A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 14, 05:30 UTC
[ROCm][Fix] Fix rocm build caused by #23791 (#23847)
Signed-off-by: charlifu <charlifu@amd.com>
Charlie Fu committed
Aug 28, 2025 at 21:52 UTC
006477e60b49babfca96352c7c648f10fff4a053
1 file changed
-1
csrc/cache_kernels.cu
-1
@@ -913,7 +913,6 @@ __global__ void cp_gather_cache(
913
const int32_t split_end = min((split + 1) * split_slots, tot_slots);
914
915
const bool is_active_split = (split_start < tot_slots);
916
- const bool is_last_split = (split_end == tot_slots);
916
917
if (!is_active_split) return;
918