Skip to content

CANN: fix acl_rstd allocation size in ggml_cann_rms_norm #15760

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
merged 1 commit into from
Sep 4, 2025
Merged

CANN: fix acl_rstd allocation size in ggml_cann_rms_norm #15760

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
11 changes: 6 additions & 5 deletions ggml/src/ggml-cann/aclnn_ops.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -975,18 +975,19 @@ void ggml_cann_rms_norm(ggml_backend_cann_context& ctx, ggml_tensor* dst) {
);

// build rstd, zero...
size_t acl_rstd_nb[GGML_MAX_DIMS];
int64_t acl_rstd_ne[] = {src->ne[1], src->ne[2], src->ne[3]};
size_t acl_rstd_nb[GGML_MAX_DIMS - 1];
acl_rstd_nb[0] = sizeof(float);
for (int i = 1; i < GGML_MAX_DIMS; i++) {
acl_rstd_nb[i] = acl_rstd_nb[i - 1] * src->ne[i - 1];
for (int i = 1; i < GGML_MAX_DIMS - 1; i++) {
acl_rstd_nb[i] = acl_rstd_nb[i - 1] * acl_rstd_ne[i - 1];
}
aclTensor* acl_rstd = get_f32_cache_acl_tensor(
ctx,
&ctx.rms_norm_zero_tensor_cache.cache,
ctx.rms_norm_zero_tensor_cache.size,
src->ne,
acl_rstd_ne,
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

You can jsut use src as rstd_ne, acl_rstd_ne array is not necessary.

Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

That approach won’t work. If we simply reuse src->ne, then for a 3D tensor the resulting dimensions would be {ne0, ne1, ne2}, whereas what we actually need are {ne1, ne2, ne3}.

acl_rstd_nb,
GGML_MAX_DIMS,
GGML_MAX_DIMS - 1,
0.0f // value
);

Expand Down