-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 10/30/2012 09:21 PM, Ulrich Pegelow wrote: > Am 30.10.2012 20:13, schrieb Jens Fendler: >> On 10/30/2012 08:43 PM, Ulrich Pegelow wrote: >> >> >> Nope - doesn't work. (Just like the easier variant using just >> gopacity*1.0f also did not..) >> >> btw, when the compilation fails, dt gives me two empty lines of >> output under a "BUILD LOG:" header (I'm using "-d opencl" only). >> Is there maybe a way to get some more output from the failing >> compilation here? >> > That seems to be one of the problems with your opencl > driver/compiler. It does not seem to have any indicative messages. > Otherwise they would be shown here. That alone is uncommon in case > of compile time errors. > > If you ask me, the only realistic option for you might be to > downgrade to an older driver. 295 has been reported by Jo to run > fine. On my box I have 304.43 also without issues.
Downgrading doesn't help. I've actually spent most of the day installing, re-installing and de-installing NVidia drivers - and tried out pretty much everything from 295.40 to 30x (from official repositories, PPAs, as well as the latest official NVidia driver. Apart from some other minor issues along the way it always boils down to the very same issue with darktable - regardless of the driver version. I'm currently running two (similar) installations on the same machine, one Mint Linux 13 with NVidia 295.40, and a Ubuntu 12.10 that currently has the latest version (some 304 I think) from NVidia's web site. They both perform exactly the same for me, i.e. they both work when changing the ...RAW kernel accordingly, and fail equally in all other cases. There are still some things I have kept constant so far. I've not moved away from Debian/Ubuntu/Mint type of installation (but would like to keep that actually), and I've been using the same versions of VirtualGL, acpi_call, bbswitch, and Bumblebee (all necessary for working with the Optimus card) so far. This leaves some room for future experiments, but for the time being I'll just stick to my highly efficient multiply-and-divide code in blendop.cl. But don't worry: I won't ask for this to be merged into master :-) My previous question regarding anybody using the same hardware (GT640M) has not yet been answered, but I would assume people are actually using (more or less) different adapters. If this is so, I believe it's a problem between this particular piece of cra.. aehm: chip and the driver/opencl compiler - which hopefully gets fixed in the near future. At least I can continue working with dt and openCL support - which is a major achievement at this point :-) Thanks a lot for your help, Ulrich! And do let me know when you need a not-so-easy testbed for some fancy opencl code. Jens -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.11 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://www.enigmail.net/ iEYEARECAAYFAlCQLbYACgkQbyTZFOIr/d08HgCg6YvQiVJLWhPLC9AEGCwagAGl a5QAoO+1x2qqQb2hDvBc91cQt77fuVBy =ks7J -----END PGP SIGNATURE----- ------------------------------------------------------------------------------ Everyone hates slow websites. So do we. Make your web apps faster with AppDynamics Download AppDynamics Lite for free today: http://p.sf.net/sfu/appdyn_sfd2d_oct _______________________________________________ darktable-devel mailing list darktable-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/darktable-devel
