Typos in documentation

[Marc_L_]

Hello,

I hope I am not polluting this forum with the following trivialities:

1. Is this the right place to report issues with MKL ?
2. There is a typo in the documentation of the page "v?LinearFrac" (in Section Vector Mathematical Functions).
   The types of scalea, scaleb, shifta and shiftb are wrong, it should be "float" instead of "float *" (and "double" instead of "double *").
3. Finally, is it really pertinent to put a "const" in front on non-pointers values (eg. MKL_INT) ?

Source: https://software.intel.com/en-us/node/521774

Best regards,

Marc Lasson.

[Gennady_F_Intel]

Hello Marc,

yes this is the right place to report the issue with MKL.  

You are right, the doc will be fixed.

[Roman_F_]

Another typo, located in function mkl_?csrcoo, at input parameter job: if job[0]=2 (left square bracket is missing).

[Gennady_F_Intel]

thanks Roman, we will check this too.

[Marc_L_]

Another one:

In the page about vslNewStreamEx https://software.intel.com/fr-fr/node/521854, the argument "params" should be of type "const unsigned int *" instead of "const unsigned int" (it is an array).

Cheers.

[Andrey_N_Intel]

Hi Marc, good catch, thanks, we will fix it in one of the next releases. Andrey

[Eugene_S_1]

I put my remark here to avoid adding a new topic. MKL Feast routines have the following declarations in (dfeast_scsrgv for example): _Mkl_Api(void,dfeast_scsrgv,(const char* uplo , const MKL_INT* n , const double* sa , const MKL_INT* isa , const MKL_INT* jsa , const double* sb , const MKL_INT* isb , const MKL_INT* jsb , MKL_INT* fpm , double* epsout , MKL_INT* loop , const double* emin , const double* emax , MKL_INT* m0 , double* e , double* x , MKL_INT* mode , double* res , MKL_INT* info)) I wonder why is it "mode" and not just "m" as the reference manual reads? Shouldn't the confusing "mode" be replaced with "m"?