Residual addition for video software techniques

摘要

According to some embodiments, a technique provides for the execution of an instruction that includes receiving residual data of a first image and decoded pixels of a second image, zero-extending a plurality of unsigned data operands of the decoded pixels producing a plurality of unpacked data operands, adding a plurality of signed data operands of the residual data to the plurality of unpacked data operands producing a plurality of signed results; and saturating the plurality of signed results producing a plurality of unsigned results.

主权项

1. A processor comprising:
a plurality of registers; and a decode and control unit to receive an instruction, the instruction to indicate a first source register of the plurality of registers and a second source register of the plurality of registers, the first source register to store a first plurality of signed operands of a first precision, the second source register to store a second plurality of unsigned operands of a second precision, the first precision to be greater than the second precision, the first plurality to be less than the second plurality; and a functional unit to perform the instruction and to store a plurality of result operands in only one of an upper half and a lower half of a destination register of the plurality of registers, the plurality of result operands to be unsigned operands of the second precision, wherein the plurality of results operands are each to include a sum of a signed operand of the first plurality of signed operands with a corresponding unsigned operand of the second plurality of unsigned operands, the corresponding unsigned operands from only one of an upper half and a lower half of the second source register, wherein the functional unit is not to store sums in another half of the destination register besides said one of the upper and the lower halves used to store the plurality of result operands.