Any hint as to how I can fix this? (other than pray for nVidia to do
something about it, of course !)
Or is it something I can ignore? Will it come to bite my behind if I
decide to go on zith my faulty ICD?
Rohit Garg a écrit :
My guess would be that your ICD (prolly nv supplied)
On Wed, Apr 14, 2010 at 9:39 PM, Hua Wong <hua.wong(a)pasteur.fr> wrote:
From what I've seen during the installation,
pyopencl-0.91.5 has been
installed without problems (no errors or warning during make...).
Most examples seems to run without any troubles.
However sometimes goes into "Abort trap" and terminate.
For example, matrix-multiply works fine with all the GPUs, but return
Abort trap when launched on the CPU.
Choose device(s) from these choices:
 <pyopencl.Device 'GeForce GT 120' at 0x1022600>
 <pyopencl.Device 'GeForce GT 120' at 0x2022600>
 <pyopencl.Device 'GeForce GT 120' at 0x3022600>
 <pyopencl.Device 'GeForce GT 120' at 0x4022600>
 <pyopencl.Device 'Intel(R) Xeon(R) CPU E5520 @ 2.27GHz'
Choice, comma-separated :4
My main question is : can I brush this error aside? Or did my
installation gone horribly wrong somewhere ?
Mac OS X 10.6.3
Python 2.6.5 (r265:79063, Apr 14 2010, 14:25:50)
[GCC 4.2.1 (Apple Inc. build 5646) (dot 1)] on darwin
everything has been made for a x86_64 architecture
PyOpenCL mailing list