Skip to content

M
miniaudio
Commit 4f96309c authored by David Reid's avatar David Reid
Browse files
Options

Add experimental loop detection to the routing system.

parent c5403669
Show whitespace changes
Inline Side-by-side
Showing with 167 additions and 87 deletions
miniaudio.h
View file @ 4f96309c
......@@ -1774,6 +1774,7 @@ typedef int ma_result;
#define MA_NO_DEVICE -104
#define MA_API_NOT_FOUND -105
#define MA_INVALID_DEVICE_CONFIG -106
#define MA_LOOP -107
/* State errors. */
#define MA_DEVICE_NOT_INITIALIZED -200
research/miniaudio_engine.h
View file @ 4f96309c
......@@ -609,7 +609,13 @@ earlier in the pipeline (data sources, for example) will be cheaper than the cos
higher level nodes, such as some kind of final post-processing endpoint. If you need to do mass
detachments, detach starting from the lowest level nodes and work your way towards the final
endpoint node (but don't try detaching the node graph's endpoint). If the audio thread is not
running, detachment will be fast and detachment in any order will be the same.
running, detachment will be fast and detachment in any order will be the same. The reason nodes
need to wait for their input attachments to complete is due to the potential for desyncs between
data sources. If the node was to terminate processing mid way through processing it's inputs,
there's a chance that some of the underlying data sources will have been read, but then others not.
That will then result in a potential desynchronization when detaching and reattaching higher-level
nodes. A possible solution to this is to have an option when detaching to terminate processing
before processing all input attachments which should be fairly simple.
Another compromise, albeit less significant, is locking when attaching and detaching nodes. This
locking is achieved by means of a spinlock in order to reduce memory overhead. A lock is present
......@@ -641,15 +647,12 @@ only be happening in a forward direction which means the "previous" pointer won'
used. The same general process applies to detachment. See `ma_node_attach_to_output/input_node()`
and `ma_node_detach_output_bus()` for the implementation of this mechanism.
One outstanding problem exists regarding attaching and detaching. It is possible for an output bus
to be detached while the audio thread is in the middle of processing it. This by itself is not a
problem because the node is still valid. The problem is that of reattaching the output bus to a new
input bus while a read is still happening on the audio thread. What *could* happen is that the node
is reattached to a new input bus which hasn't yet been iterated in the current call to
`ma_node_graph_read_pcm_frames()` thereby resulting in the node getting processed twice. This would
flow through the base data source and result in a desync because it's read from it twice in the
same call to `ma_node_graph_read_pcm_frames()`. This is an unusual scenario and would most likely
go unnoticed by the majority of people, but it's still an issue to consider.
Loop detection is achieved through the use of a counter. At the ma_node_graph level there is a
counter which is updated after each read. There is also a counter for each node which is set to the
counter of the node graph plus 1 after each time it processes data. Before anything is processed, a
check is performed that the node's counter is lower or equal to the node graph. If so, it's fine to
proceed with processing. If not, MA_LOOP is returned and nothing is output. This represents a sort
of termination point.
*/
......@@ -786,6 +789,7 @@ struct ma_node_base
ma_uint16 consumedFrameCountIn;
/* These variables are read and written between different threads. */
volatile ma_uint32 readCounter; /* For loop prevention. Compared with the current read count of the node graph. If larger, means a loop was encountered and reading is aborted and no samples read. */
volatile ma_node_state state;
ma_node_input_bus inputBuses[MA_MAX_NODE_BUS_COUNT];
ma_node_output_bus outputBuses[MA_MAX_NODE_BUS_COUNT];
......@@ -2696,15 +2700,20 @@ MA_API ma_result ma_node_graph_read_pcm_frames(ma_node_graph* pNodeGraph, void*
channels = ma_node_get_output_channels(&pNodeGraph->endpoint, 0);
ma_node_graph_set_is_reading(pNodeGraph, MA_TRUE);
{
/* We'll be nice and try to do a full read of all frameCount frames. */
totalFramesRead = 0;
while (totalFramesRead < frameCount) {
ma_uint32 framesJustRead;
ma_uint32 framesToRead = frameCount - totalFramesRead;
ma_node_graph_set_is_reading(pNodeGraph, MA_TRUE);
{
result = ma_node_read_pcm_frames(&pNodeGraph->endpoint, 0, (float*)ma_offset_pcm_frames_ptr(pFramesOut, totalFramesRead, ma_format_f32, channels), framesToRead, &framesJustRead);
}
ma_node_graph_set_is_reading(pNodeGraph, MA_FALSE);
ma_node_graph_increment_read_counter(pNodeGraph);
totalFramesRead += framesJustRead;
if (result != MA_SUCCESS) {
......@@ -2720,9 +2729,6 @@ MA_API ma_result ma_node_graph_read_pcm_frames(ma_node_graph* pNodeGraph, void*
if (pFramesRead != NULL) {
*pFramesRead = totalFramesRead;
}
}
ma_node_graph_set_is_reading(pNodeGraph, MA_FALSE);
ma_node_graph_increment_read_counter(pNodeGraph);
return result;
}
......@@ -3536,6 +3542,24 @@ MA_API float ma_node_get_output_bus_volume(const ma_node* pNode, ma_uint32 outpu
}
static ma_uint32 ma_node_set_read_counter(ma_node* pNode, ma_uint32 newReadCounter)
{
ma_node_base* pNodeBase = (ma_node_base*)pNode;
ma_uint32 oldReadCounter;
MA_ASSERT(pNodeBase != NULL);
/*
This function will be only ever be called in a controlled environment (only on the audio
thread, and never concurrently).
*/
oldReadCounter = c89atomic_load_32(&pNodeBase->readCounter);
c89atomic_exchange_32(&pNodeBase->readCounter, newReadCounter);
return oldReadCounter;
}
static void ma_node_process_pcm_frames_ex_simple(ma_node* pNode, float** ppFramesOut, ma_uint32* pFrameCountOut, const float** ppFramesIn, ma_uint32* pFrameCountIn)
{
ma_node_base* pNodeBase = (ma_node_base*)pNode;
......@@ -3575,6 +3599,7 @@ static ma_result ma_node_read_pcm_frames(ma_node* pNode, ma_uint32 outputBusInde
ma_uint32 inputBusCount;
ma_uint32 outputBusCount;
ma_uint32 totalFramesRead = 0;
ma_uint32 readCounter;
float* ppFramesIn[MA_MAX_NODE_BUS_COUNT];
float* ppFramesOut[MA_MAX_NODE_BUS_COUNT];
......@@ -3611,6 +3636,12 @@ static ma_result ma_node_read_pcm_frames(ma_node* pNode, ma_uint32 outputBusInde
inputBusCount = ma_node_get_input_bus_count(pNode);
outputBusCount = ma_node_get_output_bus_count(pNode);
/*
We need to grab the read counter at the start so we can set a new read counter first up. We
need to do this first so that recursive reads can have access to the new counter.
*/
readCounter = ma_node_set_read_counter(pNode, ma_node_graph_get_read_counter(pNodeBase->pNodeGraph) + 1);
/*
Run a simplified path when there are no inputs and one output. In this case there's nothing to
actually read and we can go straight to output. This is a very common scenario because the vast
......@@ -3621,9 +3652,16 @@ static ma_result ma_node_read_pcm_frames(ma_node* pNode, ma_uint32 outputBusInde
ma_uint32 inputFrameCount = 0;
ma_uint32 outputFrameCount = frameCount; /* Just read as much as we can. The callback will return what was actually read. */
/* Don't do anything if our read counter is ahead of the node graph. That means we're */
if (readCounter <= ma_node_graph_get_read_counter(pNodeBase->pNodeGraph)) {
ppFramesOut[0] = pFramesOut;
ma_node_process_pcm_frames_ex(pNode, ppFramesOut, &outputFrameCount, NULL, &inputFrameCount);
totalFramesRead = outputFrameCount;
} else {
/* Read counter is ahead of the node graph. Loop detected. Abort this read with nothing read. */
totalFramesRead = 0;
result = MA_LOOP; /* So the caller knows to stop attempting to read more data. */
}
} else {
/* Slow path. Need to do caching. */
ma_uint32 framesToRead;
......@@ -3660,6 +3698,8 @@ static ma_result ma_node_read_pcm_frames(ma_node* pNode, ma_uint32 outputBusInde
pNodeBase->cachedFrameCountOut = 0;
/* If our node's counter is ahead of the node graph's counter it means we've hit a loop. In this case we need to skip reading entirely. */
if (readCounter <= ma_node_graph_get_read_counter(pNodeBase->pNodeGraph)) {
/*
We need to prepare our output frame pointers for processing. In the same iteration we need
to mark every output bus as unread so that future calls to this function for different buses
......@@ -3723,6 +3763,10 @@ static ma_result ma_node_read_pcm_frames(ma_node* pNode, ma_uint32 outputBusInde
pNodeBase->consumedFrameCountIn += (ma_uint16)frameCountIn;
pNodeBase->cachedFrameCountIn -= (ma_uint16)frameCountIn;
pNodeBase->cachedFrameCountOut = (ma_uint16)frameCountOut;
} else {
/* Getting here means the loop counter of the node is ahead of the graph which means we've hit a loop. */
result = MA_LOOP; /* So the caller knows to stop attempting to read more data. */
}
} else {
/*
We're not needing to read anything from the input buffer so just read directly from our
......
research/miniaudio_routing.c
View file @ 4f96309c
......@@ -5,6 +5,7 @@
ma_node_graph g_nodeGraph;
ma_data_source_node g_dataSourceNode;
ma_splitter_node g_splitterNode;
ma_splitter_node g_loopNode; /* For testing loop detection. We're going to route one of these endpoints back to g_splitterNode to form a loop. */
void data_callback(ma_device* pDevice, void* pFramesOut, const void* pFramesIn, ma_uint32 frameCount)
{
......@@ -89,18 +90,41 @@ int main(int argc, char** argv)
}
/*
Splitter node. Note that we've already attached the data source node to another, so this section
will test that changing of attachments works as expected.
*/
splitterNodeConfig = ma_splitter_node_config_init(device.playback.channels);
/* Loop detection testing. */
result = ma_splitter_node_init(&g_nodeGraph, &splitterNodeConfig, NULL, &g_loopNode);
if (result != MA_SUCCESS) {
printf("Failed to initialize loop node.");
return -1;
}
/* Connect both outputs of the splitter to the endpoint for now. Later on we'll test effects and whatnot. */
ma_node_attach_to_output_node(&g_loopNode, 0, ma_node_graph_get_endpoint(&g_nodeGraph), 0);
ma_node_attach_to_output_node(&g_loopNode, 1, ma_node_graph_get_endpoint(&g_nodeGraph), 0);
/* Adjust the volume of the splitter node's endpoints. We'll just do it 50/50 so that both of them combine to reproduce the original signal at the endpoint. */
ma_node_set_output_bus_volume(&g_loopNode, 0, 0.5f);
ma_node_set_output_bus_volume(&g_loopNode, 1, 0.5f);
result = ma_splitter_node_init(&g_nodeGraph, &splitterNodeConfig, NULL, &g_splitterNode);
if (result != MA_SUCCESS) {
printf("Failed to initialize splitter node.");
return -1;
}
#if 0
/* Connect both outputs of the splitter to the endpoint for now. Later on we'll test effects and whatnot. */
ma_node_attach_to_output_node(&g_splitterNode, 0, ma_node_graph_get_endpoint(&g_nodeGraph), 0);
ma_node_attach_to_output_node(&g_splitterNode, 1, ma_node_graph_get_endpoint(&g_nodeGraph), 0);
......@@ -108,11 +132,22 @@ int main(int argc, char** argv)
/* Adjust the volume of the splitter node's endpoints. We'll just do it 50/50 so that both of them combine to reproduce the original signal at the endpoint. */
ma_node_set_output_bus_volume(&g_splitterNode, 0, 0.5f);
ma_node_set_output_bus_volume(&g_splitterNode, 1, 0.5f);
#else
/* Connect both outputs of the splitter to the endpoint for now. Later on we'll test effects and whatnot. */
ma_node_attach_to_output_node(&g_splitterNode, 0, &g_loopNode, 0);
ma_node_attach_to_output_node(&g_splitterNode, 1, &g_loopNode, 1);
/* Now loop back to the splitter node to form a loop. */
ma_node_attach_to_output_node(&g_loopNode, 1, &g_splitterNode, 0);
#endif
/* The data source needs to have it's connection changed from the endpoint to the splitter. */
ma_node_attach_to_output_node(&g_dataSourceNode, 0, &g_splitterNode, 0);
/* Stop the splitter node for testing. */
/*ma_node_set_state(&g_splitterNode, ma_node_state_stopped);*/
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment