vwemirror/BT411
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Audio-dropout fix: OpenAL source-pool lifecycle (uninit aliasing, atomic-delete leak, stranded partials)

Browse Source 
Field report: audio cutting in and out toward the end of the match.
Three structural defects in the port's source-pool lifecycle:

1. SourceSet.sources[] was uninitialized heap garbage until first
   acquisition -- AL names are small ints, so a recycled-heap slot could
   alias a LIVE source owned by another sound; alIsSource() skipped
   generation, two sounds shared one source, and whichever released
   first deleted the other's mid-play.  SourceSet ctor now zero-inits.
2. ReleaseSourceSet's bulk alDeleteSources(count, sources) is ATOMIC on
   invalid names (AL spec) -- one empty/-1 slot and NOTHING deleted;
   after the first pool exhaustion every partial release leaked and the
   pool never recovered.  Now per-slot guarded delete, slots parked at 0.
3. Failed acquisitions stranded partial sets forever (dropped transients
   are never released by anything).  ReleaseChannels() handback at both
   failure sites (StartRequest + dormant-resume) + a dtor backstop.

Verified: 2-node MP smoke (mission + lobby-loop relaunch clean, census
live/acquireFails 0/0) + 100s solo combat smoke (52 audio triggers,
zero AL errors).  Awaiting live playtest confirmation at scale.
KB: context/wintesla-port.md audio-dropout section.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
arcattack
2026-07-23 11:23:07 -05:00
co-authored by Claude Opus 4.8
parent cbebb0f1c1
commit a999e5c49f
4 changed files with 73 additions and 12 deletions
Download Patch File Download Diff File Expand all files Collapse all files
context/wintesla-port.md
+21
View File
@@ -144,6 +144,27 @@ on top of it. Full detail: `docs/PROGRESS_LOG.md §5b, §8`.
watcher timing encore-erratic that session. Harness kept (env-gated, additive): `[aud-tail]`
logging in `L4AUDLVL.cpp` (PLAY/STOP/fade) + `emitter.cpp` (state transitions), and
`BT_FIRE_PULSE="on,off"` single-shot trigger pulsing in `mech4.cpp`.
- **Audio-dropout fix (field 2026-07-23: "cutting in and out toward the end of the match")** [T2,
code-verified vs the AL spec; awaiting live playtest confirm]: three structural defects in the
port's OpenAL source-pool lifecycle (`L4AUDRND.cpp` / `L4AUDIO.cpp` / `L4AUDHDW.h`):
(1) `SourceSet.sources[]` was **uninitialized heap garbage** until first acquisition — AL names
are small ints, so a recycled-heap slot could alias a LIVE source owned by another sound;
`alIsSource()` then skipped generation, two sounds shared one source, and whichever released
first deleted the other's source mid-play (random sound deaths worsening with heap churn).
Fix: `SourceSet` ctor zero-inits (0 is never a valid name). (2) `ReleaseSourceSet` used bulk
`alDeleteSources(count, sources)`, which is **ATOMIC on invalid names** — one empty/-1 slot
(partial set, double release) and NOTHING deleted; after the first pool exhaustion every partial
release leaked and the pool never recovered. Fix: per-slot guarded delete, slots parked at 0.
(3) Failed acquisitions stranded partial sets forever — dropped transients are never released by
anything (the dtor didn't release; maintenance only touches running sources). Fix: handback via
`ReleaseChannels()` at both failure sites (StartRequest + dormant-resume) + a dtor backstop.
Context: the port's `RequestAudioChannels` has NO channel budget (the 1995 per-card budget is
the commented-out block) — it alGenSources until OpenAL Soft's ~256 cap; on failure a transient
is silently dropped (StartRequest early-return), a sustained goes dormant — which is exactly the
"cutting in and out" presentation. Priority-steal (`RequestAudioResources`) only engages at cap.
PERMANENT diagnostics: 30 s `[audio] source census: live/acquireFails` + an always-on
`[audio] ACQUIRE FAILED` line. Baseline (2-node soak): live ≈ 21 peak under fire, 0 at mission
end; exhaustion needs 7-player scale or the leak spiral, aliasing needs neither.
- **L4 HAL on Windows + the VS build system: DONE.** [T1]
- **Red Planet game logic: COMPLETE & buildable** — `VTV/VTVMPPR/WEAPSYS/...`. RP is at/near playable. [T1]
- **Multiplayer: the DOS `NETNUB` driver is already replaced by a ~3.5k-line WinSock2 TCP
engine/MUNGA_L4/L4AUDHDW.h
+13
View File
@@ -24,6 +24,19 @@ struct SourceSet
{
int count;
ALuint sources[AUDIO_SOURCESET_CAPACITY];
// Audio-dropout fix (field 2026-07-23): sources[] was UNINITIALIZED heap
// garbage until the first acquisition. OpenAL names are small integers --
// exactly what recycled heap is full of -- so a garbage slot could alias a
// LIVE source owned by another sound; alIsSource() then skipped generation
// and two sounds shared one AL source, and whichever released first
// deleted the other's source mid-play (sounds randomly dying, worsening
// with heap churn late in a match). 0 is never a valid source name.
SourceSet() : count(0)
{
for (int i = 0; i < AUDIO_SOURCESET_CAPACITY; i++)
sources[i] = 0;
}
};
//~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ MIDI types ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
engine/MUNGA_L4/L4AUDIO.cpp
+14
View File
@@ -719,6 +719,20 @@ void
//
L4AudioSource::~L4AudioSource()
{
// Audio-dropout fix: backstop -- if this source dies still holding OpenAL
// sources (the stop/suspend maintenance never ran for it), hand them back
// so entity churn can't drain the pool over a long match.
if (application != NULL && application->GetAudioRenderer() != NULL)
{
for (int i = 0; i < channelSet.count; i++)
{
if (alIsSource(channelSet.sources[i]))
{
application->GetAudioRenderer()->ReleaseSourceSet(channelSet);
break;
}
}
}
}
//
engine/MUNGA_L4/L4AUDRND.cpp
+25 -12
View File
@@ -558,6 +558,12 @@ void
);
if (!resources_available)
{
// Audio-dropout fix: a failed acquisition can leave a PARTIAL set
// (alGenSources succeeded for some slots before the pool ran dry).
// A dropped transient never plays and nothing else ever released it,
// so the partial sources dangled forever -- once the pool exhausted
// ONCE, every drop leaked a little more and it never recovered.
audio_source->ReleaseChannels();
if (audio_source->GetAudioRenderType() == SustainedAudioRenderType)
{
//
@@ -1024,7 +1030,12 @@ void
if (!RequestAudioResources(audio_source,
source_result)
)
{
// Audio-dropout fix: hand back any partial acquisition (see
// StartRequest) -- the source stays dormant and retries later.
audio_source->ReleaseChannels();
break;
}
//
// Remove from dormant list and request list
@@ -1445,26 +1456,28 @@ long gBTAudioAcquireFails = 0;
void L4AudioRenderer::ReleaseSourceSet(SourceSet &sourceSet)
{
// Audio-dropout fix: the old bulk alDeleteSources(count, sources) is
// ATOMIC per the AL spec -- ONE invalid name in the array (an empty slot
// of a partial set, or the old -1 sentinel on a double release) and
// NOTHING is deleted. After the first pool-exhaustion event every
// partial release leaked its real sources and the pool never recovered
// ("audio cutting in and out toward the end of the match"). Delete each
// valid source individually and park the slot at 0 (never a valid name).
extern long gBTAudioSourcesLive;
for (int i = 0; i < sourceSet.count; i++)
{
if (alIsSource(sourceSet.sources[i]))
--gBTAudioSourcesLive;
ALenum state;
alGetSourcei(sourceSet.sources[i], AL_SOURCE_STATE, &state);
if (state == AL_PLAYING)
{
alSourceStop(sourceSet.sources[i]);
ALint state = AL_STOPPED;
alGetSourcei(sourceSet.sources[i], AL_SOURCE_STATE, &state);
if (state == AL_PLAYING)
alSourceStop(sourceSet.sources[i]);
alDeleteSources(1, sourceSet.sources + i);
--gBTAudioSourcesLive;
}
sourceSet.sources[i] = 0;
}
alDeleteSources(sourceSet.count, sourceSet.sources);
for (int i = 0; i < sourceSet.count; i++)
{
sourceSet.sources[i] = -1;
}
}
//~~~~~~~~~~~~~~~~~~~~~~ L4AudioRenderer profile bits ~~~~~~~~~~~~~~~~~~~~~~~~~