From d4e5cb930dda1ff6196c438c942d500180305a3b Mon Sep 17 00:00:00 2001
From: Sergey Shlyapnikov <sergey.shlyapnikov@intel.com>
Date: Thu, 23 Jan 2025 18:01:29 +0400
Subject: [PATCH] [GPU] Fix synchronization in PagedAttention operation when
 KV-cache rotation is enabled but skipped for the current iteration

---
 src/plugins/intel_gpu/src/graph/impls/ocl/paged_attention.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/plugins/intel_gpu/src/graph/impls/ocl/paged_attention.cpp b/src/plugins/intel_gpu/src/graph/impls/ocl/paged_attention.cpp
index 1bcd4b0bb10fe2..0f285f57e9d18f 100644
--- a/src/plugins/intel_gpu/src/graph/impls/ocl/paged_attention.cpp
+++ b/src/plugins/intel_gpu/src/graph/impls/ocl/paged_attention.cpp
@@ -346,7 +346,7 @@ struct paged_attention_impl : multi_stage_primitive<paged_attention> {
 
         std::vector<event::ptr> res_events;
         std::vector<event::ptr> dep_events = events;
-        if (has_rotated_blocks) {
+        if (has_rotated_blocks && !_kernels_data[Stage::KV_CACHE_ROTATE].kernels[0].skip_execution) {
             execute_stage(dep_events, instance, res_events, Stage::KV_CACHE_ROTATE, is_mixed_mode);
             dep_events = res_events;
         }