|
| View previous topic :: View next topic |
| Author |
Message |
DAVID-SPH
Joined: 23 May 2011 Posts: 28
|
Posted: Thu Aug 23, 2012 5:18 am Post subject: |
|
|
Any other bit intrinsics not supported?... I kind of need to know because I'm programming a Radix Sort on CUDA Fortran and I will make heavy use of all of them.
Best Regards, |
|
| Back to top |
|
 |
mkcolg
Joined: 30 Jun 2004 Posts: 4995 Location: The Portland Group Inc.
|
Posted: Thu Aug 23, 2012 10:47 am Post subject: |
|
|
It will be the same for all of them (btest, iand, ibclr, ibits, ibset, ieor, ior, ishft, ishftc, not) since we don't have the GPU run time device routine. If the intrinsic gets inlined, then it's not a problem.
- Mat |
|
| Back to top |
|
 |
DAVID-SPH
Joined: 23 May 2011 Posts: 28
|
Posted: Thu Aug 23, 2012 5:10 pm Post subject: |
|
|
mmm can I use the "inline" keyword to force it to be inlined?
the ibits one is critical for the radix sort... |
|
| Back to top |
|
 |
mkcolg
Joined: 30 Jun 2004 Posts: 4995 Location: The Portland Group Inc.
|
Posted: Fri Aug 24, 2012 10:01 am Post subject: |
|
|
| Quote: | mmm can I use the "inline" keyword to force it to be inlined?
the ibits one is critical for the radix sort... | When they can be inlined, they are inlined. But the user can't effect this.
I wish I knew off hand the exact circumstance of when each are inlined or not. Though, I'll do some digging on ibits and get back to you.
- Mat |
|
| Back to top |
|
 |
mkcolg
Joined: 30 Jun 2004 Posts: 4995 Location: The Portland Group Inc.
|
Posted: Tue Sep 04, 2012 1:30 pm Post subject: |
|
|
Word from engineering is that they'll have "btest", "iand", "ibclr", "ibset", "ieor", and "iand" into 12.9. Granted, I haven't verified it myself, but that's the word.
- Mat |
|
| Back to top |
|
 |
|
|
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum
|
Powered by phpBB © 2001, 2002 phpBB Group
|