Hi Ben, Sorry for the slow reply here. Now fixed:
r5264 (longer explanation: the sphNG reader assumed that certain arrays were always in certain places, e.g. the first real*4 array was always the density. I’ve now updated the sphNG format reader so that it now identifies x,y,z,m,h,rho and Bx,By,Bz by the tags and puts them in the right order no matter which block or precision you dump them in) Thanks for reporting! Daniel > On 6 Oct 2015, at 4:36 AM, Benjamin Lewis <[email protected]> wrote: > > > Hi Daniel, hope things are well. > > I've been hacking about with sphNG to make the code full > double-precision everywhere. However, I have a problem with splash, > which is helpfully segfaulting on my new REAL*8 only dump files. > > What I've done is moved every real into 'number 6', i.e. the 'default > reals' section, but splash seems to get confused. Unsure why - if I > build with bounds checking on I get the very helpful > > forrtl: severe (408): fort: (2): Subscript #2 of the array DAT has value > 25 which is greater than the upper bound of 24 > > Which I've managed to pin down to line 1553 of read_data_sphNG.f90 with > gdb. However, I can't work out what's causing splash to run off the end > of the DAT array? > > An example dump file is attached. > > > Interestingly, if I make all the reals real*8s, i.e. 'number 8' in the > header, everything works. I'd prefer to avoid doing it this way, > however, since the flexibility of putting it into the default real > section is too tempting. > > Thanks :-) > > -- > Benjamin Lewis -- You received this message because you are subscribed to the Google Groups "SPLASH users forum" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at http://groups.google.com/group/splash-users. For more options, visit https://groups.google.com/d/optout.
