Experimental Physics and
Industrial Control System

[email protected] · Mon, 30 Jul 2012 13:57:42 -0700 (PDT)

Hey guys,

Thanks for all your help and advice. Using gdb I found my problem was not assigning the appropriate amount of memory to copy the record msg char* into a buffer char*. I fixed this by just accessing the record msg directly without copying it. Attached below is the custom conversion routine I found works well and reliably.

 // WF read routine for AGE5071C, converts Hz to MHz for P1==1
static int readWF(struct gpibDpvt *pdpvt,int P1, int P2, char**P3) {

  struct waveformRecord *pwf=(struct waveformRecord *)pdpvt->precord;
  asynUser *pasynUser = pdpvt->pasynUser;

  double* value = (double *)pwf->bptr;
  double data;
  int n;

  pwf->nord=0;

  while ((pwf->nord < pwf->nelm) && sscanf(pdpvt->msg,"%lE,%n",&data,&n)==1) {

    if (P1==1) data/=1E6;

    memcpy(value+pwf->nord,&data,sizeof(double));
    pwf->nord++;
    pdpvt->msg+=n;

    asynPrint(pasynUser,ASYN_TRACE_FLOW,"(NORD,VAL) : (%i,%E)\n",pwf->nord,data);

  }

  if (pwf->nord == 0) return -1;

  pwf->udf = 0;
  return 0;

}

I can definitely see now the the advantages of using stream device. I my system control needs become more complicated I'll switch over to using that instead of asynGpib.

Cheers,

Andrew

On Mon, 30 Jul 2012, Eric Norum wrote:

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System