Closed GoogleCodeExporter closed 9 years ago
What do you know - CUDA 2.2 seems to have fixed this bug - so I will comment
the offending syncthreads for
now (only tested on linux though - I will test on windows as well)
Original comment by shu...@gmail.com
on 25 Jun 2009 at 5:18
Verified on Linux with CUDA 2.2
Original comment by shu...@gmail.com
on 25 Jun 2009 at 5:22
Verified on Windows XP with CUDA 2.2
Original comment by shu...@gmail.com
on 25 Jun 2009 at 3:21
Reverted since behavior much more stable with the syncthreads
Original comment by shu...@gmail.com
on 25 Jun 2009 at 7:00
FIne, I'll fix it. I explained how to fix it in the description, I'll try it
myself. But we aren't marking this "wont fix".
Original comment by harr...@gmail.com
on 25 Jun 2009 at 9:38
I tried your approach and it didn't work (that is putting the syncthreads
before calls to warpSegScan)
Original comment by shu...@gmail.com
on 25 Jun 2009 at 9:41
This advisory is fixed in CUDA 2.3. There is a bug in the compiler where
moving the
__syncthreads() from the beginning of warpSegScan() to right before each call
of it
causes the code to fail. I filed an NVIDIA bug for this.
Closing this for now.
Original comment by harr...@gmail.com
on 26 Jun 2009 at 8:03
Original comment by harr...@gmail.com
on 26 Jun 2009 at 8:04
I fixed this for CUDA 2.2.
Original comment by harr...@gmail.com
on 30 Jun 2009 at 5:07
Original issue reported on code.google.com by
harr...@gmail.com
on 17 Jun 2009 at 1:38