mirror/miniaudio
1
0
Fork 0
mirror of https://github.com/mackron/miniaudio.git synced 2026-04-21 15:56:58 +02:00
Code Issues Packages Projects Releases Wiki Activity

Fix a subtle multi-threading bug.

Browse Source
This commit is contained in:
David Reid
2021-06-09 20:56:37 +10:00
parent 96bcc0075a
commit a10e763681
1 changed files with 13 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files
research/miniaudio_engine.h
+13 -8
View File
@@ -7269,8 +7269,9 @@ static ma_result ma_resource_manager_data_buffer_init_internal(ma_resource_manag
}
pDataBuffer->pResourceManager = pResourceManager;
pDataBuffer->pNode = pDataBufferNode;
pDataBuffer->flags = flags;
pDataBuffer->pNode = pDataBufferNode;
pDataBuffer->flags = flags;
pDataBuffer->result = MA_BUSY; /* Always default to MA_BUSY for safety. It'll be overwritten when loading completes or an error occurs. */
/* If we're loading asynchronously we need to post a job to the job queue to initialize the connector. */
if (async == MA_FALSE || ma_resource_manager_data_buffer_node_result(pDataBufferNode) == MA_SUCCESS) {
@@ -7290,6 +7291,13 @@ static ma_result ma_resource_manager_data_buffer_init_internal(ma_resource_manag
ma_resource_manager_inline_notification_init(pResourceManager, &initNotification);
}
/*
The status of the data buffer needs to be set to MA_BUSY before posting the job so that the
worker thread is aware of it's busy state. If the LOAD_DATA_BUFFER job sees a status other
than MA_BUSY, it'll assume an error and fall through to an early exit.
*/
c89atomic_exchange_i32(&pDataBuffer->result, MA_BUSY);
job = ma_job_init(MA_JOB_LOAD_DATA_BUFFER);
job.order = ma_resource_manager_data_buffer_next_execution_order(pDataBuffer);
job.loadDataBuffer.pDataBuffer = pDataBuffer;
@@ -7302,9 +7310,6 @@ static ma_result ma_resource_manager_data_buffer_init_internal(ma_resource_manag
/* TODO: Post an error here. */
c89atomic_exchange_i32(&pDataBuffer->result, result);
} else {
/* The job has been posted, so make sure the buffer's status is set to busy. The worker thread will */
c89atomic_exchange_i32(&pDataBuffer->result, MA_BUSY);
if ((flags & MA_DATA_SOURCE_FLAG_WAIT_INIT) != 0) {
ma_resource_manager_inline_notification_wait(&initNotification);
@@ -8717,7 +8722,7 @@ static ma_result ma_resource_manager_process_job__load_data_buffer_node(ma_resou
/* First thing we need to do is check whether or not the data buffer is getting deleted. If so we just abort. */
if (ma_resource_manager_data_buffer_node_result(pJob->loadDataBufferNode.pDataBufferNode) != MA_BUSY) {
result = MA_INVALID_OPERATION; /* The data buffer may be getting deleted before it's even been loaded. */
result = ma_resource_manager_data_buffer_node_result(pJob->loadDataBufferNode.pDataBufferNode); /* The data buffer may be getting deleted before it's even been loaded. */
goto done;
}
@@ -8874,7 +8879,7 @@ static ma_result ma_resource_manager_process_job__page_data_buffer_node(ma_resou
/* Don't do any more decoding if the data buffer has started the uninitialization process. */
if (ma_resource_manager_data_buffer_node_result(pJob->pageDataBufferNode.pDataBufferNode) != MA_BUSY) {
result = MA_UNAVAILABLE;
result = ma_resource_manager_data_buffer_node_result(pJob->pageDataBufferNode.pDataBufferNode);
goto done;
}
@@ -8952,7 +8957,7 @@ static ma_result ma_resource_manager_process_job__load_data_buffer(ma_resource_m
just abort, but making sure we increment the execution pointer.
*/
if (ma_resource_manager_data_buffer_result(pJob->loadDataBuffer.pDataBuffer) != MA_BUSY) {
result = MA_INVALID_OPERATION; /* The data buffer may be getting deleted before it's even been loaded. */
result = ma_resource_manager_data_buffer_result(pJob->loadDataBuffer.pDataBuffer); /* The data buffer may be getting deleted before it's even been loaded. */
goto done;
}