Le jeudi 10 août 2017 à 09:31 +0200, Rafael Laboissière a écrit : > > * Adrian Bunk <b...@debian.org> [2017-08-07 01:57]: > > > Source: octave-image > > Version: 2.6.1-1 > > Severity: serious > > Tags: buster sid > > > > Some recent change in unstable and buster makes octave-image FTBFS on i386: > > > > https://tests.reproducible-builds.org/debian/history/octave-image.html > > > > https://tests.reproducible-builds.org/debian/rb-pkg/unstable/i386/octave-image.html > > > > ... > > Checking CC files ... > > warning: function /build/1st/octave-image-2.6.1/inst/private/iscolormap.m > > shadows a core library function > > warning: called from > > /tmp/filePMKFm8 at line 1 column 1 > > [__spatial_filtering__] > > PASSES 21 out of 21 tests > > [graycomatrix] > > PASSES 2 out of 2 tests > > [watershed] > > ***** test > > im = [ > > 2 3 30 2 > > 3 30 3 30 > > 255 31 30 4 > > 2 255 31 30 > > 1 2 255 5]; > > > > labeled8 = [ > > 1 1 0 3 > > 1 1 0 3 > > 0 0 0 0 > > 2 2 0 4 > > 2 2 0 4]; > > assert (watershed (im), labeled8); > > assert (watershed (im, 8), labeled8); > > !!!!! test failed > > ASSERT errors for: assert (watershed (im),labeled8) > > > > Location | Observed | Expected | Reason > > (3,4) 3 0 Abs err 3 exceeds tol 0 > > (4,4) 0 4 Abs err 4 exceeds tol 0 > > [bwdist] > > Sat Sep 8 05:36:25 UTC 2018 - pbuilder was killed by timeout after 18h. > > I would guess that the culprit is the following unit test in bwdist: > > ## The quasi-euclidean method is apparently sensitive to a machine precision > ## error that happens in x86 systems only. This test will cause an endless > ## loop in case of a regression. > %!test > %! bw = [ 0 1 1 0 0 0 1 0 > %! 0 0 0 0 0 0 0 0 > %! 1 1 0 0 0 0 0 0 > %! 0 0 0 0 0 0 1 0 > %! 0 0 0 0 1 0 0 1 > %! 0 0 0 0 0 0 0 0 > %! 1 0 0 0 0 0 0 0 > %! 0 0 1 0 0 1 1 0]; > %! out = single ([ > %! 1.00000 0.00000 0.00000 1.00000 2.00000 1.00000 0.00000 > 1.00000 > %! 1.00000 1.00000 1.00000 sqrt(2) sqrt(2)+1 sqrt(2) 1.00000 > sqrt(2) > %! 0.00000 0.00000 1.00000 2.00000 2.00000 sqrt(2) 1.00000 > sqrt(2) > %! 1.00000 1.00000 sqrt(2) sqrt(2) 1.00000 1.00000 0.00000 > 1.00000 > %! 2.00000 2.00000 2.00000 1.00000 0.00000 1.00000 1.00000 > 0.00000 > %! 1.00000 sqrt(2) 2.00000 sqrt(2) 1.00000 sqrt(2) sqrt(2) > 1.00000 > %! 0.00000 1.00000 1.00000 sqrt(2) sqrt(2) 1.00000 1.00000 > sqrt(2) > %! 1.00000 1.00000 0.00000 1.00000 1.00000 0.00000 0.00000 > 1.00000 > %! ]); > %! assert (bwdist (bw, "quasi-euclidean"), out); > > Should we run the test above only on non-x86 architectures?
I would rather patch the test to allow for some tolerance margin. -- ⢀⣴⠾⠻⢶⣦ Sébastien Villemot ⣾⠁⢠⠒⠀⣿⡁ Debian Developer ⢿⡄⠘⠷⠚⠋⠀ http://sebastien.villemot.name ⠈⠳⣄ http://www.debian.org