1
1

Temporary commit to fix the ibm tests. I still have to take a deeper look in the datatype black magic ... *just kidding*

This commit was SVN r5325.
Этот коммит содержится в:
George Bosilca 2005-04-13 21:58:56 +00:00
родитель 44373667ce
Коммит 1fac78adb4

Просмотреть файл

@ -583,12 +583,20 @@ ompi_convertor_pack_no_conv_contig( ompi_convertor_t* pConv,
pSrc = pConv->pBaseBuf + pStack[0].disp;
if( 0 == length ) break;
}
/* The logic here should be quite simple. As the data is contiguous we will just copy data
* (we dont have to do any conversion). Then the only thing that is interesting is to
* be sure that the bConverted is the correct displacement. So we can always set the
* stack[1].disp to ZERO and keep the stack[1].disp equal to bConverted (by lower bound) .
*/
pStack[1].disp = 0;
#if 0
/* the number of complete datatypes still to be copied */
pStack[0].count = pConv->count - (pConv->bConverted / pData->size);
/* the amount of data (in bytes) that still have to be done on the last data */
pStack[1].count = pConv->bConverted - pData->size * (pConv->count - pStack[0].count);
pStack[1].disp = pData->size - pStack[1].count;
pStack[0].disp -= (pData->size - pStack[1].count);
#endif
/* update the return value */
*max_data = pConv->bConverted - initial_amount;
*out_size = iov_count;