diff options
| author | Jordan Henderson <jhenderson@hdfgroup.org> | 2017-04-10 14:46:06 (GMT) |
|---|---|---|
| committer | Jordan Henderson <jhenderson@hdfgroup.org> | 2017-04-10 14:46:06 (GMT) |
| commit | ca9b25fd01867dcc0161942f490910761ec66877 (patch) | |
| tree | a189693c32a4b76cdf0c18ecc568ae558e033160 | |
| parent | ccb4e9ed9a5e0af27db9f1504a628e35d7f4cf92 (diff) | |
| parent | 7883c0d01b4eadba84697c7bb3322a304d242ed6 (diff) | |
| download | hdf5-ca9b25fd01867dcc0161942f490910761ec66877.zip hdf5-ca9b25fd01867dcc0161942f490910761ec66877.tar.gz hdf5-ca9b25fd01867dcc0161942f490910761ec66877.tar.bz2 | |
Merge pull request #4 in ~JHENDERSON/hdf5 from feature/parallel_filters to develop
* commit '7883c0d01b4eadba84697c7bb3322a304d242ed6': (38 commits)
Add comments
In current implementation, list must be gathered to all
Correct error with usage of array_gather function
Make array_gather routine more general
Updated comments
Add comments to H5D_filtered_collective_io_info_t data structure
Remove debugging statements
Revert "Merge latest changes from develop"
Merge latest changes from develop
removed erroneous casts from non-rvalue H5MM_free() calls. Replaced assignment of void return values with explicit value of NULL
Fix issue with uninitialized chunk index during chunk re-insertion into index
Remove potential infinite loops caused by goto in error label region
Remove note about src_type_size vs dst_type_size
Trigger error if library tries to break to independent I/O with filters in the pipeline
Fix issue in H5D__mpio_array_gather when no one contributes anything
Iterator for initial chunk entry IO operation should use source datatype size
Fix bug where data doesn't write correctly when MEMSPACE isnt the same as FILESPACE
Check return value of H5S_GET_EXTENT_NPOINTS macro
Fix macro complaint in H5Dmpio.c
Test code for user passing one-chunk-per-process hint to library
...
| -rw-r--r-- | src/H5Dchunk.c | 5 | ||||
| -rw-r--r-- | src/H5Dint.c | 4 | ||||
| -rw-r--r-- | src/H5Dio.c | 12 | ||||
| -rw-r--r-- | src/H5Dmpio.c | 1378 | ||||
| -rw-r--r-- | src/H5Dpkg.h | 5 | ||||
| -rw-r--r-- | src/H5Dscatgath.c | 7 | ||||
| -rw-r--r-- | src/H5err.txt | 2 |
7 files changed, 1348 insertions, 65 deletions
diff --git a/src/H5Dchunk.c b/src/H5Dchunk.c index 63e692f..7e1906b 100644 --- a/src/H5Dchunk.c +++ b/src/H5Dchunk.c @@ -297,9 +297,6 @@ static herr_t H5D__chunk_unlock(const H5D_io_info_t *io_info, static herr_t H5D__chunk_cache_prune(const H5D_t *dset, hid_t dxpl_id, const H5D_dxpl_cache_t *dxpl_cache, size_t size); static herr_t H5D__chunk_prune_fill(H5D_chunk_it_ud1_t *udata, hbool_t new_unfilt_chunk); -static herr_t H5D__chunk_file_alloc(const H5D_chk_idx_info_t *idx_info, - const H5F_block_t *old_chunk, H5F_block_t *new_chunk, hbool_t *need_insert, - hsize_t scaled[]); #ifdef H5_HAVE_PARALLEL static herr_t H5D__chunk_collective_fill(const H5D_t *dset, hid_t dxpl_id, H5D_chunk_coll_info_t *chunk_info, size_t chunk_size, const void *fill_buf); @@ -6240,7 +6237,7 @@ done: * *------------------------------------------------------------------------- */ -static herr_t +herr_t H5D__chunk_file_alloc(const H5D_chk_idx_info_t *idx_info, const H5F_block_t *old_chunk, H5F_block_t *new_chunk, hbool_t *need_insert, hsize_t scaled[]) { diff --git a/src/H5Dint.c b/src/H5Dint.c index 2048e49..c9bf953 100644 --- a/src/H5Dint.c +++ b/src/H5Dint.c @@ -1210,10 +1210,6 @@ H5D__create(H5F_t *file, hid_t type_id, const H5S_t *space, hid_t dcpl_id, /* Don't allow compact datasets to allocate space later */ if(layout->type == H5D_COMPACT && fill->alloc_time != H5D_ALLOC_TIME_EARLY) HGOTO_ERROR(H5E_DATASET, H5E_BADVALUE, NULL, "compact dataset must have early space allocation") - - /* If MPI VFD is used, no filter support yet. */ - if(H5F_HAS_FEATURE(file, H5FD_FEAT_HAS_MPI) && pline->nused > 0) - HGOTO_ERROR(H5E_DATASET, H5E_UNSUPPORTED, NULL, "Parallel I/O does not support filters yet") } /* end if */ /* Set the latest version of the layout, pline & fill messages, if requested */ diff --git a/src/H5Dio.c b/src/H5Dio.c index f5087da..2122cb0 100644 --- a/src/H5Dio.c +++ b/src/H5Dio.c @@ -666,11 +666,6 @@ H5D__write(H5D_t *dataset, hid_t mem_type_id, const H5S_t *mem_space, if(H5T_get_class(type_info.mem_type, TRUE) == H5T_REFERENCE && H5T_get_ref_type(type_info.mem_type) == H5R_DATASET_REGION) HGOTO_ERROR(H5E_DATASET, H5E_UNSUPPORTED, FAIL, "Parallel IO does not support writing region reference datatypes yet") - - /* Can't write to chunked datasets with filters, in parallel */ - if(dataset->shared->layout.type == H5D_CHUNKED && - dataset->shared->dcpl_cache.pline.nused > 0) - HGOTO_ERROR(H5E_IO, H5E_WRITEERROR, FAIL, "cannot write to chunked storage with filters in parallel") } /* end if */ else { /* Collective access is not permissible without a MPI based VFD */ @@ -1159,6 +1154,13 @@ H5D__ioinfo_adjust(H5D_io_info_t *io_info, const H5D_t *dset, hid_t dxpl_id, io_info->io_ops.single_write = H5D__mpio_select_write; } /* end if */ else { + /* Check if there are any filters in the pipeline. If there are, + * we cannot break to independent I/O if this is a write operation; + * otherwise there will be metadata inconsistencies in the file. + */ + if (io_info->op_type == H5D_IO_OP_WRITE && io_info->dset->shared->dcpl_cache.pline.nused > 0) + HGOTO_ERROR(H5E_IO, H5E_NO_INDEPENDENT, FAIL, "can't perform independent write with filters in pipeline") + /* If we won't be doing collective I/O, but the user asked for * collective I/O, change the request to use independent I/O, but * mark it so that we remember to revert the change. diff --git a/src/H5Dmpio.c b/src/H5Dmpio.c index 441cc96..e6fbb6a 100644 --- a/src/H5Dmpio.c +++ b/src/H5Dmpio.c @@ -86,7 +86,6 @@ #define H5D_CHUNK_SELECT_IRREG 2 #define H5D_CHUNK_SELECT_NONE 0 - /******************/ /* Local Typedefs */ /******************/ @@ -96,6 +95,17 @@ typedef struct H5D_chunk_addr_info_t { H5D_chunk_info_t chunk_info; } H5D_chunk_addr_info_t; +/* Information about a single chunk when performing collective filtered IO */ +typedef struct H5D_filtered_collective_io_info_t { + H5D_chunk_info_t chunk_info; /* Info about this chunk, such as chunk index and file and memory dataspace */ + H5F_block_t old_chunk; /* The address in the file and size of this chunk before being filtered */ + H5F_block_t new_chunk; /* The address in the file and size of this chunk after being filtered */ + hbool_t full_overwrite; /* Whether or not this chunk is being fully overwritten */ + size_t io_size; /* Size of the I/O to this chunk */ + size_t num_writers; /* Total number of processes writing to this chunk */ + int owner; /* Process which will be writing to this chunk */ + void *buf; /* Chunk data to be written to file/that has been read from file*/ +} H5D_filtered_collective_io_info_t; /********************/ /* Local Prototypes */ @@ -105,9 +115,15 @@ static herr_t H5D__chunk_collective_io(H5D_io_info_t *io_info, static herr_t H5D__multi_chunk_collective_io(H5D_io_info_t *io_info, const H5D_type_info_t *type_info, H5D_chunk_map_t *fm, H5P_genplist_t *dx_plist); +static herr_t H5D__multi_chunk_filtered_collective_io(H5D_io_info_t *io_info, + const H5D_type_info_t *type_info, H5D_chunk_map_t *fm, + H5P_genplist_t *dx_plist); static herr_t H5D__link_chunk_collective_io(H5D_io_info_t *io_info, const H5D_type_info_t *type_info, H5D_chunk_map_t *fm, int sum_chunk, H5P_genplist_t *dx_plist); +static herr_t H5D__link_chunk_filtered_collective_io(H5D_io_info_t *io_info, + const H5D_type_info_t *type_info, H5D_chunk_map_t *fm, + H5P_genplist_t *dx_plist); static herr_t H5D__inter_collective_io(H5D_io_info_t *io_info, const H5D_type_info_t *type_info, const H5S_t *file_space, const H5S_t *mem_space); @@ -126,6 +142,21 @@ static herr_t H5D__mpio_get_min_chunk(const H5D_io_info_t *io_info, const H5D_chunk_map_t *fm, int *min_chunkf); static herr_t H5D__mpio_get_sum_chunk(const H5D_io_info_t *io_info, const H5D_chunk_map_t *fm, int *sum_chunkf); +static herr_t H5D__construct_filtered_io_info_list(const H5D_io_info_t *io_info, + const H5D_type_info_t *type_info, const H5D_chunk_map_t *fm, + H5D_filtered_collective_io_info_t **chunk_list, size_t *num_entries); +static herr_t H5D__mpio_array_gatherv(void *local_array, size_t local_array_num_entries, + size_t array_entry_size, void **gathered_array, size_t *gathered_array_num_entries, + int nprocs, hbool_t allgather, int root, MPI_Comm comm, int (*sort_func)(const void *, const void *)); +static herr_t H5D__mpio_filtered_collective_write_type( + H5D_filtered_collective_io_info_t *chunk_list, size_t num_entries, + MPI_Datatype *new_mem_type, hbool_t *mem_type_derived, + MPI_Datatype *new_file_type, hbool_t *file_type_derived); +static herr_t H5D__filtered_collective_chunk_entry_io(H5D_filtered_collective_io_info_t *chunk_entry, + const H5D_io_info_t *io_info, const H5D_type_info_t *type_info); +static int H5D__cmp_chunk_addr(const void *chunk_addr_info1, const void *chunk_addr_info2); +static int H5D__cmp_filtered_collective_io_info_entry(const void *filtered_collective_io_info_entry1, + const void *filtered_collective_io_info_entry2); /*********************/ @@ -144,7 +175,7 @@ static herr_t H5D__mpio_get_sum_chunk(const H5D_io_info_t *io_info, * Purpose: Checks if an direct I/O transfer is possible between memory and * the file. * - * Return: Sauccess: Non-negative: TRUE or FALSE + * Return: Success: Non-negative: TRUE or FALSE * Failure: Negative * * Programmer: Quincey Koziol @@ -155,11 +186,11 @@ static herr_t H5D__mpio_get_sum_chunk(const H5D_io_info_t *io_info, htri_t H5D__mpio_opt_possible(const H5D_io_info_t *io_info, const H5S_t *file_space, const H5S_t *mem_space, const H5D_type_info_t *type_info, - const H5D_chunk_map_t *fm, H5P_genplist_t *dx_plist) + const H5D_chunk_map_t H5_ATTR_UNUSED *fm, H5P_genplist_t *dx_plist) { int local_cause = 0; /* Local reason(s) for breaking collective mode */ int global_cause = 0; /* Global reason(s) for breaking collective mode */ - htri_t ret_value; /* Return value */ + htri_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_PACKAGE @@ -208,11 +239,6 @@ H5D__mpio_opt_possible(const H5D_io_info_t *io_info, const H5S_t *file_space, * use collective IO will defer until each chunk IO is reached. */ - /* Don't allow collective operations if filters need to be applied */ - if(io_info->dset->shared->layout.type == H5D_CHUNKED && - io_info->dset->shared->dcpl_cache.pline.nused > 0) - local_cause |= H5D_MPIO_FILTERS; - /* Check for independent I/O */ if(local_cause & H5D_MPIO_SET_INDEPENDENT) global_cause = local_cause; @@ -302,6 +328,100 @@ done: /*------------------------------------------------------------------------- + * Function: H5D__mpio_array_gatherv + * + * Purpose: Given arrays by MPI ranks, gathers them into a single array + * which is either gathered to the rank specified by root when + * allgather is false, or is distributed back to all ranks + * when allgather is true. If the sort_func argument is + * specified, the list is sorted before being returned. + * + * If allgather is specified as true, root is ignored. + * + * Return: Non-negative on success/Negative on failure + * + * Programmer: Jordan Henderson + * Sunday, April 9th, 2017 + * + *------------------------------------------------------------------------- + */ +static herr_t +H5D__mpio_array_gatherv(void *local_array, size_t local_array_num_entries, + size_t array_entry_size, void **_gathered_array, size_t *_gathered_array_num_entries, + int nprocs, hbool_t allgather, int root, MPI_Comm comm, int (*sort_func)(const void *, const void *)) +{ + size_t gathered_array_num_entries = 0; /* The size of the newly-constructed array */ + size_t i; + void *gathered_array = NULL; /* The newly-constructed array returned to the caller */ + int *receive_counts_array = NULL; /* Array containing number of entries each process is contributing */ + int *displacements_array = NULL; /* Array of displacements where each process places its data in the final array */ + int mpi_code; + int sendcount; + herr_t ret_value = SUCCEED; + + FUNC_ENTER_STATIC + + HDassert(_gathered_array); + HDassert(_gathered_array_num_entries); + + /* Determine the size of the end result array */ + if (MPI_SUCCESS != (mpi_code = MPI_Allreduce(&local_array_num_entries, &gathered_array_num_entries, 1, MPI_INT, MPI_SUM, comm))) + HMPI_GOTO_ERROR(FAIL, "MPI_Allreduce failed", mpi_code) + + /* If 0 entries resulted from the collective operation, no one is writing anything */ + if (gathered_array_num_entries > 0) { + if (NULL == (gathered_array = H5MM_malloc(gathered_array_num_entries * array_entry_size))) + HGOTO_ERROR(H5E_DATASET, H5E_CANTALLOC, FAIL, "couldn't allocate gathered array") + + if (NULL == (receive_counts_array = (int *) H5MM_malloc((size_t) nprocs * sizeof(*receive_counts_array)))) + HGOTO_ERROR(H5E_DATASET, H5E_CANTALLOC, FAIL, "couldn't allocate receive counts array") + + if (NULL == (displacements_array = (int *) H5MM_malloc((size_t) nprocs * sizeof(*displacements_array)))) + HGOTO_ERROR(H5E_DATASET, H5E_CANTALLOC, FAIL, "couldn't allocate receive displacements array") + + /* Inform each process of how many entries each other process is contributing to the resulting array */ + if (MPI_SUCCESS != (mpi_code = MPI_Allgather(&local_array_num_entries, 1, MPI_INT, receive_counts_array, 1, MPI_INT, comm))) + HMPI_GOTO_ERROR(FAIL, "MPI_Allgather failed", mpi_code) + + /* Multiply each receive count by the size of the array entry, since the data is sent as bytes */ + for (i = 0; i < (size_t) nprocs; i++) + H5_CHECKED_ASSIGN(receive_counts_array[i], int, (size_t) receive_counts_array[i] * array_entry_size, size_t); + + /* Set receive buffer offsets for MPI_Allgatherv */ + displacements_array[0] = 0; + for (i = 1; i < (size_t) nprocs; i++) + displacements_array[i] = displacements_array[i - 1] + receive_counts_array[i - 1]; + + H5_CHECKED_ASSIGN(sendcount, int, local_array_num_entries * array_entry_size, size_t); + + if (allgather) { + if (MPI_SUCCESS != (mpi_code = MPI_Allgatherv(local_array, sendcount, MPI_BYTE, + gathered_array, receive_counts_array, displacements_array, MPI_BYTE, comm))) + HMPI_GOTO_ERROR(FAIL, "MPI_Allgatherv failed", mpi_code) + } /* end if */ + else { + if (MPI_SUCCESS != (mpi_code = MPI_Gatherv(local_array, sendcount, MPI_BYTE, + gathered_array, receive_counts_array, displacements_array, MPI_BYTE, root, comm))) + HMPI_GOTO_ERROR(FAIL, "MPI_Allgatherv failed", mpi_code) + } /* end else */ + + if (sort_func) HDqsort(gathered_array, gathered_array_num_entries, array_entry_size, sort_func); + } /* end if */ + + *_gathered_array = gathered_array; + *_gathered_array_num_entries = gathered_array_num_entries; + +done: + if (receive_counts_array) + H5MM_free(receive_counts_array); + if (displacements_array) + H5MM_free(displacements_array); + + FUNC_LEAVE_NOAPI(ret_value) +} /* end H5D__mpio_array_gatherv() */ + + +/*------------------------------------------------------------------------- * Function: H5D__ioinfo_xfer_mode * * Purpose: Switch to between collective & independent MPI I/O @@ -400,7 +520,7 @@ H5D__mpio_get_min_chunk(const H5D_io_info_t *io_info, const H5D_chunk_map_t *fm, FUNC_ENTER_STATIC /* Get the number of chunks to perform I/O on */ - num_chunkf = H5SL_count(fm->sel_chunks); + H5_CHECKED_ASSIGN(num_chunkf, int, H5SL_count(fm->sel_chunks), size_t) /* Determine the minimum # of chunks for all processes */ if(MPI_SUCCESS != (mpi_code = MPI_Allreduce(&num_chunkf, min_chunkf, 1, MPI_INT, MPI_MIN, io_info->comm))) @@ -482,7 +602,7 @@ H5D__contig_collective_read(H5D_io_info_t *io_info, const H5D_type_info_t *type_ HGOTO_ERROR(H5E_IO, H5E_READERROR, FAIL, "couldn't finish shared collective MPI-IO") /* Obtain the data transfer properties */ - if(NULL == (dx_plist = H5I_object(io_info->raw_dxpl_id))) + if(NULL == (dx_plist = (H5P_genplist_t *)H5I_object(io_info->raw_dxpl_id))) HGOTO_ERROR(H5E_ARGS, H5E_BADTYPE, FAIL, "not a data transfer property list") /* Set the actual I/O mode property. internal_collective_io will not break to @@ -529,7 +649,7 @@ H5D__contig_collective_write(H5D_io_info_t *io_info, const H5D_type_info_t *type HGOTO_ERROR(H5E_IO, H5E_WRITEERROR, FAIL, "couldn't finish shared collective MPI-IO") /* Obtain the data transfer properties */ - if(NULL == (dx_plist = H5I_object(io_info->raw_dxpl_id))) + if(NULL == (dx_plist = (H5P_genplist_t *)H5I_object(io_info->raw_dxpl_id))) HGOTO_ERROR(H5E_ARGS, H5E_BADTYPE, FAIL, "not a data transfer property list") /* Set the actual I/O mode property. internal_collective_io will not break to @@ -601,12 +721,13 @@ H5D__chunk_collective_io(H5D_io_info_t *io_info, const H5D_type_info_t *type_inf HDassert(fm); /* Obtain the data transfer properties */ - if(NULL == (dx_plist = H5I_object(io_info->raw_dxpl_id))) + if(NULL == (dx_plist = (H5P_genplist_t *)H5I_object(io_info->raw_dxpl_id))) HGOTO_ERROR(H5E_ARGS, H5E_BADTYPE, FAIL, "not a file access property list") - /* Check the optional property list on what to do with collective chunk IO. */ + /* Check the optional property list for the collective chunk IO optimization option */ if(H5P_get(dx_plist, H5D_XFER_MPIO_CHUNK_OPT_HARD_NAME, &chunk_opt_mode) < 0) HGOTO_ERROR(H5E_IO, H5E_CANTGET, FAIL, "couldn't get chunk optimization option") + if(H5FD_MPIO_CHUNK_ONE_IO == chunk_opt_mode) io_option = H5D_ONE_LINK_CHUNK_IO; /*no opt*/ /* direct request to multi-chunk-io */ @@ -622,13 +743,13 @@ H5D__chunk_collective_io(H5D_io_info_t *io_info, const H5D_type_info_t *type_inf if((mpi_size = H5F_mpi_get_size(io_info->dset->oloc.file)) < 0) HGOTO_ERROR(H5E_IO, H5E_MPI, FAIL, "unable to obtain mpi size") - /* Get the chunk optimization option */ + /* Get the chunk optimization option threshold */ if(H5P_get(dx_plist, H5D_XFER_MPIO_CHUNK_OPT_NUM_NAME, &one_link_chunk_io_threshold) < 0) - HGOTO_ERROR(H5E_IO, H5E_CANTGET, FAIL, "couldn't get chunk optimization option") + HGOTO_ERROR(H5E_IO, H5E_CANTGET, FAIL, "couldn't get chunk optimization option threshold value") /* step 1: choose an IO option */ /* If the average number of chunk per process is greater than a threshold, we will do one link chunked IO. */ - if((unsigned)sum_chunk / mpi_size >= one_link_chunk_io_threshold) + if((unsigned)sum_chunk / (unsigned)mpi_size >= one_link_chunk_io_threshold) io_option = H5D_ONE_LINK_CHUNK_IO_MORE_OPT; #ifdef H5_HAVE_INSTRUMENTED_LIBRARY else @@ -683,19 +804,46 @@ H5D__chunk_collective_io(H5D_io_info_t *io_info, const H5D_type_info_t *type_inf #endif /* step 2: Go ahead to do IO.*/ - if(H5D_ONE_LINK_CHUNK_IO == io_option || H5D_ONE_LINK_CHUNK_IO_MORE_OPT == io_option) { - if(H5D__link_chunk_collective_io(io_info, type_info, fm, sum_chunk, dx_plist) < 0) - HGOTO_ERROR(H5E_IO, H5E_CANTGET, FAIL, "couldn't finish linked chunk MPI-IO") - } /* end if */ - /* direct request to multi-chunk-io */ - else if(H5D_MULTI_CHUNK_IO == io_option) { - if(H5D__multi_chunk_collective_io(io_info, type_info, fm, dx_plist) < 0) - HGOTO_ERROR(H5E_IO, H5E_CANTGET, FAIL, "couldn't finish optimized multiple chunk MPI-IO") - } /* end if */ - else { /* multiple chunk IO via threshold */ - if(H5D__multi_chunk_collective_io(io_info, type_info, fm, dx_plist) < 0) - HGOTO_ERROR(H5E_IO, H5E_CANTGET, FAIL, "couldn't finish optimized multiple chunk MPI-IO") - } /* end else */ + switch (io_option) { + case H5D_ONE_LINK_CHUNK_IO: + case H5D_ONE_LINK_CHUNK_IO_MORE_OPT: + /* Check if there are any filters in the pipeline */ + if(io_info->dset->shared->dcpl_cache.pline.nused > 0) { + /* For now, Multi-chunk IO must be forced for parallel filtered read, + * so that data can be unfiltered as it is received. There is significant + * complexity in unfiltering the data when it is read all at once into a + * single buffer. + */ + if (io_info->op_type == H5D_IO_OP_READ) { + if(H5D__multi_chunk_filtered_collective_io(io_info, type_info, fm, dx_plist) < 0) + HGOTO_ERROR(H5E_IO, H5E_CANTGET, FAIL, "couldn't finish optimized multiple filtered chunk MPI-IO") + } + else { + if(H5D__link_chunk_filtered_collective_io(io_info, type_info, fm, dx_plist) < 0) + HGOTO_ERROR(H5E_IO, H5E_CANTGET, FAIL, "couldn't finish filtered linked chunk MPI-IO") + } + } + else { + /* Perform unfiltered link chunk collective IO */ + if(H5D__link_chunk_collective_io(io_info, type_info, fm, sum_chunk, dx_plist) < 0) + HGOTO_ERROR(H5E_IO, H5E_CANTGET, FAIL, "couldn't finish linked chunk MPI-IO") + } + break; + + case H5D_MULTI_CHUNK_IO: /* direct request to do multi-chunk IO */ + default: /* multiple chunk IO via threshold */ + /* Check if there are any filters in the pipeline */ + if(io_info->dset->shared->dcpl_cache.pline.nused > 0) { + if(H5D__multi_chunk_filtered_collective_io(io_info, type_info, fm, dx_plist) < 0) + HGOTO_ERROR(H5E_IO, H5E_CANTGET, FAIL, "couldn't finish optimized multiple filtered chunk MPI-IO") + } + else { + /* Perform unfiltered multi chunk collective IO */ + if(H5D__multi_chunk_collective_io(io_info, type_info, fm, dx_plist) < 0) + HGOTO_ERROR(H5E_IO, H5E_CANTGET, FAIL, "couldn't finish optimized multiple chunk MPI-IO") + } + break; + } /* end switch */ done: FUNC_LEAVE_NOAPI(ret_value) @@ -1095,6 +1243,226 @@ if(H5DEBUG(D)) /*------------------------------------------------------------------------- + * Function: H5D__link_chunk_filtered_collective_io + * + * Purpose: Routine for one collective IO with one MPI derived datatype + * to link with all filtered chunks + * + * 1. Construct a list of selected chunks in the collective IO + * operation + * A. If any chunk is being written to by more than 1 + * process, the process writing the most data to the + * chunk will take ownership of the chunk (the first + * process seen that is writing the most data becomes + * the new owner in the case of ties) + * 2. If the operation is a write operation + * A. Loop through each chunk in the operation + * I. If this is not a full overwrite of the chunk + * a) Read the chunk from file and pass the chunk + * through the filter pipeline in reverse order + * (Unfilter the chunk) + * II. Update the chunk data with the modifications from + * the owning process + * III. Receive any modification data from other + * processes and update the chunk data with these + * modifications + * IV. Filter the chunk + * B. Contribute the modified chunks to an array gathered + * by all processes which contains the new sizes of + * every chunk modified in the collective IO operation + * C. All processes collectively re-allocate each chunk + * from the gathered array with their new sizes after + * the filter operation + * D. If this process has any chunks selected in the IO + * operation, create an MPI derived type for memory and + * file to write out the process' selected chunks to the + * file + * E. Perform the collective write + * F. All processes collectively re-insert each modified + * chunk from the gathered array into the chunk index + * + * + * Return: Non-negative on success/Negative on failure + * + * Programmer: Jordan Henderson + * Friday, Nov. 4th, 2016 + * + *------------------------------------------------------------------------- + */ +static herr_t +H5D__link_chunk_filtered_collective_io(H5D_io_info_t *io_info, const H5D_type_info_t *type_info, + H5D_chunk_map_t *fm, H5P_genplist_t *dx_plist) +{ + H5D_filtered_collective_io_info_t *chunk_list = NULL; /* The list of chunks being read/written */ + H5D_filtered_collective_io_info_t *collective_chunk_list = NULL; /* The list of chunks used during collective operations */ + H5D_mpio_actual_chunk_opt_mode_t actual_chunk_opt_mode = H5D_MPIO_LINK_CHUNK; /* The actual chunk IO optimization mode */ + H5D_mpio_actual_io_mode_t actual_io_mode = H5D_MPIO_CHUNK_COLLECTIVE; /* The chunk IO mode used (Independent vs Collective) */ + H5D_storage_t ctg_store; /* Chunk storage information as contiguous dataset */ + MPI_Datatype mem_type = MPI_BYTE; + MPI_Datatype file_type = MPI_BYTE; + hbool_t mem_type_is_derived = FALSE; + hbool_t file_type_is_derived = FALSE; + size_t chunk_list_num_entries; + size_t collective_chunk_list_num_entries; + size_t *num_chunks_selected_array = NULL; /* Array of number of chunks selected on each process */ + size_t i; /* Local index variable */ + int mpi_rank, mpi_size, mpi_code; + herr_t ret_value = SUCCEED; + + FUNC_ENTER_STATIC + + /* Obtain the current rank of the process and the number of processes */ + if ((mpi_rank = H5F_mpi_get_rank(io_info->dset->oloc.file)) < 0) + HGOTO_ERROR(H5E_IO, H5E_MPI, FAIL, "unable to obtain mpi rank") + if ((mpi_size = H5F_mpi_get_size(io_info->dset->oloc.file)) < 0) + HGOTO_ERROR(H5E_IO, H5E_MPI, FAIL, "unable to obtain mpi size") + + /* Set the actual-chunk-opt-mode property. */ + if (H5P_set(dx_plist, H5D_MPIO_ACTUAL_CHUNK_OPT_MODE_NAME, &actual_chunk_opt_mode) < 0) + HGOTO_ERROR(H5E_PLIST, H5E_CANTSET, FAIL, "couldn't set actual chunk opt mode property") + + /* Set the actual-io-mode property. + * Link chunk filtered I/O does not break to independent, so can set right away + */ + if (H5P_set(dx_plist, H5D_MPIO_ACTUAL_IO_MODE_NAME, &actual_io_mode) < 0) + HGOTO_ERROR(H5E_PLIST, H5E_CANTSET, FAIL, "couldn't set actual io mode property") + + /* Build a list of selected chunks in the collective io operation */ + if (H5D__construct_filtered_io_info_list(io_info, type_info, fm, &chunk_list, &chunk_list_num_entries) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_CANTINIT, FAIL, "couldn't construct filtered I/O info list") + + if (io_info->op_type == H5D_IO_OP_WRITE) { /* Filtered collective write */ + H5D_chk_idx_info_t index_info; + H5D_chunk_ud_t udata; + hsize_t mpi_buf_count; + + /* Construct chunked index info */ + index_info.f = io_info->dset->oloc.file; + index_info.dxpl_id = io_info->md_dxpl_id; + index_info.pline = &(io_info->dset->shared->dcpl_cache.pline); + index_info.layout = &(io_info->dset->shared->layout.u.chunk); + index_info.storage = &(io_info->dset->shared->layout.storage.u.chunk); + + /* Set up chunk information for insertion to chunk index */ + udata.common.layout = index_info.layout; + udata.common.storage = index_info.storage; + udata.filter_mask = 0; + + /* Iterate through all the chunks in the collective write operation, + * updating each chunk with the data modifications from other processes, + * then re-filtering the chunk. + */ + for (i = 0; i < chunk_list_num_entries; i++) + if (H5D__filtered_collective_chunk_entry_io(&chunk_list[i], io_info, type_info) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_WRITEERROR, FAIL, "couldn't process chunk entry") + + /* Gather the new chunk sizes to all processes for a collective reallocation + * of the chunks in the file. + */ + if (H5D__mpio_array_gatherv(chunk_list, chunk_list_num_entries, sizeof(*chunk_list), + (void **) &collective_chunk_list, &collective_chunk_list_num_entries, mpi_size, + true, 0, io_info->comm, NULL) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_CANTGATHER, FAIL, "couldn't gather new chunk sizes") + + /* Collectively re-allocate the modified chunks (from each process) in the file */ + for (i = 0; i < collective_chunk_list_num_entries; i++) { + hbool_t insert; + + if (H5D__chunk_file_alloc(&index_info, &collective_chunk_list[i].old_chunk, &collective_chunk_list[i].new_chunk, + &insert, collective_chunk_list[i].chunk_info.scaled) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_CANTALLOC, FAIL, "unable to allocate chunk") + } /* end for */ + + if (NULL == (num_chunks_selected_array = (size_t *) H5MM_malloc((size_t) mpi_size * sizeof(*num_chunks_selected_array)))) + HGOTO_ERROR(H5E_DATASET, H5E_CANTALLOC, FAIL, "couldn't allocate num chunks selected array") + + if (MPI_SUCCESS != (mpi_code = MPI_Allgather(&chunk_list_num_entries, 1, MPI_UNSIGNED_LONG_LONG, num_chunks_selected_array, + 1, MPI_UNSIGNED_LONG_LONG, io_info->comm))) + HMPI_GOTO_ERROR(FAIL, "MPI_Allgather failed", mpi_code) + + /* If this process has any chunks selected, create a MPI type for collectively + * writing out the chunks to file. Otherwise, the process contributes to the + * collective write with a none type. + */ + if (chunk_list_num_entries) { + size_t offset; + + /* During the collective re-allocation of chunks in the file, the record for each + * chunk is only updated in the collective array, not in the local copy of chunks on each + * process. However, each process needs the updated chunk records so that they can create + * a MPI type for the collective write that will write to the chunk's possible new locations + * in the file instead of the old ones. This ugly hack seems to be the best solution to + * copy the information back to the local array and avoid having to modify the collective + * write type function in an ugly way so that it will accept the collective array instead + * of the local array. This works correctly because the array gather function guarantees + * that the chunk data in the collective array is ordered in blocks by rank. + */ + for (i = 0, offset = 0; i < (size_t) mpi_rank; i++) + offset += num_chunks_selected_array[i]; + + HDmemcpy(chunk_list, &collective_chunk_list[offset], num_chunks_selected_array[mpi_rank] * sizeof(H5D_filtered_collective_io_info_t)); + + /* Create single MPI type encompassing each selection in the dataspace */ + if (H5D__mpio_filtered_collective_write_type(chunk_list, chunk_list_num_entries, + &mem_type, &mem_type_is_derived, &file_type, &file_type_is_derived) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_BADTYPE, FAIL, "couldn't create MPI link chunk I/O type") + + /* Override the write buffer to point to the address of the first + * chunk data buffer + */ + io_info->u.wbuf = chunk_list[0].buf; + } /* end if */ + + /* We have a single, complicated MPI datatype for both memory & file */ + mpi_buf_count = (mem_type_is_derived && file_type_is_derived) ? (hsize_t) 1 : (hsize_t) 0; + + /* Set up the base storage address for this operation */ + ctg_store.contig.dset_addr = 0; /* Write address must be set to address 0 */ + io_info->store = &ctg_store; + + /* Perform I/O */ + if (H5D__final_collective_io(io_info, type_info, mpi_buf_count, &file_type, &mem_type) < 0) + HGOTO_ERROR(H5E_IO, H5E_CANTGET, FAIL, "couldn't finish MPI-IO") + + /* Participate in the collective re-insertion of all chunks modified + * in this iteration into the chunk index + */ + for (i = 0; i < collective_chunk_list_num_entries; i++) { + udata.chunk_block = collective_chunk_list[i].new_chunk; + udata.common.scaled = collective_chunk_list[i].chunk_info.scaled; + udata.chunk_idx = collective_chunk_list[i].chunk_info.index; + + if ((index_info.storage->ops->insert)(&index_info, &udata, io_info->dset) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_CANTINSERT, FAIL, "unable to insert chunk address into index") + } /* end for */ + } /* end if */ + +done: + /* Free resources used by a process which had some selection */ + if (chunk_list) { + for (i = 0; i < chunk_list_num_entries; i++) + if (chunk_list[i].buf) + H5MM_free(chunk_list[i].buf); + + H5MM_free(chunk_list); + } + + if (num_chunks_selected_array) + H5MM_free(num_chunks_selected_array); + if (collective_chunk_list) + H5MM_free(collective_chunk_list); + + /* Free the MPI buf and file types, if they were derived */ + if (mem_type_is_derived && MPI_SUCCESS != (mpi_code = MPI_Type_free(&mem_type))) + HMPI_DONE_ERROR(FAIL, "MPI_Type_free failed", mpi_code) + if (file_type_is_derived && MPI_SUCCESS != (mpi_code = MPI_Type_free(&file_type))) + HMPI_DONE_ERROR(FAIL, "MPI_Type_free failed", mpi_code) + + FUNC_LEAVE_NOAPI(ret_value) +} /* end H5D__link_chunk_filtered_collective_io() */ + + +/*------------------------------------------------------------------------- * Function: H5D__multi_chunk_collective_io * * Purpose: To do IO per chunk according to IO mode(collective/independent/none) @@ -1227,7 +1595,7 @@ if(H5DEBUG(D)) * to ease switching between to mixed I/O without checking the current * value of the property. You can see the definition in H5Ppublic.h */ - actual_io_mode = actual_io_mode | H5D_MPIO_CHUNK_COLLECTIVE; + actual_io_mode = (H5D_mpio_actual_io_mode_t) (actual_io_mode | H5D_MPIO_CHUNK_COLLECTIVE); } /* end if */ else { @@ -1267,7 +1635,7 @@ if(H5DEBUG(D)) mspace = chunk_info->mspace; /* Update the local variable tracking the dxpl's actual io mode. */ - actual_io_mode = actual_io_mode | H5D_MPIO_CHUNK_INDEPENDENT; + actual_io_mode = (H5D_mpio_actual_io_mode_t) (actual_io_mode | H5D_MPIO_CHUNK_INDEPENDENT); } /* end if */ else { fspace = mspace = NULL; @@ -1308,6 +1676,305 @@ done: /*------------------------------------------------------------------------- + * Function: H5D__multi_chunk_filtered_collective_io + * + * Purpose: To do filtered collective IO per chunk to save on memory, + * as opposed to collective IO of every chunk at once + * + * 1. Construct a list of selected chunks in the collective IO + * operation + * A. If any chunk is being written to by more than 1 + * process, the process writing the most data to the + * chunk will take ownership of the chunk (the first + * process seen that is writing the most data becomes + * the new owner in the case of ties) + * 2. If the operation is a read operation + * A. Loop through each chunk in the operation + * I. Read the chunk from the file + * II. Unfilter the chunk + * III. Scatter the read chunk data to the user's buffer + * 3. If the operation is a write operation + * A. Loop through each chunk in the operation + * I. If this is not a full overwrite of the chunk + * a) Read the chunk from file and pass the chunk + * through the filter pipeline in reverse order + * (Unfilter the chunk) + * II. Update the chunk data with the modifications from + * the owning process + * III. Receive any modification data from other + * processes and update the chunk data with these + * modifications + * IV. Filter the chunk + * V. Contribute the chunk to an array gathered by + * all processes which contains every chunk + * modified in this iteration (up to one chunk + * per process, some processes may not have a + * selection/may have less chunks to work on than + * other processes) + * VI. All processes collectively re-allocate each + * chunk from the gathered array with their new + * sizes after the filter operation + * VII. Proceed with the collective write operation + * for the chunks modified on this iteration + * VIII. All processes collectively re-insert each + * chunk from the gathered array into the chunk + * index + * + * Return: Non-negative on success/Negative on failure + * + * Programmer: Jordan Henderson + * Friday, Dec. 2nd, 2016 + * + *------------------------------------------------------------------------- + */ +static herr_t +H5D__multi_chunk_filtered_collective_io(H5D_io_info_t *io_info, const H5D_type_info_t *type_info, + H5D_chunk_map_t *fm, H5P_genplist_t *dx_plist) +{ + H5D_filtered_collective_io_info_t *chunk_list = NULL; /* The list of chunks being read/written */ + H5D_filtered_collective_io_info_t *collective_chunk_list = NULL; /* The list of chunks used during collective operations */ + H5D_mpio_actual_chunk_opt_mode_t actual_chunk_opt_mode = H5D_MPIO_MULTI_CHUNK; /* The actual chunk IO optimization mode */ + H5D_mpio_actual_io_mode_t actual_io_mode = H5D_MPIO_CHUNK_COLLECTIVE; /* The chunk IO mode used (Independent vs Collective) */ + H5D_storage_t store; /* union of EFL and chunk pointer in file space */ + H5D_io_info_t ctg_io_info; /* Contiguous I/O info object */ + H5D_storage_t ctg_store; /* Chunk storage information as contiguous dataset */ + MPI_Datatype *file_type_array = NULL; + MPI_Datatype *mem_type_array = NULL; + hbool_t *file_type_is_derived_array = NULL; + hbool_t *mem_type_is_derived_array = NULL; + hbool_t *has_chunk_selected_array = NULL; /* Array of whether or not each process is contributing a chunk to each iteration */ + size_t chunk_list_num_entries; + size_t collective_chunk_list_num_entries; + size_t i, j; /* Local index variable */ + int mpi_rank, mpi_size, mpi_code; + herr_t ret_value = SUCCEED; + + FUNC_ENTER_STATIC + + /* Obtain the current rank of the process and the number of processes */ + if ((mpi_rank = H5F_mpi_get_rank(io_info->dset->oloc.file)) < 0) + HGOTO_ERROR(H5E_IO, H5E_MPI, FAIL, "unable to obtain mpi rank") + if ((mpi_size = H5F_mpi_get_size(io_info->dset->oloc.file)) < 0) + HGOTO_ERROR(H5E_IO, H5E_MPI, FAIL, "unable to obtain mpi size") + + /* Set the actual chunk opt mode property */ + if (H5P_set(dx_plist, H5D_MPIO_ACTUAL_CHUNK_OPT_MODE_NAME, &actual_chunk_opt_mode) < 0) + HGOTO_ERROR(H5E_PLIST, H5E_CANTSET, FAIL, "couldn't set actual chunk opt mode property") + + /* Set the actual_io_mode property. + * Multi chunk I/O does not break to independent, so can set right away + */ + if (H5P_set(dx_plist, H5D_MPIO_ACTUAL_IO_MODE_NAME, &actual_io_mode) < 0) + HGOTO_ERROR(H5E_PLIST, H5E_CANTSET, FAIL, "couldn't set actual chunk io mode property") + + /* Build a list of selected chunks in the collective IO operation */ + if (H5D__construct_filtered_io_info_list(io_info, type_info, fm, &chunk_list, &chunk_list_num_entries) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_CANTINIT, FAIL, "couldn't construct filtered I/O info list") + + /* Set up contiguous I/O info object */ + HDmemcpy(&ctg_io_info, io_info, sizeof(ctg_io_info)); + ctg_io_info.store = &ctg_store; + ctg_io_info.layout_ops = *H5D_LOPS_CONTIG; + + /* Initialize temporary contiguous storage info */ + ctg_store.contig.dset_size = (hsize_t) io_info->dset->shared->layout.u.chunk.size; + ctg_store.contig.dset_addr = 0; + + /* Set dataset storage for I/O info */ + io_info->store = &store; + + if (io_info->op_type == H5D_IO_OP_READ) { /* Filtered collective read */ + for (i = 0; i < chunk_list_num_entries; i++) + if (H5D__filtered_collective_chunk_entry_io(&chunk_list[i], io_info, type_info) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_READERROR, FAIL, "couldn't process chunk entry") + } /* end if */ + else { /* Filtered collective write */ + H5D_chk_idx_info_t index_info; + H5D_chunk_ud_t udata; + size_t max_num_chunks; + hsize_t mpi_buf_count; + + /* Construct chunked index info */ + index_info.f = io_info->dset->oloc.file; + index_info.dxpl_id = io_info->md_dxpl_id; + index_info.pline = &(io_info->dset->shared->dcpl_cache.pline); + index_info.layout = &(io_info->dset->shared->layout.u.chunk); + index_info.storage = &(io_info->dset->shared->layout.storage.u.chunk); + + /* Set up chunk information for insertion to chunk index */ + udata.common.layout = index_info.layout; + udata.common.storage = index_info.storage; + udata.filter_mask = 0; + + /* Retrieve the maximum number of chunks being written among all processes */ + if (MPI_SUCCESS != (mpi_code = MPI_Allreduce(&chunk_list_num_entries, &max_num_chunks, + 1, MPI_UNSIGNED_LONG_LONG, MPI_MAX, io_info->comm))) + HMPI_GOTO_ERROR(FAIL, "MPI_Allreduce failed", mpi_code) + + /* If no one is writing anything at all, end the operation */ + if (!(max_num_chunks > 0)) HGOTO_DONE(SUCCEED); + + /* Allocate arrays for storing MPI file and mem types and whether or not the + * types were derived. + */ + if (NULL == (file_type_array = (MPI_Datatype *) H5MM_malloc(max_num_chunks * sizeof(*file_type_array)))) + HGOTO_ERROR(H5E_DATASET, H5E_CANTALLOC, FAIL, "couldn't allocate file type array") + + if (NULL == (file_type_is_derived_array = (hbool_t *) H5MM_calloc(max_num_chunks * sizeof(*file_type_is_derived_array)))) + HGOTO_ERROR(H5E_DATASET, H5E_CANTALLOC, FAIL, "couldn't allocate file type is derived array") + + if (NULL == (mem_type_array = (MPI_Datatype *) H5MM_malloc(max_num_chunks * sizeof(*mem_type_array)))) + HGOTO_ERROR(H5E_DATASET, H5E_CANTALLOC, FAIL, "couldn't allocate mem type array") + + if (NULL == (mem_type_is_derived_array = (hbool_t *) H5MM_calloc(max_num_chunks * sizeof(*mem_type_is_derived_array)))) + HGOTO_ERROR(H5E_DATASET, H5E_CANTALLOC, FAIL, "couldn't allocate mem type is derived array") + + /* Iterate over the max number of chunks among all processes, as this process could + * have no chunks left to work on, but it still needs to participate in the collective + * re-allocation and re-insertion of chunks modified by other processes. + */ + for (i = 0; i < max_num_chunks; i++) { + /* Check if this process has a chunk to work on for this iteration */ + hbool_t have_chunk_to_process = i < chunk_list_num_entries; + + if (have_chunk_to_process) + if (H5D__filtered_collective_chunk_entry_io(&chunk_list[i], io_info, type_info) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_WRITEERROR, FAIL, "couldn't process chunk entry") + + /* Gather the new chunk sizes to all processes for a collective re-allocation + * of the chunks in the file + */ + if (H5D__mpio_array_gatherv(&chunk_list[i], have_chunk_to_process ? 1 : 0, sizeof(*chunk_list), + (void **) &collective_chunk_list, &collective_chunk_list_num_entries, mpi_size, + true, 0, io_info->comm, NULL) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_CANTGATHER, FAIL, "couldn't gather new chunk sizes") + + /* Participate in the collective re-allocation of all chunks modified + * in this iteration. + */ + for (j = 0; j < collective_chunk_list_num_entries; j++) { + hbool_t insert = FALSE; + + if (H5D__chunk_file_alloc(&index_info, &collective_chunk_list[j].old_chunk, &collective_chunk_list[j].new_chunk, + &insert, chunk_list[j].chunk_info.scaled) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_CANTALLOC, FAIL, "unable to allocate chunk") + } /* end for */ + + if (NULL == (has_chunk_selected_array = (hbool_t *) H5MM_malloc((size_t) mpi_size * sizeof(*has_chunk_selected_array)))) + HGOTO_ERROR(H5E_DATASET, H5E_CANTALLOC, FAIL, "couldn't allocate num chunks selected array") + + if (MPI_SUCCESS != (mpi_code = MPI_Allgather(&have_chunk_to_process, 1, MPI_C_BOOL, has_chunk_selected_array, + 1, MPI_C_BOOL, io_info->comm))) + HMPI_GOTO_ERROR(FAIL, "MPI_Allgather failed", mpi_code) + + /* If this process has a chunk to work on, create a MPI type for the + * memory and file for writing out the chunk + */ + if (have_chunk_to_process) { + size_t offset; + int mpi_type_count; + + for (j = 0, offset = 0; j < (size_t) mpi_rank; j++) + offset += has_chunk_selected_array[j]; + + /* Collect the new chunk info back to the local copy, since only the record in the + * collective array gets updated by the chunk re-allocation */ + HDmemcpy(&chunk_list[i].new_chunk, &collective_chunk_list[offset].new_chunk, sizeof(chunk_list[i].new_chunk)); + + H5_CHECKED_ASSIGN(mpi_type_count, int, chunk_list[i].new_chunk.length, hsize_t); + + /* Create MPI memory type for writing to chunk */ + if (MPI_SUCCESS != (mpi_code = MPI_Type_contiguous(mpi_type_count, MPI_BYTE, &mem_type_array[i]))) + HMPI_GOTO_ERROR(FAIL, "MPI_Type_contiguous failed", mpi_code) + if(MPI_SUCCESS != (mpi_code = MPI_Type_commit(&mem_type_array[i]))) + HMPI_GOTO_ERROR(FAIL, "MPI_Type_commit failed", mpi_code) + mem_type_is_derived_array[i] = TRUE; + + /* Create MPI file type for writing to chunk */ + if (MPI_SUCCESS != (mpi_code = MPI_Type_contiguous(mpi_type_count, MPI_BYTE, &file_type_array[i]))) + HMPI_GOTO_ERROR(FAIL, "MPI_Type_contiguous failed", mpi_code) + if(MPI_SUCCESS != (mpi_code = MPI_Type_commit(&file_type_array[i]))) + HMPI_GOTO_ERROR(FAIL, "MPI_Type_commit failed", mpi_code) + file_type_is_derived_array[i] = TRUE; + + mpi_buf_count = 1; + + /* Set up the base storage address for this operation */ + ctg_store.contig.dset_addr = chunk_list[i].new_chunk.offset; + + /* Override the write buffer to point to the address of the + * chunk data buffer + */ + ctg_io_info.u.wbuf = chunk_list[i].buf; + } /* end if */ + else { + mem_type_array[i] = file_type_array[i] = MPI_BYTE; + mpi_buf_count = 0; + } /* end else */ + + /* Perform the I/O */ + if (H5D__final_collective_io(&ctg_io_info, type_info, mpi_buf_count, &file_type_array[i], &mem_type_array[i]) < 0) + HGOTO_ERROR(H5E_IO, H5E_CANTGET, FAIL, "couldn't finish MPI-IO") + + /* Participate in the collective re-insertion of all chunks modified + * in this iteration into the chunk index + */ + for (j = 0; j < collective_chunk_list_num_entries; j++) { + udata.chunk_block = collective_chunk_list[j].new_chunk; + udata.common.scaled = collective_chunk_list[j].chunk_info.scaled; + udata.chunk_idx = collective_chunk_list[j].chunk_info.index; + + if ((index_info.storage->ops->insert)(&index_info, &udata, io_info->dset) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_CANTINSERT, FAIL, "unable to insert chunk address into index") + } /* end for */ + + if (collective_chunk_list){ + H5MM_free(collective_chunk_list); + collective_chunk_list = NULL; + } + if (has_chunk_selected_array){ + H5MM_free(has_chunk_selected_array); + has_chunk_selected_array = NULL; + } + } /* end for */ + + /* Free the MPI file and memory types, if they were derived */ + for (i = 0; i < max_num_chunks; i++) { + if (file_type_is_derived_array[i]) + if (MPI_SUCCESS != (mpi_code = MPI_Type_free(&file_type_array[i]))) + HMPI_DONE_ERROR(FAIL, "MPI_Type_free failed", mpi_code) + + if (mem_type_is_derived_array[i]) + if (MPI_SUCCESS != (mpi_code = MPI_Type_free(&mem_type_array[i]))) + HMPI_DONE_ERROR(FAIL, "MPI_Type_free failed", mpi_code) + } /* end for */ + } /* end else */ + +done: + if (chunk_list) { + for (i = 0; i < chunk_list_num_entries; i++) + if (chunk_list[i].buf) + H5MM_free(chunk_list[i].buf); + + H5MM_free(chunk_list); + } + + if (collective_chunk_list) + H5MM_free(collective_chunk_list); + if (file_type_array) + H5MM_free(file_type_array); + if (mem_type_array) + H5MM_free(mem_type_array); + if (file_type_is_derived_array) + H5MM_free(file_type_is_derived_array); + if (mem_type_is_derived_array) + H5MM_free(mem_type_is_derived_array); + + FUNC_LEAVE_NOAPI(ret_value) +} /* end H5D__multi_chunk_filtered_collective_io() */ + + +/*------------------------------------------------------------------------- * Function: H5D__inter_collective_io * * Purpose: Routine for the shared part of collective IO between multiple chunk @@ -1474,7 +2141,7 @@ if(H5DEBUG(D)) static int H5D__cmp_chunk_addr(const void *chunk_addr_info1, const void *chunk_addr_info2) { - haddr_t addr1, addr2; + haddr_t addr1 = HADDR_UNDEF, addr2 = HADDR_UNDEF; FUNC_ENTER_STATIC_NOERR @@ -1486,6 +2153,36 @@ H5D__cmp_chunk_addr(const void *chunk_addr_info1, const void *chunk_addr_info2) /*------------------------------------------------------------------------- + * Function: H5D__cmp_filtered_collective_io_info_entry + * + * Purpose: Routine to compare filtered collective chunk io info + * entries + * + * Description: Callback for qsort() to compare filtered collective chunk + * io info entries + * + * Return: -1, 0, 1 + * + * Programmer: Jordan Henderson + * Wednesday, Nov. 30th, 2016 + * + *------------------------------------------------------------------------- + */ +static int +H5D__cmp_filtered_collective_io_info_entry(const void *filtered_collective_io_info_entry1, const void *filtered_collective_io_info_entry2) +{ + haddr_t addr1 = HADDR_UNDEF, addr2 = HADDR_UNDEF; + + FUNC_ENTER_STATIC_NOERR + + addr1 = ((const H5D_filtered_collective_io_info_t *) filtered_collective_io_info_entry1)->new_chunk.offset; + addr2 = ((const H5D_filtered_collective_io_info_t *) filtered_collective_io_info_entry2)->new_chunk.offset; + + FUNC_LEAVE_NOAPI(H5F_addr_cmp(addr1, addr2)) +} /* end H5D__cmp_filtered_collective_io_info_entry() */ + + +/*------------------------------------------------------------------------- * Function: H5D__sort_chunk * * Purpose: Routine to sort chunks in increasing order of chunk address @@ -1559,7 +2256,7 @@ if(H5DEBUG(D)) HDfprintf(H5DEBUG(D), "Coming inside H5D_OBTAIN_ALL_CHUNK_ADDR_COL\n"); #endif /* Allocate array for chunk addresses */ - if(NULL == (total_chunk_addr_array = H5MM_malloc(sizeof(haddr_t) * (size_t)fm->layout->u.chunk.nchunks))) + if(NULL == (total_chunk_addr_array = (haddr_t *)H5MM_malloc(sizeof(haddr_t) * (size_t)fm->layout->u.chunk.nchunks))) HGOTO_ERROR(H5E_RESOURCE, H5E_NOSPACE, FAIL, "unable to allocate memory chunk address array") /* Retrieve all the chunk addresses with process 0 */ @@ -1583,7 +2280,7 @@ if(H5DEBUG(D)) /* Iterate over all chunks for this process */ while(chunk_node) { - if(NULL == (chunk_info = H5SL_item(chunk_node))) + if(NULL == (chunk_info = (H5D_chunk_info_t *)H5SL_item(chunk_node))) HGOTO_ERROR(H5E_STORAGE, H5E_CANTGET, FAIL,"couldn't get chunk info from skipped list") if(many_chunk_opt == H5D_OBTAIN_ONE_CHUNK_ADDR_IND) { @@ -1668,7 +2365,7 @@ static herr_t H5D__obtain_mpio_mode(H5D_io_info_t* io_info, H5D_chunk_map_t *fm, H5P_genplist_t *dx_plist, uint8_t assign_io_mode[], haddr_t chunk_addr[]) { - int total_chunks; + size_t total_chunks; unsigned percent_nproc_per_chunk, threshold_nproc_per_chunk; uint8_t* io_mode_info = NULL; uint8_t* recv_io_mode_info = NULL; @@ -1678,7 +2375,8 @@ H5D__obtain_mpio_mode(H5D_io_info_t* io_info, H5D_chunk_map_t *fm, H5D_chunk_info_t* chunk_info; int mpi_size, mpi_rank; MPI_Comm comm; - int ic, root; + int root; + size_t ic; int mpi_code; #ifdef H5_HAVE_INSTRUMENTED_LIBRARY int new_value; @@ -1699,7 +2397,7 @@ H5D__obtain_mpio_mode(H5D_io_info_t* io_info, H5D_chunk_map_t *fm, HGOTO_ERROR(H5E_IO, H5E_MPI, FAIL, "unable to obtain mpi size") /* Setup parameters */ - H5_CHECKED_ASSIGN(total_chunks, int, fm->layout->u.chunk.nchunks, hsize_t); + H5_CHECKED_ASSIGN(total_chunks, size_t, fm->layout->u.chunk.nchunks, hsize_t); if(H5P_get(dx_plist, H5D_XFER_MPIO_CHUNK_OPT_RATIO_NAME, &percent_nproc_per_chunk) < 0) HGOTO_ERROR(H5E_IO, H5E_CANTGET, FAIL, "couldn't get percent nproc per chunk") /* if ratio is 0, perform collective io */ @@ -1711,39 +2409,42 @@ H5D__obtain_mpio_mode(H5D_io_info_t* io_info, H5D_chunk_map_t *fm, HGOTO_DONE(SUCCEED) } /* end if */ - threshold_nproc_per_chunk = mpi_size * percent_nproc_per_chunk/100; + + threshold_nproc_per_chunk = (unsigned)mpi_size * percent_nproc_per_chunk/100; /* Allocate memory */ if(NULL == (io_mode_info = (uint8_t *)H5MM_calloc(total_chunks))) HGOTO_ERROR(H5E_DATASET, H5E_CANTALLOC, FAIL, "couldn't allocate I/O mode info buffer") - if(NULL == (mergebuf = H5MM_malloc((sizeof(haddr_t) + 1) * total_chunks))) + if(NULL == (mergebuf = (uint8_t *)H5MM_malloc((sizeof(haddr_t) + 1) * total_chunks))) HGOTO_ERROR(H5E_DATASET, H5E_CANTALLOC, FAIL, "couldn't allocate mergebuf buffer") tempbuf = mergebuf + total_chunks; if(mpi_rank == root) - if(NULL == (recv_io_mode_info = (uint8_t *)H5MM_malloc(total_chunks * mpi_size))) + if(NULL == (recv_io_mode_info = (uint8_t *)H5MM_malloc(total_chunks * (size_t)mpi_size))) HGOTO_ERROR(H5E_DATASET, H5E_CANTALLOC, FAIL, "couldn't allocate recv I/O mode info buffer") /* Obtain the regularity and selection information for all chunks in this process. */ chunk_node = H5SL_first(fm->sel_chunks); while(chunk_node) { - chunk_info = H5SL_item(chunk_node); + chunk_info = (H5D_chunk_info_t *)H5SL_item(chunk_node); - io_mode_info[chunk_info->index] = H5D_CHUNK_SELECT_REG; /* this chunk is selected and is "regular" */ + io_mode_info[chunk_info->index] = H5D_CHUNK_SELECT_REG; /* this chunk is selected and is "regular" */ chunk_node = H5SL_next(chunk_node); } /* end while */ /* Gather all the information */ - if(MPI_SUCCESS != (mpi_code = MPI_Gather(io_mode_info, total_chunks, MPI_BYTE, recv_io_mode_info, total_chunks, MPI_BYTE, root, comm))) + H5_CHECK_OVERFLOW(total_chunks, size_t, int) + if(MPI_SUCCESS != (mpi_code = MPI_Gather(io_mode_info, (int)total_chunks, MPI_BYTE, + recv_io_mode_info, (int)total_chunks, MPI_BYTE, root, comm))) HMPI_GOTO_ERROR(FAIL, "MPI_Gather failed", mpi_code) /* Calculate the mode for IO(collective, independent or none) at root process */ if(mpi_rank == root) { - int nproc; - int* nproc_per_chunk; + size_t nproc; + unsigned* nproc_per_chunk; /* pre-computing: calculate number of processes and regularity of the selection occupied in each chunk */ - if(NULL == (nproc_per_chunk = (int*)H5MM_calloc(total_chunks * sizeof(int)))) + if(NULL == (nproc_per_chunk = (unsigned*)H5MM_calloc(total_chunks * sizeof(*nproc_per_chunk)))) HGOTO_ERROR(H5E_DATASET, H5E_CANTALLOC, FAIL, "couldn't allocate nproc_per_chunk buffer") /* calculating the chunk address */ @@ -1753,7 +2454,7 @@ H5D__obtain_mpio_mode(H5D_io_info_t* io_info, H5D_chunk_map_t *fm, } /* end if */ /* checking for number of process per chunk and regularity of the selection*/ - for(nproc = 0; nproc < mpi_size; nproc++) { + for(nproc = 0; nproc < (size_t)mpi_size; nproc++) { uint8_t *tmp_recv_io_mode_info = recv_io_mode_info + (nproc * total_chunks); /* Calculate the number of process per chunk and adding irregular selection option */ @@ -1837,5 +2538,586 @@ done: FUNC_LEAVE_NOAPI(ret_value) } /* end H5D__obtain_mpio_mode() */ + + +/*------------------------------------------------------------------------- + * Function: H5D__construct_filtered_io_info_list + * + * Purpose: Constructs a list of entries which contain the necessary + * information for inter-process communication when performing + * collective io on filtered chunks. This list is used by + * every process in operations that must be collectively done + * on every chunk, such as chunk re-allocation, insertion of + * chunks into the chunk index, etc. + * + * Return: Non-negative on success/Negative on failure + * + * Programmer: Jordan Henderson + * Tuesday, January 10th, 2017 + * + *------------------------------------------------------------------------- + */ +static herr_t +H5D__construct_filtered_io_info_list(const H5D_io_info_t *io_info, const H5D_type_info_t *type_info, + const H5D_chunk_map_t *fm, H5D_filtered_collective_io_info_t **chunk_list, size_t *num_entries) +{ + H5D_filtered_collective_io_info_t *local_info_array = NULL; /* The list of initially select chunks for this process */ + H5D_filtered_collective_io_info_t *shared_chunks_info_array = NULL; /* The list of all chunks selected in the operation by all processes */ + H5S_sel_iter_t *mem_iter = NULL; /* Memory iterator for H5D__gather_mem */ + unsigned char *mod_data = NULL; /* Chunk modification data sent by a process to a chunk's owner */ + H5SL_node_t *chunk_node; + MPI_Request *send_requests = NULL; /* Array of MPI_Isend chunk modification data send requests */ + MPI_Status *send_statuses = NULL; /* Array of MPI_Isend chunk modification send statuses */ + hbool_t mem_iter_init = FALSE; + size_t num_send_requests = 0; + size_t num_chunks_selected; + size_t shared_chunks_info_array_num_entries; + size_t i; + int mpi_rank, mpi_size, mpi_code; + herr_t ret_value = SUCCEED; + + FUNC_ENTER_STATIC + + HDassert(io_info); + HDassert(type_info); + HDassert(fm); + HDassert(chunk_list); + HDassert(num_entries); + HDassert(TRUE == H5P_isa_class(io_info->raw_dxpl_id, H5P_DATASET_XFER)); + + if ((mpi_rank = H5F_mpi_get_rank(io_info->dset->oloc.file)) < 0) + HGOTO_ERROR(H5E_IO, H5E_MPI, FAIL, "unable to obtain mpi rank") + if ((mpi_size = H5F_mpi_get_size(io_info->dset->oloc.file)) < 0) + HGOTO_ERROR(H5E_IO, H5E_MPI, FAIL, "unable to obtain mpi size") + + /* Each process builds a local list of the chunks they have selected */ + if ((num_chunks_selected = H5SL_count(fm->sel_chunks))) { + H5D_chunk_info_t *chunk_info; + H5D_chunk_ud_t udata; + hssize_t select_npoints; + hssize_t chunk_npoints; + + if (NULL == (local_info_array = (H5D_filtered_collective_io_info_t *) H5MM_malloc(num_chunks_selected * sizeof(*local_info_array)))) + HGOTO_ERROR(H5E_DATASET, H5E_CANTALLOC, FAIL, "couldn't allocate local io info array buffer") + + chunk_node = H5SL_first(fm->sel_chunks); + for (i = 0; chunk_node; i++) { + chunk_info = (H5D_chunk_info_t *) H5SL_item(chunk_node); + + /* Obtain this chunk's address */ + if (H5D__chunk_lookup(io_info->dset, io_info->md_dxpl_id, chunk_info->scaled, &udata) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_CANTGET, FAIL, "error looking up chunk address") + + local_info_array[i].chunk_info = *chunk_info; + local_info_array[i].old_chunk = local_info_array[i].new_chunk = udata.chunk_block; + local_info_array[i].num_writers = 0; + local_info_array[i].owner = mpi_rank; + local_info_array[i].buf = NULL; + + if ((select_npoints = H5S_GET_SELECT_NPOINTS(chunk_info->mspace)) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_CANTCOUNT, FAIL, "dataspace is invalid") + local_info_array[i].io_size = (size_t) select_npoints * type_info->src_type_size; + + if ((chunk_npoints = H5S_GET_EXTENT_NPOINTS(chunk_info->fspace)) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_CANTCOUNT, FAIL, "dataspace is invalid") + local_info_array[i].full_overwrite = + (local_info_array[i].io_size >= (hsize_t) chunk_npoints * type_info->dst_type_size) ? TRUE : FALSE; + + chunk_node = H5SL_next(chunk_node); + } /* end for */ + } /* end if */ + + /* Redistribute shared chunks to new owners as necessary */ + if (io_info->op_type == H5D_IO_OP_WRITE) { + if (num_chunks_selected) + if (NULL == (send_requests = (MPI_Request *) H5MM_malloc(num_chunks_selected * sizeof(*send_requests)))) + HGOTO_ERROR(H5E_DATASET, H5E_CANTALLOC, FAIL, "couldn't allocate send requests buffer") + + if (NULL == (mem_iter = (H5S_sel_iter_t *) H5MM_malloc(sizeof(H5S_sel_iter_t)))) + HGOTO_ERROR(H5E_DATASET, H5E_CANTALLOC, FAIL, "couldn't allocate memory iterator") + + if (H5D__mpio_array_gatherv(local_info_array, num_chunks_selected, sizeof(*local_info_array), + (void **) &shared_chunks_info_array, &shared_chunks_info_array_num_entries, mpi_size, + true, 0, io_info->comm, H5D__cmp_filtered_collective_io_info_entry) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_CANTGATHER, FAIL, "couldn't gather array") + + for (i = 0, num_chunks_selected = 0, num_send_requests = 0; i < shared_chunks_info_array_num_entries;) { + H5D_filtered_collective_io_info_t chunk_entry; + haddr_t chunk_addr = shared_chunks_info_array[i].old_chunk.offset; + size_t total_io_size = 0; + size_t num_writers = 0; + size_t max_bytes = 0; + int new_owner = 0; + + /* Set the chunk entry's file dataspace to NULL as a sentinel value. + * Any process which is contributing modifications to this chunk will + * obtain a valid file space while processing duplicates below. Any + * process which still has a NULL file space after processing all of + * the duplicate entries for a shared chunk are assumed to not be + * contributing to the chunk and so will not try to access an invalid + * dataspace when processes are sending chunk data to new owners */ + chunk_entry.chunk_info.fspace = NULL; + + /* Process duplicate entries caused by another process writing + * to the same chunk + */ + do { + /* Store the correct chunk entry information in case this process + * becomes the new chunk's owner. The chunk entry that this process + * contributed will be the only one with a valid dataspace selection + * on that particular process + */ + if (mpi_rank == shared_chunks_info_array[i].owner) + chunk_entry = shared_chunks_info_array[i]; + + /* Add this chunk entry's IO size to the running total */ + total_io_size += shared_chunks_info_array[i].io_size; + + /* New owner of the chunk is determined by the process + * which is writing the most data to the chunk + */ + if (shared_chunks_info_array[i].io_size > max_bytes) { + max_bytes = shared_chunks_info_array[i].io_size; + new_owner = shared_chunks_info_array[i].owner; + } + + num_writers++; + } while (++i < shared_chunks_info_array_num_entries && shared_chunks_info_array[i].old_chunk.offset == chunk_addr); + + if (mpi_rank == new_owner) { + hssize_t chunk_npoints; + + /* Make sure the new owner will know how many other processes will + * be sending chunk modification data to it + */ + chunk_entry.num_writers = num_writers; + + /* Set the full chunk overwrite status. It is assumed that this is a full + * overwrite of the chunk if the total IO size is equal to the size of the + * chunk. If the IO size is greater than the size of the chunk, there is an + * overlapping write between processes, meaning there is no guarantee on + * the integrity of data in the write operation. However, this still + * represents a full overwrite of the chunk. + */ + if ((chunk_npoints = H5S_GET_EXTENT_NPOINTS(chunk_entry.chunk_info.fspace)) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_CANTCOUNT, FAIL, "dataspace is invalid") + + chunk_entry.full_overwrite = (total_io_size >= (hsize_t) chunk_npoints * type_info->dst_type_size) ? TRUE : FALSE; + + /* New owner takes possession of the chunk */ + shared_chunks_info_array[num_chunks_selected++] = chunk_entry; + } /* end if */ + else if (chunk_entry.chunk_info.fspace) { + unsigned char *mod_data_p = NULL; /* Use second pointer since H5S_encode advances pointer */ + hssize_t iter_nelmts; /* Number of points to iterate over for the send operation */ + size_t mod_data_size; + + /* Not the new owner of this chunk, encode the file space selection and + * modification data into a buffer and send it to the new chunk owner */ + + /* Determine size of serialized chunk memory dataspace plus the size + * of the data being written + */ + if (H5S_encode(chunk_entry.chunk_info.fspace, &mod_data, &mod_data_size) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_CANTENCODE, FAIL, "unable to get encoded dataspace size") + + if ((iter_nelmts = H5S_GET_SELECT_NPOINTS(chunk_entry.chunk_info.mspace)) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_CANTCOUNT, FAIL, "dataspace is invalid") + + mod_data_size += (size_t) iter_nelmts * type_info->src_type_size; + + if (NULL == (mod_data = (unsigned char *) H5MM_malloc(mod_data_size))) + HGOTO_ERROR(H5E_DATASET, H5E_CANTALLOC, FAIL, "couldn't allocate chunk modification send buffer") + + /* Serialize the chunk's file dataspace into the buffer */ + mod_data_p = mod_data; + if (H5S_encode(chunk_entry.chunk_info.fspace, &mod_data_p, &mod_data_size) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_CANTENCODE, FAIL, "unable to encode dataspace") + + /* Initialize iterator for memory selection */ + if (H5S_select_iter_init(mem_iter, chunk_entry.chunk_info.mspace, type_info->src_type_size) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_CANTINIT, FAIL, "unable to initialize memory selection information") + mem_iter_init = TRUE; + + /* Collect the modification data into the buffer */ + if (!H5D__gather_mem(io_info->u.wbuf, chunk_entry.chunk_info.mspace, mem_iter, + (size_t) iter_nelmts, io_info->dxpl_cache, mod_data_p)) + HGOTO_ERROR(H5E_IO, H5E_WRITEERROR, FAIL, "couldn't gather from write buffer") + + /* Send modification data to new owner */ + H5_CHECK_OVERFLOW(mod_data_size, size_t, int) + H5_CHECK_OVERFLOW(chunk_entry.chunk_info.index, hsize_t, int) + if (MPI_SUCCESS != (mpi_code = MPI_Isend(mod_data, (int) mod_data_size, MPI_BYTE, new_owner, + (int) chunk_entry.chunk_info.index, io_info->comm, &send_requests[num_send_requests++]))) + HMPI_GOTO_ERROR(FAIL, "MPI_Isend failed", mpi_code) + + if (mod_data) { + H5MM_free(mod_data); + mod_data = NULL; + } + if (mem_iter_init && H5S_SELECT_ITER_RELEASE(mem_iter) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_CANTFREE, FAIL, "couldn't release selection iterator") + mem_iter_init = FALSE; + } /* end else */ + } /* end for */ + + /* Rank 0 redistributes any shared chunks to new owners as necessary */ + if (mpi_rank == 0) { + + } + + /* Release old list */ + if (local_info_array) + H5MM_free(local_info_array); + + /* Local info list becomes modified (redistributed) chunk list */ + local_info_array = shared_chunks_info_array; + + /* Now that the chunks have been redistributed, each process must send its modification data + * to the new owners of any of the chunks it previously possessed + */ + for (i = 0; i < num_chunks_selected; i++) { + if (mpi_rank != local_info_array[i].owner) { + + } + } /* end for */ + + /* Wait for all async send requests to complete before returning */ + if (num_send_requests) { + if (NULL == (send_statuses = (MPI_Status *) H5MM_malloc(num_send_requests * sizeof(*send_statuses)))) + HGOTO_ERROR(H5E_DATASET, H5E_CANTALLOC, FAIL, "couldn't allocate send statuses buffer") + + H5_CHECK_OVERFLOW(num_send_requests, size_t, int); + if (MPI_SUCCESS != (mpi_code = MPI_Waitall((int) num_send_requests, send_requests, send_statuses))) + HMPI_GOTO_ERROR(FAIL, "MPI_Waitall failed", mpi_code) + } + } /* end if */ + + *chunk_list = local_info_array; + *num_entries = num_chunks_selected; + +done: + if (send_requests) + H5MM_free(send_requests); + if (send_statuses) + H5MM_free(send_statuses); + if (mod_data) + H5MM_free(mod_data); + if (mem_iter_init && H5S_SELECT_ITER_RELEASE(mem_iter) < 0) + HDONE_ERROR(H5E_DATASET, H5E_CANTFREE, FAIL, "couldn't release selection iterator") + if (mem_iter) + H5MM_free(mem_iter); + + FUNC_LEAVE_NOAPI(ret_value) +} /* end H5D__construct_filtered_io_info_list() */ + + +/*------------------------------------------------------------------------- + * Function: H5D__mpio_filtered_collective_write_type + * + * Purpose: Constructs a MPI derived datatype for both the memory and + * the file for a collective write of filtered chunks. The + * datatype contains the offsets in the file and the locations + * of the filtered chunk data buffers. + * + * Return: Non-negative on success/Negative on failure + * + * Programmer: Jordan Henderson + * Tuesday, November 22, 2016 + * + *------------------------------------------------------------------------- + */ +static herr_t +H5D__mpio_filtered_collective_write_type(H5D_filtered_collective_io_info_t *chunk_list, + size_t num_entries, MPI_Datatype *new_mem_type, hbool_t *mem_type_derived, + MPI_Datatype *new_file_type, hbool_t *file_type_derived) +{ + MPI_Aint *write_buf_array = NULL; /* Relative displacements of filtered chunk data buffers */ + MPI_Aint *file_offset_array = NULL; /* Chunk offsets in the file */ + int *length_array = NULL; /* Filtered Chunk lengths */ + herr_t ret_value = SUCCEED; + + FUNC_ENTER_STATIC + + HDassert(chunk_list); + HDassert(new_mem_type); + HDassert(mem_type_derived); + HDassert(new_file_type); + HDassert(file_type_derived); + + if (num_entries > 0) { + size_t i; + int mpi_code; + void *base_buf; + + H5_CHECK_OVERFLOW(num_entries, size_t, int); + + /* Allocate arrays */ + if (NULL == (length_array = (int *) H5MM_malloc((size_t) num_entries * sizeof(int)))) + HGOTO_ERROR(H5E_RESOURCE, H5E_CANTALLOC, FAIL, "memory allocation failed for filtered collective write length array") + if (NULL == (write_buf_array = (MPI_Aint *) H5MM_malloc((size_t) num_entries * sizeof(MPI_Aint)))) + HGOTO_ERROR(H5E_RESOURCE, H5E_CANTALLOC, FAIL, "memory allocation failed for filtered collective write buf length array") + if (NULL == (file_offset_array = (MPI_Aint *) H5MM_malloc((size_t) num_entries * sizeof(MPI_Aint)))) + HGOTO_ERROR(H5E_RESOURCE, H5E_CANTALLOC, FAIL, "memory allocation failed for collective write offset array") + + /* Ensure the list is sorted in ascending order of offset in the file */ + HDqsort(chunk_list, num_entries, sizeof(*chunk_list), H5D__cmp_filtered_collective_io_info_entry); + + base_buf = chunk_list[0].buf; + for (i = 0; i < num_entries; i++) { + /* Set up the offset in the file, the length of the chunk data, and the relative + * displacement of the chunk data write buffer + */ + file_offset_array[i] = (MPI_Aint) chunk_list[i].new_chunk.offset; + length_array[i] = (int) chunk_list[i].new_chunk.length; + write_buf_array[i] = (MPI_Aint) chunk_list[i].buf - (MPI_Aint) base_buf; + } /* end for */ + + /* Create memory MPI type */ + if (MPI_SUCCESS != (mpi_code = MPI_Type_create_hindexed((int) num_entries, length_array, write_buf_array, MPI_BYTE, new_mem_type))) + HMPI_GOTO_ERROR(FAIL, "MPI_Type_create_hindexed failed", mpi_code) + *mem_type_derived = TRUE; + if (MPI_SUCCESS != (mpi_code = MPI_Type_commit(new_mem_type))) + HMPI_GOTO_ERROR(FAIL, "MPI_Type_commit failed", mpi_code) + + /* Create file MPI type */ + if (MPI_SUCCESS != (mpi_code = MPI_Type_create_hindexed((int) num_entries, length_array, file_offset_array, MPI_BYTE, new_file_type))) + HMPI_GOTO_ERROR(FAIL, "MPI_Type_create_hindexed failed", mpi_code) + *file_type_derived = TRUE; + if (MPI_SUCCESS != (mpi_code = MPI_Type_commit(new_file_type))) + HMPI_GOTO_ERROR(FAIL, "MPI_Type_commit failed", mpi_code) + } /* end if */ + +done: + if (write_buf_array) + H5MM_free(write_buf_array); + if (file_offset_array) + H5MM_free(file_offset_array); + if (length_array) + H5MM_free(length_array); + + FUNC_LEAVE_NOAPI(ret_value) +} /* end H5D__mpio_filtered_collective_write_type() */ + + +/*------------------------------------------------------------------------- + * Function: H5D__filtered_collective_chunk_entry_io + * + * Purpose: Given an entry for a filtered chunk, performs the necessary + * steps for updating the chunk data during a collective + * write, or for reading the chunk from file during a + * collective read. + * + * Return: Non-negative on success/Negative on failure + * + * Programmer: Jordan Henderson + * Wednesday, January 18, 2017 + * + *------------------------------------------------------------------------- + */ +static herr_t +H5D__filtered_collective_chunk_entry_io(H5D_filtered_collective_io_info_t *chunk_entry, + const H5D_io_info_t *io_info, const H5D_type_info_t *type_info) +{ + H5S_sel_iter_t *mem_iter = NULL; /* Memory iterator for H5D__scatter_mem/H5D__gather_mem */ + unsigned char *mod_data = NULL; /* Chunk modification data sent by a process to a chunk's owner */ + unsigned filter_mask = 0; + hssize_t iter_nelmts; /* Number of points to iterate over for the chunk IO operation */ + hbool_t mem_iter_init = FALSE; + size_t buf_size; + size_t mod_data_alloced_bytes = 0; + H5S_t *dataspace = NULL; /* Other process' dataspace for the chunk */ + void *tmp_gath_buf = NULL; /* Temporary gather buffer for owner of the chunk to gather into from + application write buffer before scattering out to the chunk data buffer */ + herr_t ret_value = SUCCEED; + + FUNC_ENTER_STATIC + + HDassert(chunk_entry); + HDassert(io_info); + HDassert(type_info); + + /* If this is a read operation or a write operation where the chunk is not being fully + * overwritten, enough memory must be allocated to read the filtered chunk from the file. + * If this is a write operation where the chunk is being fully overwritten, enough memory + * must be allocated for the size of the unfiltered chunk. + */ + if (!chunk_entry->full_overwrite || io_info->op_type == H5D_IO_OP_READ) { + buf_size = chunk_entry->old_chunk.length; + } + else { + hssize_t extent_npoints; + + if ((extent_npoints = H5S_GET_EXTENT_NPOINTS(chunk_entry->chunk_info.fspace)) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_CANTCOUNT, FAIL, "dataspace is invalid") + + buf_size = (hsize_t) extent_npoints * type_info->src_type_size; + } + + chunk_entry->new_chunk.length = buf_size; + + if (NULL == (chunk_entry->buf = H5MM_malloc(buf_size))) + HGOTO_ERROR(H5E_DATASET, H5E_CANTALLOC, FAIL, "couldn't allocate chunk data buffer") + + /* If this is not a full chunk overwrite or this is a read operation, the chunk must be + * read from the file and unfiltered. + */ + if (!chunk_entry->full_overwrite || io_info->op_type == H5D_IO_OP_READ) { + /* XXX: Test with MPI types and collective read to improve performance */ + if (H5F_block_read(io_info->dset->oloc.file, H5FD_MEM_DRAW, chunk_entry->old_chunk.offset, + buf_size, H5AC_rawdata_dxpl_id, chunk_entry->buf) < 0) + HGOTO_ERROR(H5E_IO, H5E_READERROR, FAIL, "unable to read raw data chunk") + + if (H5Z_pipeline(&io_info->dset->shared->dcpl_cache.pline, H5Z_FLAG_REVERSE, &filter_mask, + io_info->dxpl_cache->err_detect, io_info->dxpl_cache->filter_cb, + (size_t *) &chunk_entry->new_chunk.length, &buf_size, &chunk_entry->buf) < 0) + HGOTO_ERROR(H5E_PLINE, H5E_CANTFILTER, FAIL, "couldn't unfilter chunk for modifying") + } /* end if */ + + /* Initialize iterator for memory selection */ + if (NULL == (mem_iter = (H5S_sel_iter_t *) H5MM_malloc(sizeof(*mem_iter)))) + HGOTO_ERROR(H5E_DATASET, H5E_CANTALLOC, FAIL, "couldn't allocate memory iterator") + + if (H5S_select_iter_init(mem_iter, chunk_entry->chunk_info.mspace, type_info->src_type_size) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_CANTINIT, FAIL, "unable to initialize memory selection information") + mem_iter_init = TRUE; + + if ((iter_nelmts = H5S_GET_SELECT_NPOINTS(chunk_entry->chunk_info.mspace)) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_CANTCOUNT, FAIL, "dataspace is invalid") + + /* If this is a read operation, scatter the read chunk data to the user's buffer. + * + * If this is a write operation, update the chunk data buffer with the modifications + * from the current process, then apply any modifications from other processes. Finally, + * filter the newly-updated chunk. + */ + switch (io_info->op_type) { + case H5D_IO_OP_READ: + if (H5D__scatter_mem(chunk_entry->buf, chunk_entry->chunk_info.mspace, mem_iter, + (size_t) iter_nelmts, io_info->dxpl_cache, io_info->u.rbuf) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_READERROR, FAIL, "couldn't scatter to read buffer") + break; + + case H5D_IO_OP_WRITE: + if (NULL == (tmp_gath_buf = H5MM_malloc((hsize_t) iter_nelmts * type_info->src_type_size))) + HGOTO_ERROR(H5E_DATASET, H5E_CANTALLOC, FAIL, "couldn't allocate temporary gather buffer") + + /* Gather modification data from the application write buffer into a temporary buffer */ + if (!H5D__gather_mem(io_info->u.wbuf, chunk_entry->chunk_info.mspace, mem_iter, + (size_t) iter_nelmts, io_info->dxpl_cache, tmp_gath_buf)) + HGOTO_ERROR(H5E_IO, H5E_WRITEERROR, FAIL, "couldn't gather from write buffer") + + if (H5S_SELECT_ITER_RELEASE(mem_iter) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_CANTFREE, FAIL, "couldn't release selection iterator") + mem_iter_init = FALSE; + + /* Initialize iterator for file selection */ + if (H5S_select_iter_init(mem_iter, chunk_entry->chunk_info.fspace, type_info->dst_type_size) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_CANTINIT, FAIL, "unable to initialize file selection information") + mem_iter_init = TRUE; + + if ((iter_nelmts = H5S_GET_SELECT_NPOINTS(chunk_entry->chunk_info.fspace)) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_CANTCOUNT, FAIL, "dataspace is invalid") + + /* Scatter the owner's modification data into the chunk data buffer according to + * the file space. + */ + if (H5D__scatter_mem(tmp_gath_buf, chunk_entry->chunk_info.fspace, mem_iter, + (size_t) iter_nelmts, io_info->dxpl_cache, chunk_entry->buf) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_READERROR, FAIL, "couldn't scatter to chunk data buffer") + + if (H5S_SELECT_ITER_RELEASE(mem_iter) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_CANTFREE, FAIL, "couldn't release selection iterator") + mem_iter_init = FALSE; + + /* Update the chunk data with any modifications from other processes */ + while (chunk_entry->num_writers > 1) { + const unsigned char *mod_data_p; /* Use second pointer since H5S_decode advances pointer */ + MPI_Status status; + int count; + int mpi_code; + + /* Probe for the incoming message from another process */ + H5_CHECK_OVERFLOW(chunk_entry->chunk_info.index, hsize_t, int) + if (MPI_SUCCESS != (mpi_code = MPI_Probe(MPI_ANY_SOURCE, (int) chunk_entry->chunk_info.index, + io_info->comm, &status))) + HMPI_GOTO_ERROR(FAIL, "MPI_Probe failed", mpi_code) + + /* Retrieve the message size */ + if (MPI_SUCCESS != (mpi_code = MPI_Get_count(&status, MPI_BYTE, &count))) + HMPI_GOTO_ERROR(FAIL, "MPI_Get_count failed", mpi_code) + + if ((size_t) count > mod_data_alloced_bytes) { + if (NULL == (mod_data = (unsigned char *) H5MM_realloc(mod_data, (size_t) count))) + HGOTO_ERROR(H5E_DATASET, H5E_CANTALLOC, FAIL, "couldn't allocate modification data receive buffer") + + mod_data_alloced_bytes = (size_t) count; + } + + if (MPI_SUCCESS != (mpi_code = MPI_Recv(mod_data, count, MPI_BYTE, MPI_ANY_SOURCE, + (int) chunk_entry->chunk_info.index, io_info->comm, &status))) + HMPI_GOTO_ERROR(FAIL, "MPI_Recv failed", mpi_code) + + /* Decode the process' chunk file dataspace */ + mod_data_p = mod_data; + if (NULL == (dataspace = H5S_decode(&mod_data_p))) + HGOTO_ERROR(H5E_DATASET, H5E_CANTDECODE, FAIL, "unable to decode dataspace") + + if (H5S_select_iter_init(mem_iter, dataspace, type_info->dst_type_size) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_CANTINIT, FAIL, "unable to initialize memory selection information") + mem_iter_init = TRUE; + + if ((iter_nelmts = H5S_GET_SELECT_NPOINTS(dataspace)) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_CANTCOUNT, FAIL, "dataspace is invalid") + + /* Update the chunk data with the received modification data */ + if (H5D__scatter_mem(mod_data_p, dataspace, mem_iter, (size_t) iter_nelmts, + io_info->dxpl_cache, chunk_entry->buf) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_WRITEERROR, FAIL, "couldn't scatter to write buffer") + + chunk_entry->num_writers--; + + if (H5S_SELECT_ITER_RELEASE(mem_iter) < 0) + HGOTO_ERROR(H5E_DATASET, H5E_CANTFREE, FAIL, "couldn't release selection iterator") + mem_iter_init = FALSE; + if (dataspace) { + if (H5S_close(dataspace) < 0) + HGOTO_ERROR(H5E_DATASPACE, H5E_CANTFREE, FAIL, "can't close dataspace") + dataspace = NULL; + } + } /* end while */ + + /* Filter the chunk */ + if (H5Z_pipeline(&io_info->dset->shared->dcpl_cache.pline, 0, &filter_mask, + io_info->dxpl_cache->err_detect, io_info->dxpl_cache->filter_cb, + (size_t *) &chunk_entry->new_chunk.length, &buf_size, &chunk_entry->buf) < 0) + HGOTO_ERROR(H5E_PLINE, H5E_CANTFILTER, H5_ITER_ERROR, "output pipeline failed") + +#if H5_SIZEOF_SIZE_T > 4 + /* Check for the chunk expanding too much to encode in a 32-bit value */ + if (chunk_entry->new_chunk.length > ((size_t) 0xffffffff)) + HGOTO_ERROR(H5E_DATASET, H5E_BADRANGE, FAIL, "chunk too large for 32-bit length") +#endif + + break; + default: + HGOTO_ERROR(H5E_DATASET, H5E_BADVALUE, FAIL, "invalid I/O operation") + } /* end switch */ + +done: + if (mod_data) + H5MM_free(mod_data); + if (tmp_gath_buf) + H5MM_free(tmp_gath_buf); + if (mem_iter_init && H5S_SELECT_ITER_RELEASE(mem_iter) < 0) + HDONE_ERROR(H5E_DATASET, H5E_CANTFREE, FAIL, "couldn't release selection iterator") + if (mem_iter) + H5MM_free(mem_iter); + if (dataspace) + if (H5S_close(dataspace) < 0) + HDONE_ERROR(H5E_DATASPACE, H5E_CANTFREE, FAIL, "can't close dataspace") + + FUNC_LEAVE_NOAPI(ret_value) +} /* end H5D__filtered_collective_chunk_entry_io() */ #endif /* H5_HAVE_PARALLEL */ diff --git a/src/H5Dpkg.h b/src/H5Dpkg.h index f54a9f2..bd58d38 100644 --- a/src/H5Dpkg.h +++ b/src/H5Dpkg.h @@ -618,6 +618,9 @@ H5_DLL herr_t H5D__select_write(const H5D_io_info_t *io_info, H5_DLL herr_t H5D__scatter_mem(const void *_tscat_buf, const H5S_t *space, H5S_sel_iter_t *iter, size_t nelmts, const H5D_dxpl_cache_t *dxpl_cache, void *_buf); +H5_DLL size_t H5D__gather_mem(const void *_buf, + const H5S_t *space, H5S_sel_iter_t *iter, size_t nelmts, + const H5D_dxpl_cache_t *dxpl_cache, void *_tgath_buf/*out*/); H5_DLL herr_t H5D__scatgath_read(const H5D_io_info_t *io_info, const H5D_type_info_t *type_info, hsize_t nelmts, const H5S_t *file_space, const H5S_t *mem_space); @@ -667,6 +670,8 @@ H5_DLL herr_t H5D__chunk_lookup(const H5D_t *dset, hid_t dxpl_id, const hsize_t *scaled, H5D_chunk_ud_t *udata); H5_DLL herr_t H5D__chunk_allocated(H5D_t *dset, hid_t dxpl_id, hsize_t *nbytes); H5_DLL herr_t H5D__chunk_allocate(const H5D_io_info_t *io_info, hbool_t full_overwrite, hsize_t old_dim[]); +H5_DLL herr_t H5D__chunk_file_alloc(const H5D_chk_idx_info_t *idx_info, const H5F_block_t *old_chunk, + H5F_block_t *new_chunk, hbool_t *need_insert, hsize_t scaled[]); H5_DLL herr_t H5D__chunk_update_old_edge_chunks(H5D_t *dset, hid_t dxpl_id, hsize_t old_dim[]); H5_DLL herr_t H5D__chunk_prune_by_extent(H5D_t *dset, hid_t dxpl_id, diff --git a/src/H5Dscatgath.c b/src/H5Dscatgath.c index 55111f0..929feb7 100644 --- a/src/H5Dscatgath.c +++ b/src/H5Dscatgath.c @@ -49,9 +49,6 @@ static herr_t H5D__scatter_file(const H5D_io_info_t *io_info, static size_t H5D__gather_file(const H5D_io_info_t *io_info, const H5S_t *file_space, H5S_sel_iter_t *file_iter, size_t nelmts, void *buf); -static size_t H5D__gather_mem(const void *_buf, - const H5S_t *space, H5S_sel_iter_t *iter, size_t nelmts, - const H5D_dxpl_cache_t *dxpl_cache, void *_tgath_buf/*out*/); static herr_t H5D__compound_opt_read(size_t nelmts, const H5S_t *mem_space, H5S_sel_iter_t *iter, const H5D_dxpl_cache_t *dxpl_cache, const H5D_type_info_t *type_info, void *user_buf/*out*/); @@ -305,6 +302,7 @@ H5D__scatter_mem (const void *_tscat_buf, const H5S_t *space, HDassert(space); HDassert(iter); HDassert(nelmts > 0); + HDassert(dxpl_cache); HDassert(buf); /* Allocate the vector I/O arrays */ @@ -366,7 +364,7 @@ done: * *------------------------------------------------------------------------- */ -static size_t +size_t H5D__gather_mem(const void *_buf, const H5S_t *space, H5S_sel_iter_t *iter, size_t nelmts, const H5D_dxpl_cache_t *dxpl_cache, void *_tgath_buf/*out*/) @@ -389,6 +387,7 @@ H5D__gather_mem(const void *_buf, const H5S_t *space, HDassert(space); HDassert(iter); HDassert(nelmts > 0); + HDassert(dxpl_cache); HDassert(tgath_buf); /* Allocate the vector I/O arrays */ diff --git a/src/H5err.txt b/src/H5err.txt index ffa9315..b2ff3bd 100644 --- a/src/H5err.txt +++ b/src/H5err.txt @@ -245,6 +245,8 @@ MINOR, LINK, H5E_CANTSORT, Can't sort objects MINOR, MPI, H5E_MPI, Some MPI function failed MINOR, MPI, H5E_MPIERRSTR, MPI Error String MINOR, MPI, H5E_CANTRECV, Can't receive data +MINOR, MPI, H5E_CANTGATHER, Can't gather data +MINOR, MPI, H5E_NO_INDEPENDENT, Can't perform independent IO # Heap errors MINOR, HEAP, H5E_CANTRESTORE, Can't restore condition |