fix(spiflash): USB storage cannot persist files copied

[richardclli]

The root cause is after USB unplugged, static DSTATUS spi_flash_initialize(BYTE lun) will be called again. frftl need to be initialized only once, a second init will cause memory leak and data loss.

No problem in 2.10.x, the bug is introduce from 2.11 onwards.

Summary by CodeRabbit

• Bug Fixes
  • Enhanced flash memory initialization to properly handle repeated initialization attempts and ensure reliable storage synchronization when the flash file transfer layer is enabled.

[coderabbitai]

📝 Walkthrough

Walkthrough

spi_flash_initialize gains a two-path control flow under USE_FLASH_FTL: repeated calls now invoke ftlSync and return STA_NOINIT on failure; first-time calls compute flash size in MB, call ftlInit, and set frftlInitDone only on success.

Changes

FRFTL Initialization Control Flow

Layer / File(s)	Summary
FRFTL repeat-init and first-init logic radio/src/targets/common/arm/stm32/diskio_spi_flash.cpp	Adds a branch for already-initialized state (frftlInitDone) that calls ftlSync and returns STA_NOINIT on failure. First-time path computes flash size in MB before calling ftlInit, and guards frftlInitDone = true behind successful initialization.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~5 minutes

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Description check	⚠️ Warning	PR description lacks required template structure, missing explicit issue reference and detailed summary section.	Add 'Fixes #' with issue number and expand summary under 'Summary of changes:' to match template requirements.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Docstring Coverage	✅ Passed	Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Title check	✅ Passed	The title mentions USB storage and files not persisting, which directly aligns with the PR objective of fixing a persistence bug, but it doesn't specify the technical root cause (repeated initialization of spi_flash_initialize).

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch richardclli/fix-pl18-usb-copy-data-loss

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[richardclli]

@raphaelcoeffic I am not feeling comfortable about the control flow of USB plug and unplugged, maybe you should take a look in it. I solve my problem using a simple fix, but the point is why the driver will be initialized twice that cause the problem.

[richardclli]

@pfeerick I have tested this change, it is ready to merge. I asked Raphael to review, but that is another problem.

[richardclli]

I discovered it when I trying to setup my new heli. The new image file, just didn't store properly.

[pfeerick]

I discovered it when I trying to setup my new heli. The new image file, just didn't store properly.

Thanks for confirming I'm not going nuts... I could have sworn I had similar issues with PL18EV or PA01 but it was intermittent and probably thought I'd forgotten to transfer the files or something.

[richardclli]

Should not affect PA01, only nor flash filesystem has this problem, PA01 is just onboard sdcard chip.