12.55 __usada8 intrinsic

This intrinsic inserts a USADA8 instruction into the instruction stream generated by the compiler.

It enables you to perform four unsigned 8-bit subtractions, and add the absolute values of the differences to a 32-bit accumulate operand.

Syntax

unsigned int __usada8(unsigned int val1, unsigned int val2, unsigned int val3)

Where:

val1

holds the first four 8-bit operands for the subtractions

val2

holds the second four 8-bit operands for the subtractions

val3

holds the accumulation value.

Return value

The __usada8 intrinsic returns the sum of the absolute differences of the following bytes, added to the accumulation value:

  • The subtraction of the first byte in the second operand from the first byte in the first operand.

  • The subtraction of the second byte in the second operand from the second byte in the first operand.

  • The subtraction of the third byte in the second operand from the third byte in the first operand.

  • The subtraction of the fourth byte in the second operand from the fourth byte in the first operand.

Examples

unsigned int subtract_add_diff_accumulate(unsigned int val1, unsigned int val2, unsigned int val3)
{
    unsigned int res;
    res = __usada8(val1,val2,val3); /* absdiff1 = val1[7:0] - val2[7:0]
                                       absdiff2 = val1[15:8] - val2[15:8]
                                       absdiff3 = val1[23:16] - val2[23:16]
                                       absdiff4 = val1[31:24] - val2[31:24]
                                       sum = absdiff1 + absdiff2 + absdiff3
                                        + absdiff4
                                       res[31:0] = sum[31:0] + val3[31:0]
                              */
    return res;
}
Related reference
10.147 ARMv6 SIMD intrinsics
Related information
USADA8
ARM and Thumb instruction summary
Non-ConfidentialPDF file icon PDF versionARM DUI0472J
Copyright © 2010-2013 ARM. All rights reserved.