ggml-cuda: reorder only relevant nodes (#17639)

This commit is contained in:
Aman Gupta 2025-12-02 12:36:31 +08:00 committed by GitHub
parent 7b6d745364
commit ed32089927
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
2 changed files with 62 additions and 13 deletions

View File

@ -989,6 +989,10 @@ struct ggml_cuda_concurrent_event {
int n_streams = 0;
std::unordered_map<const ggml_tensor *, int> stream_mapping;
// Original order of nodes in this concurrent region (before interleaving)
// Used to restore grouping for fusion within streams
std::vector<const ggml_tensor *> original_order;
const ggml_tensor * join_node;
ggml_cuda_concurrent_event() = default;
@ -1011,6 +1015,7 @@ struct ggml_cuda_concurrent_event {
, fork_event(other.fork_event)
, n_streams(other.n_streams)
, stream_mapping(std::move(other.stream_mapping))
, original_order(std::move(other.original_order))
, join_node(other.join_node) {
other.fork_event = nullptr;
}
@ -1121,11 +1126,9 @@ struct ggml_cuda_concurrent_event {
};
struct ggml_cuda_stream_context {
std::vector<const ggml_tensor *> original_nodes;
std::unordered_map<const ggml_tensor *, ggml_cuda_concurrent_event> concurrent_events;
void reset() {
original_nodes.clear();
concurrent_events.clear();
}
};

View File

@ -3238,9 +3238,56 @@ static void evaluate_and_capture_cuda_graph(ggml_backend_cuda_context * cuda_ctx
}
}
if (should_launch_concurrent_events) {
//Restore the original graph to enable fusion within the streams
cgraph->nodes = const_cast<ggml_tensor **>(stream_ctx.original_nodes.data());
cgraph->n_nodes = (int) stream_ctx.original_nodes.size();
// Restore original node order within each concurrent region to enable fusion within streams
std::unordered_map<const ggml_tensor *, int> node_to_idx;
node_to_idx.reserve(cgraph->n_nodes);
for (int i = 0; i < cgraph->n_nodes; ++i) {
node_to_idx[cgraph->nodes[i]] = i;
}
for (auto & [fork_node, event] : stream_ctx.concurrent_events) {
// Find positions of all nodes from this event in the current graph
std::vector<int> positions;
positions.reserve(event.original_order.size());
bool all_found = true;
for (const ggml_tensor * orig_node : event.original_order) {
auto it = node_to_idx.find(orig_node);
if (it != node_to_idx.end()) {
positions.push_back(it->second);
} else {
all_found = false;
break;
}
}
if (!all_found || positions.size() != event.original_order.size()) {
continue;
}
// Sort positions to get contiguous range
std::vector<int> sorted_positions = positions;
std::sort(sorted_positions.begin(), sorted_positions.end());
bool is_contiguous = true;
for (size_t i = 1; i < sorted_positions.size(); ++i) {
if (sorted_positions[i] != sorted_positions[i-1] + 1) {
is_contiguous = false;
break;
}
}
if (!is_contiguous) {
continue;
}
// Restore original order at the sorted positions
int start_pos = sorted_positions[0];
for (size_t i = 0; i < event.original_order.size(); ++i) {
cgraph->nodes[start_pos + i] = const_cast<ggml_tensor *>(event.original_order[i]);
}
}
}
for (int i = 0; i < cgraph->n_nodes; i++) {
@ -3805,14 +3852,6 @@ static void ggml_backend_cuda_graph_optimize(ggml_backend_t backend, ggml_cgraph
// store {fork_idx, join_idx}
std::vector<std::pair<int, int>> concurrent_node_ranges;
// save the original nodes
std::vector<const ggml_tensor *> original_nodes;
original_nodes.reserve(cgraph->n_nodes);
for (int i = 0; i < cgraph->n_nodes; ++i) {
original_nodes.push_back(cgraph->nodes[i]);
}
cuda_ctx->stream_context().original_nodes = std::move(original_nodes);
for (const auto & [root_node, count] : fan_out) {
if (count >= min_fan_out && count <= max_fan_out) {
const int root_node_idx = node_indices[root_node];
@ -3917,6 +3956,13 @@ static void ggml_backend_cuda_graph_optimize(ggml_backend_t backend, ggml_cgraph
continue;
}
// Save the original order of nodes in this region before interleaving
// This is used later to restore grouping for fusion within streams
concurrent_event.original_order.reserve(total_branch_nodes);
for (int i = fork_node_idx + 1; i < join_node_idx; ++i) {
concurrent_event.original_order.push_back(cgraph->nodes[i]);
}
std::unordered_map<const ggml_tensor *, ggml_cuda_concurrent_event> & concurrent_events = cuda_ctx->stream_context().concurrent_events;
GGML_ASSERT(concurrent_events.find(root_node) == concurrent_events.end());
concurrent_events.emplace(root_node, std::move(concurrent_event));