3 Commits

Author	SHA1	Message	Date
Siavash Sameni	5e9718aeb2	docs: incident report — SIGBUS in ART GC during audio JNI calls ... Android 16's concurrent mark-compact GC crashes when flipping thread roots on our MAX_PRIORITY audio threads during JNI calls (AudioRecord.read / AudioTrack.write). Not our code — all crash frames are in libart.so. Proposed fixes: - Short term: DirectByteBuffer to reduce JNI transitions - Long term: Oboe native audio from Rust (no JNI, no GC) Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>	2026-04-06 19:21:32 +04:00
Claude	6597b5bd86	docs: incident report + fix spec for capture thread use-after-free crash ... SIGSEGV on hangup: capture thread calls writeAudio() via JNI after teardown() has freed the native engine handle. TOCTOU race between the nativeHandle==0L check and destroy() on the ViewModel thread. Fix: CountDownLatch(2) — audio threads count down after exiting loops, teardown() awaits before destroy(). 2 Kotlin files, no Rust changes. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>	2026-04-06 09:21:35 +00:00
Claude	4d54b6f9e4	docs: incident reports for send-task crash and playout ring desync ... Two root-caused bugs documented with full evidence: 1. Send task fatal exit on QUIC congestion (FIXED in 2092245) - send_media() Err(Blocked) caused break → killed entire call - Now drops packet and continues 2. Playout ring buffer cursor desync (ROOT-CAUSED, fix pending) - AudioRing::write() mutates read_pos from producer thread on overflow - Violates SPSC contract → reader/writer fight over read_pos - Causes 12-16s bidirectional silence ~25-30s into call - Both clients affected simultaneously Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>	2026-04-06 08:52:14 +00:00