topic Re: ippsAutoScale_16s() in Intel® Integrated Performance Primitives

ippsAutoScale_16s()

pft_top — Thu, 29 Jul 2004 13:46:26 GMT

What is the primitive's exact function? How is it implemented? Please tell me the aguments' meaning, especially the last one which is always equals 3 as a int input argument, why?

Re: ippsAutoScale_16s()

Vladimir_Dudnik — Thu, 29 Jul 2004 13:56:04 GMT

Hi,

I recommend you to look at the IPP headers/manuals first, for example, you can find in ippsc.h:

/* /////////////////////////////////////////////////////////////////////////////
// Name: ippsAutoScale
// Purpose: Autoscaling: Find and normalize (shift left) the absolute
// maximal element and then do the same shift for all other elements
// Parameters:
// pSrc pointer to the input vector
// pDst pointer to the output vector
// pSrcDst pointer to the input/output vector
// len length of the input vector
// pScale pointer to the output scaling factor
// (number of bit the output vector has been shifted left)
// Returns: ippStsNoErr, if no errors
*/
IPPAPI( IppStatus, ippsAutoScale_16s, (const Ipp16s *pSrc, Ipp16s *pDst, int Len , int *pScale) )
IPPAPI( IppStatus, ippsAutoScale_16s_I, ( Ipp16s *pSrcDst, int Len , int *pScale) )

Function behaviuor is described in ippsman.pdf.

Regards,

Vladimir

Re: ippsAutoScale_16s()

pft_top — Thu, 29 Jul 2004 15:21:44 GMT

Hi,

I am using ippwmmx40, there is not any primitive named ippsAutoScale_16s() in the ippman.pdf or ippsc.h but ippSC.h under the stalled directory.Can you mail you ippman.pdf to me?I am badly in need of it.

thanks!

pft.top.

Re: ippsAutoScale_16s()

Vladimir_Dudnik — Thu, 29 Jul 2004 16:43:38 GMT

Hi,

oh, I see. It means IPP for Xscale does not provide this function. If you feel it is issue, please submit it through Premier support.

Regards,
Vladimir