192   PetscInt             gid, row, start_slice, cid;  in matmult_seqsell_tiled_kernel8()  local
201     cid = blockIdx.x * chunksperblock + iter; /* chunk id */  in matmult_seqsell_tiled_kernel8()
202     if (cid < totalchunks) {  in matmult_seqsell_tiled_kernel8()
203       start_slice = chunk_slice_map[cid]; /* starting slice at each iteration */  in matmult_seqsell_tiled_kernel8()
204       gid         = cid * BLOCKY * WARP_SIZE + threadIdx.y * WARP_SIZE + threadIdx.x;  in matmult_seqsell_tiled_kernel8()
205 …if ((cid + 1) * BLOCKY * WARP_SIZE > sliidx[start_slice + 1]) { /* this iteration covers more than…  in matmult_seqsell_tiled_kernel8()
221 …if (iter == chunksperblock - 1 || (cid + 2) * BLOCKY * WARP_SIZE > sliidx[start_slice + 1]) { /* l…  in matmult_seqsell_tiled_kernel8()
247   PetscInt             gid, row, start_slice, cid;  in matmultadd_seqsell_tiled_kernel8()  local
256     cid = blockIdx.x * chunksperblock + iter; /* chunk id */  in matmultadd_seqsell_tiled_kernel8()
257     if (cid < totalchunks) {  in matmultadd_seqsell_tiled_kernel8()
258       start_slice = chunk_slice_map[cid]; /* starting slice at each iteration */  in matmultadd_seqsell_tiled_kernel8()
259       gid         = cid * BLOCKY * WARP_SIZE + threadIdx.y * WARP_SIZE + threadIdx.x;  in matmultadd_seqsell_tiled_kernel8()
260 …if ((cid + 1) * BLOCKY * WARP_SIZE > sliidx[start_slice + 1]) { /* this iteration covers more than…  in matmultadd_seqsell_tiled_kernel8()
276 …if (iter == chunksperblock - 1 || (cid + 2) * BLOCKY * WARP_SIZE > sliidx[start_slice + 1]) { /* l…  in matmultadd_seqsell_tiled_kernel8()