We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
There was an error while loading. Please reload this page.
1 parent 868060f commit 4fe4522Copy full SHA for 4fe4522
ggml/src/ggml-vulkan/vulkan-shaders/ssm_scan.comp
@@ -111,12 +111,12 @@ void main() {
111
int warp_id = tid / WARP_SIZE;
112
int warp_offset = warp_id * WARP_SIZE;
113
114
- warp_sdata[warp_offset + lane] = y;
+ warp_sdata[tid] = y;
115
barrier();
116
117
[[unroll]] for (int offset = WARP_SIZE / 2; offset > 0; offset >>= 1) {
118
if (lane < offset) {
119
- warp_sdata[warp_offset + lane] += warp_sdata[warp_offset + lane + offset];
+ warp_sdata[tid] += warp_sdata[tid + offset];
120
}
121
122
0 commit comments