Shared memory between Host and Device How to access the structures

Accelerated Computing CUDA CUDA Programming and Performance
1

Hello,

I’ve an 8800 GTX with me and I have 3 kernels running on the card currently, one after the other.

There are number of structures defined in the C code. There are some structures which need to be accessed by both host and device memory.

Is there any shared memory location where I can allocate these structures so that BOTH - host (CPU) and device (GPU) - can access it.

If yes, then kindly mention how to allocate and access that memory.

Thanks,
Tanmay Anjaria

2

The memory spaces are different, this is not possible

3

Tanmay Anjaria, please stop posting same questions on every sub-forum.

4

okay…

5

but I posted in 2 subforums ONLY… “every subforum” is a huge number as we have numbr of subforums here…

I’ve NOT posted it to unrelated forums anyway…

I’ve taken the note though… will take care of it…