Skip to content

ggml : fix unaligned access in AMX code #16315

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
Oct 6, 2025
Merged

ggml : fix unaligned access in AMX code #16315

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
1 change: 1 addition & 0 deletions ggml/src/ggml-cpu/amx/amx.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -149,6 +149,7 @@ class extra_buffer_type : ggml::cpu::extra_buffer_type {
if (op->op == GGML_OP_MUL_MAT && is_contiguous_2d(op->src[0]) && // src0 must be contiguous
is_contiguous_2d(op->src[1]) && // src1 must be contiguous
op->src[0]->buffer && op->src[0]->buffer->buft == ggml_backend_amx_buffer_type() &&
op->src[0]->ne[0] % (TILE_K * 2 * 32) == 0 && // TODO: not sure if correct (https://github.com/ggml-org/llama.cpp/pull/16315)
Copy link
Member Author

Choose a reason for hiding this comment

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

Adding this fixes the CI workflows, though I'm not sure about the reasoning. It effectively requires the row size to be multiple of 2048.

op->ne[0] % (TILE_N * 2) == 0 && // out_features is 32x
(qtype_has_amx_kernels(op->src[0]->type) || (op->src[0]->type == GGML_TYPE_F16))) {
// src1 must be host buffer
Expand Down
Loading