[FLASH-BUGS] flash-bugs Digest, Vol 24, Issue 1

Fri Aug 26 10:08:18 CDT 2016

Dear Jason,

thank you for bringing this bug to our attention. Thankfully this bug does
not appear to propagate in computations, rather manifests itself in checks
for out-of-bounds arrays, the output of the auxiliary DIVV_VAR, and the
criterion for application of the hybrid solver, which is currently
triggered only by the X and Y velocity gradients, as the third direction is
being ignored.

The bug will be corrected in the upcoming release.

Again, thank you for bringing this up.

Best,
-- Petros

On Fri, Aug 26, 2016 at 9:59 AM, Jason Galyardt <jason.galyardt at gmail.com>
wrote:

> Dear FLASH Developers,
>
> I forgot to include the following in my last bug report:
>
> source/Grid/GridMain/Grid_putBlkData.F90: lines 400 and 407:
>      if(NDIM > 2) then ! JEG -- Should be: if (NDIM > 1) then
>         if ((startingPos(JAXIS) + dataSize(2) + gr_jguard -1) > jmax) then
>            print *, "Error: Grid_putBlkData"
>            call Driver_abortFlash("Grid_putBlkData indicies too large")
>         end if
>      end if
>
>      if(NDIM > 3) then ! JEG -- Should be: if (NDIM > 2) then
>         if ((startingPos(KAXIS) + dataSize(3) + gr_kguard -1) > kmax) then
>            print *, "Error: Grid_putBlkData"
>            call Driver_abortFlash("Grid_putBlkData indicies too large")
>         end if
>      end if
>
> Regards,
> Jason
>
> ----
> Jason Galyardt, PhD
> Department of Physics and Astronomy
> University of Georgia
>
>
> On Fri, Aug 26, 2016 at 10:49 AM, Jason Galyardt <jason.galyardt at gmail.com
> >
> wrote:
>
> > Dear FLASH Developers,
> >
> > Similar to the bug report referenced in the subject line, I found
> > dimensionality check bugs in the following places (FLASH 4.3):
> >
> > source/Grid/GridMain/Grid_getBlkData.F90: lines 416 and 423:
> >      if(NDIM > 2) then ! JEG -- Should be: if (NDIM > 1) then
> >         if ((startingPos(JAXIS) + dataSize(2) + gr_jguard -1) > jmax)
> then
> >            print *, "Error: Grid_getBlkData"
> >            call Driver_abortFlash("Grid_getBlkData indicies too large")
> >         end if
> >      end if
> >
> >      if(NDIM > 3) then ! JEG -- Should be: if (NDIM > 2) then
> >         if ((startingPos(KAXIS) + dataSize(3) + gr_kguard -1) > kmax)
> then
> >            print *, "Error: Grid_getBlkData"
> >            call Driver_abortFlash("Grid_getBlkData indicies too large")
> >         end if
> >      end if
> >
> >
> > source/physics/Hydro/HydroMain/unsplit/hy_uhd_energyFix.F90: line 316
> >                  if (NDIM > 3) then ! JEG -- Should be: if (NDIM > 2)
> then
> >                     U(DIVV_VAR,i,j,k) = U(DIVV_VAR,i,j,k) &
> >                          +(U(VELZ_VAR,i,j,k+1)-U(VELZ_
> > VAR,i,j,k-1))/del(DIR_Z)
> >                  endif
> >
> > These bugs turned up via a simple grep for "NDIM > 3" in the source code
> > used in my current simulation, and I have not evaluated their impact. It
> > might be a good idea to do something similar for the entire FLASH source
> > tree, as this is an easy mistake to make (and easily corrected).
> >
> > Cheers,
> > Jason
> >
> > ----
> > Jason Galyardt, PhD
> > Department of Physics and Astronomy
> > University of Georgia
> >
> >
> > On Wed, Aug 24, 2016 at 1:00 PM, <flash-bugs-request at flash.uchicago.edu>
> > wrote:
> >
> >> Send flash-bugs mailing list submissions to
> >>         flash-bugs at flash.uchicago.edu
> >>
> >> To subscribe or unsubscribe via the World Wide Web, visit
> >>         http://flash.uchicago.edu/mailman/listinfo/flash-bugs
> >> or, via email, send a message with subject or body 'help' to
> >>         flash-bugs-request at flash.uchicago.edu
> >>
> >> You can reach the person managing the list at
> >>         flash-bugs-owner at flash.uchicago.edu
> >>
> >> When replying, please edit your Subject line so it is more specific
> >> than "Re: Contents of flash-bugs digest..."
> >>
> >>
> >> Today's Topics:
> >>
> >>    1.  Dimension problem in hy_uhd_getRiemannState.F90 (Yingchao Lu)
> >>    2. Re:  Dimension problem in hy_uhd_getRiemannState.F90
> >>       (Dongwook Lee)
> >>
> >>
> >> ----------------------------------------------------------------------
> >>
> >> Message: 1
> >> Date: Tue, 23 Aug 2016 16:19:14 -0500
> >> From: "Yingchao Lu" <yingchao.lu at rice.edu>
> >> Subject: [FLASH-BUGS] Dimension problem in hy_uhd_getRiemannState.F90
> >> To: <flash-bugs at flash.uchicago.edu>
> >> Message-ID: <000901d1fd84$0036b230$00a41690$@rice.edu>
> >> Content-Type: text/plain;       charset="us-ascii"
> >>
> >> Dear FLASH developers,
> >>
> >>
> >>
> >> In looking at
> >> source/physics/Hydro/HydroMain/unsplit/hy_uhd_getRiemannState.F90
> >>
> >> Line 284 in the FLASH 4.3 version of this file reads as follows:
> >>
> >>
> >>
> >> if (NDIM > 3) then
> >>
> >>
> >>
> >> However, maximum dimension of simulation is 3, yielding:
> >>
> >>
> >>
> >> if (NDIM > 2) then
> >>
> >>
> >>
> >> I don't know if this will affect the simulation results much.
> >>
> >>
> >>
> >> Best Regards,
> >>
> >> Yingchao
> >>
> >>
> >>
> >> ----
> >>
> >> Yingchao Lu
> >>
> >> Department of Physics and Astronomy
> >>
> >> Rice University
> >>
> >>
> >>
> >>
> >>
> >> ------------------------------
> >>
> >> Message: 2
> >> Date: Tue, 23 Aug 2016 16:52:53 -0500
> >> From: Dongwook Lee <dongwook at flash.uchicago.edu>
> >> Subject: Re: [FLASH-BUGS] Dimension problem in
> >>         hy_uhd_getRiemannState.F90
> >> To: yingchao.lu at rice.edu
> >> Cc: flash-bugs at flash.uchicago.edu
> >> Message-ID:
> >>         <CANvkU6FeuzJx3UZFg=POnj4agQn6Xw3YdsYcCtXOEQdaGr=Kqg at mail.
> >> gmail.com>
> >> Content-Type: text/plain; charset=UTF-8
> >>
> >> Dear Yingchao,
> >>
> >> Thanks for reporting the bug. We will include the bug fix in the next
> >> release.
> >>
> >> Best,
> >> Dongwook
> >>
> >> On Tue, Aug 23, 2016 at 4:19 PM, Yingchao Lu <yingchao.lu at rice.edu>
> >> wrote:
> >>
> >> > Dear FLASH developers,
> >> >
> >> >
> >> >
> >> > In looking at
> >> > source/physics/Hydro/HydroMain/unsplit/hy_uhd_getRiemannState.F90
> >> >
> >> > Line 284 in the FLASH 4.3 version of this file reads as follows:
> >> >
> >> >
> >> >
> >> > if (NDIM > 3) then
> >> >
> >> >
> >> >
> >> > However, maximum dimension of simulation is 3, yielding:
> >> >
> >> >
> >> >
> >> > if (NDIM > 2) then
> >> >
> >> >
> >> >
> >> > I don't know if this will affect the simulation results much.
> >> >
> >> >
> >> >
> >> > Best Regards,
> >> >
> >> > Yingchao
> >> >
> >> >
> >> >
> >> > ----
> >> >
> >> > Yingchao Lu
> >> >
> >> > Department of Physics and Astronomy
> >> >
> >> > Rice University
> >> >
> >> >
> >> >
> >> >
> >>
> >>
> >> ------------------------------
> >>
> >> _______________________________________________
> >> flash-bugs mailing list
> >> flash-bugs at flash.uchicago.edu
> >> http://flash.uchicago.edu/mailman/listinfo/flash-bugs
> >>
> >>
> >> End of flash-bugs Digest, Vol 24, Issue 1
> >> *****************************************
> >>
> >
> >
>

-- 
*************************************************
Petros Tzeferacos, PhD

Research Scientist,
Associate Director & Code Group Leader,
Flash Center for Computational Science,
Department of Astronomy & Astrophysics,
University of Chicago.
5640 S. Ellis Ave, ERC 553
Chicago, IL 60637
Tel. 773-834-3236
petros.tzeferacos at flash.uchicago.edu
*************************************************