ggml-cuda: missing __syncthreads() in ssm-scan.cu
Author: FrankZhuQSCreated May 19, 2026Updated Sep 15, 2026
Commit 43e6301 The above commit remove the second __syncthreads() in ssm_scan_f32 on line 81. It seems that this sync is a must or some threads may read data from smemB/smemC of next loop.
Source: ggml-org/ggml