Closed Hurleyworks closed 4 years ago
Hi,
nvcc gets rid of unused callable program by default. You need to specify the option "Generate Relocatable Device Code" with "-rdc=true" to prevent this behavior.
Thanks,
and thank you for sharing your project! It seems really nice. I need to update the utility again since OptiX 7.2 has been announced (while it is minor update maybe) and there are still features it doesn't support yet like curve primitives, deformation blur.
Great. That fixed it. Thanks for the quick response.
On Wed, Oct 7, 2020 at 3:01 PM shocker-0x15 notifications@github.com wrote:
Hi,
nvcc gets rid of unused callable program by default. You need to specify the option "Generate Relocatable Device Code" with "-rdc=true" to prevent this behavior.
Thanks,
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/shocker-0x15/OptiX7_Utility/issues/7#issuecomment-705133083, or unsubscribe https://github.com/notifications/unsubscribe-auth/AB56PP3XAE4ESWXAZ2T5LXLSJS3IPANCNFSM4SHY47KQ .
-- -Steve http://www.hurleyworks.com
As you've probably discovered, OptiX7.2 made breaking changes to OptiX7_Utility probably due to the new bounding box automation
Error C2039 'aabbs': is not a member of 'OptixBuildInputInstanceArray'
optiX7Util
On Wed, Oct 7, 2020 at 3:49 PM Hurleyworks stevehurley33@gmail.com wrote:
Great. That fixed it. Thanks for the quick response.
On Wed, Oct 7, 2020 at 3:01 PM shocker-0x15 notifications@github.com wrote:
Hi,
nvcc gets rid of unused callable program by default. You need to specify the option "Generate Relocatable Device Code" with "-rdc=true" to prevent this behavior.
Thanks,
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/shocker-0x15/OptiX7_Utility/issues/7#issuecomment-705133083, or unsubscribe https://github.com/notifications/unsubscribe-auth/AB56PP3XAE4ESWXAZ2T5LXLSJS3IPANCNFSM4SHY47KQ .
-- -Steve http://www.hurleyworks.com
-- -Steve http://www.hurleyworks.com
Updated. I have not studied 7.2.0 features other than AABB automatic computation yet, but made it possible to be compiled with 7.2.0 for now. Automatic AABB computation is very good feature from 7.2.0.
Thanks! I've got it working here now.
Automatic AABB computation is very good feature from 7.2.0.
I'm so glad I've held off on trying to add motion blur support. I think it should be easier now. :)
On Thu, Oct 8, 2020 at 2:23 PM shocker-0x15 notifications@github.com wrote:
Updated. I have not studied 7.2.0 features other than AABB automatic computation yet, but made it possible to be compiled with 7.2.0 for now. Automatic AABB computation is very good feature from 7.2.0.
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/shocker-0x15/OptiX7_Utility/issues/7#issuecomment-705743887, or unsubscribe https://github.com/notifications/unsubscribe-auth/AB56PP2SMOD5BYMGDQLF2CLSJX7S5ANCNFSM4SHY47KQ .
-- -Steve http://www.hurleyworks.com
Hello, this is off topic. Do you use Visual Studio 2019? If so is your Visual Studio fine after installing CUDA 11.1 toolkit? My visual studio appears to get slow after around installing CUDA 11.1 (I'm not entirely sure).
https://forums.developer.nvidia.com/t/visual-studio-2019-became-too-slow-after-installing-cuda-11-1/155952 It seems many others have the same problem.
Thanks for the heads up. I was wondering what was going on. My slowdowns are not as drastic as others are reporting but they were enough to be annoying. I installed an older version of NsVSE and it seems to be fixed. NSight debugger still doesn't work correctly in Optix7 though. I was hoping it would be fixed in 7.2 https://forums.developer.nvidia.com/t/debugging-optix-7-1-with-nsight/142843/2 https://forums.developer.nvidia.com/t/debugging-optix-7-1-with-nsight/142843/2
On Tue, Oct 13, 2020 at 3:20 PM shocker-0x15 notifications@github.com wrote:
https://forums.developer.nvidia.com/t/visual-studio-2019-became-too-slow-after-installing-cuda-11-1/155952 It seems many others have the same problem.
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/shocker-0x15/OptiX7_Utility/issues/7#issuecomment-707955762, or unsubscribe https://github.com/notifications/unsubscribe-auth/AB56PPZOD6HRZKSVDZICGYLSKSSBPANCNFSM4SHY47KQ .
-- -Steve http://www.hurleyworks.com
Howdy, hope all is well!
I have been stuck on a strange problem for a while. When I compile a kernel with a direct callable program, the ptx output does not contain the program so the call to pipeline.createCallableProgramGroup() fails.
[ 2][COMPILE FEEDBACK]: COMPILE ERROR: "__direct_callable__testCall" not found in programDescriptions[0].callables.moduleDC
Here's the kernel code, the output ptx code and the pipeline creation function. This barebones example compile and runs fine if I remove the callable program
https://gist.github.com/Hurleyworks/6573dd5ebb9ccf306e1184a9d5e30154
Do you have an idea what I'm doing wrong. Everything works as expected in the uber sample so it has to be something I'm doing wrong somewhere.
Thanks again for making the project. I'm really enjoying using it ... it makes life so much easier :) https://www.youtube.com/watch?v=MPXU3avzv_s&ab_channel=HurleyworksInc