Mojo function
to_consecutive_sgprs
to_consecutive_sgprs(vgpr_one: UInt32, vgpr_two: UInt32) -> SIMD[DType.uint32, 2]
Moves data from two VGPRS to two consecutive SGPRS, e.g. s[0:1], s[2:3]. This move is performed by the first thread in a warp, and the registers are shared by all threads in the warp.
Args:
- vgpr_one (
UInt32): The register to move to a SGPR n. - vgpr_two (
UInt32): The register to move to a SGPR n+1.
Returns:
Was this page helpful?
Thank you! We'll create more content like this.
Thank you for helping us improve!