User Tools

Site Tools


simd_20signum_20function

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

simd_20signum_20function [2018/03/31 13:19] (current)
Line 1: Line 1:
 +=====SIMD signum function=====
  
 +//by Richard Russell, 30th December 2014 (based on a [[http://​blogs.msdn.com/​b/​oldnewthing/​archive/​2014/​12/​29/​10583275.aspx|blog post]] by Raymond Chen)//\\ \\  The **signum** function returns **-1** for all negative numbers, **+1** for all positive non-zero numbers and **zero** for an input of zero; it corresponds to the **SGN()** function in BBC BASIC. There is no direct equivalent in assembly language, but the following code snippets illustrate how it may be calculated on multiple data values using MMX and SSE2 code. In all cases the **input** is in mm0 or xmm0 and the **output** is in mm1 or xmm1:\\ \\ **Signum of eight 8-bit values**\\ ​
 +        pxor    mm1, mm1
 +        pxor    mm2, mm2
 +        pcmpgtb mm1, mm0 ; mm1 = pcmpgt(0, x)
 +        pcmpgtb mm0, mm2 ; mm0 = pcmpgt(x, 0)
 +        psubb   mm1, mm0 ; mm1 = signum
 +\\ **Signum of four 16-bit values**\\ ​
 +        pxor    mm1, mm1
 +        pxor    mm2, mm2
 +        pcmpgtw mm1, mm0 ; mm1 = pcmpgt(0, x)
 +        pcmpgtw mm0, mm2 ; mm0 = pcmpgt(x, 0)
 +        psubw   mm1, mm0 ; mm1 = signum
 +\\ **Signum of two 32-bit values**\\ ​
 +        pxor    mm1, mm1
 +        pxor    mm2, mm2
 +        pcmpgtd mm1, mm0 ; mm1 = pcmpgt(0, x)
 +        pcmpgtd mm0, mm2 ; mm0 = pcmpgt(x, 0)
 +        psubd   mm1, mm0 ; mm1 = signum
 +\\ **Signum of sixteen 8-bit values (requires ASMLIB2 library)**\\ ​
 +        pxor    xmm1, xmm1
 +        pxor    xmm2, xmm2
 +        pcmpgtb xmm1, xmm0 ; xmm1 = pcmpgt(0, x)
 +        pcmpgtb xmm0, xmm2 ; xmm0 = pcmpgt(x, 0)
 +        psubb   xmm1, xmm0 ; xmm1 = signum
 +\\ **Signum of eight 16-bit values (requires ASMLIB2 library)**\\ ​
 +        pxor    xmm1, xmm1
 +        pxor    xmm2, xmm2
 +        pcmpgtw xmm1, xmm0 ; xmm1 = pcmpgt(0, x)
 +        pcmpgtw xmm0, xmm2 ; xmm0 = pcmpgt(x, 0)
 +        psubw   xmm1, xmm0 ; xmm1 = signum
 +\\ **Signum of four 32-bit values (requires ASMLIB2 library)**\\ ​
 +        pxor    xmm1, xmm1
 +        pxor    xmm2, xmm2
 +        pcmpgtd xmm1, xmm0 ; xmm1 = pcmpgt(0, x)
 +        pcmpgtd xmm0, xmm2 ; xmm0 = pcmpgt(x, 0)
 +        psubd   xmm1, xmm0 ; xmm1 = signum
simd_20signum_20function.txt ยท Last modified: 2018/03/31 13:19 (external edit)