tvm_tirx.tirx.op.gpu_warp_shuffle_up

tvm_tirx.tirx.op.gpu_warp_shuffle_up#

tvm_tirx.tirx.op.gpu_warp_shuffle_up(mask, value, offset, width, warp_size, *, ty=None, loc=UNKNOWN_LOC)#

Copy value from a lane with lower (by offset) index relative to caller.

Parameters:

  • mask (Expr) – The warp mask indicates active threads inside warp.

  • value (Expr) – The value to exchange.

  • offset (Expr) – The difference between source lane index and destination lane index: offset = dst_lane_idx - src_lane_idx

  • width (Expr) – The width of sub-sections to perform warp shuffle.

  • warp_size (Expr) – The warp size.

  • loc (Location)

Returns:

call (Expr) – The call expression.