[petsc-dev] Is there a reason for short arch and host names?

Jed Brown jedbrown at mcs.anl.gov
Sat Jul 28 11:37:47 CDT 2012


http://petsc.cs.iit.edu/petsc/releases/petsc-3.3/rev/93f374d090c4

On Fri, Jul 27, 2012 at 6:04 PM, Barry Smith <bsmith at mcs.anl.gov> wrote:

>
>     I absolutely hate sizeof arch.   That is one of the most perverse
> things people can do in C.   What is wrong with good-old sizeof(arch)?
>
>     Needs to go in the PETSc style guide :-)
>
>
>     Barry
>
>
> On Jul 27, 2012, at 2:35 PM, Jed Brown wrote:
>
> > I'm sick of truncated PETSC_ARCH in log_summary. Does anyone object to
> my pushing the following to petsc-3.3? Should we just make all the fields
> huge so we're guaranteed to get everything? Since pname can be as long as
> PETSC_MAX_PATH_LEN, this is not a line we can reasonably limit/align.
> >
> > --- a/src/sys/plog/plog.c
> > +++ b/src/sys/plog/plog.c
> > @@ -1269,7 +1269,7 @@
> >    PetscStageInfo     *stageInfo = PETSC_NULL;
> >    PetscEventPerfInfo *eventInfo = PETSC_NULL;
> >    PetscClassPerfInfo *classInfo;
> > -  char               arch[10], hostname[64], username[16],
> pname[PETSC_MAX_PATH_LEN], date[64];
> > +  char               arch[64], hostname[128], username[16],
> pname[PETSC_MAX_PATH_LEN], date[64];
> >    const char         *name;
> >    PetscLogDouble     locTotalTime, TotalTime, TotalFlops;
> >    PetscLogDouble     numMessages, messageLength, avgMessLen,
> numReductions;
> > @@ -1311,12 +1311,12 @@
> >    ierr = PetscFPrintf(comm, fd, "***             WIDEN YOUR WINDOW TO
> 120 CHARACTERS.  Use 'enscript -r -fCourier9' to print this document
>      ***\n");CHKERRQ(ierr);
> >    ierr = PetscFPrintf(comm, fd,
> "************************************************************************************************************************\n");CHKERRQ(ierr);
> >    ierr = PetscFPrintf(comm, fd,
> "\n---------------------------------------------- PETSc Performance
> Summary: ----------------------------------------------\n\n");CHKERRQ(ierr);
> > -  ierr = PetscGetArchType(arch, 10);CHKERRQ(ierr);
> > -  ierr = PetscGetHostName(hostname, 64);CHKERRQ(ierr);
> > -  ierr = PetscGetUserName(username, 16);CHKERRQ(ierr);
> > +  ierr = PetscGetArchType(arch,sizeof arch);CHKERRQ(ierr);
> > +  ierr = PetscGetHostName(hostname,sizeof hostname);CHKERRQ(ierr);
> > +  ierr = PetscGetUserName(username,sizeof username);CHKERRQ(ierr);
> >    ierr = PetscGetProgramName(pname, PETSC_MAX_PATH_LEN);CHKERRQ(ierr);
> > -  ierr = PetscGetDate(date, 64);CHKERRQ(ierr);
> > -  ierr = PetscGetVersion(version,256);CHKERRQ(ierr);
> > +  ierr = PetscGetDate(date,sizeof date);CHKERRQ(ierr);
> > +  ierr = PetscGetVersion(version,sizeof version);CHKERRQ(ierr);
> >    if (size == 1) {
> >      ierr = PetscFPrintf(comm,fd,"%s on a %s named %s with %d processor,
> by %s %s\n", pname, arch, hostname, size, username, date);CHKERRQ(ierr);
> >    } else {
> >
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20120728/e36fd526/attachment.html>


More information about the petsc-dev mailing list