Lumpiasty/llama.cpp
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

metal : extend l2_norm support for non-cont src0 (#19502)

Browse Source
This commit is contained in:
Georgi Gerganov
2026-02-11 14:53:19 +02:00
committed by GitHub
parent ada90bf2ba
commit 9ab072ebbe
5 changed files with 75 additions and 30 deletions
Download Patch File Download Diff File Expand all files Collapse all files
ggml/src/ggml-metal/ggml-metal.metal
+20 -10
View File
@@ -2706,26 +2706,32 @@ template [[host_name("kernel_rms_norm_f32_4")]] kernel kernel_rms_norm_f
template [[host_name("kernel_rms_norm_mul_f32_4")]] kernel kernel_rms_norm_fuse_t kernel_rms_norm_fuse_impl<float4, 2>;
template [[host_name("kernel_rms_norm_mul_add_f32_4")]] kernel kernel_rms_norm_fuse_t kernel_rms_norm_fuse_impl<float4, 3>;
kernel void kernel_l2_norm_f32(
template <typename T0, typename T>
kernel void kernel_l2_norm_impl(
constant ggml_metal_kargs_l2_norm & args,
device const char * src0,
device char * dst,
threadgroup float * shmem_f32 [[threadgroup(0)]],
uint tgpig[[threadgroup_position_in_grid]],
ushort tpitg[[thread_position_in_threadgroup]],
ushort sgitg[[simdgroup_index_in_threadgroup]],
ushort tiisg[[thread_index_in_simdgroup]],
ushort ntg[[threads_per_threadgroup]]) {
uint3 tgpig[[threadgroup_position_in_grid]],
ushort3 tpitg[[thread_position_in_threadgroup]],
ushort sgitg[[simdgroup_index_in_threadgroup]],
ushort tiisg[[thread_index_in_simdgroup]],
ushort3 ntg[[threads_per_threadgroup]]) {
const int i03 = tgpig.z;
const int i02 = tgpig.y;
const int i01 = tgpig.x;
if (sgitg == 0) {
shmem_f32[tiisg] = 0.0f;
}
device const float4 * x = (device const float4 *) (src0 + tgpig*args.nb01);
device const T0 * x = (device const T0 *) (src0 + i03*args.nb03 + i02*args.nb02 + i01*args.nb01);
device T * y = (device T *) (dst + i03*args.nb3 + i02*args.nb2 + i01*args.nb1);
float sumf = 0.0f;
// parallel sum
for (int i00 = tpitg; i00 < args.ne00_4; i00 += ntg) {
for (int i00 = tpitg.x; i00 < args.ne00; i00 += ntg.x) {
sumf += dot(x[i00], x[i00]);
}
sumf = simd_sum(sumf);
@@ -2743,12 +2749,16 @@ kernel void kernel_l2_norm_f32(
const float scale = 1.0f/sqrt(max(sumf, args.eps));
device float4 * y = (device float4 *) dst + tgpig*args.ne00_4;
for (int i00 = tpitg; i00 < args.ne00_4; i00 += ntg) {
for (int i00 = tpitg.x; i00 < args.ne00; i00 += ntg.x) {
y[i00] = x[i00] * scale;
}
}
typedef decltype(kernel_l2_norm_impl<float, float>) kernel_l2_norm_t;
template [[host_name("kernel_l2_norm_f32_f32")]] kernel kernel_l2_norm_t kernel_l2_norm_impl<float, float>;
template [[host_name("kernel_l2_norm_f32_f32_4")]] kernel kernel_l2_norm_t kernel_l2_norm_impl<float4, float4>;
kernel void kernel_group_norm_f32(
constant ggml_metal_kargs_group_norm & args,
device const float * src0,