EngineX-Ascend/enginex-ascend-910-llama.cpp
10
0
Fork 0
Code Issues Pull Requests Actions 4 Projects Releases Wiki Activity

sycl: add usage of enqueue_functions extension (#14244)

Browse Source 
* Add header and namespace to use enqueue_functions extension

* Convert submit and parallel_for to use new extension in convert.cpp

* Convert submit and parallel_for to use extension in ggml-sycl.cpp

* Convert submit and parallel_for to use extension in gla.cpp

* Convert submit and parallel_for in mmq.cpp

* Convert submit and parallel_for in mmvq.cpp

* Convert submit and parallel_for in remaining files

* Convert all simple parallel_for to nd_launch from enqueue_functions
extension

* Wrapping extension in general function

Create a general function that enable the enqueue_functions extension if
it is enable in the compiler, otherwise call the general SYCL function
to launch kernels.

---------

Signed-off-by: nscipione <nicolo.scipione@codeplay.com>
This commit is contained in:
Nicolò Scipione
2025-06-20 15:07:21 +02:00
committed by GitHub
parent 6369be0735
commit 8308f98c7f
19 changed files with 750 additions and 986 deletions
Download Patch File Download Diff File Expand all files Collapse all files

166
ggml/src/ggml-sycl/cpy.cpp
View File

@@ -413,7 +413,8 @@ static void ggml_cpy_f16_f32_sycl(const char * cx, char * cdst, const int ne, co
{
dpct::has_capability_or_fail(stream->get_device(), { sycl::aspect::fp16 });
stream->parallel_for(
sycl_parallel_for(
stream,
sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks) * sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE),
sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE)),
[=](sycl::nd_item<3> item_ct1) {
@@ -431,7 +432,8 @@ static void ggml_cpy_f32_f32_sycl(const char * cx, char * cdst, const int ne, co
{
dpct::has_capability_or_fail(stream->get_device(), { sycl::aspect::fp16 });
stream->parallel_for(
sycl_parallel_for(
stream,
sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks) * sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE),
sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE)),
[=](sycl::nd_item<3> item_ct1) {
@@ -449,7 +451,8 @@ static void ggml_cpy_f32_f16_sycl(const char * cx, char * cdst, const int ne, co
{
dpct::has_capability_or_fail(stream->get_device(), { sycl::aspect::fp16 });
stream->parallel_for(
sycl_parallel_for(
stream,
sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks) * sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE),
sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE)),
[=](sycl::nd_item<3> item_ct1) {
@@ -465,11 +468,11 @@ static void ggml_cpy_f32_q8_0_sycl(const char * cx, char * cdst, const int ne, c
const int nb12, const int nb13, queue_ptr stream) {
GGML_ASSERT(ne % QK8_0 == 0);
const int num_blocks = ne / QK8_0;
stream->parallel_for(sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks), sycl::range<3>(1, 1, 1)),
[=](sycl::nd_item<3> item_ct1) {
cpy_f32_q<cpy_blck_f32_q8_0, QK8_0>(cx, cdst, ne, ne00, ne01, ne02, nb00, nb01, nb02, nb03,
ne10, ne11, ne12, nb10, nb11, nb12, nb13, item_ct1);
});
sycl_parallel_for(stream, sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks), sycl::range<3>(1, 1, 1)),
[=](sycl::nd_item<3> item_ct1) {
cpy_f32_q<cpy_blck_f32_q8_0, QK8_0>(cx, cdst, ne, ne00, ne01, ne02, nb00, nb01, nb02, nb03,
ne10, ne11, ne12, nb10, nb11, nb12, nb13, item_ct1);
});
}
static void ggml_cpy_q8_0_f32_sycl(const char * cx, char * cdst, const int ne, const int ne00, const int ne01,
@@ -477,11 +480,11 @@ static void ggml_cpy_q8_0_f32_sycl(const char * cx, char * cdst, const int ne, c
const int ne10, const int ne11, const int ne12, const int nb10, const int nb11,
const int nb12, const int nb13, queue_ptr stream) {
const int num_blocks = ne;
stream->parallel_for(sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks), sycl::range<3>(1, 1, 1)),
[=](sycl::nd_item<3> item_ct1) {
cpy_q_f32<cpy_blck_q8_0_f32, QK8_0>(cx, cdst, ne, ne00, ne01, ne02, nb00, nb01, nb02, nb03,
ne10, ne11, ne12, nb10, nb11, nb12, nb13, item_ct1);
});
sycl_parallel_for(stream, sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks), sycl::range<3>(1, 1, 1)),
[=](sycl::nd_item<3> item_ct1) {
cpy_q_f32<cpy_blck_q8_0_f32, QK8_0>(cx, cdst, ne, ne00, ne01, ne02, nb00, nb01, nb02, nb03,
ne10, ne11, ne12, nb10, nb11, nb12, nb13, item_ct1);
});
}
static void ggml_cpy_f32_q4_0_sycl(const char * cx, char * cdst, const int ne, const int ne00, const int ne01,
@@ -490,11 +493,11 @@ static void ggml_cpy_f32_q4_0_sycl(const char * cx, char * cdst, const int ne, c
const int nb12, const int nb13, queue_ptr stream) {
GGML_ASSERT(ne % QK4_0 == 0);
const int num_blocks = ne / QK4_0;
stream->parallel_for(sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks), sycl::range<3>(1, 1, 1)),
[=](sycl::nd_item<3> item_ct1) {
cpy_f32_q<cpy_blck_f32_q4_0, QK4_0>(cx, cdst, ne, ne00, ne01, ne02, nb00, nb01, nb02, nb03,
ne10, ne11, ne12, nb10, nb11, nb12, nb13, item_ct1);
});
sycl_parallel_for(stream, sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks), sycl::range<3>(1, 1, 1)),
[=](sycl::nd_item<3> item_ct1) {
cpy_f32_q<cpy_blck_f32_q4_0, QK4_0>(cx, cdst, ne, ne00, ne01, ne02, nb00, nb01, nb02, nb03,
ne10, ne11, ne12, nb10, nb11, nb12, nb13, item_ct1);
});
}
static void ggml_cpy_q4_0_f32_sycl(const char * cx, char * cdst, const int ne, const int ne00, const int ne01,
@@ -502,8 +505,9 @@ static void ggml_cpy_q4_0_f32_sycl(const char * cx, char * cdst, const int ne, c
const int ne10, const int ne11, const int ne12, const int nb10, const int nb11,
const int nb12, const int nb13, queue_ptr stream) {
const int num_blocks = ne;
stream->parallel_for(
sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks), sycl::range<3>(1, 1, 1)), [=](sycl::nd_item<3> item_ct1) {
sycl_parallel_for(
stream, sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks), sycl::range<3>(1, 1, 1)),
[=](sycl::nd_item<3> item_ct1) {
cpy_q_f32<cpy_blck_q_f32<dequantize_q4_0, QK4_0>, QK4_0>(cx, cdst, ne, ne00, ne01, ne02, nb00, nb01, nb02,
nb03, ne10, ne11, ne12, nb10, nb11, nb12, nb13,
item_ct1);
@@ -516,11 +520,11 @@ static void ggml_cpy_f32_q4_1_sycl(const char * cx, char * cdst, const int ne, c
const int nb12, const int nb13, queue_ptr stream) {
GGML_ASSERT(ne % QK4_1 == 0);
const int num_blocks = ne / QK4_1;
stream->parallel_for(sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks), sycl::range<3>(1, 1, 1)),
[=](sycl::nd_item<3> item_ct1) {
cpy_f32_q<cpy_blck_f32_q4_1, QK4_1>(cx, cdst, ne, ne00, ne01, ne02, nb00, nb01, nb02, nb03,
ne10, ne11, ne12, nb10, nb11, nb12, nb13, item_ct1);
});
sycl_parallel_for(stream, sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks), sycl::range<3>(1, 1, 1)),
[=](sycl::nd_item<3> item_ct1) {
cpy_f32_q<cpy_blck_f32_q4_1, QK4_1>(cx, cdst, ne, ne00, ne01, ne02, nb00, nb01, nb02, nb03,
ne10, ne11, ne12, nb10, nb11, nb12, nb13, item_ct1);
});
}
static void ggml_cpy_q4_1_f32_sycl(const char * cx, char * cdst, const int ne, const int ne00, const int ne01,
@@ -528,8 +532,9 @@ static void ggml_cpy_q4_1_f32_sycl(const char * cx, char * cdst, const int ne, c
const int ne10, const int ne11, const int ne12, const int nb10, const int nb11,
const int nb12, const int nb13, queue_ptr stream) {
const int num_blocks = ne;
stream->parallel_for(
sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks), sycl::range<3>(1, 1, 1)), [=](sycl::nd_item<3> item_ct1) {
sycl_parallel_for(
stream, sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks), sycl::range<3>(1, 1, 1)),
[=](sycl::nd_item<3> item_ct1) {
cpy_q_f32<cpy_blck_q_f32<dequantize_q4_1, QK4_1>, QK4_1>(cx, cdst, ne, ne00, ne01, ne02, nb00, nb01, nb02,
nb03, ne10, ne11, ne12, nb10, nb11, nb12, nb13,
item_ct1);
@@ -542,11 +547,11 @@ static void ggml_cpy_f32_q5_0_sycl(const char * cx, char * cdst, const int ne, c
const int nb12, const int nb13, queue_ptr stream) {
GGML_ASSERT(ne % QK5_0 == 0);
const int num_blocks = ne / QK5_0;
stream->parallel_for(sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks), sycl::range<3>(1, 1, 1)),
[=](sycl::nd_item<3> item_ct1) {
cpy_f32_q<cpy_blck_f32_q5_0, QK5_0>(cx, cdst, ne, ne00, ne01, ne02, nb00, nb01, nb02, nb03,
ne10, ne11, ne12, nb10, nb11, nb12, nb13, item_ct1);
});
sycl_parallel_for(stream, sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks), sycl::range<3>(1, 1, 1)),
[=](sycl::nd_item<3> item_ct1) {
cpy_f32_q<cpy_blck_f32_q5_0, QK5_0>(cx, cdst, ne, ne00, ne01, ne02, nb00, nb01, nb02, nb03,
ne10, ne11, ne12, nb10, nb11, nb12, nb13, item_ct1);
});
}
static void ggml_cpy_q5_0_f32_sycl(const char * cx, char * cdst, const int ne, const int ne00, const int ne01,
@@ -554,8 +559,9 @@ static void ggml_cpy_q5_0_f32_sycl(const char * cx, char * cdst, const int ne, c
const int ne10, const int ne11, const int ne12, const int nb10, const int nb11,
const int nb12, const int nb13, queue_ptr stream) {
const int num_blocks = ne;
stream->parallel_for(
sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks), sycl::range<3>(1, 1, 1)), [=](sycl::nd_item<3> item_ct1) {
sycl_parallel_for(
stream, sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks), sycl::range<3>(1, 1, 1)),
[=](sycl::nd_item<3> item_ct1) {
cpy_q_f32<cpy_blck_q_f32<dequantize_q5_0, QK5_0>, QK5_0>(cx, cdst, ne, ne00, ne01, ne02, nb00, nb01, nb02,
nb03, ne10, ne11, ne12, nb10, nb11, nb12, nb13,
item_ct1);
@@ -568,11 +574,11 @@ static void ggml_cpy_f32_q5_1_sycl(const char * cx, char * cdst, const int ne, c
const int nb12, const int nb13, queue_ptr stream) {
GGML_ASSERT(ne % QK5_1 == 0);
const int num_blocks = ne / QK5_1;
stream->parallel_for(sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks), sycl::range<3>(1, 1, 1)),
[=](sycl::nd_item<3> item_ct1) {
cpy_f32_q<cpy_blck_f32_q5_1, QK5_1>(cx, cdst, ne, ne00, ne01, ne02, nb00, nb01, nb02, nb03,
ne10, ne11, ne12, nb10, nb11, nb12, nb13, item_ct1);
});
sycl_parallel_for(stream, sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks), sycl::range<3>(1, 1, 1)),
[=](sycl::nd_item<3> item_ct1) {
cpy_f32_q<cpy_blck_f32_q5_1, QK5_1>(cx, cdst, ne, ne00, ne01, ne02, nb00, nb01, nb02, nb03,
ne10, ne11, ne12, nb10, nb11, nb12, nb13, item_ct1);
});
}
static void ggml_cpy_q5_1_f32_sycl(const char * cx, char * cdst, const int ne, const int ne00, const int ne01,
@@ -580,8 +586,9 @@ static void ggml_cpy_q5_1_f32_sycl(const char * cx, char * cdst, const int ne, c
const int ne10, const int ne11, const int ne12, const int nb10, const int nb11,
const int nb12, const int nb13, queue_ptr stream) {
const int num_blocks = ne;
stream->parallel_for(
sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks), sycl::range<3>(1, 1, 1)), [=](sycl::nd_item<3> item_ct1) {
sycl_parallel_for(
stream, sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks), sycl::range<3>(1, 1, 1)),
[=](sycl::nd_item<3> item_ct1) {
cpy_q_f32<cpy_blck_q_f32<dequantize_q5_1, QK5_1>, QK5_1>(cx, cdst, ne, ne00, ne01, ne02, nb00, nb01, nb02,
nb03, ne10, ne11, ne12, nb10, nb11, nb12, nb13,
item_ct1);
@@ -594,11 +601,11 @@ static void ggml_cpy_f32_iq4_nl_sycl(const char * cx, char * cdst, const int ne,
const int nb12, const int nb13, queue_ptr stream) {
GGML_ASSERT(ne % QK4_NL == 0);
const int num_blocks = ne / QK4_NL;
stream->parallel_for(
sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks), sycl::range<3>(1, 1, 1)), [=](sycl::nd_item<3> item_ct1) {
cpy_f32_q<cpy_blck_f32_iq4_nl, QK4_NL>(cx, cdst, ne, ne00, ne01, ne02, nb00, nb01, nb02, nb03, ne10, ne11,
ne12, nb10, nb11, nb12, nb13, item_ct1);
});
sycl_parallel_for(stream, sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks), sycl::range<3>(1, 1, 1)),
[=](sycl::nd_item<3> item_ct1) {
cpy_f32_q<cpy_blck_f32_iq4_nl, QK4_NL>(cx, cdst, ne, ne00, ne01, ne02, nb00, nb01, nb02, nb03,
ne10, ne11, ne12, nb10, nb11, nb12, nb13, item_ct1);
});
}
static void ggml_cpy_f16_f16_sycl(const char * cx, char * cdst, const int ne, const int ne00, const int ne01,
@@ -609,7 +616,8 @@ static void ggml_cpy_f16_f16_sycl(const char * cx, char * cdst, const int ne, co
{
dpct::has_capability_or_fail(stream->get_device(), { sycl::aspect::fp16 });
stream->parallel_for(
sycl_parallel_for(
stream,
sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks) * sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE),
sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE)),
[=](sycl::nd_item<3> item_ct1) {
@@ -628,7 +636,8 @@ static void ggml_cpy_i16_i16_sycl(const char * cx, char * cdst, const int ne, co
// dpct::has_capability_or_fail(stream->get_device(),
// {sycl::aspect::fp16});
stream->parallel_for(
sycl_parallel_for(
stream,
sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks) * sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE),
sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE)),
[=](sycl::nd_item<3> item_ct1) {
@@ -647,7 +656,8 @@ static void ggml_cpy_i32_i32_sycl(const char * cx, char * cdst, const int ne, co
// dpct::has_capability_or_fail(stream->get_device(),
// {sycl::aspect::fp16});
stream->parallel_for(
sycl_parallel_for(
stream,
sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks) * sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE),
sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE)),
[=](sycl::nd_item<3> item_ct1) {
@@ -662,11 +672,13 @@ static void ggml_cpy_q8_0_q8_0(const char * cx, char * cdst, const int ne, const
const int ne10, const int ne11, const int ne12, const int nb10, const int nb11,
const int nb12, const int nb13, queue_ptr stream) {
const int num_blocks = ceil_div(ne, SYCL_CPY_BLOCK_SIZE);
stream->parallel_for(
sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks) * sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE),
sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE)), [=](sycl::nd_item<3> item_ct1) {
cpy_q_q<block_q8_0, QK8_0>(cx, cdst, ne, ne00, ne01, ne02, nb00, nb01, nb02, nb03, ne10, ne11, ne12, nb10, nb11, nb12, nb13, item_ct1);
});
sycl_parallel_for(stream,
sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks) * sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE),
sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE)),
[=](sycl::nd_item<3> item_ct1) {
cpy_q_q<block_q8_0, QK8_0>(cx, cdst, ne, ne00, ne01, ne02, nb00, nb01, nb02, nb03, ne10, ne11,
ne12, nb10, nb11, nb12, nb13, item_ct1);
});
}
@@ -675,11 +687,13 @@ static void ggml_cpy_q5_0_q5_0(const char * cx, char * cdst, const int ne, const
const int ne10, const int ne11, const int ne12, const int nb10, const int nb11,
const int nb12, const int nb13, queue_ptr stream) {
const int num_blocks = ceil_div(ne, SYCL_CPY_BLOCK_SIZE);
stream->parallel_for(
sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks) * sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE),
sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE)), [=](sycl::nd_item<3> item_ct1) {
cpy_q_q<block_q5_0, QK5_0>(cx, cdst, ne, ne00, ne01, ne02, nb00, nb01, nb02, nb03, ne10, ne11, ne12, nb10, nb11, nb12, nb13, item_ct1);
});
sycl_parallel_for(stream,
sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks) * sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE),
sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE)),
[=](sycl::nd_item<3> item_ct1) {
cpy_q_q<block_q5_0, QK5_0>(cx, cdst, ne, ne00, ne01, ne02, nb00, nb01, nb02, nb03, ne10, ne11,
ne12, nb10, nb11, nb12, nb13, item_ct1);
});
}
@@ -689,11 +703,13 @@ static void ggml_cpy_q5_1_q5_1(const char * cx, char * cdst, const int ne, const
const int nb12, const int nb13, queue_ptr stream) {
const int num_blocks = ceil_div(ne, SYCL_CPY_BLOCK_SIZE);
stream->parallel_for(
sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks) * sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE),
sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE)), [=](sycl::nd_item<3> item_ct1) {
cpy_q_q<block_q5_1, QK5_1>(cx, cdst, ne, ne00, ne01, ne02, nb00, nb01, nb02, nb03, ne10, ne11, ne12, nb10, nb11, nb12, nb13, item_ct1);
});
sycl_parallel_for(stream,
sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks) * sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE),
sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE)),
[=](sycl::nd_item<3> item_ct1) {
cpy_q_q<block_q5_1, QK5_1>(cx, cdst, ne, ne00, ne01, ne02, nb00, nb01, nb02, nb03, ne10, ne11,
ne12, nb10, nb11, nb12, nb13, item_ct1);
});
}
@@ -702,10 +718,13 @@ static void ggml_cpy_q4_0_q4_0(const char * cx, char * cdst, const int ne, const
const int ne10, const int ne11, const int ne12, const int nb10, const int nb11,
const int nb12, const int nb13, queue_ptr stream) {
const int num_blocks = ceil_div(ne, SYCL_CPY_BLOCK_SIZE);
stream->parallel_for(
sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks) * sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE), sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE)), [=](sycl::nd_item<3> item_ct1) {
cpy_q_q<block_q4_0, QK4_0>(cx, cdst, ne, ne00, ne01, ne02, nb00, nb01, nb02, nb03, ne10, ne11, ne12, nb10, nb11, nb12, nb13, item_ct1);
});
sycl_parallel_for(stream,
sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks) * sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE),
sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE)),
[=](sycl::nd_item<3> item_ct1) {
cpy_q_q<block_q4_0, QK4_0>(cx, cdst, ne, ne00, ne01, ne02, nb00, nb01, nb02, nb03, ne10, ne11,
ne12, nb10, nb11, nb12, nb13, item_ct1);
});
}
@@ -715,10 +734,13 @@ static void ggml_cpy_q4_1_q4_1(const char * cx, char * cdst, const int ne, const
const int nb12, const int nb13, queue_ptr stream) {
const int num_blocks = ceil_div(ne, SYCL_CPY_BLOCK_SIZE);
stream->parallel_for(
sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks) * sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE), sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE)), [=](sycl::nd_item<3> item_ct1) {
cpy_q_q<block_q4_1, QK4_1>(cx, cdst, ne, ne00, ne01, ne02, nb00, nb01, nb02, nb03, ne10, ne11, ne12, nb10, nb11, nb12, nb13, item_ct1);
});
sycl_parallel_for(stream,
sycl::nd_range<3>(sycl::range<3>(1, 1, num_blocks) * sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE),
sycl::range<3>(1, 1, SYCL_CPY_BLOCK_SIZE)),
[=](sycl::nd_item<3> item_ct1) {
cpy_q_q<block_q4_1, QK4_1>(cx, cdst, ne, ne00, ne01, ne02, nb00, nb01, nb02, nb03, ne10, ne11,
ne12, nb10, nb11, nb12, nb13, item_ct1);
});
}
void ggml_sycl_cpy(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1) try {