Control: severity -1 important
I've now filed #947148 against openmpi 4 which seems to break lt_dlopen()
Andreas
Processing control commands:
> severity -1 important
Bug #946422 [src:silx] silx: autopkgtest regression: pocl error
Severity set to 'important' from 'serious'
--
946422: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=946422
Debian Bug Tracking System
Contact ow...@bugs.debian.org with
Processing control commands:
> severity -1 important
Bug #946582 [pocl] silx: autopkgtest regression: pocl error
Bug #946583 [pocl] silx: autopkgtest regression: pocl error
Severity set to 'important' from 'serious'
Severity set to 'important' from 'serious'
--
946582:
On 19/12/2019 17.21, Andreas Beckmann wrote:
> I can reproduce the original problem if I take some OpenCL hello world
> program (e.g. https://gist.github.com/ddemidov/2925717 with s/GPU/CPU/)
> and insert MPI_Init(, - no python needed
But only with OpenMPI 4 in sid, not with mpich.
I cannot
On 19/12/2019 14.33, Graham Inggs wrote:
> On Thu, 19 Dec 2019 at 15:17, Andreas Beckmann wrote:
>> I'm suspecting openmpi (that gets loaded by the io import) somehow messes up
>> some state,
>> causing the lt_*() failures.
>
> I wonder if this is related to #946986 ?
Downgrading to
On Thu, 19 Dec 2019 at 15:17, Andreas Beckmann wrote:
> I'm suspecting openmpi (that gets loaded by the io import) somehow messes up
> some state,
> causing the lt_*() failures.
I wonder if this is related to #946986 ?
On 19/12/2019 11.59, PICCA Frederic-Emmanuel wrote:
> I found that commenting this line
>
> # self.d_array_5 = pyopencl.array.zeros_like(self.d_array_img) - 5
>
> remove the pocl issue.
I think that's a red herring. Without that line I get python errors because
d_array_5 is missing.
That's
looking in picca@sixs7:~/Debian/silx/silx/silx/opencl/test/test_addition.py
def setUp(self):
if ocl is None:
return
self.shape = 4096
self.data = numpy.random.random(self.shape).astype(numpy.float32)
self.d_array_img =
I decided to concentrate myself on one opencl test (addition)
So I deactivated all other test by commenting the test in
silx/opencl/__init__.py
If I do not import silxs.io, this test works
(sid_amd64-dchroot)picca@barriere:~$ PYOPENCL_COMPILER_OUTPUT=1
With the silx.io import I have this
(sid_amd64-dchroot)picca@barriere:~$
PYTHONPATH=silx-0.11.0+dfsg/.pybuild/cpython3_3.7_silx/build python3 test.py
pocl error: lt_dlopen("(null)") or lt_dlsym() failed with 'can't close resident
module'.
note: missing symbols in the kernel binary might be
I've minimized it to
=8<=
import unittest
def suite():
from silx.io import test as test_io
from silx.opencl import test as test_ocl
test_suite = unittest.TestSuite()
test_suite.addTest(test_ocl.suite())
return test_suite
if __name__ == '__main__':
runner =
not better
test cpp engine for medfilt2d ... ok
testOpenCLMedFilt2d (silx.image.test.test_medianfilter.TestMedianFilterEngines)
test cpp engine for medfilt2d ... pocl error: lt_dlopen("(null)") or lt_dlsym()
failed with 'can't close resident module'.
note: missing symbols in the kernel binary
It seems that this test does not PASS
@unittest.skipUnless(ocl, "PyOpenCl is missing")
def testOpenCLMedFilt2d(self):
"""test cpp engine for medfilt2d"""
res = medianfilter.medfilt2d(
image=TestMedianFilterEngines.IMG,
On 12/12/2019 16.33, Graham Inggs wrote:
> On 2019/12/12 16:25, Andreas Beckmann wrote:
>> I can reproduce the error in the silx tests in sid. This could be
>> related to pocl being rebuilt for the hwloc 1->2 transition.
>
> Silx's autopkgtests started failing in unstable on 2019-11-15 [1].
>
On 2019/12/12 16:25, Andreas Beckmann wrote:
I can reproduce the error in the silx tests in sid. This could be
related to pocl being rebuilt for the hwloc 1->2 transition.
Silx's autopkgtests started failing in unstable on 2019-11-15 [1].
Pocl was binNMU'd for hwloc15 on 2019-11-27 [2].
[1]
Followup-For: Bug #946422
Hi,
I can reproduce the error in the silx tests in sid. This could be
related to pocl being rebuilt for the hwloc 1->2 transition.
This looks like it should be reproducible with just pyopencl.
Can you minimize the failing test to a simple script using pyopencl
only?
Source: silx
Version: 0.11.0+dfsg-2
Severity: serious
X-Debbugs-CC: debian...@lists.debian.org
User: debian...@lists.debian.org
Usertags: regression
Hi Maintainer
The autopkgtests of silx have regressed in testing [1].
I have copied what I hope is the relevant part of the log below.
Regards
17 matches
Mail list logo
