'long long' type is not supported

Accelerated Computing CUDA CUDA Programming and Performance
1

Hi,

I got the following error message : ‘long long’ type is not supported

Here is the code :

n ^= scramble; return (float)n / (float)0x100000000LL;

So, how can I solve this ?

2

I have never heard of ‘long long’. Could you not just get rid of that 2nd ‘L’?

3

Thanks for your answer,

I have try several options, but none works !!!

In this post : [url=“http://forums.nvidia.com/index.php?showtopic=176212”]http://forums.nvidia.com/index.php?showtopic=176212[/url] I have try several other solutions but I got some errors !!!

NB: long long is a C99 standard, so I think it should be supported (It is on the AMD SDK).

4

OpenCL isn’t C99 and “long long” isn’t part of OpenCL. Anyway, the “long” type of OpenCL satisfies the requirements for “long long” in C99, so you might as well just use it.

See [url=“Scalar Data Types”]http://www.khronos.org/registry/cl/sdk/1.0...rDataTypes.html[/url]

5

OpenCL isn’t C99 and “long long” isn’t part of OpenCL. Anyway, the “long” type of OpenCL satisfies the requirements for “long long” in C99, so you might as well just use it.

See [url=“Scalar Data Types”]http://www.khronos.org/registry/cl/sdk/1.0...rDataTypes.html[/url]