CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-11-14 11:07:10 +00:00
Code Issues Packages Projects Releases Wiki Activity

SYCL: Remove misleading ggml_sycl_op_flatten function (#12387)

Browse Source 
* SYCL: Remove misleading ggml_sycl_op_flatten function

* remove trailing whitespace

* Fix L2 norm from rebase

* remove try catch block from element_wise.cpp

* remove comment from common.hp

* ggml-sycl.cpp: Add try catch sycl::exception block in compute_forward

* norm.cpp: remove try catch exception block
This commit is contained in:
Akarshan Biswas
2025-03-31 14:55:24 +05:30
committed by GitHub
parent f52d59d771
commit 6c02a032fa
12 changed files with 369 additions and 586 deletions
Download Patch File Download Diff File Expand all files Collapse all files

35
ggml/src/ggml-sycl/common.cpp
View File

@@ -66,41 +66,6 @@ int64_t downsample_sycl_global_range(int64_t accumulate_block_num, int64_t block
return sycl_down_blk_size;
}
void ggml_sycl_op_flatten(ggml_backend_sycl_context & ctx, const ggml_tensor *src0,
const ggml_tensor *src1, ggml_tensor *dst,
const ggml_sycl_op_flatten_t op) try {
const bool use_src1 = src1 != nullptr;
if(use_src1)
GGML_ASSERT(strcmp(src1->buffer->buft->iface.get_name(src1->buffer->buft), GGML_SYCL_NAME "_Split") != 0);
GGML_ASSERT(strcmp(dst->buffer->buft->iface.get_name(dst->buffer->buft), GGML_SYCL_NAME "_Split") != 0);
// dd = data device
float * src0_ddf = (float *) src0->data;
float * src1_ddf = use_src1 ? (float *) src1->data : nullptr;
float * dst_ddf = (float *) dst->data;
ggml_sycl_pool_alloc<float> src0_f(ctx.pool());
ggml_sycl_pool_alloc<float> src1_f(ctx.pool());
ggml_sycl_pool_alloc<float> dst_f(ctx.pool());
ggml_sycl_set_device(ctx.device);
queue_ptr main_stream = ctx.stream();
// GGML_SYCL_DEBUG("ctx.device=%d, main_stream=%p src0_on_device=%d, src1_on_device=%d, dst_on_device=%d\n",
// ctx.device, main_stream, src0_on_device, src1_on_device, dst_on_device);
// do the computation
op(ctx, src0, src1, dst, src0_ddf, src1_ddf, dst_ddf, main_stream);
// print_ggml_tensor("tensor", dst);
}
catch (sycl::exception const &exc) {
std::cerr << exc.what() << "Exception caught at file:" << __FILE__
<< ", line:" << __LINE__ << std::endl;
std::exit(1);
}
void release_extra_gpu(ggml_tensor_extra_gpu * extra, std::vector<queue_ptr> streams) {
for (int i = 0; i < ggml_sycl_info().device_count; ++i) {
for (int64_t is = 0; is < GGML_SYCL_MAX_STREAMS; ++is) {