[Insight-users] Resampling 64-bit crash

[Atwood, Robert C]

 Oops, the file size  bug differs depending on output. If I just read a
.mhd bigger than 2.1 Gb it gets error due to requesting a negative byte
count (SIGNED 32 Bit integer must be used somewhere in there) and writes
a zero-size file.


But if I select a different output eg .hdr , then 4gb works but still 5
gb does not. However, during processing it  uses up the memory according
to 'top' so perhaps its primarily an output problem ?



./ResampleBig s.mhd ss.hdr 10
 NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND
   0 9587m 4.2g 3300 R  400 27.0   1:34.83 ResampleBig
(later)

   0 9587m 9.3g 3320 R  100 59.6   3:40.35 ResampleBig
-rw-r--r--  1 rcatwood staff8 1.4G Feb 14 19:26 ss.img
-rw-r--r--  1 rcatwood staff8  348 Feb 14 19:26 ss.hdr

 ./ResampleBig s.mhd ss.hdr 4
-rw-r--r--  1 rcatwood staff8 3.8G Feb 14 19:22 ss.img
-rw-r--r--  1 rcatwood staff8  348 Feb 14 19:22 ss.hdr

./ResampleBig s.mhd ss.hdr 5
-rw-r--r--  1 rcatwood staff8  348 Feb 14 19:23 ss.hdr
-rw-r--r--  1 rcatwood staff8 673M Feb 14 19:23 ss.img