[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[netCDF #XJX-664489]: NCVPTG illegal stride with version 4.3.3.1, works for version 4.1.1



Hello,

Thank you for the bug report regarding this regression; my apologies for the 
delayed response, as I have been traveling this week for AGU.  I will return to 
the office next week where I can investigate this fully and will follow up with 
you. That you especially for the test program; that should help speed up 
identifying the cause for the bug.

Have a great weekend,

-Ward

> Hello,
> I am seeing some strange behavior with netcdf version 4.3.3.1.  A call to 
> NCVPTG with a stride of 1 produces an "Illegal stride" error.  The identical 
> call works fine with version 4.1.1 (these are the two versions I have 
> immediately available.)
> 
> Further, adding 1 to the stride in the 4.3.3.1 version produces the correct 
> behavior.  That is, in version 4.3.3.1 a stride of 2 is used to set every 
> value in the array dimension, a stride of 3 sets every other value, etc.
> 
> I've attached a test program that reproduces this behavior.  There is a 
> logical flag in the source (ADD_ONE_TO_STRIDE) that if set to .FALSE. 
> produces the correct behavior for 4.1.1. but the incorrect behavior for 
> 4.3.3.1.  Similarly, ADD_ONE_TO_STRIDE == .TRUE. will work for 4.3.3.1 but 
> not 4.1.1.
> 
> Best regard,
> Jeremy
> 
> Dr. Jeremy Lore
> Fusion and Materials for Nuclear Systems Division
> Oak Ridge National Laboratory
> Bldg. 5600, MS-6169
> Oak Ridge, TN 37831-6169
> (865) 574-0982
> 
> 
> 


Ticket Details
===================
Ticket ID: XJX-664489
Department: Support netCDF
Priority: Normal
Status: Closed