summaryrefslogtreecommitdiffstats
path: root/testpar/t_coll_chunk.c
diff options
context:
space:
mode:
Diffstat (limited to 'testpar/t_coll_chunk.c')
-rw-r--r--testpar/t_coll_chunk.c110
1 files changed, 55 insertions, 55 deletions
diff --git a/testpar/t_coll_chunk.c b/testpar/t_coll_chunk.c
index 5fe6207..1476fc7 100644
--- a/testpar/t_coll_chunk.c
+++ b/testpar/t_coll_chunk.c
@@ -35,7 +35,7 @@ static void coll_chunktest(const char* filename,int chunk_factor,int select_fact
/*-------------------------------------------------------------------------
* Function: coll_chunk1
*
- * Purpose: Wrapper to test the collective chunk IO for regular JOINT
+ * Purpose: Wrapper to test the collective chunk IO for regular JOINT
selection with a single chunk
*
* Return: Success: 0
@@ -52,7 +52,7 @@ static void coll_chunktest(const char* filename,int chunk_factor,int select_fact
/* ------------------------------------------------------------------------
* Descriptions for the selection: One big singluar selection inside one chunk
- * Two dimensions,
+ * Two dimensions,
*
* dim1 = SPACE_DIM1(5760)*mpi_size
* dim2 = SPACE_DIM2(3)
@@ -82,7 +82,7 @@ coll_chunk1(void)
/*-------------------------------------------------------------------------
* Function: coll_chunk2
*
- * Purpose: Wrapper to test the collective chunk IO for regular DISJOINT
+ * Purpose: Wrapper to test the collective chunk IO for regular DISJOINT
selection with a single chunk
*
* Return: Success: 0
@@ -99,7 +99,7 @@ coll_chunk1(void)
/* ------------------------------------------------------------------------
* Descriptions for the selection: many disjoint selections inside one chunk
- * Two dimensions,
+ * Two dimensions,
*
* dim1 = SPACE_DIM1*mpi_size(5760)
* dim2 = SPACE_DIM2(3)
@@ -111,7 +111,7 @@ coll_chunk1(void)
* count1 = SPACE_DIM2/stride(3/3 = 1)
* start0 = mpi_rank*SPACE_DIM1
* start1 = 0
- *
+ *
* ------------------------------------------------------------------------
*/
void
@@ -129,7 +129,7 @@ coll_chunk2(void)
/*-------------------------------------------------------------------------
* Function: coll_chunk3
*
- * Purpose: Wrapper to test the collective chunk IO for regular JOINT
+ * Purpose: Wrapper to test the collective chunk IO for regular JOINT
selection with at least number of 2*mpi_size chunks
*
* Return: Success: 0
@@ -158,7 +158,7 @@ coll_chunk2(void)
* count1 = SPACE_DIM2(3)
* start0 = mpi_rank*SPACE_DIM1
* start1 = 0
- *
+ *
* ------------------------------------------------------------------------
*/
@@ -180,7 +180,7 @@ coll_chunk3(void)
/*-------------------------------------------------------------------------
* Function: coll_chunk4
*
- * Purpose: Wrapper to test the collective chunk IO for regular JOINT
+ * Purpose: Wrapper to test the collective chunk IO for regular JOINT
selection with at least number of 2*mpi_size chunks
*
* Return: Success: 0
@@ -209,7 +209,7 @@ coll_chunk3(void)
* count1 = SPACE_DIM2(3)
* start0 = mpi_rank*SPACE_DIM1
* start1 = 0
- *
+ *
* ------------------------------------------------------------------------
*/
@@ -230,7 +230,7 @@ coll_chunk4(void)
/*-------------------------------------------------------------------------
* Function: coll_chunk4
*
- * Purpose: Wrapper to test the collective chunk IO for regular JOINT
+ * Purpose: Wrapper to test the collective chunk IO for regular JOINT
selection with at least number of 2*mpi_size chunks
*
* Return: Success: 0
@@ -259,7 +259,7 @@ coll_chunk4(void)
* count1 = SPACE_DIM2(3)
* start0 = mpi_rank*SPACE_DIM1
* start1 = 0
- *
+ *
* ------------------------------------------------------------------------
*/
@@ -280,7 +280,7 @@ coll_chunk5(void)
/*-------------------------------------------------------------------------
* Function: coll_chunk6
*
- * Purpose: Wrapper to test the collective chunk IO for regular JOINT
+ * Purpose: Wrapper to test the collective chunk IO for regular JOINT
selection with at least number of 2*mpi_size chunks
*
* Return: Success: 0
@@ -309,7 +309,7 @@ coll_chunk5(void)
* count1 = SPACE_DIM2(3)
* start0 = mpi_rank*SPACE_DIM1
* start1 = 0
- *
+ *
* ------------------------------------------------------------------------
*/
@@ -330,7 +330,7 @@ coll_chunk6(void)
/*-------------------------------------------------------------------------
* Function: coll_chunk7
*
- * Purpose: Wrapper to test the collective chunk IO for regular JOINT
+ * Purpose: Wrapper to test the collective chunk IO for regular JOINT
selection with at least number of 2*mpi_size chunks
*
* Return: Success: 0
@@ -359,7 +359,7 @@ coll_chunk6(void)
* count1 = SPACE_DIM2(3)
* start0 = mpi_rank*SPACE_DIM1
* start1 = 0
- *
+ *
* ------------------------------------------------------------------------
*/
@@ -380,7 +380,7 @@ coll_chunk7(void)
/*-------------------------------------------------------------------------
* Function: coll_chunk8
*
- * Purpose: Wrapper to test the collective chunk IO for regular JOINT
+ * Purpose: Wrapper to test the collective chunk IO for regular JOINT
selection with at least number of 2*mpi_size chunks
*
* Return: Success: 0
@@ -409,7 +409,7 @@ coll_chunk7(void)
* count1 = SPACE_DIM2(3)
* start0 = mpi_rank*SPACE_DIM1
* start1 = 0
- *
+ *
* ------------------------------------------------------------------------
*/
@@ -430,7 +430,7 @@ coll_chunk8(void)
/*-------------------------------------------------------------------------
* Function: coll_chunk9
*
- * Purpose: Wrapper to test the collective chunk IO for regular JOINT
+ * Purpose: Wrapper to test the collective chunk IO for regular JOINT
selection with at least number of 2*mpi_size chunks
*
* Return: Success: 0
@@ -459,7 +459,7 @@ coll_chunk8(void)
* count1 = SPACE_DIM2(3)
* start0 = mpi_rank*SPACE_DIM1
* start1 = 0
- *
+ *
* ------------------------------------------------------------------------
*/
@@ -480,7 +480,7 @@ coll_chunk9(void)
/*-------------------------------------------------------------------------
* Function: coll_chunk10
*
- * Purpose: Wrapper to test the collective chunk IO for regular JOINT
+ * Purpose: Wrapper to test the collective chunk IO for regular JOINT
selection with at least number of 2*mpi_size chunks
*
* Return: Success: 0
@@ -509,7 +509,7 @@ coll_chunk9(void)
* count1 = SPACE_DIM2(3)
* start0 = mpi_rank*SPACE_DIM1
* start1 = 0
- *
+ *
* ------------------------------------------------------------------------
*/
@@ -534,7 +534,7 @@ coll_chunk10(void)
* Purpose: The real testing routine for regular selection of collective
chunking storage
testing both write and read,
- If anything fails, it may be read or write. There is no
+ If anything fails, it may be read or write. There is no
separation test between read and write.
*
* Return: Success: 0
@@ -613,7 +613,7 @@ coll_chunktest(const char* filename,
chunk_dims[0] = dims[0]/chunk_factor;
/* to decrease the testing time, maintain bigger chunk size */
-
+
(chunk_factor == 1) ? (chunk_dims[1] = SPACE_DIM2) : (chunk_dims[1] = SPACE_DIM2/2);
status = H5Pset_chunk(crp_plist, 2, chunk_dims);
VRFY((status >= 0),"chunk creation property list succeeded");
@@ -671,8 +671,8 @@ coll_chunktest(const char* filename,
break;
default:
;
- }
-
+ }
+
#ifdef H5_HAVE_INSTRUMENTED_LIBRARY
if(facc_type == FACC_MPIO) {
switch(api_option){
@@ -683,44 +683,44 @@ coll_chunktest(const char* filename,
VRFY((status >= 0),"testing property list inserted succeeded");
break;
case API_MULTI_HARD:
- prop_value = H5D_XFER_COLL_CHUNK_DEF;
+ prop_value = H5D_XFER_COLL_CHUNK_DEF;
status = H5Pinsert(xfer_plist,H5D_XFER_COLL_CHUNK_MULTI_HARD_NAME,H5D_XFER_COLL_CHUNK_SIZE,&prop_value,
NULL,NULL,NULL,NULL,NULL,NULL);
VRFY((status >= 0),"testing property list inserted succeeded");
break;
case API_LINK_TRUE:
- prop_value = H5D_XFER_COLL_CHUNK_DEF;
+ prop_value = H5D_XFER_COLL_CHUNK_DEF;
status = H5Pinsert(xfer_plist,H5D_XFER_COLL_CHUNK_LINK_NUM_TRUE_NAME,H5D_XFER_COLL_CHUNK_SIZE,&prop_value,
NULL,NULL,NULL,NULL,NULL,NULL);
VRFY((status >= 0),"testing property list inserted succeeded");
break;
case API_LINK_FALSE:
- prop_value = H5D_XFER_COLL_CHUNK_DEF;
+ prop_value = H5D_XFER_COLL_CHUNK_DEF;
status = H5Pinsert(xfer_plist,H5D_XFER_COLL_CHUNK_LINK_NUM_FALSE_NAME,H5D_XFER_COLL_CHUNK_SIZE,&prop_value,
NULL,NULL,NULL,NULL,NULL,NULL);
VRFY((status >= 0),"testing property list inserted succeeded");
-
+
break;
case API_MULTI_COLL:
- prop_value = H5D_XFER_COLL_CHUNK_DEF;
+ prop_value = H5D_XFER_COLL_CHUNK_DEF;
status = H5Pinsert(xfer_plist,H5D_XFER_COLL_CHUNK_MULTI_RATIO_COLL_NAME,H5D_XFER_COLL_CHUNK_SIZE,&prop_value,
NULL,NULL,NULL,NULL,NULL,NULL);
VRFY((status >= 0),"testing property list inserted succeeded");
-
+
break;
case API_MULTI_IND:
- prop_value = H5D_XFER_COLL_CHUNK_DEF;
+ prop_value = H5D_XFER_COLL_CHUNK_DEF;
status = H5Pinsert(xfer_plist,H5D_XFER_COLL_CHUNK_MULTI_RATIO_IND_NAME,H5D_XFER_COLL_CHUNK_SIZE,&prop_value,
NULL,NULL,NULL,NULL,NULL,NULL);
VRFY((status >= 0),"testing property list inserted succeeded");
-
+
break;
default:
;
- }
+ }
}
#endif
-
+
/* write data collectively */
status = H5Dwrite(dataset, H5T_NATIVE_INT, H5S_ALL, file_dataspace,
xfer_plist, data_array1);
@@ -761,7 +761,7 @@ coll_chunktest(const char* filename,
break;
default:
;
- }
+ }
}
#endif
@@ -779,7 +779,7 @@ coll_chunktest(const char* filename,
if (data_array1) HDfree(data_array1);
-
+
/* Use collective read to verify the correctness of collective write. */
/* allocate memory for data buffer */
@@ -851,12 +851,12 @@ coll_chunktest(const char* filename,
/* Set up the selection */
static void
-ccslab_set(int mpi_rank,
- int mpi_size,
- hsize_t start[],
+ccslab_set(int mpi_rank,
+ int mpi_size,
+ hsize_t start[],
hsize_t count[],
- hsize_t stride[],
- hsize_t block[],
+ hsize_t stride[],
+ hsize_t block[],
int mode)
{
@@ -905,7 +905,7 @@ ccslab_set(int mpi_rank,
case BYROW_SELECTUNBALANCE:
/* The first one-third of the number of processes only
select top half of the domain, The rest will select the bottom
- half of the domain. */
+ half of the domain. */
block[0] = 1;
count[0] = 2;
@@ -915,9 +915,9 @@ ccslab_set(int mpi_rank,
start[1] = 0;
stride[1] = 1;
if((mpi_rank *3)<(mpi_size*2)) start[0] = mpi_rank;
- else start[0] = 1 + SPACE_DIM1*mpi_size/2 + (mpi_rank-2*mpi_size/3);
+ else start[0] = 1 + SPACE_DIM1*mpi_size/2 + (mpi_rank-2*mpi_size/3);
break;
-
+
case BYROW_SELECTINCHUNK:
/* Each process will only select one chunk */
@@ -959,10 +959,10 @@ ccslab_set(int mpi_rank,
* Assume dimension rank is 2.
*/
static void
-ccdataset_fill(hsize_t start[],
+ccdataset_fill(hsize_t start[],
hsize_t stride[],
- hsize_t count[],
- hsize_t block[],
+ hsize_t count[],
+ hsize_t block[],
DATATYPE * dataset)
{
DATATYPE *dataptr = dataset;
@@ -994,8 +994,8 @@ ccdataset_fill(hsize_t start[],
* Print the first block of the content of the dataset.
*/
static void
-ccdataset_print(hsize_t start[],
- hsize_t block[],
+ccdataset_print(hsize_t start[],
+ hsize_t block[],
DATATYPE * dataset)
{
@@ -1025,11 +1025,11 @@ ccdataset_print(hsize_t start[],
* Print the content of the dataset.
*/
static int
-ccdataset_vrfy(hsize_t start[],
- hsize_t count[],
- hsize_t stride[],
- hsize_t block[],
- DATATYPE *dataset,
+ccdataset_vrfy(hsize_t start[],
+ hsize_t count[],
+ hsize_t stride[],
+ hsize_t block[],
+ DATATYPE *dataset,
DATATYPE *original)
{
hsize_t i, j,k1,k2;