, tests/platform/nucleo_l476rg) plus a self-documenting error in mcu_config.rs cross-linking this tracker. Blocks Support STM32F030C8T6 FastLED#750 (F0) and Problem adapting FastLED library for stm32l476rg board FastLED#975 (L4). Adding the stm32f0.json and stm32l4.json MCU configs is a small, well-scoped follow-up that does not need its own tracker - it can be picked up directly off the fixture failures.

Per the issue's "do not over-engineer" guidance, the actual F0/L4 mcu_config implementation is intentionally out of scope here. The fixtures and error wiring landed in PR #396 make sure that work cannot regress silently.

See the mapping table comment above for evidence and per-issue resolution status. All seven source FastLED issues have been commented on with the fbuild status and the appropriate next-step owner.