Opened 12 years ago
Closed 12 years ago
#4323 closed defect (fixed)
autotest/warp/warp.py: TEST warp_27 fails
Reported by: | peifer | Owned by: | warmerdam |
---|---|---|---|
Priority: | normal | Milestone: | |
Component: | Autotest | Version: | svn-trunk |
Severity: | normal | Keywords: | |
Cc: |
Description
The test checks if its result is the same as gdalwarp -rb -t_srs EPSG:4326 ../gcore/data/byte.tif tmp/warp_27.tif
. This is true AFAICT, as the Checksum() is in both cases 4586. However, the test fails as it checks for a checksum value of 4583:
if cs != 4583: return 'fail'
(I was testing on 64bit Linux.)
Change History (3)
comment:1 by , 12 years ago
comment:2 by , 12 years ago
The configure script tells me:
GDAL is now configured for x86_64-unknown-linux-gnu C compiler: gcc -g -O2 C++ compiler: g++ -g -O2 ... $ g++ --version | head -1 g++ (Debian 4.6.1-13) 4.6.1
comment:3 by , 12 years ago
Resolution: | → fixed |
---|---|
Status: | new → closed |
I've tried to reproduce with a self-compiled gcc 4.6.1, but it worked for me, so I've changed the test so as to compare effectively with the result of 'gdalwarp -rb -t_srs EPSG:4326 ../gcore/data/byte.tif tmp/warp_27_ref.tif'. Hopefully the difference you see is some subtle rounding difference without much impact, but it is still weird I can't reproduce
r23305 /trunk/autotest/warp/warp.py: Make warp_27 more robust (#4323)
Which compiler version do you use ? GDAL compiled with -O2 or not ?