Skip to content

ggml-cpu: restore the src1 guard on the F16 Hadamard matmul path #147

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
khosravipasha merged 2 commits into from
Sep 4, 2026
Merged

ggml-cpu: restore the src1 guard on the F16 Hadamard matmul path #147

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
dedd235
ggml-cpu: restore the src1 guard on the F16 Hadamard matmul path
bri-prism Sep 3, 2026
7be8db7
ggml-cpu: keep the F16 src1 invariant comment on one line
bri-prism Sep 4, 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
11 changes: 5 additions & 6 deletions ggml/src/ggml-cpu/ggml-cpu.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,6 +5,7 @@
#include "ggml-backend.h"
#include "traits.h"
#include "ggml-cpu-impl.h"
#include "ggml-cpu.h"
#include "ggml-impl.h"
#include "quants.h"
#include "ggml-threading.h"
Expand Down Expand Up @@ -1340,7 +1341,9 @@ UseGgmlGemm1:;
const size_t nbw3 = nbw2*ne12;

assert(params->wsize >= ne13*nbw3);
GGML_ASSERT(src1->type == GGML_TYPE_F32 || src1->type == GGML_TYPE_F16);
// F16 src1 converts straight to float, so wdata rows must be floats; a quantized vec_dot_type would overrun the buffer
GGML_ASSERT(src1->type == GGML_TYPE_F32 ||
(src1->type == GGML_TYPE_F16 && vec_dot_type == GGML_TYPE_F32));

#if 0
for (int64_t i13 = 0; i13 < ne13; ++i13) {
Expand All @@ -1366,11 +1369,7 @@ UseGgmlGemm1:;
if (src1->type == GGML_TYPE_F32) {
from_float((const float *) src1_block, dst_block, n_block);
} else {
const ggml_fp16_t * src_f16 = (const ggml_fp16_t *) src1_block;
float * dst_f32 = (float *) dst_block;
for (int64_t i = 0; i < n_block; ++i) {
dst_f32[i] = GGML_CPU_FP16_TO_FP32(src_f16[i]);
}
ggml_cpu_fp16_to_fp32((const ggml_fp16_t *) src1_block, (float *) dst_block, n_block);
}
}
}
Expand Down