fix TEST_CASE("esp_littlefs_info returns used_bytes > total_bytes", "[littlefs]")

[BrianPugh]

Fix infinite loop in "used_bytes > total_bytes" test

When the filesystem is nearly full, littlefs internally recovers from NOSPC during metadata directory splits by falling back to in-place compaction (lfs_dir_splittingcompact). This means fwrite/fclose never return errors even though the allocator logs "No more free space" — the operation actually succeeds.

The test relied on I/O errors to exit its fill loop. This has broken before (ESP-IDF v5.0 changed stdio buffering so fwrite stopped failing; fixed by also checking fclose). Now neither fails.

Fix: detect disk-full by stopping once used_bytes from esp_littlefs_info stops growing for 10 consecutive iterations. The used <= total invariant is still checked every iteration.

[Copilot]

Pull request overview

Updates the LittleFS test esp_littlefs_info returns used_bytes > total_bytes to avoid an infinite loop when the filesystem recovers internally from NOSPC and stdio calls don’t return errors, by adding a secondary “disk full” detection based on used_bytes no longer increasing.

Changes:

• Add tracking (prev_used/stall_count) to break out of the fill loop when used_bytes stops growing for 10 iterations.
• Hoist total/used declarations outside the loop and reset them each iteration before calling esp_littlefs_info.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.