LC to overlap with following kernels

ORippler · ORippler · commit b508845ca5b9 · 2026-06-03T18:19:46.000+02:00
diff --git a/ggml/src/ggml-cuda/mmvq.cu b/ggml/src/ggml-cuda/mmvq.cu
@@ -731,6 +731,8 @@ static __global__ void mul_mat_vec_q_moe(
         }
     }
 
+    ggml_cuda_pdl_lc();
+
     // Warp-level reduction only - no shared memory needed
 #pragma unroll
     for (int i = 0; i < c_rows_per_block; ++i) {

Original file line number	Diff line number	Diff line change
`@@ -731,6 +731,8 @@ static __global__ void mul_mat_vec_q_moe(`
`731`	`731`	`}`
`732`	`732`	`}`
`733`	`733`
	`734`	`+ ggml_cuda_pdl_lc();`
	`735`	`+`
`734`	`736`	`// Warp-level reduction only - no shared memory needed`
`735`	`737`	`#pragma unroll`
`736`	`738`	`for (int i = 0; i < c_rows_per_block; ++i) {`