scottk/UltraGrid
1
0
Fork 0
mirror of https://github.com/outbackdingo/UltraGrid.git synced 2026-03-21 10:40:21 +00:00
Code Issues Packages Projects Releases Wiki Activity

ring_buffer: expose the zero-copy interface

Browse Source
This commit is contained in:
Martin Piatka
2021-08-30 14:52:49 +02:00
parent 59f547a261
commit 250eafdb48
2 changed files with 59 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
src/utils/ring_buffer.cpp
View File

@@ -93,7 +93,7 @@ static int calculate_avail_write(int start, int end, int buf_len) {
}
static int ring_get_read_regions(struct ring_buffer *ring, int max_len,
int ring_get_read_regions(struct ring_buffer *ring, int max_len,
void **ptr1, int *size1,
void **ptr2, int *size2)
{
@@ -124,7 +124,7 @@ static int ring_get_read_regions(struct ring_buffer *ring, int max_len,
return read_len;
}
static void ring_advance_read_idx(struct ring_buffer *ring, int amount) {
void ring_advance_read_idx(struct ring_buffer *ring, int amount) {
// start index is modified only by this (reader) thread, so relaxed is enough
int start = std::atomic_load_explicit(&ring->start, std::memory_order_relaxed);
@@ -161,7 +161,7 @@ void ring_buffer_flush(struct ring_buffer * buf) {
buf->end = 0;
}
static bool ring_get_write_regions(struct ring_buffer *ring, int requested_len,
int ring_get_write_regions(struct ring_buffer *ring, int requested_len,
void **ptr1, int *size1,
void **ptr2, int *size2)
{
@@ -174,7 +174,7 @@ static bool ring_get_write_regions(struct ring_buffer *ring, int requested_len,
int end = std::atomic_load_explicit(&ring->end, std::memory_order_relaxed);
if(requested_len > ring->len) {
return false;
return 0;
}
int end_idx = end % ring->len;
@@ -186,10 +186,10 @@ static bool ring_get_write_regions(struct ring_buffer *ring, int requested_len,
*size2 = requested_len - *size1;
}
return true;
return *size1 + *size2;
}
static bool ring_advance_write_idx(struct ring_buffer *ring, int amount) {
bool ring_advance_write_idx(struct ring_buffer *ring, int amount) {
const int start = std::atomic_load_explicit(&ring->start, std::memory_order_acquire);
// end index is modified only by this (writer) thread, so relaxed is enough
const int end = std::atomic_load_explicit(&ring->end, std::memory_order_relaxed);