[GRASS-dev] [GRASS GIS] #2764: corrupt data written to FCELL and DCELL rasters, hard to re-produce

[GRASS GIS]

#2764: corrupt data written to FCELL and DCELL rasters, hard to re-produce
--------------------+-------------------------
 Reporter:  dylan   |      Owner:  grass-dev@…
     Type:  defect  |     Status:  new
 Priority:  normal  |  Milestone:  7.1.0
Component:  Raster  |    Version:  unspecified
 Keywords:          |        CPU:  x86-64
 Platform:  Linux   |
--------------------+-------------------------
 I think that I may have made some progress after a long battle with some
 very hard to reproduce [http://osgeo-org.1560.x6.nabble.com/Error-reading-
 raster-data-for-row-xxx-only-when-using-r-series-and-t-rast-series-
 td5229569.html#a5230287 errors]. The attached bash scripts can be used in
 any location (LL and projected give the same results on my machine) and
 will test the following:

 1. generate a new DCELL map via `r.surf.gauss`
 2. test resulting map via `r.univar`
 3. repeat process, overwriting previous DCELL map

 I have noticed that running these tests serially and in parallel produce
 the same results: ~2% of the maps generated generate an error message when
 read via `r.univar`. This is the same error message [http://osgeo-
 org.1560.x6.nabble.com/Error-reading-raster-data-for-row-xxx-only-when-
 using-r-series-and-t-rast-series-td5229569.html that started my recent
 string of postings to GRASS-dev]. After applying a patch (#2762) to add
 some more information to the error message, I see:

 {{{
 ERROR: Error reading raster data for row 1949 of <testing_gauss_7>
        (buffer size mismatch)
 }}}

 Tests were performed using r66514 of grass_trunk.

--
Ticket URL: <https://trac.osgeo.org/grass/ticket/2764>
GRASS GIS <https://grass.osgeo.org>