Skip to content

ggml: fix backend split scheduler race condition #26040

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
0cc4m merged 2 commits into from
Aug 20, 2026
Merged

ggml: fix backend split scheduler race condition #26040

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
16a2011
ggml: fix backend split scheduler race condition
0cc4m Jul 23, 2026
0f1f40e
only sync when split has no inputs
0cc4m Aug 17, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
22 changes: 17 additions & 5 deletions ggml/src/ggml-backend.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1546,11 +1546,23 @@ static enum ggml_status ggml_backend_sched_compute_splits(ggml_backend_sched_t s
std::vector<int32_t> ids;
std::vector<ggml_bitset_t> used_ids;

int prev_backend_id = -1;

for (int split_id = 0; split_id < sched->n_splits; split_id++) {
struct ggml_backend_sched_split * split = &splits[split_id];
int split_backend_id = split->backend_id;
ggml_backend_t split_backend = sched->backends[split_backend_id];

// ensure the previous split's async work has completed before we start
// this split, the allocator may have reused buffer regions across splits
if (split->n_inputs == 0 && prev_backend_id >= 0 && prev_backend_id != split_backend_id) {
if (sched->events[prev_backend_id][sched->cur_copy] != NULL) {
ggml_backend_event_synchronize(sched->events[prev_backend_id][sched->cur_copy]);
} else {
ggml_backend_synchronize(sched->backends[prev_backend_id]);
Comment thread
0cc4m marked this conversation as resolved.
}
}

// copy the input tensors to the split backend
for (int input_id = 0; input_id < split->n_inputs; input_id++) {
ggml_backend_t input_backend = ggml_backend_sched_get_tensor_backend(sched, split->inputs[input_id]);
Expand Down Expand Up @@ -1713,12 +1725,12 @@ static enum ggml_status ggml_backend_sched_compute_splits(ggml_backend_sched_t s
}
}

// record the event of this copy
if (split->n_inputs > 0) {
if (sched->events[split_backend_id][sched->cur_copy] != NULL) {
ggml_backend_event_record(sched->events[split_backend_id][sched->cur_copy], split_backend);
}
// record the event of this split
if (sched->events[split_backend_id][sched->cur_copy] != NULL) {
ggml_backend_event_record(sched->events[split_backend_id][sched->cur_copy], split_backend);
}

prev_backend_id = split_backend_id;
}

return GGML_STATUS_SUCCESS;
Expand Down
Loading