Skip to content

Fix meta backend tensor reads for split tensors during state serialization #22063

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
JohannesGaessler merged 1 commit into from
Apr 18, 2026
Merged

Fix meta backend tensor reads for split tensors during state serialization #22063

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
40 changes: 39 additions & 1 deletion ggml/src/ggml-backend-meta.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1270,7 +1270,45 @@ static void ggml_backend_meta_buffer_get_tensor(ggml_backend_buffer_t buffer, co
GGML_ASSERT(ggml_is_contiguous(tensor));

const ggml_backend_meta_split_state split_state = ggml_backend_meta_get_split_state(tensor, /*assume_sync =*/ false);
GGML_ASSERT(split_state.n_segments == 1);

if (split_state.n_segments != 1) {
GGML_ASSERT(split_state.axis >= 0 && split_state.axis < GGML_MAX_DIMS);
GGML_ASSERT(offset == 0);
GGML_ASSERT(size == ggml_nbytes(tensor));
GGML_ASSERT(tensor->ne[3] == 1);
size_t offset_data = 0;
std::vector<size_t> simple_offsets(n_bufs, 0);
if (split_state.axis == GGML_BACKEND_SPLIT_AXIS_0) {
GGML_ASSERT(tensor->ne[2] == 1);
const int64_t blck_size = ggml_blck_size(tensor->type);
for (size_t s = 0; s < split_state.n_segments; s++) {
for (size_t j = 0; j < n_bufs; j++) {
const ggml_tensor * simple_tensor = ggml_backend_meta_buffer_simple_tensor(tensor, j);
GGML_ASSERT(split_state.ne[s*n_bufs + j] % blck_size == 0);
const size_t nbytes = split_state.ne[s*n_bufs + j]/blck_size * tensor->nb[0];
ggml_backend_tensor_get_2d(simple_tensor, (char *) data + offset_data, simple_offsets[j], nbytes,
tensor->ne[1], simple_tensor->nb[1], tensor->nb[1]);
offset_data += nbytes;
simple_offsets[j] += nbytes;
}
}
GGML_ASSERT(offset_data*tensor->ne[1] == size);
return;
}
GGML_ASSERT(split_state.axis == GGML_BACKEND_SPLIT_AXIS_1);
for (size_t s = 0; s < split_state.n_segments; s++) {
for (size_t j = 0; j < n_bufs; j++) {
const ggml_tensor * simple_tensor = ggml_backend_meta_buffer_simple_tensor(tensor, j);
const size_t nbytes = split_state.ne[s*n_bufs + j] * tensor->nb[1];
ggml_backend_tensor_get_2d(simple_tensor, (char *) data + offset_data, simple_offsets[j], nbytes,
tensor->ne[2], simple_tensor->nb[2], tensor->nb[2]);
offset_data += nbytes;
simple_offsets[j] += nbytes;
}
}
GGML_ASSERT(offset_data*tensor->ne[2] == size);
return;
}

switch (split_state.axis) {
case GGML_BACKEND_SPLIT_AXIS_0:
Expand Down
Loading