[mpich-discuss] COMM_SELF and collective IO

Rajeev Thakur thakur at mcs.anl.gov
Fri Oct 16 19:41:34 CDT 2009


You will have the same if you use COMM_SELF for one and COMM_WORLD for
the other. You don't want to use COMM_SELF for both. 

> -----Original Message-----
> From: mpich-discuss-bounces at mcs.anl.gov 
> [mailto:mpich-discuss-bounces at mcs.anl.gov] On Behalf Of burlen
> Sent: Friday, October 16, 2009 7:22 PM
> To: mpich-discuss at mcs.anl.gov
> Subject: Re: [mpich-discuss] COMM_SELF and collective IO
> 
> I could then I will have to have two codes with only a one 
> line difference.
> 
> Rajeev Thakur wrote:
> > Why don't you just use MPI_File_read instead?
> >
> > Rajeev
> >
> >   
> >> -----Original Message-----
> >> From: mpich-discuss-bounces at mcs.anl.gov 
> >> [mailto:mpich-discuss-bounces at mcs.anl.gov] On Behalf Of burlen
> >> Sent: Friday, October 16, 2009 5:39 PM
> >> To: mpich-discuss at mcs.anl.gov
> >> Subject: [mpich-discuss] COMM_SELF and collective IO
> >>
> >> I have an io code that uses the subarray type. I use the 
> collective 
> >> IO function, MPI_File_read_all. Now, I have a use case where each 
> >> process needs to read independently of the others. What I 
> am tempted 
> >> to do is change the communicator to MPI_COMM_SELF and use the same 
> >> code. I suspect that this is a bad idea. Is it?
> >>
> >> _______________________________________________
> >> mpich-discuss mailing list
> >> mpich-discuss at mcs.anl.gov
> >> https://lists.mcs.anl.gov/mailman/listinfo/mpich-discuss
> >>
> >>     
> >
> > _______________________________________________
> > mpich-discuss mailing list
> > mpich-discuss at mcs.anl.gov
> > https://lists.mcs.anl.gov/mailman/listinfo/mpich-discuss
> >   
> 
> _______________________________________________
> mpich-discuss mailing list
> mpich-discuss at mcs.anl.gov
> https://lists.mcs.anl.gov/mailman/listinfo/mpich-discuss
> 



More information about the mpich-discuss mailing list