Databases
StackOverflow2013
Postsdbo
PO
 

Note that there are some explanatory texts on larger screens.

plurals
  1. PO
    primarykey
    data
    text
    <p>You have to be aware where your memory resides. malloc allocates host memory, cudaMalloc allocates memory on the device and returns a pointer to that memory back. However, this pointer is only valid in device functions. </p> <p>What you want could be achived as followed:</p> <pre><code>typedef struct { int width; int height; float* elements; } Matrix; int main void() { int rows, cols, numMat = 2; // These are actually determined at run-time Matrix* data = (Matrix*)malloc(numMat * sizeof(Matrix)); // ... Successfully read from file into "data" ... Matrix* h_data = (Matrix*)malloc(numMat * sizeof(Matrix)); memcpy(h_data, data, numMat * sizeof(Matrix); for (int i=0; i&lt;numMat; i++){ cudaMalloc(&amp;(h_data[i].elements), rows*cols*sizeof(float)); cudaMemcpy(h_data[i].elements, data[i].elements, rows*cols*sizeof(float)), cudaMemcpyHostToDevice); }// matrix data is now on the gpu, now copy the "meta" data to gpu Matrix* d_data; cudaMalloc(&amp;d_data, numMat*sizeof(Matrix)); cudaMemcpy(d_data, h_data, numMat*sizeof(Matrix)); // ... Do other things ... } </code></pre> <p>To make things clear: <code>Matrix* data</code> contains the data on the host. <code>Matrix* h_data</code> contains a pointer to the device memory in elements which can be passed to the kernels as parameters. The memory is on the GPU. <code>Matrix* d_data</code> is completly on the GPU and can be used like data on the host.</p> <p>in your kernel code you kann now access the matrix values, e.g.,</p> <pre><code>__global__ void doThings(Matrix* matrices) { matrices[i].elements[0] = 42; } </code></pre>
    singulars
    1. This table or related slice is empty.
    1. POHow to use cudaMalloc / cudaMemcpy for a pointer to a structure containing pointers?
      singulars
      1. PTQuestion
    1. PTAnswer
    1. This table or related slice is empty.
    1. USMichael Haidl
    plurals
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. POHow to use cudaMalloc / cudaMemcpy for a pointer to a structure containing pointers?
      singulars
      1. PTQuestion
    1. This table or related slice is empty.
    1. VO
      singulars
      1. PO
      1. This table or related slice is empty.
      1. VTUpMod
    2. VO
      singulars
      1. PO
      1. This table or related slice is empty.
      1. VTAcceptedByOriginator
    1. COthanks @kronos, I will try this. Is the idea of having the "intermediate pointer" `h_data` the standard way to do this?
      singulars
      1. PO
      1. USt_carn
    2. COWell that depends. It represents an overhead in your host code, because you have to store your other struct data twice. This could be leads to bugs. You can do 2 things: pack the device pointers into an array and pass the array to a kernel (the array must be allocated in device memory as well), or you add a field to your struct that holds the device pointer. With the second oppinion you can use the same structure on host and device side but accessing host data via elements and device data, lets say, via d_elements.
      singulars
      1. PO
      1. USMichael Haidl
    3. COThis is very helpful. It makes me wonder if I need to put anything entirely onto the device at all - if I can just call the kernel with host pointers to device memory. Thanks again!
      singulars
      1. PO
      1. USt_carn
 

Querying!

 
Guidance
A row detail

Detail views are divided into sections. All the information in the data section comes from columns in the selected row. The other sections display data from other, related rows.

Related data can be related in a to-one or a to-many fashion. Captions of data related in a to-many fashion link to a list view showing a filtered view of the table.

Try moving around until you find a non-empty to-many entry and click on the label to get to one. You can move back to the root by clicking on the database name in the header.

SQuiL has stopped working due to an internal error.

If you are curious you may find further information in the browser console, which is accessible through the devtools (F12).

Reload