[mpich-discuss] MPI_Init() and console output

[Pavan Balaji]

Hi Thomas,

I looked through the code.  Our internal code seems to be buffering data 
in this case and not writing it out.  I wrote that piece of code, but 
don't remember why it is this way.  Maybe it's just a bug.  But I'll 
need to think through it some more.

Can you file a ticket so we don't lose track of this issue?

https://trac.mcs.anl.gov/projects/mpich2/newticket

  -- Pavan

On 12/14/2011 06:43 AM, Thomas Leissing wrote:
> Dear all,
>
> Somewhere in my boundary element solver I have a console progress bar
> which prints to screen the progress of an operation.
> It looks like this:
> INFO : 50% [------------> ]
>
> The function is very simple:
> //i is the current iteration number
> //n is the total number of iterations
> void progressBar(int x, int n)
> {
>
> if ( x % (n/100) != 0 ) return;
>
> // Calculate the ratio of complete-to-incomplete.
> float ratio = x/(float)n;
> int c = ratio * 80
>
> // Show the percentage complete.
> printf("INFO :\t%3d%% [", (int)(ratio*100) );
>
> // Show the load bar.
> for (int x=0; x<c-1; x++)
> printf("-");
>
> printf(">");
>
> for (int x=c; x<pBarWidth; x++)
> printf(" ");
>
> // ANSI Control codes to go back to the
> // previous line and clear it.
> printf("]\n\033[F\033[J");
> }
>
>
> My problem is that the progress bar doesn't print to screen as soon as I
> call the MPI_Init() function (it works fine otherwise).
> I tried to play with fflush(stdout) commands without any success.
>
> Any ideas ?
>
> Thanks,
> Thomas Leissing
>
>
> _______________________________________________
> mpich-discuss mailing list     mpich-discuss at mcs.anl.gov
> To manage subscription options or unsubscribe:
> https://lists.mcs.anl.gov/mailman/listinfo/mpich-discuss

-- 
Pavan Balaji
http://www.mcs.anl.gov/~balaji