Skip to content

ROCm fix: replace host-side rsqrtf with 1/sqrtf for HIP compatibility #117

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

Open
reffdev wants to merge 1 commit into
base: rocm
Choose a base branch
from
+2 −2
Open

ROCm fix: replace host-side rsqrtf with 1/sqrtf for HIP compatibility #117

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
4 changes: 2 additions & 2 deletions ds4_cuda.cu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6168,7 +6168,7 @@ extern "C" int ds4_gpu_attention_prefill_raw_heads_tensor(ds4_gpu_tensor *heads,
if (!tmp) return 0;
float *scores = tmp;
float *out_tmp = (float *)((char *)tmp + out_offset);
const float alpha = rsqrtf((float)head_dim);
const float alpha = 1.0f/sqrtf((float)head_dim);
const float beta = 0.0f;
cublasStatus_t st = cublasSgemmStridedBatched(g_cublas,
CUBLAS_OP_T,
Expand Down Expand Up @@ -6538,7 +6538,7 @@ static int attention_prefill_mixed_launch(
n_comp,
head_dim);
if (!cuda_ok(cudaGetLastError(), "attention mixed kv pack launch")) return 0;
const float alpha = rsqrtf((float)head_dim);
const float alpha = 1.0f/sqrtf((float)head_dim);
const float beta = 0.0f;
cublasStatus_t st = cublasSgemmStridedBatched(g_cublas,
CUBLAS_OP_T,
Expand Down