What is wrong? please help me,thanks

Accelerated Computing CUDA CUDA Programming and Performance
1

I used cuda-memcheck to check my programm. The following message was presented. Why?

========= Invalid global write of size 4
========= at 0x00000268 in gpuGenerateVerletList(float3*, int*, unsigned long, int*, float*, float, bool*)
========= by thread (31,0,0) in block (0,0,0)
========= Address 0x1f06aa007c is out of bounds
========= Saved host backtrace up to driver entry point at kernel launch time
========= Host Frame:/usr/lib64/libcuda.so.1 (cuLaunchKernel + 0x2c5) [0x204235]
========= Host Frame:t10 [0x1b9a1]
========= Host Frame:t10 [0x39543]
========= Host Frame:t10 [0x5df4]
========= Host Frame:t10 [0x5021]
========= Host Frame:t10 [0x508c]
========= Host Frame:t10 [0x3bdb]
========= Host Frame:/lib64/libc.so.6 (__libc_start_main + 0xfd) [0x1ed1d]
========= Host Frame:t10 [0x2d69]

2

it means that your kernel code is writing to an invalid location

You can use the methodology outlined here:

[url]cuda - Unspecified launch failure on Memcpy - Stack Overflow

to narrow this down to a single line of your kernel code that is causing the problem. Of course, this will require some debugging of your code.

3

Thank you very much.