/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * Copyright by The HDF Group. * * Copyright by the Board of Trustees of the University of Illinois. * * All rights reserved. * * * * This file is part of HDF5. The full HDF5 copyright notice, including * * terms governing use, modification, and redistribution, is contained in * * the files COPYING and Copyright.html. COPYING can be found at the root * * of the source code distribution tree; Copyright.html can be found at the * * root level of an installed copy of the electronic HDF5 document set and * * is linked from the top-level documents page. It can also be found at * * http://hdfgroup.org/HDF5/doc/Copyright.html. If you do not have * * access to either file, you may request a copy from help@hdfgroup.org. * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */ /*------------------------------------------------------------------------- * * Created: H5C.c * June 1 2004 * John Mainzer * * Purpose: Functions in this file implement a generic cache for * things which exist on disk, and which may be * unambiguously referenced by their disk addresses. * * The code in this module was initially written in * support of a complete re-write of the metadata cache * in H5AC.c However, other uses for the cache code * suggested themselves, and thus this file was created * in an attempt to support re-use. * * For a detailed overview of the cache, please see the * header comment for H5C_t in H5Cpkg.h. * * Modifications: * * QAK - 11/27/2004 * Switched over to using skip list routines instead of TBBT * routines. * * JRM - 12/15/04 * Added code supporting manual and automatic cache resizing. * See the header for H5C_auto_size_ctl_t in H5Cprivate.h for * an overview. * * Some elements of the automatic cache resize code depend on * the LRU list. Thus if we ever choose to support a new * replacement policy, we will either have to disable those * elements of the auto resize code when running the new * policy, or modify them to make use of similar information * maintained by the new policy code. * *------------------------------------------------------------------------- */ /************************************************************************** * * To Do: * * Code Changes: * * - Remove extra functionality in H5C_flush_single_entry()? * * - Change protect/unprotect to lock/unlock. * * - Change the way the dirty flag is set. Probably pass it in * as a parameter in unprotect & insert. * * - Size should also be passed in as a parameter in insert and * unprotect -- or some other way should be found to advise the * cache of changes in entry size. * * - Flush entries in increasing address order in * H5C_make_space_in_cache(). * * - Also in H5C_make_space_in_cache(), use high and low water marks * to reduce the number of I/O calls. * * - When flushing, attempt to combine contiguous entries to reduce * I/O overhead. Can't do this just yet as some entries are not * contiguous. Do this in parallel only or in serial as well? * * - Create MPI type for dirty objects when flushing in parallel. * * - Now that TBBT routines aren't used, fix nodes in memory to * point directly to the skip list node from the LRU list, eliminating * skip list lookups when evicting objects from the cache. * * Tests: * * - Trim execution time. (This is no longer a major issue with the * shift from the TBBT to a hash table for indexing.) * * - Add random tests. * **************************************************************************/ #define H5C_PACKAGE /*suppress error about including H5Cpkg */ #define H5F_PACKAGE /*suppress error about including H5Fpkg */ #include "H5private.h" /* Generic Functions */ #include "H5Cpkg.h" /* Cache */ #include "H5Dprivate.h" /* Dataset functions */ #include "H5Eprivate.h" /* Error handling */ #include "H5Fpkg.h" /* Files */ #include "H5FDprivate.h" /* File drivers */ #include "H5FLprivate.h" /* Free Lists */ #include "H5Iprivate.h" /* IDs */ #include "H5MMprivate.h" /* Memory management */ #include "H5Pprivate.h" /* Property lists */ #include "H5SLprivate.h" /* Skip lists */ /* * Private file-scope variables. */ /* Declare a free list to manage the H5C_t struct */ H5FL_DEFINE_STATIC(H5C_t); /* * Private file-scope function declarations: */ static herr_t H5C__auto_adjust_cache_size(H5F_t * f, hid_t primary_dxpl_id, hid_t secondary_dxpl_id, hbool_t write_permitted, hbool_t * first_flush_ptr); static herr_t H5C__autoadjust__ageout(H5F_t * f, double hit_rate, enum H5C_resize_status * status_ptr, size_t * new_max_cache_size_ptr, hid_t primary_dxpl_id, hid_t secondary_dxpl_id, hbool_t write_permitted, hbool_t * first_flush_ptr); static herr_t H5C__autoadjust__ageout__cycle_epoch_marker(H5C_t * cache_ptr); static herr_t H5C__autoadjust__ageout__evict_aged_out_entries(H5F_t * f, hid_t primary_dxpl_id, hid_t secondary_dxpl_id, hbool_t write_permitted, hbool_t * first_flush_ptr); static herr_t H5C__autoadjust__ageout__insert_new_marker(H5C_t * cache_ptr); static herr_t H5C__autoadjust__ageout__remove_all_markers(H5C_t * cache_ptr); static herr_t H5C__autoadjust__ageout__remove_excess_markers(H5C_t * cache_ptr); static herr_t H5C__flash_increase_cache_size(H5C_t * cache_ptr, size_t old_entry_size, size_t new_entry_size); static herr_t H5C_flush_single_entry(H5F_t * f, hid_t primary_dxpl_id, hid_t secondary_dxpl_id, const H5C_class_t * type_ptr, haddr_t addr, unsigned flags, hbool_t * first_flush_ptr, hbool_t del_entry_from_slist_on_destroy); static herr_t H5C_flush_invalidate_cache(H5F_t * f, hid_t primary_dxpl_id, hid_t secondary_dxpl_id, unsigned flags); static void * H5C_load_entry(H5F_t * f, hid_t dxpl_id, const H5C_class_t * type, haddr_t addr, void * udata, hbool_t skip_file_checks); static herr_t H5C_make_space_in_cache(H5F_t * f, hid_t primary_dxpl_id, hid_t secondary_dxpl_id, size_t space_needed, hbool_t write_permitted, hbool_t * first_flush_ptr); #if H5C_DO_EXTREME_SANITY_CHECKS static herr_t H5C_validate_lru_list(H5C_t * cache_ptr); static herr_t H5C_verify_not_in_index(H5C_t * cache_ptr, H5C_cache_entry_t * entry_ptr); #endif /* H5C_DO_EXTREME_SANITY_CHECKS */ /**************************************************************************** * * #defines and declarations for epoch marker cache entries. * * As a strategy for automatic cache size reduction, the cache may insert * marker entries in the LRU list at the end of each epoch. These markers * are then used to identify entries that have not been accessed for n * epochs so that they can be evicted from the cache. * ****************************************************************************/ /* Note that H5C__MAX_EPOCH_MARKERS is defined in H5Cpkg.h, not here because * it is needed to dimension arrays in H5C_t. */ #define H5C__EPOCH_MARKER_TYPE H5C__MAX_NUM_TYPE_IDS static void *H5C_epoch_marker_load(H5F_t *f, hid_t dxpl_id, haddr_t addr, void *udata); static herr_t H5C_epoch_marker_flush(H5F_t *f, hid_t dxpl_id, hbool_t dest, haddr_t addr, void *thing, unsigned *flags_ptr); static herr_t H5C_epoch_marker_dest(H5F_t *f, void *thing); static herr_t H5C_epoch_marker_clear(H5F_t *f, void *thing, hbool_t dest); static herr_t H5C_epoch_marker_notify(H5C_notify_action_t action, void *thing); static herr_t H5C_epoch_marker_size(const H5F_t *f, const void *thing, size_t *size_ptr); const H5C_class_t epoch_marker_class = { /* id = */ H5C__EPOCH_MARKER_TYPE, /* load = */ &H5C_epoch_marker_load, /* flush = */ &H5C_epoch_marker_flush, /* dest = */ &H5C_epoch_marker_dest, /* clear = */ &H5C_epoch_marker_clear, /* notify = */&H5C_epoch_marker_notify, /* size = */ &H5C_epoch_marker_size }; /*************************************************************************** * Class functions for H5C__EPOCH_MAKER_TYPE: * * None of these functions should ever be called, so there is no point in * documenting them separately. * JRM - 11/16/04 * ***************************************************************************/ static void * H5C_epoch_marker_load(H5F_t UNUSED * f, hid_t UNUSED dxpl_id, haddr_t UNUSED addr, void UNUSED * udata) { void * ret_value = NULL; /* Return value */ FUNC_ENTER_NOAPI(H5C_epoch_marker_load, NULL) HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, NULL, "called unreachable fcn.") done: FUNC_LEAVE_NOAPI(ret_value) } static herr_t H5C_epoch_marker_flush(H5F_t UNUSED *f, hid_t UNUSED dxpl_id, hbool_t UNUSED dest, haddr_t UNUSED addr, void UNUSED *thing, unsigned UNUSED * flags_ptr) { herr_t ret_value = FAIL; /* Return value */ FUNC_ENTER_NOAPI(H5C_epoch_marker_flush, FAIL) HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "called unreachable fcn.") done: FUNC_LEAVE_NOAPI(ret_value) } static herr_t H5C_epoch_marker_dest(H5F_t UNUSED * f, void UNUSED * thing) { herr_t ret_value = FAIL; /* Return value */ FUNC_ENTER_NOAPI(H5C_epoch_marker_dest, FAIL) HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "called unreachable fcn.") done: FUNC_LEAVE_NOAPI(ret_value) } static herr_t H5C_epoch_marker_clear(H5F_t UNUSED * f, void UNUSED * thing, hbool_t UNUSED dest) { herr_t ret_value = FAIL; /* Return value */ FUNC_ENTER_NOAPI(H5C_epoch_marker_clear, FAIL) HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "called unreachable fcn.") done: FUNC_LEAVE_NOAPI(ret_value) } static herr_t H5C_epoch_marker_notify(H5C_notify_action_t UNUSED action, void UNUSED * thing) { herr_t ret_value = FAIL; /* Return value */ FUNC_ENTER_NOAPI_NOINIT(H5C_epoch_marker_notify) HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "called unreachable fcn.") done: FUNC_LEAVE_NOAPI(ret_value) } static herr_t H5C_epoch_marker_size(const H5F_t UNUSED * f, const void UNUSED * thing, size_t UNUSED * size_ptr) { herr_t ret_value = FAIL; /* Return value */ FUNC_ENTER_NOAPI(H5C_epoch_marker_size, FAIL) HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "called unreachable fcn.") done: FUNC_LEAVE_NOAPI(ret_value) } /*------------------------------------------------------------------------- * Function: H5C_create * * Purpose: Allocate, initialize, and return the address of a new * instance of H5C_t. * * In general, the max_cache_size parameter must be positive, * and the min_clean_size parameter must lie in the closed * interval [0, max_cache_size]. * * The check_write_permitted parameter must either be NULL, * or point to a function of type H5C_write_permitted_func_t. * If it is NULL, the cache will use the write_permitted * flag to determine whether writes are permitted. * * Return: Success: Pointer to the new instance. * * Failure: NULL * * Programmer: John Mainzer * 6/2/04 * * Modifications: * * JRM -- 7/20/04 * Updated for the addition of the hash table. * * JRM -- 10/5/04 * Added call to H5C_reset_cache_hit_rate_stats(). Also * added initialization for cache_is_full flag and for * resize_ctl. * * JRM -- 11/12/04 * Added initialization for the new size_decreased field. * * JRM -- 11/17/04 * Added/updated initialization for the automatic cache * size control data structures. * * JRM -- 6/24/05 * Added support for the new write_permitted field of * the H5C_t structure. * * JRM -- 7/5/05 * Added the new log_flush parameter and supporting code. * * JRM -- 9/21/05 * Added the new aux_ptr parameter and supporting code. * * JRM -- 1/20/06 * Added initialization of the new prefix field in H5C_t. * * JRM -- 3/16/06 * Added initialization for the pinned entry related fields. * * JRM -- 5/31/06 * Added initialization for the trace_file_ptr field. * * JRM -- 8/19/06 * Added initialization for the flush_in_progress field. * * JRM -- 8/25/06 * Added initialization for the slist_len_increase and * slist_size_increase fields. These fields are used * for sanity checking in the flush process, and are not * compiled in unless H5C_DO_SANITY_CHECKS is TRUE. * * JRM -- 3/28/07 * Added initialization for the new is_read_only and * ro_ref_count fields. * * JRM -- 7/27/07 * Added initialization for the new evictions_enabled * field of H5C_t. * * JRM -- 12/31/07 * Added initialization for the new flash cache size increase * related fields of H5C_t. * * JRM -- 11/5/08 * Added initialization for the new clean_index_size and * dirty_index_size fields of H5C_t. * *------------------------------------------------------------------------- */ H5C_t * H5C_create(size_t max_cache_size, size_t min_clean_size, int max_type_id, const char * (* type_name_table_ptr), H5C_write_permitted_func_t check_write_permitted, hbool_t write_permitted, H5C_log_flush_func_t log_flush, void * aux_ptr) { int i; H5C_t * cache_ptr = NULL; H5C_t * ret_value = NULL; /* Return value */ FUNC_ENTER_NOAPI(H5C_create, NULL) HDassert( max_cache_size >= H5C__MIN_MAX_CACHE_SIZE ); HDassert( max_cache_size <= H5C__MAX_MAX_CACHE_SIZE ); HDassert( min_clean_size <= max_cache_size ); HDassert( max_type_id >= 0 ); HDassert( max_type_id < H5C__MAX_NUM_TYPE_IDS ); HDassert( type_name_table_ptr ); HDassert( ( write_permitted == TRUE ) || ( write_permitted == FALSE ) ); for ( i = 0; i <= max_type_id; i++ ) { HDassert( (type_name_table_ptr)[i] ); HDassert( HDstrlen(( type_name_table_ptr)[i]) > 0 ); } if ( NULL == (cache_ptr = H5FL_CALLOC(H5C_t)) ) { HGOTO_ERROR(H5E_RESOURCE, H5E_NOSPACE, NULL, \ "memory allocation failed") } if ( (cache_ptr->slist_ptr = H5SL_create(H5SL_TYPE_HADDR)) == NULL ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTCREATE, NULL, "can't create skip list.") } /* If we get this far, we should succeed. Go ahead and initialize all * the fields. */ cache_ptr->magic = H5C__H5C_T_MAGIC; cache_ptr->flush_in_progress = FALSE; cache_ptr->trace_file_ptr = NULL; cache_ptr->aux_ptr = aux_ptr; cache_ptr->max_type_id = max_type_id; cache_ptr->type_name_table_ptr = type_name_table_ptr; cache_ptr->max_cache_size = max_cache_size; cache_ptr->min_clean_size = min_clean_size; cache_ptr->check_write_permitted = check_write_permitted; cache_ptr->write_permitted = write_permitted; cache_ptr->log_flush = log_flush; cache_ptr->evictions_enabled = TRUE; cache_ptr->index_len = 0; cache_ptr->index_size = (size_t)0; cache_ptr->clean_index_size = (size_t)0; cache_ptr->dirty_index_size = (size_t)0; cache_ptr->slist_len = 0; cache_ptr->slist_size = (size_t)0; #if H5C_DO_SANITY_CHECKS cache_ptr->slist_len_increase = 0; cache_ptr->slist_size_increase = 0; #endif /* H5C_DO_SANITY_CHECKS */ for ( i = 0; i < H5C__HASH_TABLE_LEN; i++ ) { (cache_ptr->index)[i] = NULL; } cache_ptr->pl_len = 0; cache_ptr->pl_size = (size_t)0; cache_ptr->pl_head_ptr = NULL; cache_ptr->pl_tail_ptr = NULL; cache_ptr->pel_len = 0; cache_ptr->pel_size = (size_t)0; cache_ptr->pel_head_ptr = NULL; cache_ptr->pel_tail_ptr = NULL; cache_ptr->LRU_list_len = 0; cache_ptr->LRU_list_size = (size_t)0; cache_ptr->LRU_head_ptr = NULL; cache_ptr->LRU_tail_ptr = NULL; cache_ptr->cLRU_list_len = 0; cache_ptr->cLRU_list_size = (size_t)0; cache_ptr->cLRU_head_ptr = NULL; cache_ptr->cLRU_tail_ptr = NULL; cache_ptr->dLRU_list_len = 0; cache_ptr->dLRU_list_size = (size_t)0; cache_ptr->dLRU_head_ptr = NULL; cache_ptr->dLRU_tail_ptr = NULL; cache_ptr->size_increase_possible = FALSE; cache_ptr->flash_size_increase_possible = FALSE; cache_ptr->flash_size_increase_threshold = 0; cache_ptr->size_decrease_possible = FALSE; cache_ptr->resize_enabled = FALSE; cache_ptr->cache_full = FALSE; cache_ptr->size_decreased = FALSE; (cache_ptr->resize_ctl).version = H5C__CURR_AUTO_SIZE_CTL_VER; (cache_ptr->resize_ctl).rpt_fcn = NULL; (cache_ptr->resize_ctl).set_initial_size = FALSE; (cache_ptr->resize_ctl).initial_size = H5C__DEF_AR_INIT_SIZE; (cache_ptr->resize_ctl).min_clean_fraction = H5C__DEF_AR_MIN_CLEAN_FRAC; (cache_ptr->resize_ctl).max_size = H5C__DEF_AR_MAX_SIZE; (cache_ptr->resize_ctl).min_size = H5C__DEF_AR_MIN_SIZE; (cache_ptr->resize_ctl).epoch_length = H5C__DEF_AR_EPOCH_LENGTH; (cache_ptr->resize_ctl).incr_mode = H5C_incr__off; (cache_ptr->resize_ctl).lower_hr_threshold = H5C__DEF_AR_LOWER_THRESHHOLD; (cache_ptr->resize_ctl).increment = H5C__DEF_AR_INCREMENT; (cache_ptr->resize_ctl).apply_max_increment = TRUE; (cache_ptr->resize_ctl).max_increment = H5C__DEF_AR_MAX_INCREMENT; (cache_ptr->resize_ctl).flash_incr_mode = H5C_flash_incr__off; (cache_ptr->resize_ctl).flash_multiple = 1.0; (cache_ptr->resize_ctl).flash_threshold = 0.25; (cache_ptr->resize_ctl).decr_mode = H5C_decr__off; (cache_ptr->resize_ctl).upper_hr_threshold = H5C__DEF_AR_UPPER_THRESHHOLD; (cache_ptr->resize_ctl).decrement = H5C__DEF_AR_DECREMENT; (cache_ptr->resize_ctl).apply_max_decrement = TRUE; (cache_ptr->resize_ctl).max_decrement = H5C__DEF_AR_MAX_DECREMENT; (cache_ptr->resize_ctl).epochs_before_eviction = H5C__DEF_AR_EPCHS_B4_EVICT; (cache_ptr->resize_ctl).apply_empty_reserve = TRUE; (cache_ptr->resize_ctl).empty_reserve = H5C__DEF_AR_EMPTY_RESERVE; cache_ptr->epoch_markers_active = 0; /* no need to initialize the ring buffer itself */ cache_ptr->epoch_marker_ringbuf_first = 1; cache_ptr->epoch_marker_ringbuf_last = 0; cache_ptr->epoch_marker_ringbuf_size = 0; /* Initialize all epoch marker entries' fields to zero/FALSE/NULL */ HDmemset(cache_ptr->epoch_markers, 0, sizeof(cache_ptr->epoch_markers)); /* Set non-zero/FALSE/NULL fields for epoch markers */ for ( i = 0; i < H5C__MAX_EPOCH_MARKERS; i++ ) { (cache_ptr->epoch_marker_active)[i] = FALSE; #ifndef NDEBUG ((cache_ptr->epoch_markers)[i]).magic = H5C__H5C_CACHE_ENTRY_T_MAGIC; #endif /* NDEBUG */ ((cache_ptr->epoch_markers)[i]).addr = (haddr_t)i; ((cache_ptr->epoch_markers)[i]).type = &epoch_marker_class; } if ( H5C_reset_cache_hit_rate_stats(cache_ptr) != SUCCEED ) { /* this should be impossible... */ HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, NULL, \ "H5C_reset_cache_hit_rate_stats failed.") } H5C_stats__reset(cache_ptr); cache_ptr->skip_file_checks = FALSE; cache_ptr->skip_dxpl_id_checks = FALSE; cache_ptr->prefix[0] = '\0'; /* empty string */ /* Set return value */ ret_value = cache_ptr; done: if ( ret_value == 0 ) { if ( cache_ptr != NULL ) { if ( cache_ptr->slist_ptr != NULL ) H5SL_close(cache_ptr->slist_ptr); cache_ptr->magic = 0; cache_ptr = H5FL_FREE(H5C_t, cache_ptr); } /* end if */ } /* end if */ FUNC_LEAVE_NOAPI(ret_value) } /* H5C_create() */ /*------------------------------------------------------------------------- * Function: H5C_def_auto_resize_rpt_fcn * * Purpose: Print results of a automatic cache resize. * * This function should only be used where HDprintf() behaves * well -- i.e. not on Windows. * * Return: void * * Programmer: John Mainzer * 10/27/04 * * Modifications: * * JRM -- 11/22/04 * Reworked function to adapt it to the addition of the * ageout method of cache size reduction. * * JRM -- 1/19/06 * Updated function for display the new prefix field of * H5C_t in output. * * JRM 12/31/07 * Updated function to handle flash size increases. * *------------------------------------------------------------------------- */ void H5C_def_auto_resize_rpt_fcn(H5C_t * cache_ptr, #ifndef NDEBUG int32_t version, #else /* NDEBUG */ int32_t UNUSED version, #endif /* NDEBUG */ double hit_rate, enum H5C_resize_status status, size_t old_max_cache_size, size_t new_max_cache_size, size_t old_min_clean_size, size_t new_min_clean_size) { HDassert( cache_ptr != NULL ); HDassert( cache_ptr->magic == H5C__H5C_T_MAGIC ); HDassert( version == H5C__CURR_AUTO_RESIZE_RPT_FCN_VER ); switch ( status ) { case in_spec: HDfprintf(stdout, "%sAuto cache resize -- no change. (hit rate = %lf)\n", cache_ptr->prefix, hit_rate); break; case increase: HDassert( hit_rate < (cache_ptr->resize_ctl).lower_hr_threshold ); HDassert( old_max_cache_size < new_max_cache_size ); HDfprintf(stdout, "%sAuto cache resize -- hit rate (%lf) out of bounds low (%6.5lf).\n", cache_ptr->prefix, hit_rate, (cache_ptr->resize_ctl).lower_hr_threshold); HDfprintf(stdout, "%s cache size increased from (%Zu/%Zu) to (%Zu/%Zu).\n", cache_ptr->prefix, old_max_cache_size, old_min_clean_size, new_max_cache_size, new_min_clean_size); break; case flash_increase: HDassert( old_max_cache_size < new_max_cache_size ); HDfprintf(stdout, "%sflash cache resize(%d) -- size threshold = %Zu.\n", cache_ptr->prefix, (int)((cache_ptr->resize_ctl).flash_incr_mode), cache_ptr->flash_size_increase_threshold); HDfprintf(stdout, "%s cache size increased from (%Zu/%Zu) to (%Zu/%Zu).\n", cache_ptr->prefix, old_max_cache_size, old_min_clean_size, new_max_cache_size, new_min_clean_size); break; case decrease: HDassert( old_max_cache_size > new_max_cache_size ); switch ( (cache_ptr->resize_ctl).decr_mode ) { case H5C_decr__off: HDfprintf(stdout, "%sAuto cache resize -- decrease off. HR = %lf\n", cache_ptr->prefix, hit_rate); break; case H5C_decr__threshold: HDassert( hit_rate > (cache_ptr->resize_ctl).upper_hr_threshold ); HDfprintf(stdout, "%sAuto cache resize -- decrease by threshold. HR = %lf > %6.5lf\n", cache_ptr->prefix, hit_rate, (cache_ptr->resize_ctl).upper_hr_threshold); HDfprintf(stdout, "%sout of bounds high (%6.5lf).\n", cache_ptr->prefix, (cache_ptr->resize_ctl).upper_hr_threshold); break; case H5C_decr__age_out: HDfprintf(stdout, "%sAuto cache resize -- decrease by ageout. HR = %lf\n", cache_ptr->prefix, hit_rate); break; case H5C_decr__age_out_with_threshold: HDassert( hit_rate > (cache_ptr->resize_ctl).upper_hr_threshold ); HDfprintf(stdout, "%sAuto cache resize -- decrease by ageout with threshold. HR = %lf > %6.5lf\n", cache_ptr->prefix, hit_rate, (cache_ptr->resize_ctl).upper_hr_threshold); break; default: HDfprintf(stdout, "%sAuto cache resize -- decrease by unknown mode. HR = %lf\n", cache_ptr->prefix, hit_rate); } HDfprintf(stdout, "%s cache size decreased from (%Zu/%Zu) to (%Zu/%Zu).\n", cache_ptr->prefix, old_max_cache_size, old_min_clean_size, new_max_cache_size, new_min_clean_size); break; case at_max_size: HDfprintf(stdout, "%sAuto cache resize -- hit rate (%lf) out of bounds low (%6.5lf).\n", cache_ptr->prefix, hit_rate, (cache_ptr->resize_ctl).lower_hr_threshold); HDfprintf(stdout, "%s cache already at maximum size so no change.\n", cache_ptr->prefix); break; case at_min_size: HDfprintf(stdout, "%sAuto cache resize -- hit rate (%lf) -- can't decrease.\n", cache_ptr->prefix, hit_rate); HDfprintf(stdout, "%s cache already at minimum size.\n", cache_ptr->prefix); break; case increase_disabled: HDfprintf(stdout, "%sAuto cache resize -- increase disabled -- HR = %lf.", cache_ptr->prefix, hit_rate); break; case decrease_disabled: HDfprintf(stdout, "%sAuto cache resize -- decrease disabled -- HR = %lf.\n", cache_ptr->prefix, hit_rate); break; case not_full: HDassert( hit_rate < (cache_ptr->resize_ctl).lower_hr_threshold ); HDfprintf(stdout, "%sAuto cache resize -- hit rate (%lf) out of bounds low (%6.5lf).\n", cache_ptr->prefix, hit_rate, (cache_ptr->resize_ctl).lower_hr_threshold); HDfprintf(stdout, "%s cache not full so no increase in size.\n", cache_ptr->prefix); break; default: HDfprintf(stdout, "%sAuto cache resize -- unknown status code.\n", cache_ptr->prefix); break; } return; } /* H5C_def_auto_resize_rpt_fcn() */ /*------------------------------------------------------------------------- * Function: H5C_dest * * Purpose: Flush all data to disk and destroy the cache. * * This function fails if any object are protected since the * resulting file might not be consistent. * * The primary_dxpl_id and secondary_dxpl_id parameters * specify the dxpl_ids used on the first write occasioned * by the destroy (primary_dxpl_id), and on all subsequent * writes (secondary_dxpl_id). This is useful in the metadata * cache, but may not be needed elsewhere. If so, just use the * same dxpl_id for both parameters. * * Note that *cache_ptr has been freed upon successful return. * * Return: Non-negative on success/Negative on failure * * Programmer: John Mainzer * 6/2/04 * *------------------------------------------------------------------------- */ herr_t H5C_dest(H5F_t * f, hid_t primary_dxpl_id, hid_t secondary_dxpl_id) { H5C_t * cache_ptr = f->shared->cache; herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI(H5C_dest, FAIL) /* Sanity check */ HDassert(cache_ptr); HDassert(cache_ptr->magic == H5C__H5C_T_MAGIC); HDassert(cache_ptr->skip_file_checks || f); /* Flush and invalidate all cache entries */ if(H5C_flush_invalidate_cache(f, primary_dxpl_id, secondary_dxpl_id, H5C__NO_FLAGS_SET) < 0 ) HGOTO_ERROR(H5E_CACHE, H5E_CANTFLUSH, FAIL, "unable to flush cache") if(cache_ptr->slist_ptr != NULL) { H5SL_close(cache_ptr->slist_ptr); cache_ptr->slist_ptr = NULL; } /* end if */ cache_ptr->magic = 0; cache_ptr = H5FL_FREE(H5C_t, cache_ptr); done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C_dest() */ /*------------------------------------------------------------------------- * * Function: H5C_expunge_entry * * Purpose: Use this function to tell the cache to expunge an entry * from the cache without writing it to disk even if it is * dirty. The entry may not be either pinned or protected. * * Return: Non-negative on success/Negative on failure * * Programmer: John Mainzer * 6/29/06 * * Modifications: * * None. * *------------------------------------------------------------------------- */ herr_t H5C_expunge_entry(H5F_t * f, hid_t primary_dxpl_id, hid_t secondary_dxpl_id, const H5C_class_t * type, haddr_t addr, unsigned flags) { H5C_t * cache_ptr; herr_t result; hbool_t first_flush = TRUE; H5C_cache_entry_t * entry_ptr = NULL; herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI(H5C_expunge_entry, FAIL) HDassert( f ); HDassert( f->shared ); cache_ptr = f->shared->cache; HDassert( cache_ptr ); HDassert( cache_ptr->magic == H5C__H5C_T_MAGIC ); HDassert( type ); HDassert( type->clear ); HDassert( type->dest ); HDassert( H5F_addr_defined(addr) ); #if H5C_DO_EXTREME_SANITY_CHECKS if ( H5C_validate_lru_list(cache_ptr) < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "LRU sanity check failed.\n"); } #endif /* H5C_DO_EXTREME_SANITY_CHECKS */ H5C__SEARCH_INDEX(cache_ptr, addr, entry_ptr, FAIL) if ( ( entry_ptr == NULL ) || ( entry_ptr->type != type ) ) { /* the target doesn't exist in the cache, so we are done. */ HGOTO_DONE(SUCCEED) } HDassert( entry_ptr->addr == addr ); HDassert( entry_ptr->type == type ); if ( entry_ptr->is_protected ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTEXPUNGE, FAIL, \ "Target entry is protected.") } if ( entry_ptr->is_pinned ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTEXPUNGE, FAIL, \ "Target entry is pinned.") } /* Pass along 'free file space' flag to cache client */ entry_ptr->free_file_space_on_destroy = ( (flags & H5C__FREE_FILE_SPACE_FLAG) != 0 ); /* If we get this far, call H5C_flush_single_entry() with the * H5C__FLUSH_INVALIDATE_FLAG and the H5C__FLUSH_CLEAR_ONLY_FLAG. * This will clear the entry, and then delete it from the cache. */ result = H5C_flush_single_entry(f, primary_dxpl_id, secondary_dxpl_id, entry_ptr->type, entry_ptr->addr, H5C__FLUSH_INVALIDATE_FLAG | H5C__FLUSH_CLEAR_ONLY_FLAG, &first_flush, TRUE); if ( result < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTEXPUNGE, FAIL, \ "H5C_flush_single_entry() failed.") } done: #if H5C_DO_EXTREME_SANITY_CHECKS if ( H5C_validate_lru_list(cache_ptr) < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "LRU sanity check failed.\n"); } #endif /* H5C_DO_EXTREME_SANITY_CHECKS */ FUNC_LEAVE_NOAPI(ret_value) } /* H5C_expunge_entry() */ /*------------------------------------------------------------------------- * Function: H5C_flush_cache * * Purpose: Flush (and possibly destroy) the entries contained in the * specified cache. * * If the cache contains protected entries, the function will * fail, as protected entries cannot be flushed. However * all unprotected entries should be flushed before the * function returns failure. * * The primary_dxpl_id and secondary_dxpl_id parameters * specify the dxpl_ids used on the first write occasioned * by the flush (primary_dxpl_id), and on all subsequent * writes (secondary_dxpl_id). This is useful in the metadata * cache, but may not be needed elsewhere. If so, just use the * same dxpl_id for both parameters. * * Return: Non-negative on success/Negative on failure or if there was * a request to flush all items and something was protected. * * Programmer: John Mainzer * 6/2/04 * * Modifications: * * JRM -- 7/20/04 * Modified the function for the addition of the hash table. * * JRM -- 11/22/04 * Added code to remove all epoch markers (if any) from the * LRU list before a destroy. Strictly speaking, this isn't * necessary, as the marker entries reside only in the LRU * list, never in the index or in the tree. However, it * never hurts to tidy up. * * JRM -- 1/6/05 * Reworked code to support the new * H5C__FLUSH_MARKED_ENTRIES_FLAG, and for the replacement of * H5F_FLUSH_INVALIDATE flag with H5C__FLUSH_INVALIDATE_FLAG. * * Note that the H5C__FLUSH_INVALIDATE_FLAG takes precidence * over the H5C__FLUSH_MARKED_ENTRIES_FLAG. Thus if both are * set, the functions behaves as if just the * H5C__FLUSH_INVALIDATE_FLAG was set. * * The H5C__FLUSH_CLEAR_ONLY_FLAG flag can co-exist with * either the H5C__FLUSH_MARKED_ENTRIES_FLAG, or the * H5C__FLUSH_INVALIDATE_FLAG. In all cases, it is simply * passed along to H5C_flush_single_entry(). In the case of * H5C__FLUSH_MARKED_ENTRIES_FLAG, it will only apply to * the marked entries. * * JRM -- 10/15/05 * Added code supporting the new * H5C__FLUSH_IGNORE_PROTECTED_FLAG. We need this flag, as * we now use this function to flush large number of entries * in increasing address order. We do this by marking the * entries to be flushed, calling this function to flush them, * and then restoring LRU order. * * However, it is possible that the cache will contain other, * unmarked protected entries, when we make this call. This * new flag allows us to ignore them. * * Note that even with this flag set, it is still an error * to try to flush a protected entry. * * JRM -- 3/25/06 * Updated function to handle pinned entries. * * JRM -- 8/19/06 * Added code managing the new flush_in_progress field of * H5C_t. * * Also reworked function to allow for the possibility that * entries will be dirtied, resized, or moved during flush * callbacks. As a result, we may have to make multiple * passes through the skip list before the cache is flushed. * * JRM -- 10/13/07 * Added code to detect and manage the case in which a * flush callback changes the s-list out from under * the function. The only way I can think of in which this * can happen is if a flush function loads an entry * into the cache that isn't there already. Quincey tells * me that this will never happen, but I'm not sure I * believe him. * * Note that this is a pretty bad scenario if it ever * happens. The code I have added should allow us to * handle the situation under all but the worst conditions, * but one can argue that I should just scream and die if I * ever detect the condidtion. * *------------------------------------------------------------------------- */ herr_t H5C_flush_cache(H5F_t *f, hid_t primary_dxpl_id, hid_t secondary_dxpl_id, unsigned flags) { H5C_t * cache_ptr = f->shared->cache; herr_t status; herr_t ret_value = SUCCEED; hbool_t destroy; hbool_t flushed_entries_last_pass; hbool_t flush_marked_entries; hbool_t first_flush = TRUE; hbool_t ignore_protected; hbool_t tried_to_flush_protected_entry = FALSE; int32_t passes = 0; int32_t protected_entries = 0; H5SL_node_t * node_ptr = NULL; H5C_cache_entry_t * entry_ptr = NULL; H5C_cache_entry_t * next_entry_ptr = NULL; #if H5C_DO_SANITY_CHECKS int64_t flushed_entries_count; size_t flushed_entries_size; int64_t initial_slist_len; size_t initial_slist_size; #endif /* H5C_DO_SANITY_CHECKS */ FUNC_ENTER_NOAPI(H5C_flush_cache, FAIL) HDassert( cache_ptr ); HDassert( cache_ptr->magic == H5C__H5C_T_MAGIC ); HDassert( cache_ptr->skip_file_checks || f); HDassert( cache_ptr->slist_ptr ); ignore_protected = ( (flags & H5C__FLUSH_IGNORE_PROTECTED_FLAG) != 0 ); destroy = ( (flags & H5C__FLUSH_INVALIDATE_FLAG) != 0 ); /* note that flush_marked_entries is set to FALSE if destroy is TRUE */ flush_marked_entries = ( ( (flags & H5C__FLUSH_MARKED_ENTRIES_FLAG) != 0 ) && ( ! destroy ) ); HDassert( ! ( destroy && ignore_protected ) ); HDassert( ! ( cache_ptr->flush_in_progress ) ); cache_ptr->flush_in_progress = TRUE; if ( destroy ) { status = H5C_flush_invalidate_cache(f, primary_dxpl_id, secondary_dxpl_id, flags); if ( status < 0 ) { /* This shouldn't happen -- if it does, we are toast so * just scream and die. */ HGOTO_ERROR(H5E_CACHE, H5E_CANTFLUSH, FAIL, \ "flush invalidate failed.") } } else { /* When we are only flushing marked entries, the slist will usually * still contain entries when we have flushed everything we should. * Thus we track whether we have flushed any entries in the last * pass, and terminate if we haven't. */ flushed_entries_last_pass = TRUE; while ( ( passes < H5C__MAX_PASSES_ON_FLUSH ) && ( cache_ptr->slist_len != 0 ) && ( protected_entries == 0 ) && ( flushed_entries_last_pass ) ) { unsigned curr_flush_dep_height = 0; unsigned flush_dep_passes = 0; flushed_entries_last_pass = FALSE; /* Loop over all flush dependency heights of entries */ while((curr_flush_dep_height <= H5C__NUM_FLUSH_DEP_HEIGHTS) && (cache_ptr->slist_len != 0) && (flush_dep_passes < H5C__MAX_PASSES_ON_FLUSH) ) { hbool_t flushed_during_dep_loop = FALSE; /* Start at beginning of skip list each time */ node_ptr = H5SL_first(cache_ptr->slist_ptr); HDassert( node_ptr != NULL ); /* Get cache entry for this node */ next_entry_ptr = (H5C_cache_entry_t *)H5SL_item(node_ptr); if ( NULL == next_entry_ptr ) HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "next_entry_ptr == NULL ?!?!") HDassert( next_entry_ptr->magic == H5C__H5C_CACHE_ENTRY_T_MAGIC ); HDassert( next_entry_ptr->is_dirty ); HDassert( next_entry_ptr->in_slist ); #if H5C_DO_SANITY_CHECKS /* For sanity checking, try to verify that the skip list has * the expected size and number of entries at the end of each * internal while loop (see below). * * Doing this get a bit tricky, as depending on flags, we may * or may not flush all the entries in the slist. * * To make things more entertaining, with the advent of the * fractal heap, the entry flush callback can cause entries * to be dirtied, resized, and/or moved. * * To deal with this, we first make note of the initial * skip list length and size: */ initial_slist_len = cache_ptr->slist_len; initial_slist_size = cache_ptr->slist_size; /* We then zero counters that we use to track the number * and total size of entries flushed: */ flushed_entries_count = 0; flushed_entries_size = 0; /* As mentioned above, there is the possibility that * entries will be dirtied, resized, and/or flushed during * our pass through the skip list. To capture the number * of entries added, and the skip list size delta, * zero the slist_len_increase and slist_size_increase of * the cache's instance of H5C_t. These fields will be * updated elsewhere to account for slist insertions and/or * dirty entry size changes. */ cache_ptr->slist_len_increase = 0; cache_ptr->slist_size_increase = 0; /* at the end of the loop, use these values to compute the * expected slist length and size and compare this with the * value recorded in the cache's instance of H5C_t. */ #endif /* H5C_DO_SANITY_CHECKS */ while ( node_ptr != NULL ) { entry_ptr = next_entry_ptr; /* With the advent of the fractal heap, it is possible * that the flush callback will dirty and/or resize * other entries in the cache. In particular, while * Quincey has promised me that this will never happen, * it is possible that the flush callback for an * entry may protect an entry that is not in the cache, * perhaps causing the cache to flush and possibly * evict the entry associated with node_ptr to make * space for the new entry. * * Thus we do a bit of extra sanity checking on entry_ptr, * and break out of this scan of the skip list if we * detect minor problems. We have a bit of leaway on the * number of passes though the skip list, so this shouldn't * be an issue in the flush in and of itself, as it should * be all but impossible for this to happen more than once * in any flush. * * Observe that that breaking out of the scan early * shouldn't break the sanity checks just after the end * of this while loop. * * If an entry has merely been marked clean and removed from * the s-list, we simply break out of the scan. * * If the entry has been evicted, we flag an error and * exit. */ #ifndef NDEBUG if ( entry_ptr->magic != H5C__H5C_CACHE_ENTRY_T_MAGIC ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "entry_ptr->magic is invalid ?!?!") } else #endif /* NDEBUG */ if ( ( ! entry_ptr->is_dirty ) || ( ! entry_ptr->in_slist ) ) { /* the s-list has been modified out from under us. * break out of the loop. */ goto end_of_inner_loop;; } /* increment node pointer now, before we delete its target * from the slist. */ node_ptr = H5SL_next(node_ptr); if ( node_ptr != NULL ) { next_entry_ptr = (H5C_cache_entry_t *)H5SL_item(node_ptr); if ( NULL == next_entry_ptr ) HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "next_entry_ptr == NULL ?!?!") HDassert( next_entry_ptr->magic == H5C__H5C_CACHE_ENTRY_T_MAGIC ); HDassert( next_entry_ptr->is_dirty ); HDassert( next_entry_ptr->in_slist ); } else { next_entry_ptr = NULL; } HDassert( entry_ptr != NULL ); HDassert( entry_ptr->in_slist ); if ( ( ! flush_marked_entries ) || ( entry_ptr->flush_marker ) ) { if ( entry_ptr->is_protected ) { /* we probably have major problems -- but lets flush * everything we can before we decide whether to flag * an error. */ tried_to_flush_protected_entry = TRUE; protected_entries++; } else if ( entry_ptr->is_pinned ) { /* Test to see if we are can flush the entry now. * If we can, go ahead and flush. Note that we * aren't trying to do a destroy here, so that * is not an issue. */ if(entry_ptr->flush_dep_height == curr_flush_dep_height ) { #if H5C_DO_SANITY_CHECKS flushed_entries_count++; flushed_entries_size += entry_ptr->size; #endif /* H5C_DO_SANITY_CHECKS */ status = H5C_flush_single_entry(f, primary_dxpl_id, secondary_dxpl_id, NULL, entry_ptr->addr, flags, &first_flush, FALSE); if ( status < 0 ) { /* This shouldn't happen -- if it does, we are toast * so just scream and die. */ HGOTO_ERROR(H5E_CACHE, H5E_CANTFLUSH, FAIL, \ "dirty pinned entry flush failed.") } /* end if */ flushed_during_dep_loop = TRUE; } /* end if */ else if(entry_ptr->flush_dep_height < curr_flush_dep_height) /* This shouldn't happen -- if it does, just scream and die. */ HGOTO_ERROR(H5E_CACHE, H5E_CANTFLUSH, FAIL, "dirty entry below current flush dep. height.") } /* end if */ else { /* Test to see if we are can flush the entry now. * If we can, go ahead and flush. Note that we * aren't trying to do a destroy here, so that * is not an issue. */ if(entry_ptr->flush_dep_height == curr_flush_dep_height ){ #if H5C_DO_SANITY_CHECKS flushed_entries_count++; flushed_entries_size += entry_ptr->size; #endif /* H5C_DO_SANITY_CHECKS */ status = H5C_flush_single_entry(f, primary_dxpl_id, secondary_dxpl_id, NULL, entry_ptr->addr, flags, &first_flush, FALSE); if ( status < 0 ) { /* This shouldn't happen -- if it does, we are * toast so just scream and die. */ HGOTO_ERROR(H5E_CACHE, H5E_CANTFLUSH, FAIL, \ "Can't flush entry.") } flushed_during_dep_loop = TRUE; } /* end if */ else if(entry_ptr->flush_dep_height < curr_flush_dep_height) /* This shouldn't happen -- if it does, just scream and die. */ HGOTO_ERROR(H5E_CACHE, H5E_CANTFLUSH, FAIL, "dirty entry below current flush dep. height.") } /* end else */ } /* end if */ } /* while ( node_ptr != NULL ) */ /* Check for incrementing flush dependency height */ if(flushed_during_dep_loop) { /* If we flushed an entry at this flush dependency height * start over at the bottom level of the flush dependencies */ curr_flush_dep_height = 0; /* Make certain we don't get stuck in an infinite loop */ flush_dep_passes++; /* Set flag for outer loop */ flushed_entries_last_pass = TRUE; } /* end if */ else curr_flush_dep_height++; } /* while ( curr_flush_dep_height <= H5C__NUM_FLUSH_DEP_HEIGHTS) */ end_of_inner_loop: #if H5C_DO_SANITY_CHECKS /* Verify that the slist size and length are as expected. */ HDassert( (initial_slist_len + cache_ptr->slist_len_increase - flushed_entries_count) == cache_ptr->slist_len ); HDassert( (initial_slist_size + cache_ptr->slist_size_increase - flushed_entries_size) == cache_ptr->slist_size ); #endif /* H5C_DO_SANITY_CHECKS */ passes++; } /* while */ HDassert( protected_entries <= cache_ptr->pl_len ); if ( ( ( cache_ptr->pl_len > 0 ) && ( !ignore_protected ) ) || ( tried_to_flush_protected_entry ) ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTFLUSH, FAIL, \ "cache has protected items") } if ( ( cache_ptr->slist_len != 0 ) && ( passes >= H5C__MAX_PASSES_ON_FLUSH ) ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTFLUSH, FAIL, \ "flush pass limit exceeded.") } #if H5C_DO_SANITY_CHECKS if ( ! flush_marked_entries ) { HDassert( cache_ptr->slist_len == 0 ); HDassert( cache_ptr->slist_size == 0 ); } #endif /* H5C_DO_SANITY_CHECKS */ } done: cache_ptr->flush_in_progress = FALSE; FUNC_LEAVE_NOAPI(ret_value) } /* H5C_flush_cache() */ /*------------------------------------------------------------------------- * Function: H5C_flush_to_min_clean * * Purpose: Flush dirty entries until the caches min clean size is * attained. * * This function is used in the implementation of the * metadata cache in PHDF5. To avoid "messages from the * future", the cache on process 0 can't be allowed to * flush entries until the other processes have reached * the same point in the calculation. If this constraint * is not met, it is possible that the other processes will * read metadata generated at a future point in the * computation. * * * Return: Non-negative on success/Negative on failure or if * write is not permitted. * * Programmer: John Mainzer * 9/16/05 * * Modifications: * * Re-wrote function to flush dirty entries in increasing * address order, while maintaining LRU order in the LRU list * upon return. * * Do this by scanning up the dirty LRU list for entries to * flush to reach min clean size, setting their flush_marker * flags, and recording their addresses in the order * encountered. * * Then call H5C_flush_cache() to flush the marked entries. * * Finally, use the list of marked entries to force the * correct LRU list order after the flush. * * JRM - 10/13/05 * * This change had the oposite of the desired effect. Lets * leave it in (albeit commented out for now). If we can't * find a case where it helps, lets get rid of it. * * * Added some sanity checks to the change which verify the * expected values of the new is_read_only and ro_ref_count * fields. * JRM - 3/29/07 * *------------------------------------------------------------------------- */ herr_t H5C_flush_to_min_clean(H5F_t * f, hid_t primary_dxpl_id, hid_t secondary_dxpl_id) { H5C_t * cache_ptr; herr_t result; hbool_t first_flush = TRUE; hbool_t write_permitted; #if 0 /* modified code -- commented out for now */ int i; int flushed_entries_count = 0; size_t flushed_entries_size = 0; size_t space_needed = 0; haddr_t * flushed_entries_list = NULL; H5C_cache_entry_t * entry_ptr = NULL; #endif /* JRM */ herr_t ret_value = SUCCEED; FUNC_ENTER_NOAPI(H5C_flush_to_min_clean, FAIL) HDassert( f ); HDassert( f->shared ); cache_ptr = f->shared->cache; HDassert( cache_ptr ); HDassert( cache_ptr->magic == H5C__H5C_T_MAGIC ); HDassert( cache_ptr->skip_file_checks || f ); if ( cache_ptr->check_write_permitted != NULL ) { result = (cache_ptr->check_write_permitted)(f, primary_dxpl_id, &write_permitted); if ( result < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "Can't get write_permitted") } } else { write_permitted = cache_ptr->write_permitted; } if ( ! write_permitted ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "cache write is not permitted!?!\n"); } #if 1 /* original code */ result = H5C_make_space_in_cache(f, primary_dxpl_id, secondary_dxpl_id, (size_t)0, write_permitted, &first_flush); if ( result < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "H5C_make_space_in_cache failed.") } #else /* modified code -- commented out for now */ if ( cache_ptr->max_cache_size > cache_ptr->index_size ) { if ( ((cache_ptr->max_cache_size - cache_ptr->index_size) + cache_ptr->cLRU_list_size) >= cache_ptr->min_clean_size ) { space_needed = 0; } else { space_needed = cache_ptr->min_clean_size - ((cache_ptr->max_cache_size - cache_ptr->index_size) + cache_ptr->cLRU_list_size); } } else { if ( cache_ptr->min_clean_size <= cache_ptr->cLRU_list_size ) { space_needed = 0; } else { space_needed = cache_ptr->min_clean_size - cache_ptr->cLRU_list_size; } } if ( space_needed > 0 ) { /* we have work to do */ HDassert( cache_ptr->slist_len > 0 ); /* allocate an array to keep a list of the entries that we * mark for flush. We need this list to touch up the LRU * list after the flush. */ flushed_entries_list = (haddr_t *)H5MM_malloc(sizeof(haddr_t) * (size_t)(cache_ptr->slist_len)); if ( flushed_entries_list == NULL ) { HGOTO_ERROR(H5E_RESOURCE, H5E_NOSPACE, FAIL, \ "memory allocation failed for flushed entries list") } /* Scan the dirty LRU list from tail forward and mark sufficient * entries to free up the necessary space. Keep a list of the * entries marked in the order in which they are encountered. */ entry_ptr = cache_ptr->dLRU_tail_ptr; while ( ( flushed_entries_size < space_needed ) && ( flushed_entries_count < cache_ptr->slist_len ) && ( entry_ptr != NULL ) ) { HDassert( ! (entry_ptr->is_protected) ); HDassert( ! (entry_ptr->is_read_only) ); HDassert( entry_ptr->ro_ref_count == 0 ); HDassert( entry_ptr->is_dirty ); HDassert( entry_ptr->in_slist ); entry_ptr->flush_marker = TRUE; flushed_entries_size += entry_ptr->size; flushed_entries_list[flushed_entries_count] = entry_ptr->addr; flushed_entries_count++; entry_ptr = entry_ptr->aux_prev; } HDassert( flushed_entries_count <= cache_ptr->slist_len ); HDassert( flushed_entries_size >= space_needed ); /* Flush the marked entries */ result = H5C_flush_cache(f, primary_dxpl_id, secondary_dxpl_id, H5C__FLUSH_MARKED_ENTRIES_FLAG | H5C__FLUSH_IGNORE_PROTECTED_FLAG); if ( result < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "H5C_flush_cache failed.") } /* Now touch up the LRU list so as to place the flushed entries in * the order they they would be in if we had flushed them in the * order we encountered them in. */ i = 0; while ( i < flushed_entries_count ) { H5C__SEARCH_INDEX_NO_STATS(cache_ptr, flushed_entries_list[i], \ entry_ptr, FAIL) /* At present, the above search must always succeed. However, * that may change. Write the code so we need only remove the * following assert in that event. */ HDassert( entry_ptr != NULL ); H5C__FAKE_RP_FOR_MOST_RECENT_ACCESS(cache_ptr, entry_ptr, FAIL) i++; } } /* if ( space_needed > 0 ) */ #endif /* end modified code -- commented out for now */ done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C_flush_to_min_clean() */ /*------------------------------------------------------------------------- * Function: H5C_get_cache_auto_resize_config * * Purpose: Copy the current configuration of the cache automatic * re-sizing function into the instance of H5C_auto_size_ctl_t * pointed to by config_ptr. * * Return: SUCCEED on success, and FAIL on failure. * * Programmer: John Mainzer * 10/8/04 * * Modifications: * * None. * *------------------------------------------------------------------------- */ herr_t H5C_get_cache_auto_resize_config(const H5C_t * cache_ptr, H5C_auto_size_ctl_t *config_ptr) { herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI(H5C_get_cache_auto_resize_config, FAIL) if ( ( cache_ptr == NULL ) || ( cache_ptr->magic != H5C__H5C_T_MAGIC ) ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Bad cache_ptr on entry.") } if ( config_ptr == NULL ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Bad config_ptr on entry.") } *config_ptr = cache_ptr->resize_ctl; config_ptr->set_initial_size = FALSE; config_ptr->initial_size = cache_ptr->max_cache_size; done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C_get_cache_auto_resize_config() */ /*------------------------------------------------------------------------- * Function: H5C_get_cache_size * * Purpose: Return the cache maximum size, the minimum clean size, the * current size, and the current number of entries in * *max_size_ptr, *min_clean_size_ptr, *cur_size_ptr, and * *cur_num_entries_ptr respectively. If any of these * parameters are NULL, skip that value. * * Return: SUCCEED on success, and FAIL on failure. * * Programmer: John Mainzer * 10/8/04 * * Modifications: * * None. * *------------------------------------------------------------------------- */ herr_t H5C_get_cache_size(H5C_t * cache_ptr, size_t * max_size_ptr, size_t * min_clean_size_ptr, size_t * cur_size_ptr, int32_t * cur_num_entries_ptr) { herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI(H5C_get_cache_size, FAIL) if ( ( cache_ptr == NULL ) || ( cache_ptr->magic != H5C__H5C_T_MAGIC ) ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Bad cache_ptr on entry.") } if ( max_size_ptr != NULL ) { *max_size_ptr = cache_ptr->max_cache_size; } if ( min_clean_size_ptr != NULL ) { *min_clean_size_ptr = cache_ptr->min_clean_size; } if ( cur_size_ptr != NULL ) { *cur_size_ptr = cache_ptr->index_size; } if ( cur_num_entries_ptr != NULL ) { *cur_num_entries_ptr = cache_ptr->index_len; } done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C_get_cache_size() */ /*------------------------------------------------------------------------- * Function: H5C_get_cache_hit_rate * * Purpose: Compute and return the current cache hit rate in * *hit_rate_ptr. If there have been no accesses since the * last time the cache hit rate stats were reset, set * *hit_rate_ptr to 0.0. On error, *hit_rate_ptr is * undefined. * * Return: SUCCEED on success, and FAIL on failure. * * Programmer: John Mainzer * 10/7/04 * * Modifications: * * None. * *------------------------------------------------------------------------- */ herr_t H5C_get_cache_hit_rate(H5C_t * cache_ptr, double * hit_rate_ptr) { herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI(H5C_get_cache_hit_rate, FAIL) if ( ( cache_ptr == NULL ) || ( cache_ptr->magic != H5C__H5C_T_MAGIC ) ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Bad cache_ptr on entry.") } if ( hit_rate_ptr == NULL ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Bad hit_rate_ptr on entry.") } HDassert( cache_ptr->cache_hits >= 0 ); HDassert( cache_ptr->cache_accesses >= cache_ptr->cache_hits ); if ( cache_ptr->cache_accesses > 0 ) { *hit_rate_ptr = ((double)(cache_ptr->cache_hits)) / ((double)(cache_ptr->cache_accesses)); } else { *hit_rate_ptr = 0.0; } done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C_get_cache_hit_rate() */ /*------------------------------------------------------------------------- * * Function: H5C_get_entry_status * * Purpose: This function is used to determine whether the cache * contains an entry with the specified base address. If * the entry exists, it also reports some status information * on the entry. * * Status information is reported in the locations pointed * to by the size_ptr, in_cache_ptr, is_dirty_ptr, and * is_protected_ptr. While in_cache_ptr must be defined, * the remaining pointers may be NULL, in which case the * associated data is not reported. * * Return: Non-negative on success/Negative on failure * * Programmer: John Mainzer * 7/1/05 * * Modifications: * * JRM -- 4/26/06 * Added the is_pinned_ptr parameter and supporting code. * *------------------------------------------------------------------------- */ herr_t H5C_get_entry_status(const H5F_t *f, haddr_t addr, size_t * size_ptr, hbool_t * in_cache_ptr, hbool_t * is_dirty_ptr, hbool_t * is_protected_ptr, hbool_t * is_pinned_ptr, hbool_t * is_flush_dep_parent_ptr, hbool_t * is_flush_dep_child_ptr) { H5C_t * cache_ptr; H5C_cache_entry_t * entry_ptr = NULL; herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI(H5C_get_entry_status, FAIL) HDassert( f ); HDassert( f->shared ); cache_ptr = f->shared->cache; HDassert( cache_ptr != NULL ); HDassert( cache_ptr->magic == H5C__H5C_T_MAGIC ); HDassert( H5F_addr_defined(addr) ); HDassert( in_cache_ptr != NULL ); /* this test duplicates two of the above asserts, but we need an * invocation of HGOTO_ERROR to keep the compiler happy. */ if ( ( cache_ptr == NULL ) || ( cache_ptr->magic != H5C__H5C_T_MAGIC ) ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Bad cache_ptr on entry.") } H5C__SEARCH_INDEX(cache_ptr, addr, entry_ptr, FAIL) if ( entry_ptr == NULL ) { /* the entry doesn't exist in the cache -- report this * and quit. */ *in_cache_ptr = FALSE; } else { *in_cache_ptr = TRUE; if ( size_ptr != NULL ) { *size_ptr = entry_ptr->size; } if ( is_dirty_ptr != NULL ) { *is_dirty_ptr = entry_ptr->is_dirty; } if ( is_protected_ptr != NULL ) { *is_protected_ptr = entry_ptr->is_protected; } if ( is_pinned_ptr != NULL ) { *is_pinned_ptr = entry_ptr->is_pinned; } if ( is_flush_dep_parent_ptr != NULL ) { *is_flush_dep_parent_ptr = (entry_ptr->flush_dep_height > 0); } if ( is_flush_dep_child_ptr != NULL ) { *is_flush_dep_child_ptr = (entry_ptr->flush_dep_parent != NULL); } } done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C_get_entry_status() */ /*------------------------------------------------------------------------- * Function: H5C_get_evictions_enabled() * * Purpose: Copy the current value of cache_ptr->evictions_enabled into * *evictions_enabled_ptr. * * Return: SUCCEED on success, and FAIL on failure. * * Programmer: John Mainzer * 7/27/07 * * Modifications: * * None. * *------------------------------------------------------------------------- */ herr_t H5C_get_evictions_enabled(const H5C_t *cache_ptr, hbool_t * evictions_enabled_ptr) { herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI(H5C_get_evictions_enabled, FAIL) if ( ( cache_ptr == NULL ) || ( cache_ptr->magic != H5C__H5C_T_MAGIC ) ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Bad cache_ptr on entry.") } if ( evictions_enabled_ptr == NULL ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "Bad evictions_enabled_ptr on entry.") } *evictions_enabled_ptr = cache_ptr->evictions_enabled; done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C_get_evictions_enabled() */ /*------------------------------------------------------------------------- * Function: H5C_get_trace_file_ptr * * Purpose: Get the trace_file_ptr field from the cache. * * This field will either be NULL (which indicates that trace * file logging is turned off), or contain a pointer to the * open file to which trace file data is to be written. * * Return: Non-negative on success/Negative on failure * * Programmer: John Mainzer * 1/20/06 * *------------------------------------------------------------------------- */ herr_t H5C_get_trace_file_ptr(const H5C_t *cache_ptr, FILE **trace_file_ptr_ptr) { FUNC_ENTER_NOAPI_NOFUNC(H5C_get_trace_file_ptr) HDassert(cache_ptr); HDassert(cache_ptr->magic == H5C__H5C_T_MAGIC); HDassert(trace_file_ptr_ptr); *trace_file_ptr_ptr = cache_ptr->trace_file_ptr; FUNC_LEAVE_NOAPI(SUCCEED) } /* H5C_get_trace_file_ptr() */ /*------------------------------------------------------------------------- * Function: H5C_get_trace_file_ptr_from_entry * * Purpose: Get the trace_file_ptr field from the cache, via an entry. * * This field will either be NULL (which indicates that trace * file logging is turned off), or contain a pointer to the * open file to which trace file data is to be written. * * Return: Non-negative on success/Negative on failure * * Programmer: Quincey Koziol * 6/9/08 * *------------------------------------------------------------------------- */ herr_t H5C_get_trace_file_ptr_from_entry(const H5C_cache_entry_t *entry_ptr, FILE **trace_file_ptr_ptr) { FUNC_ENTER_NOAPI_NOFUNC(H5C_get_trace_file_ptr_from_entry) /* Sanity checks */ HDassert(entry_ptr); HDassert(entry_ptr->cache_ptr); H5C_get_trace_file_ptr(entry_ptr->cache_ptr, trace_file_ptr_ptr); FUNC_LEAVE_NOAPI(SUCCEED) } /* H5C_get_trace_file_ptr_from_entry() */ /*------------------------------------------------------------------------- * Function: H5C_insert_entry * * Purpose: Adds the specified thing to the cache. The thing need not * exist on disk yet, but it must have an address and disk * space reserved. * * The primary_dxpl_id and secondary_dxpl_id parameters * specify the dxpl_ids used on the first write occasioned * by the insertion (primary_dxpl_id), and on all subsequent * writes (secondary_dxpl_id). This is useful in the * metadata cache, but may not be needed elsewhere. If so, * just use the same dxpl_id for both parameters. * * The primary_dxpl_id is the dxpl_id passed to the * check_write_permitted function if such a function has been * provided. * * Observe that this function cannot occasion a read. * * Return: Non-negative on success/Negative on failure * * Programmer: John Mainzer * 6/2/04 * * Modifications: * * JRM -- 7/21/04 * Updated function for the addition of the hash table. * * JRM -- 10/28/04 * Added code to set the cache_full flag to TRUE when ever * we need to make space in the cache. * * JRM -- 11/22/04 * Updated function for the addition of the first_flush_ptr * parameter to H5C_make_space_in_cache(). * * JRM -- 1/6/05 * Added the flags parameter, and code supporting * H5C__SET_FLUSH_MARKER_FLAG. Note that this flag is * ignored unless the new entry is dirty. * * JRM -- 6/6/05 * Added code to force all inserted entries to be dirty. * This is part of a set of changes moving management of the * is_dirty field of H5C_cache_entry_t into the H5C code. * * JRM -- 6/24/05 * Added support for the new write_permitted field of * the H5C_t structure. * * JRM -- 3/16/06 * Added initialization for the new is_pinned field of the * H5C_cache_entry_t structure. * * JRM -- 5/3/06 * Added initialization for the new dirtied field of the * H5C_cache_entry_t structure. * * JRM -- 8/9/06 * Added code supporting insertion of pinned entries. * * JRM -- 8/21/06 * Added initialization for the new flush_in_progress and * destroy_in_progress fields. * * JRM -- 3/29/07 * Added initialization for the new is_read_only and * ro_ref_count fields. * * JRM -- 8/1/07 * Added code to disable evictions when the new * evictions_enabled field is FALSE. * * JRM -- 12/31/07 * Added code supporting flash cache size increases. * * QAK -- 1/31/08 * Added initialization for the new free_file_space_on_destroy * field. * * JRM -- 11/13/08 * Moved test to see if we already have an entry with the * specified address in the cache. This was necessary as * we used to modify some fields in the entry to be inserted * priort to this test, which got the cache confused if the * insertion failed because the entry was already present. * * Also revised the function to call H5C_make_space_in_cache() * if the min_clean_size is not met at present, not just if * there is insufficient space in the cache for the new * entry. * * The purpose of this modification is to avoid "metadata * blizzards" in the write only case. In such instances, * the cache was allowed to fill with dirty metadata. When * we finally needed to evict an entry to make space, we had * to flush out a whole cache full of metadata -- which has * interesting performance effects. We hope to avoid (or * perhaps more accurately hide) this effect by maintaining * the min_clean_size, which should force us to start flushing * entries long before we actually have to evict something * to make space. * *------------------------------------------------------------------------- */ herr_t H5C_insert_entry(H5F_t * f, hid_t primary_dxpl_id, hid_t secondary_dxpl_id, const H5C_class_t * type, haddr_t addr, void * thing, unsigned int flags) { H5C_t * cache_ptr; herr_t result; hbool_t first_flush = TRUE; hbool_t insert_pinned; hbool_t set_flush_marker; hbool_t write_permitted = TRUE; size_t empty_space; H5C_cache_entry_t * entry_ptr; H5C_cache_entry_t * test_entry_ptr; unsigned u; /* Local index variable */ herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI(H5C_insert_entry, FAIL) HDassert( f ); HDassert( f->shared ); cache_ptr = f->shared->cache; HDassert( cache_ptr ); HDassert( cache_ptr->magic == H5C__H5C_T_MAGIC ); HDassert( cache_ptr->skip_file_checks || f ); HDassert( type ); HDassert( type->flush ); HDassert( type->size ); HDassert( H5F_addr_defined(addr) ); HDassert( thing ); #if H5C_DO_EXTREME_SANITY_CHECKS if ( H5C_verify_not_in_index(cache_ptr, (H5C_cache_entry_t *)thing) < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "thing already in index.\n"); } #endif /* H5C_DO_SANITY_CHECKS */ #if H5C_DO_EXTREME_SANITY_CHECKS if ( H5C_validate_lru_list(cache_ptr) < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "LRU sanity check failed.\n"); } #endif /* H5C_DO_EXTREME_SANITY_CHECKS */ set_flush_marker = ( (flags & H5C__SET_FLUSH_MARKER_FLAG) != 0 ); insert_pinned = ( (flags & H5C__PIN_ENTRY_FLAG) != 0 ); entry_ptr = (H5C_cache_entry_t *)thing; /* verify that the new entry isn't already in the hash table -- scream * and die if it is. */ H5C__SEARCH_INDEX(cache_ptr, addr, test_entry_ptr, FAIL) if ( test_entry_ptr != NULL ) { if ( test_entry_ptr == entry_ptr ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTINS, FAIL, \ "entry already in cache.") } else { HGOTO_ERROR(H5E_CACHE, H5E_CANTINS, FAIL, \ "duplicate entry in cache.") } } #ifndef NDEBUG entry_ptr->magic = H5C__H5C_CACHE_ENTRY_T_MAGIC; #endif /* NDEBUG */ entry_ptr->cache_ptr = cache_ptr; entry_ptr->addr = addr; entry_ptr->type = type; entry_ptr->is_protected = FALSE; entry_ptr->is_read_only = FALSE; entry_ptr->ro_ref_count = 0; entry_ptr->is_pinned = insert_pinned; entry_ptr->pinned_from_client = insert_pinned; /* newly inserted entries are assumed to be dirty */ entry_ptr->is_dirty = TRUE; /* not protected, so can't be dirtied */ entry_ptr->dirtied = FALSE; if ( (type->size)(f, thing, &(entry_ptr->size)) < 0 ) { HGOTO_ERROR(H5E_RESOURCE, H5E_CANTGETSIZE, FAIL, \ "Can't get size of thing") } HDassert( entry_ptr->size < H5C_MAX_ENTRY_SIZE ); entry_ptr->in_slist = FALSE; #ifdef H5_HAVE_PARALLEL entry_ptr->clear_on_unprotect = FALSE; #endif /* H5_HAVE_PARALLEL */ entry_ptr->flush_in_progress = FALSE; entry_ptr->destroy_in_progress = FALSE; entry_ptr->free_file_space_on_destroy = FALSE; /* Initialize flush dependency height fields */ entry_ptr->flush_dep_parent = NULL; for(u = 0; u < H5C__NUM_FLUSH_DEP_HEIGHTS; u++) entry_ptr->child_flush_dep_height_rc[u] = 0; entry_ptr->flush_dep_height = 0; entry_ptr->ht_next = NULL; entry_ptr->ht_prev = NULL; entry_ptr->next = NULL; entry_ptr->prev = NULL; entry_ptr->aux_next = NULL; entry_ptr->aux_prev = NULL; H5C__RESET_CACHE_ENTRY_STATS(entry_ptr) if ( ( cache_ptr->flash_size_increase_possible ) && ( entry_ptr->size > cache_ptr->flash_size_increase_threshold ) ) { result = H5C__flash_increase_cache_size(cache_ptr, 0, entry_ptr->size); if ( result < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTINS, FAIL, \ "H5C__flash_increase_cache_size failed.") } } if ( cache_ptr->index_size >= cache_ptr->max_cache_size ) { empty_space = 0; } else { empty_space = cache_ptr->max_cache_size - cache_ptr->index_size; } if ( ( cache_ptr->evictions_enabled ) && ( ( (cache_ptr->index_size + entry_ptr->size) > cache_ptr->max_cache_size ) || ( ( ( empty_space + cache_ptr->clean_index_size ) < cache_ptr->min_clean_size ) ) ) ) { size_t space_needed; if ( empty_space <= entry_ptr->size ) { cache_ptr->cache_full = TRUE; } if ( cache_ptr->check_write_permitted != NULL ) { result = (cache_ptr->check_write_permitted)(f, primary_dxpl_id, &write_permitted); if ( result < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTINS, FAIL, \ "Can't get write_permitted") } } else { write_permitted = cache_ptr->write_permitted; } HDassert( entry_ptr->size <= H5C_MAX_ENTRY_SIZE ); space_needed = entry_ptr->size; if ( space_needed > cache_ptr->max_cache_size ) { space_needed = cache_ptr->max_cache_size; } /* Note that space_needed is just the amount of space that * needed to insert the new entry without exceeding the cache * size limit. The subsequent call to H5C_make_space_in_cache() * may evict the entries required to free more or less space * depending on conditions. It MAY be less if the cache is * currently undersized, or more if the cache is oversized. * * The cache can exceed its maximum size limit via the following * mechanisms: * * First, it is possible for the cache to grow without * bound as long as entries are protected and not unprotected. * * Second, when writes are not permitted it is also possible * for the cache to grow without bound. * * Finally, we usually don't check to see if the cache is * oversized at the end of an unprotect. As a result, it is * possible to have a vastly oversized cache with no protected * entries as long as all the protects preceed the unprotects. * * Since items 1 and 2 are not changing any time soon, I see * no point in worrying about the third. */ result = H5C_make_space_in_cache(f, primary_dxpl_id, secondary_dxpl_id, space_needed, write_permitted, &first_flush); if ( result < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTINS, FAIL, \ "H5C_make_space_in_cache failed.") } } H5C__INSERT_IN_INDEX(cache_ptr, entry_ptr, FAIL) /* New entries are presumed to be dirty, so this if statement is * unnecessary. Rework it once the rest of the code changes are * in and tested. -- JRM */ if ( entry_ptr->is_dirty ) { entry_ptr->flush_marker = set_flush_marker; H5C__INSERT_ENTRY_IN_SLIST(cache_ptr, entry_ptr, FAIL) } else { entry_ptr->flush_marker = FALSE; } H5C__UPDATE_RP_FOR_INSERTION(cache_ptr, entry_ptr, FAIL) #if H5C_DO_EXTREME_SANITY_CHECKS if ( H5C_validate_lru_list(cache_ptr) < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "LRU sanity check failed.\n"); } #endif /* H5C_DO_EXTREME_SANITY_CHECKS */ /* If the entry's type has a 'notify' callback send a 'after insertion' * notice now that the entry is fully integrated into the cache. */ if(entry_ptr->type->notify && (entry_ptr->type->notify)(H5C_NOTIFY_ACTION_AFTER_INSERT, entry_ptr) < 0) HGOTO_ERROR(H5E_CACHE, H5E_CANTNOTIFY, FAIL, "can't notify client about entry inserted into cache") H5C__UPDATE_STATS_FOR_INSERTION(cache_ptr, entry_ptr) done: #if H5C_DO_EXTREME_SANITY_CHECKS if ( H5C_validate_lru_list(cache_ptr) < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "LRU sanity check failed.\n"); } #endif /* H5C_DO_EXTREME_SANITY_CHECKS */ FUNC_LEAVE_NOAPI(ret_value) } /* H5C_insert_entry() */ /*------------------------------------------------------------------------- * * Function: H5C_mark_entries_as_clean * * Purpose: When the H5C code is used to implement the metadata caches * in PHDF5, only the cache with MPI_rank 0 is allowed to * actually write entries to disk -- all other caches must * retain dirty entries until they are advised that the * entries are clean. * * This function exists to allow the H5C code to receive these * notifications. * * The function receives a list of entry base addresses * which must refer to dirty entries in the cache. If any * of the entries are either clean or don't exist, the * function flags an error. * * The function scans the list of entries and flushes all * those that are currently unprotected with the * H5C__FLUSH_CLEAR_ONLY_FLAG. Those that are currently * protected are flagged for clearing when they are * unprotected. * * Return: Non-negative on success/Negative on failure * * Programmer: John Mainzer * 7/5/05 * * Modifications: * * Reworked function to flush entries in LRU order instead * of increasing address order. The hope is that this will * improve the hit rate on the slave caches. * * JRM - 10/13/05 * * Leave the old code in place for now (commented out) for * benchmarking. * * JRM -- 4/13/06 * Updated function to deal with pinned entries. * *------------------------------------------------------------------------- */ #ifdef H5_HAVE_PARALLEL herr_t H5C_mark_entries_as_clean(H5F_t * f, hid_t primary_dxpl_id, hid_t secondary_dxpl_id, int32_t ce_array_len, haddr_t * ce_array_ptr) { H5C_t * cache_ptr; hbool_t first_flush = TRUE; int entries_cleared; int entries_examined; int i; int initial_list_len; haddr_t addr; #if H5C_DO_SANITY_CHECKS int pinned_entries_marked = 0; int protected_entries_marked = 0; int other_entries_marked = 0; haddr_t last_addr; #endif /* H5C_DO_SANITY_CHECKS */ H5C_cache_entry_t * clear_ptr = NULL; H5C_cache_entry_t * entry_ptr = NULL; herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI(H5C_mark_entries_as_clean, FAIL) HDassert( f ); HDassert( f->shared ); cache_ptr = f->shared->cache; HDassert( cache_ptr ); HDassert( cache_ptr->magic == H5C__H5C_T_MAGIC ); HDassert( cache_ptr->skip_file_checks || f ); HDassert( ce_array_len > 0 ); HDassert( ce_array_ptr != NULL ); #if H5C_DO_EXTREME_SANITY_CHECKS if ( H5C_validate_lru_list(cache_ptr) < 0 ) { HDassert(0); HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "LRU sanity check failed.\n"); } #endif /* H5C_DO_EXTREME_SANITY_CHECKS */ for ( i = 0; i < ce_array_len; i++ ) { addr = ce_array_ptr[i]; #if H5C_DO_SANITY_CHECKS if ( i == 0 ) { last_addr = addr; } else { if ( last_addr == addr ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "Duplicate entry in cleaned list.\n"); } else if ( last_addr > addr ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "cleaned list not sorted.\n"); } } #if H5C_DO_EXTREME_SANITY_CHECKS if ( H5C_validate_lru_list(cache_ptr) < 0 ) { HDassert(0); HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "LRU sanity check failed.\n"); } #endif /* H5C_DO_EXTREME_SANITY_CHECKS */ #endif /* H5C_DO_SANITY_CHECKS */ HDassert( H5F_addr_defined(addr) ); H5C__SEARCH_INDEX(cache_ptr, addr, entry_ptr, FAIL) if ( entry_ptr == NULL ) { #if H5C_DO_SANITY_CHECKS HDfprintf(stdout, "H5C_mark_entries_as_clean: entry[%d] = %ld not in cache.\n", (int)i, (long)addr); #endif /* H5C_DO_SANITY_CHECKS */ HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "Listed entry not in cache?!?!?.") } else if ( ! entry_ptr->is_dirty ) { #if H5C_DO_SANITY_CHECKS HDfprintf(stdout, "H5C_mark_entries_as_clean: entry %ld is not dirty!?!\n", (long)addr); #endif /* H5C_DO_SANITY_CHECKS */ HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "Listed entry not dirty?!?!?.") #if 0 /* original code */ } else if ( entry_ptr->is_protected ) { entry_ptr->clear_on_unprotect = TRUE; } else { if ( H5C_flush_single_entry(f, primary_dxpl_id, secondary_dxpl_id, entry_ptr->type, addr, H5C__FLUSH_CLEAR_ONLY_FLAG, &first_flush, TRUE) < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Can't clear entry.") } } #else /* modified code */ } else { /* Mark the entry to be cleared on unprotect. We will * scan the LRU list shortly, and clear all those entries * not currently protected. */ entry_ptr->clear_on_unprotect = TRUE; #if H5C_DO_SANITY_CHECKS if ( entry_ptr->is_protected ) { protected_entries_marked++; } else if ( entry_ptr->is_pinned ) { pinned_entries_marked++; } else { other_entries_marked++; } #endif /* H5C_DO_SANITY_CHECKS */ } #endif /* end modified code */ } #if 1 /* modified code */ /* Scan through the LRU list from back to front, and flush the * entries whose clear_on_unprotect flags are set. Observe that * any protected entries will not be on the LRU, and therefore * will not be flushed at this time. */ entries_cleared = 0; entries_examined = 0; initial_list_len = cache_ptr->LRU_list_len; entry_ptr = cache_ptr->LRU_tail_ptr; while ( ( entry_ptr != NULL ) && ( entries_examined <= initial_list_len ) && ( entries_cleared < ce_array_len ) ) { if ( entry_ptr->clear_on_unprotect ) { entry_ptr->clear_on_unprotect = FALSE; clear_ptr = entry_ptr; entry_ptr = entry_ptr->prev; entries_cleared++; if ( H5C_flush_single_entry(f, primary_dxpl_id, secondary_dxpl_id, clear_ptr->type, clear_ptr->addr, H5C__FLUSH_CLEAR_ONLY_FLAG, &first_flush, TRUE) < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Can't clear entry.") } } else { entry_ptr = entry_ptr->prev; } entries_examined++; } #if H5C_DO_SANITY_CHECKS HDassert( entries_cleared == other_entries_marked ); #endif /* H5C_DO_SANITY_CHECKS */ /* It is also possible that some of the cleared entries are on the * pinned list. Must scan that also. */ entry_ptr = cache_ptr->pel_head_ptr; while ( entry_ptr != NULL ) { if ( entry_ptr->clear_on_unprotect ) { entry_ptr->clear_on_unprotect = FALSE; clear_ptr = entry_ptr; entry_ptr = entry_ptr->next; entries_cleared++; if ( H5C_flush_single_entry(f, primary_dxpl_id, secondary_dxpl_id, clear_ptr->type, clear_ptr->addr, H5C__FLUSH_CLEAR_ONLY_FLAG, &first_flush, TRUE) < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Can't clear entry.") } } else { entry_ptr = entry_ptr->next; } } #if H5C_DO_SANITY_CHECKS HDassert( entries_cleared == pinned_entries_marked + other_entries_marked ); HDassert( entries_cleared + protected_entries_marked == ce_array_len ); #endif /* H5C_DO_SANITY_CHECKS */ HDassert( ( entries_cleared == ce_array_len ) || ( (ce_array_len - entries_cleared) <= cache_ptr->pl_len ) ); #if H5C_DO_SANITY_CHECKS i = 0; entry_ptr = cache_ptr->pl_head_ptr; while ( entry_ptr != NULL ) { if ( entry_ptr->clear_on_unprotect ) { i++; } entry_ptr = entry_ptr->next; } HDassert( (entries_cleared + i) == ce_array_len ); #endif /* H5C_DO_SANITY_CHECKS */ #endif /* modified code */ done: #if H5C_DO_EXTREME_SANITY_CHECKS if ( H5C_validate_lru_list(cache_ptr) < 0 ) { HDassert(0); HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "LRU sanity check failed.\n"); } #endif /* H5C_DO_EXTREME_SANITY_CHECKS */ FUNC_LEAVE_NOAPI(ret_value) } /* H5C_mark_entries_as_clean() */ #endif /* H5_HAVE_PARALLEL */ /*------------------------------------------------------------------------- * Function: H5C_mark_entry_dirty * * Purpose: Mark a pinned or protected entry as dirty. The target entry * MUST be either pinned or protected, and MAY be both. * * In the protected case, this call is the functional * equivalent of setting the H5C__DIRTIED_FLAG on an unprotect * call. * * In the pinned but not protected case, if the entry is not * already dirty, the function places function marks the entry * dirty and places it on the skip list. * * Return: Non-negative on success/Negative on failure * * Programmer: John Mainzer * 5/15/06 * * JRM -- 11/5/08 * Added call to H5C__UPDATE_INDEX_FOR_ENTRY_DIRTY() to * update the new clean_index_size and dirty_index_size * fields of H5C_t in the case that the entry was clean * prior to this call, and is pinned and not protected. * *------------------------------------------------------------------------- */ herr_t H5C_mark_entry_dirty(void *thing) { H5C_t * cache_ptr; H5C_cache_entry_t * entry_ptr = (H5C_cache_entry_t *)thing; herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI(H5C_mark_entry_dirty, FAIL) /* Sanity checks */ HDassert(entry_ptr); HDassert(H5F_addr_defined(entry_ptr->addr)); cache_ptr = entry_ptr->cache_ptr; HDassert(cache_ptr); HDassert(cache_ptr->magic == H5C__H5C_T_MAGIC); if ( entry_ptr->is_protected ) { HDassert( ! ((entry_ptr)->is_read_only) ); /* set the dirtied flag */ entry_ptr->dirtied = TRUE; } else if ( entry_ptr->is_pinned ) { hbool_t was_pinned_unprotected_and_clean; was_pinned_unprotected_and_clean = ! ( entry_ptr->is_dirty ); /* mark the entry as dirty if it isn't already */ entry_ptr->is_dirty = TRUE; if ( was_pinned_unprotected_and_clean ) { H5C__UPDATE_INDEX_FOR_ENTRY_DIRTY(cache_ptr, entry_ptr); } if ( ! (entry_ptr->in_slist) ) { H5C__INSERT_ENTRY_IN_SLIST(cache_ptr, entry_ptr, FAIL) } H5C__UPDATE_STATS_FOR_DIRTY_PIN(cache_ptr, entry_ptr) } else { HGOTO_ERROR(H5E_CACHE, H5E_CANTMARKDIRTY, FAIL, \ "Entry is neither pinned nor protected??") } done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C_mark_entry_dirty() */ /*------------------------------------------------------------------------- * * Function: H5C_move_entry * * Purpose: Use this function to notify the cache that an entry's * file address changed. * * Return: Non-negative on success/Negative on failure * * Programmer: John Mainzer * 6/2/04 * * JRM -- 11/5/08 * On review this function looks like no change is needed to * support the new clean_index_size and dirty_index_size * fields of H5C_t. * *------------------------------------------------------------------------- */ herr_t H5C_move_entry(H5C_t * cache_ptr, const H5C_class_t * type, haddr_t old_addr, haddr_t new_addr) { hbool_t was_dirty; H5C_cache_entry_t * entry_ptr = NULL; H5C_cache_entry_t * test_entry_ptr = NULL; #if H5C_DO_SANITY_CHECKS hbool_t removed_entry_from_slist = FALSE; #endif /* H5C_DO_SANITY_CHECKS */ herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI(H5C_move_entry, FAIL) HDassert( cache_ptr ); HDassert( cache_ptr->magic == H5C__H5C_T_MAGIC ); HDassert( type ); HDassert( H5F_addr_defined(old_addr) ); HDassert( H5F_addr_defined(new_addr) ); HDassert( H5F_addr_ne(old_addr, new_addr) ); #if H5C_DO_EXTREME_SANITY_CHECKS if ( H5C_validate_lru_list(cache_ptr) < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "LRU sanity check failed.\n"); } #endif /* H5C_DO_EXTREME_SANITY_CHECKS */ H5C__SEARCH_INDEX(cache_ptr, old_addr, entry_ptr, FAIL) if ( ( entry_ptr == NULL ) || ( entry_ptr->type != type ) ) { /* the old item doesn't exist in the cache, so we are done. */ HGOTO_DONE(SUCCEED) } HDassert( entry_ptr->addr == old_addr ); HDassert( entry_ptr->type == type ); if ( entry_ptr->is_protected ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTMOVE, FAIL, \ "Target entry is protected.") } H5C__SEARCH_INDEX(cache_ptr, new_addr, test_entry_ptr, FAIL) if ( test_entry_ptr != NULL ) { /* we are hosed */ if ( test_entry_ptr->type == type ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTMOVE, FAIL, \ "Target already moved & reinserted???.") } else { HGOTO_ERROR(H5E_CACHE, H5E_CANTMOVE, FAIL, \ "New address already in use?.") } } /* If we get this far we have work to do. Remove *entry_ptr from * the hash table (and skip list if necessary), change its address to the * new address, mark it as dirty (if it isn't already) and then re-insert. * * Update the replacement policy for a hit to avoid an eviction before * the moved entry is touched. Update stats for a move. * * Note that we do not check the size of the cache, or evict anything. * Since this is a simple re-name, cache size should be unaffected. * * Check to see if the target entry is in the process of being destroyed * before we delete from the index, etc. If it is, all we do is * change the addr. If the entry is only in the process of being flushed, * don't mark it as dirty either, lest we confuse the flush call back. */ if ( ! ( entry_ptr->destroy_in_progress ) ) { H5C__DELETE_FROM_INDEX(cache_ptr, entry_ptr) if ( entry_ptr->in_slist ) { HDassert( cache_ptr->slist_ptr ); H5C__REMOVE_ENTRY_FROM_SLIST(cache_ptr, entry_ptr) #if H5C_DO_SANITY_CHECKS removed_entry_from_slist = TRUE; #endif /* H5C_DO_SANITY_CHECKS */ } } entry_ptr->addr = new_addr; if ( ! ( entry_ptr->destroy_in_progress ) ) { was_dirty = entry_ptr->is_dirty; if ( ! ( entry_ptr->flush_in_progress ) ) { entry_ptr->is_dirty = TRUE; } H5C__INSERT_IN_INDEX(cache_ptr, entry_ptr, FAIL) if ( ! ( entry_ptr->flush_in_progress ) ) { H5C__INSERT_ENTRY_IN_SLIST(cache_ptr, entry_ptr, FAIL) #if H5C_DO_SANITY_CHECKS if ( removed_entry_from_slist ) { /* we just removed the entry from the slist. Thus we * must touch up cache_ptr->slist_len_increase and * cache_ptr->slist_size_increase to keep from skewing * the sanity checks. */ HDassert( cache_ptr->slist_len_increase > 1 ); HDassert( cache_ptr->slist_size_increase > entry_ptr->size ); cache_ptr->slist_len_increase -= 1; cache_ptr->slist_size_increase -= entry_ptr->size; } #endif /* H5C_DO_SANITY_CHECKS */ H5C__UPDATE_RP_FOR_MOVE(cache_ptr, entry_ptr, was_dirty, FAIL) } } H5C__UPDATE_STATS_FOR_MOVE(cache_ptr, entry_ptr) done: #if H5C_DO_EXTREME_SANITY_CHECKS if ( H5C_validate_lru_list(cache_ptr) < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "LRU sanity check failed.\n"); } #endif /* H5C_DO_EXTREME_SANITY_CHECKS */ FUNC_LEAVE_NOAPI(ret_value) } /* H5C_move_entry() */ /*------------------------------------------------------------------------- * Function: H5C_resize_pinned_entry * * Purpose: Resize a pinned entry. The target entry MUST be * be pinned, and MUST be unprotected. * * Resizing an entry dirties it, so if the entry is not * already dirty, the function places the entry on the * skip list. * * Return: Non-negative on success/Negative on failure * * Programmer: John Mainzer * 7/5/06 * *------------------------------------------------------------------------- */ herr_t H5C_resize_pinned_entry(void *thing, size_t new_size) { H5C_t * cache_ptr; H5C_cache_entry_t * entry_ptr = (H5C_cache_entry_t *)thing; size_t size_increase; hbool_t was_clean; herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI(H5C_resize_pinned_entry, FAIL) /* Sanity checks */ HDassert(entry_ptr); HDassert(H5F_addr_defined(entry_ptr->addr)); cache_ptr = entry_ptr->cache_ptr; HDassert(cache_ptr); HDassert(cache_ptr->magic == H5C__H5C_T_MAGIC); /* Check for usage errors */ if(new_size <= 0) HGOTO_ERROR(H5E_CACHE, H5E_CANTRESIZE, FAIL, "New size is non-positive.") if(!entry_ptr->is_pinned) HGOTO_ERROR(H5E_CACHE, H5E_CANTRESIZE, FAIL, "Entry isn't pinned??") if(entry_ptr->is_protected) HGOTO_ERROR(H5E_CACHE, H5E_CANTRESIZE, FAIL, "Entry is protected??") /* make note of whether the entry was clean to begin with */ was_clean = ! ( entry_ptr->is_dirty ); /* resizing dirties entries -- mark the entry as dirty if it * isn't already */ entry_ptr->is_dirty = TRUE; /* update for change in entry size if necessary */ if ( entry_ptr->size != new_size ) { /* do a flash cache size increase if appropriate */ if ( cache_ptr->flash_size_increase_possible ) { if ( new_size > entry_ptr->size ) { size_increase = new_size - entry_ptr->size; if(size_increase >= cache_ptr->flash_size_increase_threshold) { if(H5C__flash_increase_cache_size(cache_ptr, entry_ptr->size, new_size) < 0) HGOTO_ERROR(H5E_CACHE, H5E_CANTUNPROTECT, FAIL, "flash cache increase failed") } } } /* update the protected entry list */ H5C__DLL_UPDATE_FOR_SIZE_CHANGE((cache_ptr->pel_len), \ (cache_ptr->pel_size), \ (entry_ptr->size), (new_size)); /* update the hash table */ H5C__UPDATE_INDEX_FOR_SIZE_CHANGE((cache_ptr), (entry_ptr->size),\ (new_size), (entry_ptr), (was_clean)); /* if the entry is in the skip list, update that too */ if ( entry_ptr->in_slist ) { H5C__UPDATE_SLIST_FOR_SIZE_CHANGE((cache_ptr), (entry_ptr->size),\ (new_size)); } /* update statistics just before changing the entry size */ H5C__UPDATE_STATS_FOR_ENTRY_SIZE_CHANGE((cache_ptr), (entry_ptr), \ (new_size)); /* finally, update the entry size proper */ entry_ptr->size = new_size; } else if ( was_clean ) { H5C__UPDATE_INDEX_FOR_ENTRY_DIRTY(cache_ptr, entry_ptr) } if ( ! (entry_ptr->in_slist) ) { H5C__INSERT_ENTRY_IN_SLIST(cache_ptr, entry_ptr, FAIL) } H5C__UPDATE_STATS_FOR_DIRTY_PIN(cache_ptr, entry_ptr) done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C_resize_pinned_entry() */ /*------------------------------------------------------------------------- * Function: H5C_pin_entry_from_client() * * Purpose: Internal routine to pin a cache entry from a client action. * * Return: Non-negative on success/Negative on failure * * Programmer: Quincey Koziol * 3/26/09 * *------------------------------------------------------------------------- */ #ifndef NDEBUG static herr_t H5C_pin_entry_from_client(H5C_t * cache_ptr, H5C_cache_entry_t * entry_ptr) #else static herr_t H5C_pin_entry_from_client(H5C_t UNUSED * cache_ptr, H5C_cache_entry_t * entry_ptr) #endif { herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI_NOINIT(H5C_pin_entry_from_client) /* Sanity checks */ HDassert( cache_ptr ); HDassert( entry_ptr ); /* Check if the entry is already pinned */ if(entry_ptr->is_pinned) { /* Check if the entry was pinned through an explicit pin from a client */ if(entry_ptr->pinned_from_client) HGOTO_ERROR(H5E_CACHE, H5E_CANTPIN, FAIL, "Entry is already pinned") } /* end if */ else { entry_ptr->is_pinned = TRUE; H5C__UPDATE_STATS_FOR_PIN(cache_ptr, entry_ptr) } /* end else */ /* Mark that the entry was pinned through an explicit pin from a client */ entry_ptr->pinned_from_client = TRUE; done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C_pin_entry_from_client() */ /*------------------------------------------------------------------------- * Function: H5C_pin_protected_entry() * * Purpose: Pin a protected cache entry. The entry must be protected * at the time of call, and must be unpinned. * * Return: Non-negative on success/Negative on failure * * Programmer: John Mainzer * 4/26/06 * *------------------------------------------------------------------------- */ herr_t H5C_pin_protected_entry(void *thing) { H5C_t * cache_ptr; H5C_cache_entry_t * entry_ptr = (H5C_cache_entry_t *)thing; /* Pointer to entry to pin */ herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI(H5C_pin_protected_entry, FAIL) /* Sanity checks */ HDassert(entry_ptr); HDassert(H5F_addr_defined(entry_ptr->addr)); cache_ptr = entry_ptr->cache_ptr; HDassert(cache_ptr); HDassert(cache_ptr->magic == H5C__H5C_T_MAGIC); /* Only protected entries can be pinned */ if(!entry_ptr->is_protected) HGOTO_ERROR(H5E_CACHE, H5E_CANTPIN, FAIL, "Entry isn't protected") /* Pin the entry from a client */ if(H5C_pin_entry_from_client(cache_ptr, entry_ptr) < 0) HGOTO_ERROR(H5E_CACHE, H5E_CANTPIN, FAIL, "Can't pin entry by client") done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C_pin_protected_entry() */ /*------------------------------------------------------------------------- * Function: H5C_protect * * Purpose: If the target entry is not in the cache, load it. If * necessary, attempt to evict one or more entries to keep * the cache within its maximum size. * * Mark the target entry as protected, and return its address * to the caller. The caller must call H5C_unprotect() when * finished with the entry. * * While it is protected, the entry may not be either evicted * or flushed -- nor may it be accessed by another call to * H5C_protect. Any attempt to do so will result in a failure. * * The primary_dxpl_id and secondary_dxpl_id parameters * specify the dxpl_ids used on the first write occasioned * by the insertion (primary_dxpl_id), and on all subsequent * writes (secondary_dxpl_id). This is useful in the * metadata cache, but may not be needed elsewhere. If so, * just use the same dxpl_id for both parameters. * * All reads are performed with the primary_dxpl_id. * * Similarly, the primary_dxpl_id is passed to the * check_write_permitted function if it is called. * * Return: Success: Ptr to the desired entry * * Failure: NULL * * Programmer: John Mainzer - 6/2/04 * * Modifications: * * JRM -- 7/21/04 * Updated for the addition of the hash table. * * JRM -- 10/28/04 * Added code to set cache_full to TRUE whenever we try to * make space in the cache. * * JRM -- 11/12/04 * Added code to call to H5C_make_space_in_cache() * after the call to H5C__auto_adjust_cache_size() if that * function sets the size_decreased flag is TRUE. * * JRM -- 4/25/05 * The size_decreased flag can also be set to TRUE in * H5C_set_cache_auto_resize_config() if a new configuration * forces an immediate reduction in cache size. Modified * the code to deal with this eventuallity. * * JRM -- 6/24/05 * Added support for the new write_permitted field of H5C_t. * * JRM -- 10/22/05 * Hand optimizations. * * JRM -- 5/3/06 * Added code to set the new dirtied field in * H5C_cache_entry_t to FALSE prior to return. * * JRM -- 6/23/06 * Modified code to allow dirty entries to be loaded from * disk. This is necessary as a bug fix in the object * header code requires us to modify a header as it is read. * * JRM -- 3/28/07 * Added the flags parameter and supporting code. At least * for now, this parameter is used to allow the entry to * be protected read only, thus allowing multiple protects. * * Also added code to allow multiple read only protects * of cache entries. * * JRM -- 7/27/07 * Added code supporting the new evictions_enabled field * in H5C_t. * * JRM -- 1/3/08 * Added to do a flash cache size increase if appropriate * when a large entry is loaded. * * JRM -- 11/13/08 * Modified function to call H5C_make_space_in_cache() when * the min_clean_size is violated, not just when there isn't * enough space for and entry that has just been loaded. * * The purpose of this modification is to avoid "metadata * blizzards" in the write only case. In such instances, * the cache was allowed to fill with dirty metadata. When * we finally needed to evict an entry to make space, we had * to flush out a whole cache full of metadata -- which has * interesting performance effects. We hope to avoid (or * perhaps more accurately hide) this effect by maintaining * the min_clean_size, which should force us to start flushing * entries long before we actually have to evict something * to make space. * *------------------------------------------------------------------------- */ void * H5C_protect(H5F_t * f, hid_t primary_dxpl_id, hid_t secondary_dxpl_id, const H5C_class_t * type, haddr_t addr, void * udata, unsigned flags) { H5C_t * cache_ptr; hbool_t hit; hbool_t first_flush; hbool_t have_write_permitted = FALSE; hbool_t read_only = FALSE; hbool_t write_permitted; herr_t result; size_t empty_space; void * thing; H5C_cache_entry_t * entry_ptr; void * ret_value; /* Return value */ FUNC_ENTER_NOAPI(H5C_protect, NULL) /* check args */ HDassert( f ); HDassert( f->shared ); cache_ptr = f->shared->cache; HDassert( cache_ptr ); HDassert( cache_ptr->magic == H5C__H5C_T_MAGIC ); HDassert( cache_ptr->skip_file_checks || f ); HDassert( type ); HDassert( type->flush ); HDassert( type->load ); HDassert( H5F_addr_defined(addr) ); #if H5C_DO_EXTREME_SANITY_CHECKS if ( H5C_validate_lru_list(cache_ptr) < 0 ) { HDassert(0); HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, NULL, \ "LRU sanity check failed.\n"); } #endif /* H5C_DO_EXTREME_SANITY_CHECKS */ if ( (flags & H5C__READ_ONLY_FLAG) != 0 ) { read_only = TRUE; } /* first check to see if the target is in cache */ H5C__SEARCH_INDEX(cache_ptr, addr, entry_ptr, NULL) if ( entry_ptr != NULL ) { /* Check for trying to load the wrong type of entry from an address */ if(entry_ptr->type != type) HGOTO_ERROR(H5E_CACHE, H5E_BADTYPE, NULL, "incorrect cache entry type") hit = TRUE; thing = (void *)entry_ptr; } else { /* must try to load the entry from disk. */ hit = FALSE; thing = H5C_load_entry(f, primary_dxpl_id, type, addr, udata, cache_ptr->skip_file_checks); if ( thing == NULL ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTLOAD, NULL, "can't load entry") } entry_ptr = (H5C_cache_entry_t *)thing; /* If the entry is very large, and we are configured to allow it, * we may wish to perform a flash cache size increase. */ if ( ( cache_ptr->flash_size_increase_possible ) && ( entry_ptr->size > cache_ptr->flash_size_increase_threshold ) ) { result = H5C__flash_increase_cache_size(cache_ptr, 0, entry_ptr->size); if ( result < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTPROTECT, NULL, \ "H5C__flash_increase_cache_size failed.") } } if ( cache_ptr->index_size >= cache_ptr->max_cache_size ) { empty_space = 0; } else { empty_space = cache_ptr->max_cache_size - cache_ptr->index_size; } /* try to free up if necceary and if evictions are permitted. Note * that if evictions are enabled, we will call H5C_make_space_in_cache() * regardless if the min_free_space requirement is not met. */ if ( ( cache_ptr->evictions_enabled ) && ( ( (cache_ptr->index_size + entry_ptr->size) > cache_ptr->max_cache_size) || ( ( empty_space + cache_ptr->clean_index_size ) < cache_ptr->min_clean_size ) ) ) { size_t space_needed; if ( empty_space <= entry_ptr->size ) { cache_ptr->cache_full = TRUE; } if ( cache_ptr->check_write_permitted != NULL ) { result = (cache_ptr->check_write_permitted)(f, primary_dxpl_id, &write_permitted); if ( result < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTPROTECT, NULL, \ "Can't get write_permitted 1") } else { have_write_permitted = TRUE; first_flush = TRUE; } } else { write_permitted = cache_ptr->write_permitted; have_write_permitted = TRUE; first_flush = TRUE; } HDassert( entry_ptr->size <= H5C_MAX_ENTRY_SIZE ); space_needed = entry_ptr->size; if ( space_needed > cache_ptr->max_cache_size ) { space_needed = cache_ptr->max_cache_size; } /* Note that space_needed is just the amount of space that * needed to insert the new entry without exceeding the cache * size limit. The subsequent call to H5C_make_space_in_cache() * may evict the entries required to free more or less space * depending on conditions. It MAY be less if the cache is * currently undersized, or more if the cache is oversized. * * The cache can exceed its maximum size limit via the following * mechanisms: * * First, it is possible for the cache to grow without * bound as long as entries are protected and not unprotected. * * Second, when writes are not permitted it is also possible * for the cache to grow without bound. * * Third, the user may choose to disable evictions -- causing * the cache to grow without bound until evictions are * re-enabled. * * Finally, we usually don't check to see if the cache is * oversized at the end of an unprotect. As a result, it is * possible to have a vastly oversized cache with no protected * entries as long as all the protects preceed the unprotects. * * Since items 1, 2, and 3 are not changing any time soon, I * see no point in worrying about the fourth. */ result = H5C_make_space_in_cache(f, primary_dxpl_id, secondary_dxpl_id, space_needed, write_permitted, &first_flush); if ( result < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTPROTECT, NULL, \ "H5C_make_space_in_cache failed 1.") } } /* Insert the entry in the hash table. It can't be dirty yet, so * we don't even check to see if it should go in the skip list. * * This is no longer true -- due to a bug fix, we may modify * data on load to repair a file. */ H5C__INSERT_IN_INDEX(cache_ptr, entry_ptr, NULL) if ( ( entry_ptr->is_dirty ) && ( ! (entry_ptr->in_slist) ) ) { H5C__INSERT_ENTRY_IN_SLIST(cache_ptr, entry_ptr, NULL) } /* insert the entry in the data structures used by the replacement * policy. We are just going to take it out again when we update * the replacement policy for a protect, but this simplifies the * code. If we do this often enough, we may want to optimize this. */ H5C__UPDATE_RP_FOR_INSERTION(cache_ptr, entry_ptr, NULL) /* If the entry's type has a 'notify' callback send a 'after insertion' * notice now that the entry is fully integrated into the cache. */ if(entry_ptr->type->notify && (entry_ptr->type->notify)(H5C_NOTIFY_ACTION_AFTER_INSERT, entry_ptr) < 0) HGOTO_ERROR(H5E_CACHE, H5E_CANTNOTIFY, NULL, "can't notify client about entry inserted into cache") } HDassert( entry_ptr->addr == addr ); HDassert( entry_ptr->type == type ); if ( entry_ptr->is_protected ) { if ( ( read_only ) && ( entry_ptr->is_read_only ) ) { HDassert( entry_ptr->ro_ref_count > 0 ); (entry_ptr->ro_ref_count)++; } else { HGOTO_ERROR(H5E_CACHE, H5E_CANTPROTECT, NULL, \ "Target already protected & not read only?!?.") } } else { H5C__UPDATE_RP_FOR_PROTECT(cache_ptr, entry_ptr, NULL) entry_ptr->is_protected = TRUE; if ( read_only ) { entry_ptr->is_read_only = TRUE; entry_ptr->ro_ref_count = 1; } entry_ptr->dirtied = FALSE; } H5C__UPDATE_CACHE_HIT_RATE_STATS(cache_ptr, hit) H5C__UPDATE_STATS_FOR_PROTECT(cache_ptr, entry_ptr, hit) ret_value = thing; if ( ( cache_ptr->evictions_enabled ) && ( ( cache_ptr->size_decreased ) || ( ( cache_ptr->resize_enabled ) && ( cache_ptr->cache_accesses >= (cache_ptr->resize_ctl).epoch_length ) ) ) ) { if ( ! have_write_permitted ) { if ( cache_ptr->check_write_permitted != NULL ) { result = (cache_ptr->check_write_permitted)(f, primary_dxpl_id, &write_permitted); if ( result < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTPROTECT, NULL, \ "Can't get write_permitted 2") } else { have_write_permitted = TRUE; first_flush = TRUE; } } else { write_permitted = cache_ptr->write_permitted; have_write_permitted = TRUE; first_flush = TRUE; } } if ( ( cache_ptr->resize_enabled ) && ( cache_ptr->cache_accesses >= (cache_ptr->resize_ctl).epoch_length ) ) { result = H5C__auto_adjust_cache_size(f, primary_dxpl_id, secondary_dxpl_id, write_permitted, &first_flush); if ( result != SUCCEED ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTPROTECT, NULL, \ "Cache auto-resize failed.") } } if ( cache_ptr->size_decreased ) { cache_ptr->size_decreased = FALSE; /* check to see if the cache is now oversized due to the cache * size reduction. If it is, try to evict enough entries to * bring the cache size down to the current maximum cache size. * * Also, if the min_clean_size requirement is not met, we * should also call H5C_make_space_in_cache() to bring us * into complience. */ if ( cache_ptr->index_size >= cache_ptr->max_cache_size ) { empty_space = 0; } else { empty_space = cache_ptr->max_cache_size - cache_ptr->index_size; } if ( ( cache_ptr->index_size > cache_ptr->max_cache_size ) || ( ( empty_space + cache_ptr->clean_index_size ) < cache_ptr->min_clean_size) ) { if ( cache_ptr->index_size > cache_ptr->max_cache_size ) { cache_ptr->cache_full = TRUE; } result = H5C_make_space_in_cache(f, primary_dxpl_id, secondary_dxpl_id, (size_t)0, write_permitted, &first_flush); if ( result < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTPROTECT, NULL, \ "H5C_make_space_in_cache failed 2.") } } } } done: #if H5C_DO_EXTREME_SANITY_CHECKS if ( H5C_validate_lru_list(cache_ptr) < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, NULL, \ "LRU sanity check failed.\n"); } #endif /* H5C_DO_EXTREME_SANITY_CHECKS */ FUNC_LEAVE_NOAPI(ret_value) } /* H5C_protect() */ /*------------------------------------------------------------------------- * * Function: H5C_reset_cache_hit_rate_stats() * * Purpose: Reset the cache hit rate computation fields. * * Return: SUCCEED on success, and FAIL on failure. * * Programmer: John Mainzer, 10/5/04 * * Modifications: * * None. * *------------------------------------------------------------------------- */ herr_t H5C_reset_cache_hit_rate_stats(H5C_t * cache_ptr) { herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI(H5C_reset_cache_hit_rate_stats, FAIL) if ( ( cache_ptr == NULL ) || ( cache_ptr->magic != H5C__H5C_T_MAGIC ) ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Bad cache_ptr on entry.") } cache_ptr->cache_hits = 0; cache_ptr->cache_accesses = 0; done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C_reset_cache_hit_rate_stats() */ /*------------------------------------------------------------------------- * Function: H5C_set_cache_auto_resize_config * * Purpose: Set the cache automatic resize configuration to the * provided values if they are in range, and fail if they * are not. * * If the new configuration enables automatic cache resizing, * coerce the cache max size and min clean size into agreement * with the new policy and re-set the full cache hit rate * stats. * * Return: SUCCEED on success, and FAIL on failure. * * Programmer: John Mainzer * 10/8/04 * * Modifications: * * JRM -- 11/18/04 * Reworked function to match major changes in * H5C_auto_size_ctl_t. * * JRM -- 4/25/05 * Added code to set cache_ptr->size_decreased to TRUE * if the new configuration forces an immediate reduction * in cache size. * * JRM -- 12/31/07 * Added code supporting the new flash cache size increase * code. * *------------------------------------------------------------------------- */ herr_t H5C_set_cache_auto_resize_config(H5C_t *cache_ptr, H5C_auto_size_ctl_t *config_ptr) { herr_t result; size_t new_max_cache_size; size_t new_min_clean_size; herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI(H5C_set_cache_auto_resize_config, FAIL) if ( ( cache_ptr == NULL ) || ( cache_ptr->magic != H5C__H5C_T_MAGIC ) ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Bad cache_ptr on entry.") } if ( config_ptr == NULL ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "NULL config_ptr on entry.") } if ( config_ptr->version != H5C__CURR_AUTO_SIZE_CTL_VER ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Unknown config version.") } /* check general configuration section of the config: */ if ( SUCCEED != H5C_validate_resize_config(config_ptr, H5C_RESIZE_CFG__VALIDATE_GENERAL) ) { HGOTO_ERROR(H5E_ARGS, H5E_BADRANGE, FAIL, \ "error in general configuration fields of new config.") } /* check size increase control fields of the config: */ if ( SUCCEED != H5C_validate_resize_config(config_ptr, H5C_RESIZE_CFG__VALIDATE_INCREMENT) ) { HGOTO_ERROR(H5E_ARGS, H5E_BADRANGE, FAIL, \ "error in the size increase control fields of new config.") } /* check size decrease control fields of the config: */ if ( SUCCEED != H5C_validate_resize_config(config_ptr, H5C_RESIZE_CFG__VALIDATE_DECREMENT) ) { HGOTO_ERROR(H5E_ARGS, H5E_BADRANGE, FAIL, \ "error in the size decrease control fields of new config.") } /* check for conflicts between size increase and size decrease controls: */ if ( SUCCEED != H5C_validate_resize_config(config_ptr, H5C_RESIZE_CFG__VALIDATE_INTERACTIONS) ) { HGOTO_ERROR(H5E_ARGS, H5E_BADRANGE, FAIL, \ "conflicting threshold fields in new config.") } /* will set the increase possible fields to FALSE later if needed */ cache_ptr->size_increase_possible = TRUE; cache_ptr->flash_size_increase_possible = TRUE; cache_ptr->size_decrease_possible = TRUE; switch ( config_ptr->incr_mode ) { case H5C_incr__off: cache_ptr->size_increase_possible = FALSE; break; case H5C_incr__threshold: if ( ( config_ptr->lower_hr_threshold <= 0.0 ) || ( config_ptr->increment <= 1.0 ) || ( ( config_ptr->apply_max_increment ) && ( config_ptr->max_increment <= 0 ) ) ) { cache_ptr->size_increase_possible = FALSE; } break; default: /* should be unreachable */ HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Unknown incr_mode?!?!?.") } /* logically, this is were configuration for flash cache size increases * should go. However, this configuration depends on max_cache_size, so * we wait until the end of the function, when this field is set. */ switch ( config_ptr->decr_mode ) { case H5C_decr__off: cache_ptr->size_decrease_possible = FALSE; break; case H5C_decr__threshold: if ( ( config_ptr->upper_hr_threshold >= 1.0 ) || ( config_ptr->decrement >= 1.0 ) || ( ( config_ptr->apply_max_decrement ) && ( config_ptr->max_decrement <= 0 ) ) ) { cache_ptr->size_decrease_possible = FALSE; } break; case H5C_decr__age_out: if ( ( ( config_ptr->apply_empty_reserve ) && ( config_ptr->empty_reserve >= 1.0 ) ) || ( ( config_ptr->apply_max_decrement ) && ( config_ptr->max_decrement <= 0 ) ) ) { cache_ptr->size_decrease_possible = FALSE; } break; case H5C_decr__age_out_with_threshold: if ( ( ( config_ptr->apply_empty_reserve ) && ( config_ptr->empty_reserve >= 1.0 ) ) || ( ( config_ptr->apply_max_decrement ) && ( config_ptr->max_decrement <= 0 ) ) || ( config_ptr->upper_hr_threshold >= 1.0 ) ) { cache_ptr->size_decrease_possible = FALSE; } break; default: /* should be unreachable */ HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Unknown decr_mode?!?!?.") } if ( config_ptr->max_size == config_ptr->min_size ) { cache_ptr->size_increase_possible = FALSE; cache_ptr->flash_size_increase_possible = FALSE; cache_ptr->size_decrease_possible = FALSE; } /* flash_size_increase_possible is intentionally omitted from the * following: */ cache_ptr->resize_enabled = cache_ptr->size_increase_possible || cache_ptr->size_decrease_possible; cache_ptr->resize_ctl = *config_ptr; /* Resize the cache to the supplied initial value if requested, or as * necessary to force it within the bounds of the current automatic * cache resizing configuration. * * Note that the min_clean_fraction may have changed, so we * go through the exercise even if the current size is within * range and an initial size has not been provided. */ if ( (cache_ptr->resize_ctl).set_initial_size ) { new_max_cache_size = (cache_ptr->resize_ctl).initial_size; } else if ( cache_ptr->max_cache_size > (cache_ptr->resize_ctl).max_size ) { new_max_cache_size = (cache_ptr->resize_ctl).max_size; } else if ( cache_ptr->max_cache_size < (cache_ptr->resize_ctl).min_size ) { new_max_cache_size = (cache_ptr->resize_ctl).min_size; } else { new_max_cache_size = cache_ptr->max_cache_size; } new_min_clean_size = (size_t) ((double)new_max_cache_size * ((cache_ptr->resize_ctl).min_clean_fraction)); /* since new_min_clean_size is of type size_t, we have * * ( 0 <= new_min_clean_size ) * * by definition. */ HDassert( new_min_clean_size <= new_max_cache_size ); HDassert( (cache_ptr->resize_ctl).min_size <= new_max_cache_size ); HDassert( new_max_cache_size <= (cache_ptr->resize_ctl).max_size ); if ( new_max_cache_size < cache_ptr->max_cache_size ) { cache_ptr->size_decreased = TRUE; } cache_ptr->max_cache_size = new_max_cache_size; cache_ptr->min_clean_size = new_min_clean_size; if ( H5C_reset_cache_hit_rate_stats(cache_ptr) != SUCCEED ) { /* this should be impossible... */ HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "H5C_reset_cache_hit_rate_stats failed.") } /* remove excess epoch markers if any */ if ( ( config_ptr->decr_mode == H5C_decr__age_out_with_threshold ) || ( config_ptr->decr_mode == H5C_decr__age_out ) ) { if ( cache_ptr->epoch_markers_active > (cache_ptr->resize_ctl).epochs_before_eviction ) { result = H5C__autoadjust__ageout__remove_excess_markers(cache_ptr); if ( result != SUCCEED ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "can't remove excess epoch markers.") } } } else if ( cache_ptr->epoch_markers_active > 0 ) { result = H5C__autoadjust__ageout__remove_all_markers(cache_ptr); if ( result != SUCCEED ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "error removing all epoch markers.") } } /* configure flash size increase facility. We wait until the * end of the function, as we need the max_cache_size set before * we start to keep things simple. * * If we haven't already ruled out flash cache size increases above, * go ahead and configure it. */ if ( cache_ptr->flash_size_increase_possible ) { switch ( config_ptr->flash_incr_mode ) { case H5C_flash_incr__off: cache_ptr->flash_size_increase_possible = FALSE; break; case H5C_flash_incr__add_space: cache_ptr->flash_size_increase_possible = TRUE; cache_ptr->flash_size_increase_threshold = (size_t) (((double)(cache_ptr->max_cache_size)) * ((cache_ptr->resize_ctl).flash_threshold)); break; default: /* should be unreachable */ HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "Unknown flash_incr_mode?!?!?.") break; } } done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C_set_cache_auto_resize_config() */ /*------------------------------------------------------------------------- * Function: H5C_set_evictions_enabled() * * Purpose: Set cache_ptr->evictions_enabled to the value of the * evictions enabled parameter. * * Return: SUCCEED on success, and FAIL on failure. * * Programmer: John Mainzer * 7/27/07 * * Modifications: * * None. * *------------------------------------------------------------------------- */ herr_t H5C_set_evictions_enabled(H5C_t *cache_ptr, hbool_t evictions_enabled) { herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI(H5C_set_evictions_enabled, FAIL) if ( ( cache_ptr == NULL ) || ( cache_ptr->magic != H5C__H5C_T_MAGIC ) ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Bad cache_ptr on entry.") } if ( ( evictions_enabled != TRUE ) && ( evictions_enabled != FALSE ) ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "Bad evictions_enabled on entry.") } /* There is no fundamental reason why we should not permit * evictions to be disabled while automatic resize is enabled. * However, I can't think of any good reason why one would * want to, and allowing it would greatly complicate testing * the feature. Hence the following: */ if ( ( evictions_enabled != TRUE ) && ( ( cache_ptr->resize_ctl.incr_mode != H5C_incr__off ) || ( cache_ptr->resize_ctl.decr_mode != H5C_decr__off ) ) ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "Can't disable evictions when auto resize enabled.") } cache_ptr->evictions_enabled = evictions_enabled; done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C_set_evictions_enabled() */ /*------------------------------------------------------------------------- * Function: H5C_set_prefix * * Purpose: Set the values of the prefix field of H5C_t. This * filed is used to label some debugging output. * * Return: Non-negative on success/Negative on failure * * Programmer: John Mainzer * 1/20/06 * * Modifications: * *------------------------------------------------------------------------- */ herr_t H5C_set_prefix(H5C_t * cache_ptr, char * prefix) { herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI(H5C_set_prefix, FAIL) /* This would normally be an assert, but we need to use an HGOTO_ERROR * call to shut up the compiler. */ if ( ( ! cache_ptr ) || ( cache_ptr->magic != H5C__H5C_T_MAGIC ) ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Bad cache_ptr") } HDassert( prefix ); HDassert( HDstrlen(prefix) < H5C__PREFIX_LEN ) ; HDstrcpy(&(cache_ptr->prefix[0]), prefix); done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C_set_prefix() */ /*------------------------------------------------------------------------- * Function: H5C_set_skip_flags * * Purpose: Set the values of the skip sanity check flags. * * This function and the skip sanity check flags were created * for the convenience of the test bed. However it is * possible that there may be other uses for the flags. * * Return: Non-negative on success/Negative on failure * * Programmer: John Mainzer * 6/11/04 * * Modifications: * *------------------------------------------------------------------------- */ herr_t H5C_set_skip_flags(H5C_t * cache_ptr, hbool_t skip_file_checks, hbool_t skip_dxpl_id_checks) { herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI(H5C_set_skip_flags, FAIL) /* This would normally be an assert, but we need to use an HGOTO_ERROR * call to shut up the compiler. */ if ( ( ! cache_ptr ) || ( cache_ptr->magic != H5C__H5C_T_MAGIC ) ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Bad cache_ptr") } cache_ptr->skip_file_checks = skip_file_checks; cache_ptr->skip_dxpl_id_checks = skip_dxpl_id_checks; done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C_set_skip_flags() */ /*------------------------------------------------------------------------- * Function: H5C_set_trace_file_ptr * * Purpose: Set the trace_file_ptr field for the cache. * * This field must either be NULL (which turns of trace * file logging), or be a pointer to an open file to which * trace file data is to be written. * * Return: Non-negative on success/Negative on failure * * Programmer: John Mainzer * 1/20/06 * * Modifications: * *------------------------------------------------------------------------- */ herr_t H5C_set_trace_file_ptr(H5C_t * cache_ptr, FILE * trace_file_ptr) { herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI(H5C_set_trace_file_ptr, FAIL) /* This would normally be an assert, but we need to use an HGOTO_ERROR * call to shut up the compiler. */ if ( ( ! cache_ptr ) || ( cache_ptr->magic != H5C__H5C_T_MAGIC ) ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Bad cache_ptr") } cache_ptr->trace_file_ptr = trace_file_ptr; done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C_set_trace_file_ptr() */ /*------------------------------------------------------------------------- * Function: H5C_stats * * Purpose: Prints statistics about the cache. * * Return: Non-negative on success/Negative on failure * * Programmer: John Mainzer * 6/2/04 * * Modifications: * * JRM -- 7/21/04 * Updated function for the addition of the hash table. * * JRM -- 9/8/05 * Updated function for the addition of cache entry size * change statistics. * * JRM -- 1/13/06 * Added code to use the prefix field of H5C_t to allow * tagging of statistics output. * * JRM -- 3/21/06 * Added code supporting the pinned entry related stats. * * JRM -- 8/9/06 * More code supporting pinned entry related stats. * * JRM -- 8/23/06 * Added code supporting new flush related statistics. * * JRM -- 3/31/07 * Added code supporting the new write_protects, * read_protects, and max_read_protects fields. * * JRM -- 11/13/08 * Added code displaying the max_clean_index_size and * max_dirty_index_size. * * MAM -- 01/06/09 * Added code displaying the calls_to_msic, * total_entries_skipped_in_msic, total_entries_scanned_in_msic, * and max_entries_skipped_in_msic fields. * *------------------------------------------------------------------------- */ herr_t H5C_stats(H5C_t * cache_ptr, const char * cache_name, hbool_t #if !H5C_COLLECT_CACHE_STATS UNUSED #endif /* H5C_COLLECT_CACHE_STATS */ display_detailed_stats) { herr_t ret_value = SUCCEED; /* Return value */ #if H5C_COLLECT_CACHE_STATS int i; int64_t total_hits = 0; int64_t total_misses = 0; int64_t total_write_protects = 0; int64_t total_read_protects = 0; int64_t max_read_protects = 0; int64_t total_insertions = 0; int64_t total_pinned_insertions = 0; int64_t total_clears = 0; int64_t total_flushes = 0; int64_t total_evictions = 0; int64_t total_moves = 0; int64_t total_entry_flush_moves = 0; int64_t total_cache_flush_moves = 0; int64_t total_size_increases = 0; int64_t total_size_decreases = 0; int64_t total_entry_flush_size_changes = 0; int64_t total_cache_flush_size_changes = 0; int64_t total_pins = 0; int64_t total_unpins = 0; int64_t total_dirty_pins = 0; int64_t total_pinned_flushes = 0; int64_t total_pinned_clears = 0; int32_t aggregate_max_accesses = 0; int32_t aggregate_min_accesses = 1000000; int32_t aggregate_max_clears = 0; int32_t aggregate_max_flushes = 0; size_t aggregate_max_size = 0; int32_t aggregate_max_pins = 0; double hit_rate; double average_successful_search_depth = 0.0; double average_failed_search_depth = 0.0; double average_entries_skipped_per_calls_to_msic = 0.0; double average_entries_scanned_per_calls_to_msic = 0.0; #endif /* H5C_COLLECT_CACHE_STATS */ FUNC_ENTER_NOAPI(H5C_stats, FAIL) /* This would normally be an assert, but we need to use an HGOTO_ERROR * call to shut up the compiler. */ if ( ( ! cache_ptr ) || ( cache_ptr->magic != H5C__H5C_T_MAGIC ) || ( !cache_name ) ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Bad cache_ptr or cache_name") } #if H5C_COLLECT_CACHE_STATS for ( i = 0; i <= cache_ptr->max_type_id; i++ ) { total_hits += cache_ptr->hits[i]; total_misses += cache_ptr->misses[i]; total_write_protects += cache_ptr->write_protects[i]; total_read_protects += cache_ptr->read_protects[i]; if ( max_read_protects < cache_ptr->max_read_protects[i] ) { max_read_protects = cache_ptr->max_read_protects[i]; } total_insertions += cache_ptr->insertions[i]; total_pinned_insertions += cache_ptr->pinned_insertions[i]; total_clears += cache_ptr->clears[i]; total_flushes += cache_ptr->flushes[i]; total_evictions += cache_ptr->evictions[i]; total_moves += cache_ptr->moves[i]; total_entry_flush_moves += cache_ptr->entry_flush_moves[i]; total_cache_flush_moves += cache_ptr->cache_flush_moves[i]; total_size_increases += cache_ptr->size_increases[i]; total_size_decreases += cache_ptr->size_decreases[i]; total_entry_flush_size_changes += cache_ptr->entry_flush_size_changes[i]; total_cache_flush_size_changes += cache_ptr->cache_flush_size_changes[i]; total_pins += cache_ptr->pins[i]; total_unpins += cache_ptr->unpins[i]; total_dirty_pins += cache_ptr->dirty_pins[i]; total_pinned_flushes += cache_ptr->pinned_flushes[i]; total_pinned_clears += cache_ptr->pinned_clears[i]; #if H5C_COLLECT_CACHE_ENTRY_STATS if ( aggregate_max_accesses < cache_ptr->max_accesses[i] ) aggregate_max_accesses = cache_ptr->max_accesses[i]; if ( aggregate_min_accesses > aggregate_max_accesses ) aggregate_min_accesses = aggregate_max_accesses; if ( aggregate_min_accesses > cache_ptr->min_accesses[i] ) aggregate_min_accesses = cache_ptr->min_accesses[i]; if ( aggregate_max_clears < cache_ptr->max_clears[i] ) aggregate_max_clears = cache_ptr->max_clears[i]; if ( aggregate_max_flushes < cache_ptr->max_flushes[i] ) aggregate_max_flushes = cache_ptr->max_flushes[i]; if ( aggregate_max_size < cache_ptr->max_size[i] ) aggregate_max_size = cache_ptr->max_size[i]; if ( aggregate_max_pins < cache_ptr->max_pins[i] ) aggregate_max_pins = cache_ptr->max_pins[i]; #endif /* H5C_COLLECT_CACHE_ENTRY_STATS */ } if ( ( total_hits > 0 ) || ( total_misses > 0 ) ) { hit_rate = 100.0 * ((double)(total_hits)) / ((double)(total_hits + total_misses)); } else { hit_rate = 0.0; } if ( cache_ptr->successful_ht_searches > 0 ) { average_successful_search_depth = ((double)(cache_ptr->total_successful_ht_search_depth)) / ((double)(cache_ptr->successful_ht_searches)); } if ( cache_ptr->failed_ht_searches > 0 ) { average_failed_search_depth = ((double)(cache_ptr->total_failed_ht_search_depth)) / ((double)(cache_ptr->failed_ht_searches)); } HDfprintf(stdout, "\n%sH5C: cache statistics for %s\n", cache_ptr->prefix, cache_name); HDfprintf(stdout, "\n"); HDfprintf(stdout, "%s hash table insertion / deletions = %ld / %ld\n", cache_ptr->prefix, (long)(cache_ptr->total_ht_insertions), (long)(cache_ptr->total_ht_deletions)); HDfprintf(stdout, "%s HT successful / failed searches = %ld / %ld\n", cache_ptr->prefix, (long)(cache_ptr->successful_ht_searches), (long)(cache_ptr->failed_ht_searches)); HDfprintf(stdout, "%s Av. HT suc / failed search depth = %f / %f\n", cache_ptr->prefix, average_successful_search_depth, average_failed_search_depth); HDfprintf(stdout, "%s current (max) index size / length = %ld (%ld) / %ld (%ld)\n", cache_ptr->prefix, (long)(cache_ptr->index_size), (long)(cache_ptr->max_index_size), (long)(cache_ptr->index_len), (long)(cache_ptr->max_index_len)); HDfprintf(stdout, "%s current (max) clean/dirty idx size = %ld (%ld) / %ld (%ld)\n", cache_ptr->prefix, (long)(cache_ptr->clean_index_size), (long)(cache_ptr->max_clean_index_size), (long)(cache_ptr->dirty_index_size), (long)(cache_ptr->max_dirty_index_size)); HDfprintf(stdout, "%s current (max) slist size / length = %ld (%ld) / %ld (%ld)\n", cache_ptr->prefix, (long)(cache_ptr->slist_size), (long)(cache_ptr->max_slist_size), (long)(cache_ptr->slist_len), (long)(cache_ptr->max_slist_len)); HDfprintf(stdout, "%s current (max) PL size / length = %ld (%ld) / %ld (%ld)\n", cache_ptr->prefix, (long)(cache_ptr->pl_size), (long)(cache_ptr->max_pl_size), (long)(cache_ptr->pl_len), (long)(cache_ptr->max_pl_len)); HDfprintf(stdout, "%s current (max) PEL size / length = %ld (%ld) / %ld (%ld)\n", cache_ptr->prefix, (long)(cache_ptr->pel_size), (long)(cache_ptr->max_pel_size), (long)(cache_ptr->pel_len), (long)(cache_ptr->max_pel_len)); HDfprintf(stdout, "%s current LRU list size / length = %ld / %ld\n", cache_ptr->prefix, (long)(cache_ptr->LRU_list_size), (long)(cache_ptr->LRU_list_len)); HDfprintf(stdout, "%s current clean LRU size / length = %ld / %ld\n", cache_ptr->prefix, (long)(cache_ptr->cLRU_list_size), (long)(cache_ptr->cLRU_list_len)); HDfprintf(stdout, "%s current dirty LRU size / length = %ld / %ld\n", cache_ptr->prefix, (long)(cache_ptr->dLRU_list_size), (long)(cache_ptr->dLRU_list_len)); HDfprintf(stdout, "%s Total hits / misses / hit_rate = %ld / %ld / %f\n", cache_ptr->prefix, (long)total_hits, (long)total_misses, hit_rate); HDfprintf(stdout, "%s Total write / read (max) protects = %ld / %ld (%ld)\n", cache_ptr->prefix, (long)total_write_protects, (long)total_read_protects, (long)max_read_protects); HDfprintf(stdout, "%s Total clears / flushes / evictions = %ld / %ld / %ld\n", cache_ptr->prefix, (long)total_clears, (long)total_flushes, (long)total_evictions); HDfprintf(stdout, "%s Total insertions(pinned) / moves = %ld(%ld) / %ld\n", cache_ptr->prefix, (long)total_insertions, (long)total_pinned_insertions, (long)total_moves); HDfprintf(stdout, "%s Total entry / cache flush moves = %ld / %ld\n", cache_ptr->prefix, (long)total_entry_flush_moves, (long)total_cache_flush_moves); HDfprintf(stdout, "%s Total entry size incrs / decrs = %ld / %ld\n", cache_ptr->prefix, (long)total_size_increases, (long)total_size_decreases); HDfprintf(stdout, "%s Ttl entry/cache flush size changes = %ld / %ld\n", cache_ptr->prefix, (long)total_entry_flush_size_changes, (long)total_cache_flush_size_changes); HDfprintf(stdout, "%s Total entry pins (dirty) / unpins = %ld (%ld) / %ld\n", cache_ptr->prefix, (long)total_pins, (long)total_dirty_pins, (long)total_unpins); HDfprintf(stdout, "%s Total pinned flushes / clears = %ld / %ld\n", cache_ptr->prefix, (long)total_pinned_flushes, (long)total_pinned_clears); HDfprintf(stdout, "%s MSIC: (make space in cache) calls = %lld\n", cache_ptr->prefix, (long long)(cache_ptr->calls_to_msic)); if (cache_ptr->calls_to_msic > 0) { average_entries_skipped_per_calls_to_msic = (((double)(cache_ptr->total_entries_skipped_in_msic)) / ((double)(cache_ptr->calls_to_msic))); } HDfprintf(stdout, "%s MSIC: Average/max entries skipped = %lf / %ld\n", cache_ptr->prefix, (float)average_entries_skipped_per_calls_to_msic, (long)(cache_ptr->max_entries_skipped_in_msic)); if (cache_ptr->calls_to_msic > 0) { average_entries_scanned_per_calls_to_msic = (((double)(cache_ptr->total_entries_scanned_in_msic)) / ((double)(cache_ptr->calls_to_msic))); } HDfprintf(stdout, "%s MSIC: Average/max entries scanned = %lf / %ld\n", cache_ptr->prefix, (float)average_entries_scanned_per_calls_to_msic, (long)(cache_ptr->max_entries_scanned_in_msic)); HDfprintf(stdout, "%s MSIC: Scanned to make space(evict) = %lld\n", cache_ptr->prefix, (long long)(cache_ptr->entries_scanned_to_make_space)); HDfprintf(stdout, "%s MSIC: Scanned to satisfy min_clean = %lld\n", cache_ptr->prefix, (long long)(cache_ptr->total_entries_scanned_in_msic - cache_ptr->entries_scanned_to_make_space)); #if H5C_COLLECT_CACHE_ENTRY_STATS HDfprintf(stdout, "%s aggregate max / min accesses = %d / %d\n", cache_ptr->prefix, (int)aggregate_max_accesses, (int)aggregate_min_accesses); HDfprintf(stdout, "%s aggregate max_clears / max_flushes = %d / %d\n", cache_ptr->prefix, (int)aggregate_max_clears, (int)aggregate_max_flushes); HDfprintf(stdout, "%s aggregate max_size / max_pins = %d / %d\n", cache_ptr->prefix, (int)aggregate_max_size, (int)aggregate_max_pins); #endif /* H5C_COLLECT_CACHE_ENTRY_STATS */ if ( display_detailed_stats ) { for ( i = 0; i <= cache_ptr->max_type_id; i++ ) { HDfprintf(stdout, "\n"); HDfprintf(stdout, "%s Stats on %s:\n", cache_ptr->prefix, ((cache_ptr->type_name_table_ptr))[i]); if ( ( cache_ptr->hits[i] > 0 ) || ( cache_ptr->misses[i] > 0 ) ) { hit_rate = 100.0 * ((double)(cache_ptr->hits[i])) / ((double)(cache_ptr->hits[i] + cache_ptr->misses[i])); } else { hit_rate = 0.0; } HDfprintf(stdout, "%s hits / misses / hit_rate = %ld / %ld / %f\n", cache_ptr->prefix, (long)(cache_ptr->hits[i]), (long)(cache_ptr->misses[i]), hit_rate); HDfprintf(stdout, "%s write / read (max) protects = %ld / %ld (%d)\n", cache_ptr->prefix, (long)(cache_ptr->write_protects[i]), (long)(cache_ptr->read_protects[i]), (int)(cache_ptr->max_read_protects[i])); HDfprintf(stdout, "%s clears / flushes / evictions = %ld / %ld / %ld\n", cache_ptr->prefix, (long)(cache_ptr->clears[i]), (long)(cache_ptr->flushes[i]), (long)(cache_ptr->evictions[i])); HDfprintf(stdout, "%s insertions(pinned) / moves = %ld(%ld) / %ld\n", cache_ptr->prefix, (long)(cache_ptr->insertions[i]), (long)(cache_ptr->pinned_insertions[i]), (long)(cache_ptr->moves[i])); HDfprintf(stdout, "%s entry / cache flush moves = %ld / %ld\n", cache_ptr->prefix, (long)(cache_ptr->entry_flush_moves[i]), (long)(cache_ptr->cache_flush_moves[i])); HDfprintf(stdout, "%s size increases / decreases = %ld / %ld\n", cache_ptr->prefix, (long)(cache_ptr->size_increases[i]), (long)(cache_ptr->size_decreases[i])); HDfprintf(stdout, "%s entry/cache flush size changes = %ld / %ld\n", cache_ptr->prefix, (long)(cache_ptr->entry_flush_size_changes[i]), (long)(cache_ptr->cache_flush_size_changes[i])); HDfprintf(stdout, "%s entry pins / unpins = %ld / %ld\n", cache_ptr->prefix, (long)(cache_ptr->pins[i]), (long)(cache_ptr->unpins[i])); HDfprintf(stdout, "%s entry dirty pins/pin'd flushes = %ld / %ld\n", cache_ptr->prefix, (long)(cache_ptr->dirty_pins[i]), (long)(cache_ptr->pinned_flushes[i])); #if H5C_COLLECT_CACHE_ENTRY_STATS HDfprintf(stdout, "%s entry max / min accesses = %d / %d\n", cache_ptr->prefix, cache_ptr->max_accesses[i], cache_ptr->min_accesses[i]); HDfprintf(stdout, "%s entry max_clears / max_flushes = %d / %d\n", cache_ptr->prefix, cache_ptr->max_clears[i], cache_ptr->max_flushes[i]); HDfprintf(stdout, "%s entry max_size / max_pins = %d / %d\n", cache_ptr->prefix, (int)(cache_ptr->max_size[i]), (int)(cache_ptr->max_pins[i])); #endif /* H5C_COLLECT_CACHE_ENTRY_STATS */ } } HDfprintf(stdout, "\n"); #endif /* H5C_COLLECT_CACHE_STATS */ done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C_stats() */ /*------------------------------------------------------------------------- * * Function: H5C_stats__reset * * Purpose: Reset the stats fields to their initial values. * * Return: void * * Programmer: John Mainzer, 4/28/04 * * Modifications: * * JRM - 7/21/04 * Updated for hash table related statistics. * * JRM - 9/8/05 * Updated for size increase / decrease statistics. * * JRM - 3/20/06 * Updated for pin / unpin related statistics. * * JRM - 8/9/06 * Further updates for pin related statistics. * * JRM 8/23/06 * Added initialization code for new flush related statistics. * * JRM 2/16/07 * Added conditional compile code to avoid unused parameter * warning in the production build. * * JRM 3/31/07 * Added initialization for the new write_protects, * read_protects, and max_read_protects fields. * * JRM 11/13/08 * Added initialization for the new max_clean_index_size and * max_dirty_index_size fields. * * MAM -- 01/06/09 * Added code to initalize the calls_to_msic, * total_entries_skipped_in_msic, total_entries_scanned_in_msic, * and max_entries_skipped_in_msic fields. * *------------------------------------------------------------------------- */ void #ifndef NDEBUG H5C_stats__reset(H5C_t * cache_ptr) #else /* NDEBUG */ #if H5C_COLLECT_CACHE_STATS H5C_stats__reset(H5C_t * cache_ptr) #else /* H5C_COLLECT_CACHE_STATS */ H5C_stats__reset(H5C_t UNUSED * cache_ptr) #endif /* H5C_COLLECT_CACHE_STATS */ #endif /* NDEBUG */ { #if H5C_COLLECT_CACHE_STATS int i; #endif /* H5C_COLLECT_CACHE_STATS */ HDassert( cache_ptr ); HDassert( cache_ptr->magic == H5C__H5C_T_MAGIC ); #if H5C_COLLECT_CACHE_STATS for ( i = 0; i <= cache_ptr->max_type_id; i++ ) { cache_ptr->hits[i] = 0; cache_ptr->misses[i] = 0; cache_ptr->write_protects[i] = 0; cache_ptr->read_protects[i] = 0; cache_ptr->max_read_protects[i] = 0; cache_ptr->insertions[i] = 0; cache_ptr->pinned_insertions[i] = 0; cache_ptr->clears[i] = 0; cache_ptr->flushes[i] = 0; cache_ptr->evictions[i] = 0; cache_ptr->moves[i] = 0; cache_ptr->entry_flush_moves[i] = 0; cache_ptr->cache_flush_moves[i] = 0; cache_ptr->pins[i] = 0; cache_ptr->unpins[i] = 0; cache_ptr->dirty_pins[i] = 0; cache_ptr->pinned_flushes[i] = 0; cache_ptr->pinned_clears[i] = 0; cache_ptr->size_increases[i] = 0; cache_ptr->size_decreases[i] = 0; cache_ptr->entry_flush_size_changes[i] = 0; cache_ptr->cache_flush_size_changes[i] = 0; } cache_ptr->total_ht_insertions = 0; cache_ptr->total_ht_deletions = 0; cache_ptr->successful_ht_searches = 0; cache_ptr->total_successful_ht_search_depth = 0; cache_ptr->failed_ht_searches = 0; cache_ptr->total_failed_ht_search_depth = 0; cache_ptr->max_index_len = 0; cache_ptr->max_index_size = (size_t)0; cache_ptr->max_clean_index_size = (size_t)0; cache_ptr->max_dirty_index_size = (size_t)0; cache_ptr->max_slist_len = 0; cache_ptr->max_slist_size = (size_t)0; cache_ptr->max_pl_len = 0; cache_ptr->max_pl_size = (size_t)0; cache_ptr->max_pel_len = 0; cache_ptr->max_pel_size = (size_t)0; cache_ptr->calls_to_msic = 0; cache_ptr->total_entries_skipped_in_msic = 0; cache_ptr->total_entries_scanned_in_msic = 0; cache_ptr->max_entries_skipped_in_msic = 0; cache_ptr->max_entries_scanned_in_msic = 0; cache_ptr->entries_scanned_to_make_space = 0; #if H5C_COLLECT_CACHE_ENTRY_STATS for ( i = 0; i <= cache_ptr->max_type_id; i++ ) { cache_ptr->max_accesses[i] = 0; cache_ptr->min_accesses[i] = 1000000; cache_ptr->max_clears[i] = 0; cache_ptr->max_flushes[i] = 0; cache_ptr->max_size[i] = (size_t)0; cache_ptr->max_pins[i] = 0; } #endif /* H5C_COLLECT_CACHE_ENTRY_STATS */ #endif /* H5C_COLLECT_CACHE_STATS */ return; } /* H5C_stats__reset() */ /*------------------------------------------------------------------------- * Function: H5C_unpin_entry_from_client() * * Purpose: Internal routine to unpin a cache entry from a client action. * * Return: Non-negative on success/Negative on failure * * Programmer: Quincey Koziol * 3/24/09 * *------------------------------------------------------------------------- */ static herr_t H5C_unpin_entry_from_client(H5C_t * cache_ptr, H5C_cache_entry_t * entry_ptr, hbool_t update_rp) { herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI_NOINIT(H5C_unpin_entry_from_client) /* Sanity checking */ HDassert( cache_ptr ); HDassert( entry_ptr ); /* Error checking (should be sanity checks?) */ if(!entry_ptr->is_pinned) HGOTO_ERROR(H5E_CACHE, H5E_CANTUNPIN, FAIL, "Entry isn't pinned") if(!entry_ptr->pinned_from_client) HGOTO_ERROR(H5E_CACHE, H5E_CANTUNPIN, FAIL, "Entry wasn't pinned by cache client") /* Check if the entry is not pinned from a flush dependency */ if(!entry_ptr->pinned_from_cache) { /* If requested, update the replacement policy if the entry is not protected */ if(update_rp && !entry_ptr->is_protected) H5C__UPDATE_RP_FOR_UNPIN(cache_ptr, entry_ptr, FAIL) /* Unpin the entry now */ entry_ptr->is_pinned = FALSE; /* Update the stats for an unpin operation */ H5C__UPDATE_STATS_FOR_UNPIN(cache_ptr, entry_ptr) } /* end if */ /* Mark the entry as explicitly unpinned by the client */ entry_ptr->pinned_from_client = FALSE; done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C_unpin_entry_from_client() */ /*------------------------------------------------------------------------- * Function: H5C_unpin_entry() * * Purpose: Unpin a cache entry. The entry can be either protected or * unprotected at the time of call, but must be pinned. * * Return: Non-negative on success/Negative on failure * * Programmer: John Mainzer * 3/22/06 * *------------------------------------------------------------------------- */ herr_t H5C_unpin_entry(void *_entry_ptr) { H5C_t * cache_ptr; H5C_cache_entry_t * entry_ptr = (H5C_cache_entry_t *)_entry_ptr; /* Pointer to entry to unpin */ herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI(H5C_unpin_entry, FAIL) /* Sanity check */ HDassert(entry_ptr); cache_ptr = entry_ptr->cache_ptr; HDassert(cache_ptr); HDassert(cache_ptr->magic == H5C__H5C_T_MAGIC); /* Unpin the entry */ if(H5C_unpin_entry_from_client(cache_ptr, entry_ptr, TRUE) < 0) HGOTO_ERROR(H5E_CACHE, H5E_CANTUNPIN, FAIL, "Can't unpin entry from client") done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C_unpin_entry() */ /*------------------------------------------------------------------------- * Function: H5C_unprotect * * Purpose: Undo an H5C_protect() call -- specifically, mark the * entry as unprotected, remove it from the protected list, * and give it back to the replacement policy. * * The TYPE and ADDR arguments must be the same as those in * the corresponding call to H5C_protect() and the THING * argument must be the value returned by that call to * H5C_protect(). * * The primary_dxpl_id and secondary_dxpl_id parameters * specify the dxpl_ids used on the first write occasioned * by the unprotect (primary_dxpl_id), and on all subsequent * writes (secondary_dxpl_id). Since an uprotect cannot * occasion a write at present, all this is moot for now. * However, things change, and in any case, * H5C_flush_single_entry() needs primary_dxpl_id and * secondary_dxpl_id in its parameter list. * * The function can't cause a read either, so the dxpl_id * parameters are moot in this case as well. * * Return: Non-negative on success/Negative on failure * * If the deleted flag is TRUE, simply remove the target entry * from the cache, clear it, and free it without writing it to * disk. * * Return: Non-negative on success/Negative on failure * * Programmer: John Mainzer * 6/2/04 * * Modifications: * * JRM -- 7/21/04 * Updated the function for the addition of the hash table. * In particular, we now add dirty entries to the tree if * they aren't in the tree already. * * JRM -- 1/6/05 * Added the flags parameter, and code supporting * H5C__SET_FLUSH_MARKER_FLAG. Note that this flag is * ignored unless the new entry is dirty. Also note that * once the flush_marker field of an entry is set, the * only way it can be reset is by being flushed. * * JRM -- 6/3/05 * Added the dirtied parameter and supporting code. This * is part of an effort to move management of the is_dirty * field into the cache code. This has become necessary * to repair a cache coherency bug in PHDF5. * * JRM -- 7/5/05 * Added code supporting the new clear_on_unprotect field * of H5C_cache_entry_t. This change is also part of the * above mentioned cache coherency bug fix in PHDF5. * * JRM -- 9/8/05 * Added the size_changed and new_size parameters and the * supporting code. Since the metadata cache synchronizes * on dirty bytes creation in the PHDF5 case, we must now * track changes in entry size. * * Note that the new_size parameter is ignored unless the * size_changed parameter is TRUE. In this case, the new_size * must be positive. * * Also observe that if size_changed is TRUE, dirtied must be * TRUE. * * JRM -- 9/23/05 * Moved the size_changed parameter into flags. * * JRM -- 3/21/06 * Unpdated function to pin and unpin entries as directed via * the new H5C__PIN_ENTRY_FLAG and H5C__UNPIN_ENTRY_FLAG flags. * * JRM -- 5/3/06 * Added code to make use of the new dirtied field in * H5C_cache_entry_t. If this field is TRUE, it is the * equivalent of setting the H5C__DIRTIED_FLAG. * * JRM -- 3/29/07 * Modified function to allow a entry to be protected * more than once if the entry is protected read only. * * Also added sanity checks using the new is_read_only and * ro_ref_count parameters. * * JRM -- 12/31/07 * Modified function to support flash cache resizes. * * QAK -- 1/31/08 * Modified function to support freeing file space in client's * 'dest' callback routine. * * QAK -- 2/07/08 * Separated "destroy entry" concept from "remove entry from * cache" concept, by adding the 'take_ownership' flag. * * JRM -- 11/5/08 * Added code to update the clean_index_size and * dirty_index_size fields of H5C_t in cases where the * the entry was clean on protect, was marked dirty on * unprotect, and did not change its size. Do this via * a call to H5C__UPDATE_INDEX_FOR_ENTRY_DIRTY(). * * If the size changed, this case is already dealt with by * by the pre-existing call to * H5C__UPDATE_INDEX_FOR_SIZE_CHANGE(). * *------------------------------------------------------------------------- */ herr_t H5C_unprotect(H5F_t * f, hid_t primary_dxpl_id, hid_t secondary_dxpl_id, const H5C_class_t * type, haddr_t addr, void * thing, unsigned int flags, size_t new_size) { H5C_t * cache_ptr; hbool_t deleted; hbool_t dirtied; hbool_t set_flush_marker; hbool_t size_changed; hbool_t pin_entry; hbool_t unpin_entry; hbool_t free_file_space; hbool_t take_ownership; hbool_t was_clean; #ifdef H5_HAVE_PARALLEL hbool_t clear_entry = FALSE; #endif /* H5_HAVE_PARALLEL */ herr_t result; size_t size_increase = 0; H5C_cache_entry_t * entry_ptr; H5C_cache_entry_t * test_entry_ptr; herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI(H5C_unprotect, FAIL) deleted = ( (flags & H5C__DELETED_FLAG) != 0 ); dirtied = ( (flags & H5C__DIRTIED_FLAG) != 0 ); set_flush_marker = ( (flags & H5C__SET_FLUSH_MARKER_FLAG) != 0 ); size_changed = ( (flags & H5C__SIZE_CHANGED_FLAG) != 0 ); pin_entry = ( (flags & H5C__PIN_ENTRY_FLAG) != 0 ); unpin_entry = ( (flags & H5C__UNPIN_ENTRY_FLAG) != 0 ); free_file_space = ( (flags & H5C__FREE_FILE_SPACE_FLAG) != 0 ); take_ownership = ( (flags & H5C__TAKE_OWNERSHIP_FLAG) != 0 ); /* Changing the size of an entry dirties it. Thus, set the * dirtied flag if the size_changed flag is set. */ dirtied |= size_changed; HDassert( f ); HDassert( f->shared ); cache_ptr = f->shared->cache; HDassert( cache_ptr ); HDassert( cache_ptr->magic == H5C__H5C_T_MAGIC ); HDassert( cache_ptr->skip_file_checks || f ); HDassert( type ); HDassert( type->clear ); HDassert( type->flush ); HDassert( H5F_addr_defined(addr) ); HDassert( thing ); HDassert( ( size_changed == TRUE ) || ( size_changed == FALSE ) ); HDassert( ( ! size_changed ) || ( dirtied ) ); HDassert( ( ! size_changed ) || ( new_size > 0 ) ); HDassert( ! ( pin_entry && unpin_entry ) ); HDassert( ( ! free_file_space ) || ( deleted ) ); /* deleted flag must accompany free_file_space */ HDassert( ( ! take_ownership ) || ( deleted ) ); /* deleted flag must accompany take_ownership */ HDassert( ! ( free_file_space && take_ownership ) ); /* can't have both free_file_space & take_ownership */ entry_ptr = (H5C_cache_entry_t *)thing; HDassert( entry_ptr->addr == addr ); HDassert( entry_ptr->type == type ); /* also set the dirtied variable if the dirtied field is set in * the entry. */ dirtied |= entry_ptr->dirtied; was_clean = ! ( entry_ptr->is_dirty ); #if H5C_DO_EXTREME_SANITY_CHECKS if ( H5C_validate_lru_list(cache_ptr) < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "LRU sanity check failed.\n"); } #endif /* H5C_DO_EXTREME_SANITY_CHECKS */ /* if the entry has multiple read only protects, just decrement * the ro_ref_counter. Don't actually unprotect until the ref count * drops to zero. */ if ( entry_ptr->ro_ref_count > 1 ) { HDassert( entry_ptr->is_protected ); HDassert( entry_ptr->is_read_only ); if ( dirtied ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTUNPROTECT, FAIL, \ "Read only entry modified(1)??") } (entry_ptr->ro_ref_count)--; /* Pin or unpin the entry as requested. */ if ( pin_entry ) { /* Pin the entry from a client */ if(H5C_pin_entry_from_client(cache_ptr, entry_ptr) < 0) HGOTO_ERROR(H5E_CACHE, H5E_CANTPIN, FAIL, "Can't pin entry by client") } else if ( unpin_entry ) { /* Unpin the entry from a client */ if(H5C_unpin_entry_from_client(cache_ptr, entry_ptr, FALSE) < 0) HGOTO_ERROR(H5E_CACHE, H5E_CANTUNPIN, FAIL, "Can't unpin entry by client") } } else { if ( entry_ptr->is_read_only ) { HDassert( entry_ptr->ro_ref_count == 1 ); if ( dirtied ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTUNPROTECT, FAIL, \ "Read only entry modified(2)??") } entry_ptr->is_read_only = FALSE; entry_ptr->ro_ref_count = 0; } #ifdef H5_HAVE_PARALLEL /* When the H5C code is used to implement the metadata cache in the * PHDF5 case, only the cache on process 0 is allowed to write to file. * All the other metadata caches must hold dirty entries until they * are told that the entries are clean. * * The clear_on_unprotect flag in the H5C_cache_entry_t structure * exists to deal with the case in which an entry is protected when * its cache receives word that the entry is now clean. In this case, * the clear_on_unprotect flag is set, and the entry is flushed with * the H5C__FLUSH_CLEAR_ONLY_FLAG. * * All this is a bit awkward, but until the metadata cache entries * are contiguous, with only one dirty flag, we have to let the supplied * functions deal with the reseting the is_dirty flag. */ if ( entry_ptr->clear_on_unprotect ) { HDassert( entry_ptr->is_dirty ); entry_ptr->clear_on_unprotect = FALSE; if ( ! dirtied ) { clear_entry = TRUE; } } #endif /* H5_HAVE_PARALLEL */ if ( ! (entry_ptr->is_protected) ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTUNPROTECT, FAIL, \ "Entry already unprotected??") } /* mark the entry as dirty if appropriate */ entry_ptr->is_dirty = ( (entry_ptr->is_dirty) || dirtied ); /* update for change in entry size if necessary */ if ( ( size_changed ) && ( entry_ptr->size != new_size ) ) { /* do a flash cache size increase if appropriate */ if ( cache_ptr->flash_size_increase_possible ) { if ( new_size > entry_ptr->size ) { size_increase = new_size - entry_ptr->size; if ( size_increase >= cache_ptr->flash_size_increase_threshold ) { result = H5C__flash_increase_cache_size(cache_ptr, entry_ptr->size, new_size); if ( result < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTUNPROTECT, FAIL, \ "H5C__flash_increase_cache_size failed.") } } } } /* update the protected list */ H5C__DLL_UPDATE_FOR_SIZE_CHANGE((cache_ptr->pl_len), \ (cache_ptr->pl_size), \ (entry_ptr->size), (new_size)); /* update the hash table */ H5C__UPDATE_INDEX_FOR_SIZE_CHANGE((cache_ptr), (entry_ptr->size), \ (new_size), (entry_ptr), \ (was_clean)); /* if the entry is in the skip list, update that too */ if ( entry_ptr->in_slist ) { H5C__UPDATE_SLIST_FOR_SIZE_CHANGE((cache_ptr), \ (entry_ptr->size),\ (new_size)); } /* update statistics just before changing the entry size */ H5C__UPDATE_STATS_FOR_ENTRY_SIZE_CHANGE((cache_ptr), (entry_ptr), \ (new_size)); /* finally, update the entry size proper */ entry_ptr->size = new_size; } else if ( ( was_clean ) && ( entry_ptr->is_dirty ) ) { H5C__UPDATE_INDEX_FOR_ENTRY_DIRTY(cache_ptr, entry_ptr) } /* Pin or unpin the entry as requested. */ if ( pin_entry ) { /* Pin the entry from a client */ if(H5C_pin_entry_from_client(cache_ptr, entry_ptr) < 0) HGOTO_ERROR(H5E_CACHE, H5E_CANTPIN, FAIL, "Can't pin entry by client") } else if ( unpin_entry ) { /* Unpin the entry from a client */ if(H5C_unpin_entry_from_client(cache_ptr, entry_ptr, FALSE) < 0) HGOTO_ERROR(H5E_CACHE, H5E_CANTUNPIN, FAIL, "Can't unpin entry by client") } /* H5C__UPDATE_RP_FOR_UNPROTECT will places the unprotected entry on * the pinned entry list if entry_ptr->is_pined is TRUE. */ H5C__UPDATE_RP_FOR_UNPROTECT(cache_ptr, entry_ptr, FAIL) entry_ptr->is_protected = FALSE; /* if the entry is dirty, 'or' its flush_marker with the set flush flag, * and then add it to the skip list if it isn't there already. */ if ( entry_ptr->is_dirty ) { entry_ptr->flush_marker |= set_flush_marker; if ( ! (entry_ptr->in_slist) ) { H5C__INSERT_ENTRY_IN_SLIST(cache_ptr, entry_ptr, FAIL) } } /* this implementation of the "deleted" option is a bit inefficient, as * we re-insert the entry to be deleted into the replacement policy * data structures, only to remove them again. Depending on how often * we do this, we may want to optimize a bit. * * On the other hand, this implementation is reasonably clean, and * makes good use of existing code. * JRM - 5/19/04 */ if ( deleted ) { /* the following first flush flag will never be used as we are * calling H5C_flush_single_entry with both the * H5C__FLUSH_CLEAR_ONLY_FLAG and H5C__FLUSH_INVALIDATE_FLAG flags. * However, it is needed for the function call. */ hbool_t dummy_first_flush = TRUE; unsigned flush_flags = (H5C__FLUSH_CLEAR_ONLY_FLAG | H5C__FLUSH_INVALIDATE_FLAG); /* we can't delete a pinned entry */ HDassert ( ! (entry_ptr->is_pinned ) ); /* verify that the target entry is in the cache. */ H5C__SEARCH_INDEX(cache_ptr, addr, test_entry_ptr, FAIL) if ( test_entry_ptr == NULL ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTUNPROTECT, FAIL, \ "entry not in hash table?!?.") } else if ( test_entry_ptr != entry_ptr ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTUNPROTECT, FAIL, \ "hash table contains multiple entries for addr?!?.") } /* Pass along 'free file space' flag to cache client */ entry_ptr->free_file_space_on_destroy = free_file_space; /* Set the "take ownership" flag for the flush, if needed */ if ( take_ownership) { flush_flags |= H5C__TAKE_OWNERSHIP_FLAG; } if ( H5C_flush_single_entry(f, primary_dxpl_id, secondary_dxpl_id, type, addr, flush_flags, &dummy_first_flush, TRUE) < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTUNPROTECT, FAIL, "Can't flush.") } } #ifdef H5_HAVE_PARALLEL else if ( clear_entry ) { /* the following first flush flag will never be used as we are * calling H5C_flush_single_entry with the * H5C__FLUSH_CLEAR_ONLY_FLAG flag. However, it is needed for * the function call. */ hbool_t dummy_first_flush = TRUE; /* verify that the target entry is in the cache. */ H5C__SEARCH_INDEX(cache_ptr, addr, test_entry_ptr, FAIL) if ( test_entry_ptr == NULL ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTUNPROTECT, FAIL, \ "entry not in hash table?!?.") } else if ( test_entry_ptr != entry_ptr ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTUNPROTECT, FAIL, \ "hash table contains multiple entries for addr?!?.") } if ( H5C_flush_single_entry(f, primary_dxpl_id, secondary_dxpl_id, type, addr, H5C__FLUSH_CLEAR_ONLY_FLAG, &dummy_first_flush, TRUE) < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTUNPROTECT, FAIL, "Can't clear.") } } #endif /* H5_HAVE_PARALLEL */ } H5C__UPDATE_STATS_FOR_UNPROTECT(cache_ptr) done: #if H5C_DO_EXTREME_SANITY_CHECKS if ( H5C_validate_lru_list(cache_ptr) < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "LRU sanity check failed.\n"); } #endif /* H5C_DO_EXTREME_SANITY_CHECKS */ FUNC_LEAVE_NOAPI(ret_value) } /* H5C_unprotect() */ /*------------------------------------------------------------------------- * Function: H5C_validate_resize_config() * * Purpose: Run a sanity check on the specified sections of the * provided instance of struct H5C_auto_size_ctl_t. * * Do nothing and return SUCCEED if no errors are detected, * and flag an error and return FAIL otherwise. * * Return: Non-negative on success/Negative on failure * * Programmer: John Mainzer * 3/23/05 * * Modifications: * * Added validation for the flash increment fields. * * JRM -- 12/31/07 * *------------------------------------------------------------------------- */ herr_t H5C_validate_resize_config(H5C_auto_size_ctl_t * config_ptr, unsigned int tests) { herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI(H5C_validate_resize_config, FAIL) if ( config_ptr == NULL ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "NULL config_ptr on entry.") } if ( config_ptr->version != H5C__CURR_AUTO_SIZE_CTL_VER ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Unknown config version.") } if ( (tests & H5C_RESIZE_CFG__VALIDATE_GENERAL) != 0 ) { if ( ( config_ptr->set_initial_size != TRUE ) && ( config_ptr->set_initial_size != FALSE ) ) { HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, \ "set_initial_size must be either TRUE or FALSE"); } if ( config_ptr->max_size > H5C__MAX_MAX_CACHE_SIZE ) { HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, "max_size too big"); } if ( config_ptr->min_size < H5C__MIN_MAX_CACHE_SIZE ) { HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, "min_size too small"); } if ( config_ptr->min_size > config_ptr->max_size ) { HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, "min_size > max_size"); } if ( ( config_ptr->set_initial_size ) && ( ( config_ptr->initial_size < config_ptr->min_size ) || ( config_ptr->initial_size > config_ptr->max_size ) ) ) { HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, \ "initial_size must be in the interval [min_size, max_size]"); } if ( ( config_ptr->min_clean_fraction < 0.0 ) || ( config_ptr->min_clean_fraction > 1.0 ) ) { HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, \ "min_clean_fraction must be in the interval [0.0, 1.0]"); } if ( config_ptr->epoch_length < H5C__MIN_AR_EPOCH_LENGTH ) { HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, "epoch_length too small"); } if ( config_ptr->epoch_length > H5C__MAX_AR_EPOCH_LENGTH ) { HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, "epoch_length too big"); } } /* H5C_RESIZE_CFG__VALIDATE_GENERAL */ if ( (tests & H5C_RESIZE_CFG__VALIDATE_INCREMENT) != 0 ) { if ( ( config_ptr->incr_mode != H5C_incr__off ) && ( config_ptr->incr_mode != H5C_incr__threshold ) ) { HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, "Invalid incr_mode"); } if ( config_ptr->incr_mode == H5C_incr__threshold ) { if ( ( config_ptr->lower_hr_threshold < 0.0 ) || ( config_ptr->lower_hr_threshold > 1.0 ) ) { HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, \ "lower_hr_threshold must be in the range [0.0, 1.0]"); } if ( config_ptr->increment < 1.0 ) { HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, \ "increment must be greater than or equal to 1.0"); } if ( ( config_ptr->apply_max_increment != TRUE ) && ( config_ptr->apply_max_increment != FALSE ) ) { HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, \ "apply_max_increment must be either TRUE or FALSE"); } /* no need to check max_increment, as it is a size_t, * and thus must be non-negative. */ } /* H5C_incr__threshold */ switch ( config_ptr->flash_incr_mode ) { case H5C_flash_incr__off: /* nothing to do here */ break; case H5C_flash_incr__add_space: if ( ( config_ptr->flash_multiple < 0.1 ) || ( config_ptr->flash_multiple > 10.0 ) ) { HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, \ "flash_multiple must be in the range [0.1, 10.0]"); } if ( ( config_ptr->flash_threshold < 0.1 ) || ( config_ptr->flash_threshold > 1.0 ) ) { HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, \ "flash_threshold must be in the range [0.1, 1.0]"); } break; default: HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, \ "Invalid flash_incr_mode"); break; } } /* H5C_RESIZE_CFG__VALIDATE_INCREMENT */ if ( (tests & H5C_RESIZE_CFG__VALIDATE_DECREMENT) != 0 ) { if ( ( config_ptr->decr_mode != H5C_decr__off ) && ( config_ptr->decr_mode != H5C_decr__threshold ) && ( config_ptr->decr_mode != H5C_decr__age_out ) && ( config_ptr->decr_mode != H5C_decr__age_out_with_threshold ) ) { HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, "Invalid decr_mode"); } if ( config_ptr->decr_mode == H5C_decr__threshold ) { if ( config_ptr->upper_hr_threshold > 1.0 ) { HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, \ "upper_hr_threshold must be <= 1.0"); } if ( ( config_ptr->decrement > 1.0 ) || ( config_ptr->decrement < 0.0 ) ) { HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, \ "decrement must be in the interval [0.0, 1.0]"); } /* no need to check max_decrement as it is a size_t * and thus must be non-negative. */ } /* H5C_decr__threshold */ if ( ( config_ptr->decr_mode == H5C_decr__age_out ) || ( config_ptr->decr_mode == H5C_decr__age_out_with_threshold ) ) { if ( config_ptr->epochs_before_eviction < 1 ) { HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, \ "epochs_before_eviction must be positive"); } if ( config_ptr->epochs_before_eviction > H5C__MAX_EPOCH_MARKERS ) { HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, \ "epochs_before_eviction too big"); } if ( ( config_ptr->apply_empty_reserve != TRUE ) && ( config_ptr->apply_empty_reserve != FALSE ) ) { HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, \ "apply_empty_reserve must be either TRUE or FALSE"); } if ( ( config_ptr->apply_empty_reserve ) && ( ( config_ptr->empty_reserve > 1.0 ) || ( config_ptr->empty_reserve < 0.0 ) ) ) { HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, \ "empty_reserve must be in the interval [0.0, 1.0]"); } /* no need to check max_decrement as it is a size_t * and thus must be non-negative. */ } /* H5C_decr__age_out || H5C_decr__age_out_with_threshold */ if ( config_ptr->decr_mode == H5C_decr__age_out_with_threshold ) { if ( ( config_ptr->upper_hr_threshold > 1.0 ) || ( config_ptr->upper_hr_threshold < 0.0 ) ) { HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, \ "upper_hr_threshold must be in the interval [0.0, 1.0]"); } } /* H5C_decr__age_out_with_threshold */ } /* H5C_RESIZE_CFG__VALIDATE_DECREMENT */ if ( (tests & H5C_RESIZE_CFG__VALIDATE_INTERACTIONS) != 0 ) { if ( ( config_ptr->incr_mode == H5C_incr__threshold ) && ( ( config_ptr->decr_mode == H5C_decr__threshold ) || ( config_ptr->decr_mode == H5C_decr__age_out_with_threshold ) ) && ( config_ptr->lower_hr_threshold >= config_ptr->upper_hr_threshold ) ) { HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, \ "conflicting threshold fields in config.") } } /* H5C_RESIZE_CFG__VALIDATE_INTERACTIONS */ done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C_validate_resize_config() */ /*------------------------------------------------------------------------- * Function: H5C_adjust_flush_dependency_rc() * * Purpose: "Atomicly" adjust flush dependency ref. counts for an entry, * as a result of a flush dependency child's height changing. * * Note: Entry will remain in flush dependency relationship with its * child entry (i.e. it's not going to get unpinned as a result * of this change), but change could trickle upward, if this * entry's height changes and it has a flush dependency parent. * * Return: Non-negative on success/Negative on failure * * Programmer: Quincey Koziol * 3/05/09 * *------------------------------------------------------------------------- */ static void H5C_adjust_flush_dependency_rc(H5C_cache_entry_t * cache_entry, unsigned old_child_height, unsigned new_child_height) { FUNC_ENTER_NOAPI_NOINIT_NOFUNC(H5C_adjust_flush_dependency_rc) /* Sanity checks */ HDassert(cache_entry); HDassert(cache_entry->is_pinned); HDassert(cache_entry->flush_dep_height > 0); HDassert(cache_entry->flush_dep_height < H5C__NUM_FLUSH_DEP_HEIGHTS); HDassert(cache_entry->child_flush_dep_height_rc[old_child_height] > 0); HDassert(old_child_height < H5C__NUM_FLUSH_DEP_HEIGHTS); HDassert(old_child_height != new_child_height); HDassert(new_child_height < H5C__NUM_FLUSH_DEP_HEIGHTS); /* Adjust ref. counts for entry's flush dependency children heights */ cache_entry->child_flush_dep_height_rc[new_child_height]++; cache_entry->child_flush_dep_height_rc[old_child_height]--; /* Check for flush dependency height of entry increasing */ if((new_child_height + 1) > cache_entry->flush_dep_height) { /* Check if entry has _its_ own parent flush dependency entry */ if(NULL != cache_entry->flush_dep_parent) { /* Adjust flush dependency ref. counts on entry's parent */ H5C_adjust_flush_dependency_rc(cache_entry->flush_dep_parent, cache_entry->flush_dep_height, new_child_height + 1); } /* end if */ /* Set new flush dependency height of entry */ cache_entry->flush_dep_height = new_child_height + 1; } /* end if */ else { /* Check for child's flush dep. height decreasing and ref. count of * old child height going to zero, it could mean the parent's * flush dependency height dropped. */ if((new_child_height < old_child_height) && ((old_child_height + 1) == cache_entry->flush_dep_height) && (0 == cache_entry->child_flush_dep_height_rc[old_child_height])) { int i; /* Local index variable */ /* Re-scan child flush dependency height ref. counts to determine * this entry's height. */ #ifndef NDEBUG for(i = (H5C__NUM_FLUSH_DEP_HEIGHTS - 1); i > (int)new_child_height; i--) HDassert(0 == cache_entry->child_flush_dep_height_rc[i]); #endif /* NDEBUG */ for(i = (int)new_child_height; i >= 0; i--) /* Check for child flush dependencies of this height */ if(cache_entry->child_flush_dep_height_rc[i] > 0) break; /* Sanity checks */ HDassert((unsigned)(i + 1) < cache_entry->flush_dep_height); /* Check if entry has _its_ own parent flush dependency entry */ if(NULL != cache_entry->flush_dep_parent) { /* Adjust flush dependency ref. counts on entry's parent */ H5C_adjust_flush_dependency_rc(cache_entry->flush_dep_parent, cache_entry->flush_dep_height, (unsigned)(i + 1)); } /* end if */ /* Set new flush dependency height of entry */ cache_entry->flush_dep_height = (unsigned)(i + 1); } /* end if */ } /* end else */ /* Post-conditions, for successful operation */ HDassert(cache_entry->is_pinned); HDassert(cache_entry->flush_dep_height > 0); HDassert(cache_entry->flush_dep_height <= H5C__NUM_FLUSH_DEP_HEIGHTS); HDassert(cache_entry->child_flush_dep_height_rc[new_child_height] > 0); FUNC_LEAVE_NOAPI_VOID } /* H5C_adjust_flush_dependency_rc() */ /*------------------------------------------------------------------------- * Function: H5C_create_flush_dependency() * * Purpose: Initiates a parent<->child entry flush dependency. The parent * entry must be pinned or protected at the time of call, and must * have all dependencies removed before the cache can shut down. * * Note: Flush dependencies in the cache indicate that a child entry * must be flushed to the file before its parent. (This is * currently used to implement Single-Writer/Multiple-Reader (SWMR) * I/O access for data structures in the file). * * Each child entry can have only one parent entry, but parent * entries can have >1 child entries. The flush dependency * height of a parent entry is one greater than the max. flush * dependency height of its children. * * Creating a flush dependency between two entries will also pin * the parent entry. * * Return: Non-negative on success/Negative on failure * * Programmer: Quincey Koziol * 3/05/09 * *------------------------------------------------------------------------- */ herr_t H5C_create_flush_dependency(void * parent_thing, void * child_thing) { H5C_t * cache_ptr; H5C_cache_entry_t * parent_entry = (H5C_cache_entry_t *)parent_thing; /* Ptr to parent thing's entry */ H5C_cache_entry_t * child_entry = (H5C_cache_entry_t *)child_thing; /* Ptr to child thing's entry */ #ifndef NDEBUG unsigned prev_flush_dep_height = parent_entry->flush_dep_height; /* Previous flush height for parent entry */ #endif /* NDEBUG */ herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI(H5C_create_flush_dependency, FAIL) /* Sanity checks */ HDassert(parent_entry); HDassert(parent_entry->magic == H5C__H5C_CACHE_ENTRY_T_MAGIC); HDassert(parent_entry->flush_dep_height <= H5C__NUM_FLUSH_DEP_HEIGHTS); HDassert(H5F_addr_defined(parent_entry->addr)); HDassert(child_entry); HDassert(child_entry->magic == H5C__H5C_CACHE_ENTRY_T_MAGIC); HDassert(H5F_addr_defined(child_entry->addr)); HDassert(child_entry->flush_dep_height <= H5C__NUM_FLUSH_DEP_HEIGHTS); cache_ptr = parent_entry->cache_ptr; HDassert(cache_ptr); HDassert(cache_ptr->magic == H5C__H5C_T_MAGIC); HDassert(cache_ptr == child_entry->cache_ptr); /* More sanity checks */ if(child_entry == parent_entry) HGOTO_ERROR(H5E_CACHE, H5E_CANTDEPEND, FAIL, "Child entry flush dependency parent can't be itself") if(!(parent_entry->is_protected || parent_entry->is_pinned)) HGOTO_ERROR(H5E_CACHE, H5E_CANTDEPEND, FAIL, "Parent entry isn't pinned or protected") if(NULL != child_entry->flush_dep_parent) HGOTO_ERROR(H5E_CACHE, H5E_CANTDEPEND, FAIL, "Child entry already has flush dependency parent") { H5C_cache_entry_t *tmp_entry = parent_entry; /* Temporary cache entry in flush dependency chain */ unsigned tmp_flush_height = 0; /* Different in heights of parent entry */ /* Find the top entry in the flush dependency list */ while(NULL != tmp_entry->flush_dep_parent) { tmp_flush_height++; tmp_entry = tmp_entry->flush_dep_parent; } /* end while */ /* Check if we will make the dependency chain too long */ if((tmp_flush_height + child_entry->flush_dep_height + 1) > H5C__NUM_FLUSH_DEP_HEIGHTS) HGOTO_ERROR(H5E_CACHE, H5E_CANTDEPEND, FAIL, "Combined flush dependency height too large") } /* Check for parent not pinned */ if(!parent_entry->is_pinned) { /* Sanity check */ HDassert(parent_entry->flush_dep_height == 0); HDassert(!parent_entry->pinned_from_client); HDassert(!parent_entry->pinned_from_cache); /* Pin the parent entry */ parent_entry->is_pinned = TRUE; H5C__UPDATE_STATS_FOR_PIN(cache_ptr, parent_entry) } /* end else */ /* Mark the entry as pinned from the cache's action (possibly redundantly) */ parent_entry->pinned_from_cache = TRUE; /* Increment ref. count for parent's flush dependency children heights */ parent_entry->child_flush_dep_height_rc[child_entry->flush_dep_height]++; /* Check for increasing parent flush dependency height */ if((child_entry->flush_dep_height + 1) > parent_entry->flush_dep_height) { /* Check if parent entry has _its_ own parent flush dependency entry */ if(NULL != parent_entry->flush_dep_parent) { /* Adjust flush dependency ref. counts on parent entry's parent */ H5C_adjust_flush_dependency_rc(parent_entry->flush_dep_parent, parent_entry->flush_dep_height, (child_entry->flush_dep_height + 1)); } /* end if */ /* Increase flush dependency height of parent entry */ parent_entry->flush_dep_height = child_entry->flush_dep_height + 1; } /* end if */ /* Set parent for child entry */ child_entry->flush_dep_parent = parent_entry; /* Post-conditions, for successful operation */ HDassert(parent_entry->is_pinned); HDassert(parent_entry->flush_dep_height > 0); HDassert(parent_entry->flush_dep_height < H5C__NUM_FLUSH_DEP_HEIGHTS); HDassert(prev_flush_dep_height <= parent_entry->flush_dep_height); HDassert(parent_entry->child_flush_dep_height_rc[child_entry->flush_dep_height] > 0); HDassert(NULL != child_entry->flush_dep_parent); done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C_create_flush_dependency() */ /*------------------------------------------------------------------------- * Function: H5C_destroy_flush_dependency() * * Purpose: Terminates a parent<-> child entry flush dependency. The * parent entry must be pinned and have a positive flush * dependency height (which could go to zero as a result of * this operation). * * Return: Non-negative on success/Negative on failure * * Programmer: Quincey Koziol * 3/05/09 * *------------------------------------------------------------------------- */ herr_t H5C_destroy_flush_dependency(void *parent_thing, void * child_thing) { H5C_t * cache_ptr; H5C_cache_entry_t * parent_entry = (H5C_cache_entry_t *)parent_thing; /* Ptr to parent entry */ H5C_cache_entry_t * child_entry = (H5C_cache_entry_t *)child_thing; /* Ptr to child entry */ #ifndef NDEBUG unsigned prev_flush_dep_height = parent_entry->flush_dep_height; /* Previous flush height for parent entry */ #endif /* NDEBUG */ herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI(H5C_destroy_flush_dependency, FAIL) /* Sanity checks */ HDassert(parent_entry); HDassert(parent_entry->magic == H5C__H5C_CACHE_ENTRY_T_MAGIC); HDassert(H5F_addr_defined(parent_entry->addr)); HDassert(parent_entry->flush_dep_height <= H5C__NUM_FLUSH_DEP_HEIGHTS); HDassert(child_entry); HDassert(child_entry->flush_dep_parent != child_entry); HDassert(child_entry->magic == H5C__H5C_CACHE_ENTRY_T_MAGIC); HDassert(H5F_addr_defined(child_entry->addr)); cache_ptr = parent_entry->cache_ptr; HDassert(cache_ptr); HDassert(cache_ptr->magic == H5C__H5C_T_MAGIC); HDassert(cache_ptr == child_entry->cache_ptr); /* Usage checks */ if(!parent_entry->is_pinned) HGOTO_ERROR(H5E_CACHE, H5E_CANTUNDEPEND, FAIL, "Parent entry isn't pinned") if(0 == parent_entry->flush_dep_height) HGOTO_ERROR(H5E_CACHE, H5E_CANTUNDEPEND, FAIL, "Parent entry isn't a flush dependency parent") if(NULL == child_entry->flush_dep_parent) HGOTO_ERROR(H5E_CACHE, H5E_CANTUNDEPEND, FAIL, "Child entry doesn't have a flush dependency parent") if(0 == parent_entry->child_flush_dep_height_rc[child_entry->flush_dep_height]) HGOTO_ERROR(H5E_CACHE, H5E_CANTUNDEPEND, FAIL, "Parent entry flush dependency ref. count has no child entries of this height") if(child_entry->flush_dep_parent != parent_entry) HGOTO_ERROR(H5E_CACHE, H5E_CANTUNDEPEND, FAIL, "Parent entry isn't flush dependency parent for child entry") /* Decrement the ref. count for flush dependency height of children for parent entry */ parent_entry->child_flush_dep_height_rc[child_entry->flush_dep_height]--; /* Check for flush dependency ref. count at this height going to zero and * parent entry flush dependency height dropping */ if(((child_entry->flush_dep_height + 1) == parent_entry->flush_dep_height) && 0 == parent_entry->child_flush_dep_height_rc[child_entry->flush_dep_height]) { int i; /* Local index variable */ /* Reverse scan for new flush dependency height of parent */ #ifndef NDEBUG for(i = (H5C__NUM_FLUSH_DEP_HEIGHTS - 1); i > (int)child_entry->flush_dep_height; i--) HDassert(0 == parent_entry->child_flush_dep_height_rc[i]); #endif /* NDEBUG */ for(i = (int)child_entry->flush_dep_height; i >= 0; i--) /* Check for child flush dependencies of this height */ if(parent_entry->child_flush_dep_height_rc[i] > 0) break; /* Sanity check */ HDassert((unsigned)(i + 1) < parent_entry->flush_dep_height); /* Check if parent entry is a child in another flush dependency relationship */ if(NULL != parent_entry->flush_dep_parent) { /* Change flush dependency ref. counts of parent's parent */ H5C_adjust_flush_dependency_rc(parent_entry->flush_dep_parent, parent_entry->flush_dep_height, (unsigned)(i + 1)); } /* end if */ /* Increase flush dependency height of parent entry */ parent_entry->flush_dep_height = (unsigned)(i + 1); /* Check for height of parent dropping to zero (i.e. no longer a * parent of _any_ child flush dependencies). */ if(0 == parent_entry->flush_dep_height) { /* Sanity check */ HDassert(parent_entry->pinned_from_cache); /* Check if we should unpin parent entry now */ if(!parent_entry->pinned_from_client) { /* Update the replacement policy if the entry is not protected */ if(!parent_entry->is_protected) H5C__UPDATE_RP_FOR_UNPIN(cache_ptr, parent_entry, FAIL) /* Unpin the entry now */ parent_entry->is_pinned = FALSE; /* Update the stats for an unpin operation */ H5C__UPDATE_STATS_FOR_UNPIN(cache_ptr, parent_entry) } /* end if */ /* Mark the entry as unpinned from the cache's action */ parent_entry->pinned_from_cache = FALSE; } /* end if */ } /* end if */ /* Reset parent of child entry */ child_entry->flush_dep_parent = NULL; /* Post-conditions, for successful operation */ HDassert(prev_flush_dep_height >= parent_entry->flush_dep_height); HDassert(NULL == child_entry->flush_dep_parent); done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C_destroy_flush_dependency() */ /*************************************************************************/ /**************************** Private Functions: *************************/ /*************************************************************************/ /*------------------------------------------------------------------------- * * Function: H5C__auto_adjust_cache_size * * Purpose: Obtain the current full cache hit rate, and compare it * with the hit rate thresholds for modifying cache size. * If one of the thresholds has been crossed, adjusts the * size of the cache accordingly. * * The function then resets the full cache hit rate * statistics, and exits. * * Return: Non-negative on success/Negative on failure or if there was * an attempt to flush a protected item. * * * Programmer: John Mainzer, 10/7/04 * * Modifications: * * JRM -- 11/18/04 * Major re-write to support ageout method of cache size * reduction, and to adjust to changes in the * H5C_auto_size_ctl_t structure. * * JRM -- 1/5/08 * Added support for flash cache size increases. * *------------------------------------------------------------------------- */ static herr_t H5C__auto_adjust_cache_size(H5F_t * f, hid_t primary_dxpl_id, hid_t secondary_dxpl_id, hbool_t write_permitted, hbool_t * first_flush_ptr) { H5C_t * cache_ptr = f->shared->cache; herr_t result; hbool_t inserted_epoch_marker = FALSE; size_t new_max_cache_size = 0; size_t old_max_cache_size = 0; size_t new_min_clean_size = 0; size_t old_min_clean_size = 0; double hit_rate; enum H5C_resize_status status = in_spec; /* will change if needed */ herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI_NOINIT(H5C__auto_adjust_cache_size) HDassert( f ); HDassert( cache_ptr ); HDassert( cache_ptr->magic == H5C__H5C_T_MAGIC ); HDassert( cache_ptr->cache_accesses >= (cache_ptr->resize_ctl).epoch_length ); HDassert( 0.0 <= (cache_ptr->resize_ctl).min_clean_fraction ); HDassert( (cache_ptr->resize_ctl).min_clean_fraction <= 100.0 ); if ( !cache_ptr->resize_enabled ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Auto cache resize disabled.") } HDassert( ( (cache_ptr->resize_ctl).incr_mode != H5C_incr__off ) || \ ( (cache_ptr->resize_ctl).decr_mode != H5C_decr__off ) ); if ( H5C_get_cache_hit_rate(cache_ptr, &hit_rate) != SUCCEED ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Can't get hit rate.") } HDassert( ( 0.0 <= hit_rate ) && ( hit_rate <= 1.0 ) ); switch ( (cache_ptr->resize_ctl).incr_mode ) { case H5C_incr__off: if ( cache_ptr->size_increase_possible ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "size_increase_possible but H5C_incr__off?!?!?") } break; case H5C_incr__threshold: if ( hit_rate < (cache_ptr->resize_ctl).lower_hr_threshold ) { if ( ! cache_ptr->size_increase_possible ) { status = increase_disabled; } else if ( cache_ptr->max_cache_size >= (cache_ptr->resize_ctl).max_size ) { HDassert( cache_ptr->max_cache_size == \ (cache_ptr->resize_ctl).max_size ); status = at_max_size; } else if ( ! cache_ptr->cache_full ) { status = not_full; } else { new_max_cache_size = (size_t) (((double)(cache_ptr->max_cache_size)) * (cache_ptr->resize_ctl).increment); /* clip to max size if necessary */ if ( new_max_cache_size > (cache_ptr->resize_ctl).max_size ) { new_max_cache_size = (cache_ptr->resize_ctl).max_size; } /* clip to max increment if necessary */ if ( ( (cache_ptr->resize_ctl).apply_max_increment ) && ( (cache_ptr->max_cache_size + (cache_ptr->resize_ctl).max_increment) < new_max_cache_size ) ) { new_max_cache_size = cache_ptr->max_cache_size + (cache_ptr->resize_ctl).max_increment; } status = increase; } } break; default: HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "unknown incr_mode.") } /* If the decr_mode is either age out or age out with threshold, we * must run the marker maintenance code, whether we run the size * reduction code or not. We do this in two places -- here we * insert a new marker if the number of active epoch markers is * is less than the the current epochs before eviction, and after * the ageout call, we cycle the markers. * * However, we can't call the ageout code or cycle the markers * unless there was a full complement of markers in place on * entry. The inserted_epoch_marker flag is used to track this. */ if ( ( ( (cache_ptr->resize_ctl).decr_mode == H5C_decr__age_out ) || ( (cache_ptr->resize_ctl).decr_mode == H5C_decr__age_out_with_threshold ) ) && ( cache_ptr->epoch_markers_active < (cache_ptr->resize_ctl).epochs_before_eviction ) ) { result = H5C__autoadjust__ageout__insert_new_marker(cache_ptr); if ( result != SUCCEED ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "can't insert new epoch marker.") } else { inserted_epoch_marker = TRUE; } } /* don't run the cache size decrease code unless the cache size * increase code is disabled, or the size increase code sees no need * for action. In either case, status == in_spec at this point. */ if ( status == in_spec ) { switch ( (cache_ptr->resize_ctl).decr_mode ) { case H5C_decr__off: break; case H5C_decr__threshold: if ( hit_rate > (cache_ptr->resize_ctl).upper_hr_threshold ) { if ( ! cache_ptr->size_decrease_possible ) { status = decrease_disabled; } else if ( cache_ptr->max_cache_size <= (cache_ptr->resize_ctl).min_size ) { HDassert( cache_ptr->max_cache_size == (cache_ptr->resize_ctl).min_size ); status = at_min_size; } else { new_max_cache_size = (size_t) (((double)(cache_ptr->max_cache_size)) * (cache_ptr->resize_ctl).decrement); /* clip to min size if necessary */ if ( new_max_cache_size < (cache_ptr->resize_ctl).min_size ) { new_max_cache_size = (cache_ptr->resize_ctl).min_size; } /* clip to max decrement if necessary */ if ( ( (cache_ptr->resize_ctl).apply_max_decrement ) && ( ((cache_ptr->resize_ctl).max_decrement + new_max_cache_size) < cache_ptr->max_cache_size ) ) { new_max_cache_size = cache_ptr->max_cache_size - (cache_ptr->resize_ctl).max_decrement; } status = decrease; } } break; case H5C_decr__age_out_with_threshold: case H5C_decr__age_out: if ( ! inserted_epoch_marker ) { if ( ! cache_ptr->size_decrease_possible ) { status = decrease_disabled; } else { result = H5C__autoadjust__ageout(f, hit_rate, &status, &new_max_cache_size, primary_dxpl_id, secondary_dxpl_id, write_permitted, first_flush_ptr); if ( result != SUCCEED ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "ageout code failed.") } } } break; default: HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "unknown incr_mode.") } } /* cycle the epoch markers here if appropriate */ if ( ( ( (cache_ptr->resize_ctl).decr_mode == H5C_decr__age_out ) || ( (cache_ptr->resize_ctl).decr_mode == H5C_decr__age_out_with_threshold ) ) && ( ! inserted_epoch_marker ) ) { /* move last epoch marker to the head of the LRU list */ result = H5C__autoadjust__ageout__cycle_epoch_marker(cache_ptr); if ( result != SUCCEED ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "error cycling epoch marker.") } } if ( ( status == increase ) || ( status == decrease ) ) { old_max_cache_size = cache_ptr->max_cache_size; old_min_clean_size = cache_ptr->min_clean_size; new_min_clean_size = (size_t) ((double)new_max_cache_size * ((cache_ptr->resize_ctl).min_clean_fraction)); /* new_min_clean_size is of size_t, and thus must be non-negative. * Hence we have * * ( 0 <= new_min_clean_size ). * * by definition. */ HDassert( new_min_clean_size <= new_max_cache_size ); HDassert( (cache_ptr->resize_ctl).min_size <= new_max_cache_size ); HDassert( new_max_cache_size <= (cache_ptr->resize_ctl).max_size ); cache_ptr->max_cache_size = new_max_cache_size; cache_ptr->min_clean_size = new_min_clean_size; if ( status == increase ) { cache_ptr->cache_full = FALSE; } else if ( status == decrease ) { cache_ptr->size_decreased = TRUE; } /* update flash cache size increase fields as appropriate */ if ( cache_ptr->flash_size_increase_possible ) { switch ( (cache_ptr->resize_ctl).flash_incr_mode ) { case H5C_flash_incr__off: HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "flash_size_increase_possible but H5C_flash_incr__off?!") break; case H5C_flash_incr__add_space: cache_ptr->flash_size_increase_threshold = (size_t) (((double)(cache_ptr->max_cache_size)) * ((cache_ptr->resize_ctl).flash_threshold)); break; default: /* should be unreachable */ HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "Unknown flash_incr_mode?!?!?.") break; } } } if ( (cache_ptr->resize_ctl).rpt_fcn != NULL ) { (*((cache_ptr->resize_ctl).rpt_fcn)) (cache_ptr, H5C__CURR_AUTO_RESIZE_RPT_FCN_VER, hit_rate, status, old_max_cache_size, new_max_cache_size, old_min_clean_size, new_min_clean_size); } if ( H5C_reset_cache_hit_rate_stats(cache_ptr) != SUCCEED ) { /* this should be impossible... */ HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "H5C_reset_cache_hit_rate_stats failed.") } done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C__auto_adjust_cache_size() */ /*------------------------------------------------------------------------- * * Function: H5C__autoadjust__ageout * * Purpose: Implement the ageout automatic cache size decrement * algorithm. Note that while this code evicts aged out * entries, the code does not change the maximum cache size. * Instead, the function simply computes the new value (if * any change is indicated) and reports this value in * *new_max_cache_size_ptr. * * Return: Non-negative on success/Negative on failure or if there was * an attempt to flush a protected item. * * * Programmer: John Mainzer, 11/18/04 * * Modifications: * * None. * *------------------------------------------------------------------------- */ static herr_t H5C__autoadjust__ageout(H5F_t * f, double hit_rate, enum H5C_resize_status * status_ptr, size_t * new_max_cache_size_ptr, hid_t primary_dxpl_id, hid_t secondary_dxpl_id, hbool_t write_permitted, hbool_t * first_flush_ptr) { H5C_t * cache_ptr = f->shared->cache; herr_t result; size_t test_size; herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI_NOINIT(H5C__autoadjust__ageout) HDassert( f ); HDassert( cache_ptr ); HDassert( cache_ptr->magic == H5C__H5C_T_MAGIC ); HDassert( ( status_ptr ) && ( *status_ptr == in_spec ) ); HDassert( ( new_max_cache_size_ptr ) && ( *new_max_cache_size_ptr == 0 ) ); /* remove excess epoch markers if any */ if ( cache_ptr->epoch_markers_active > (cache_ptr->resize_ctl).epochs_before_eviction ) { result = H5C__autoadjust__ageout__remove_excess_markers(cache_ptr); if ( result != SUCCEED ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "can't remove excess epoch markers.") } } if ( ( (cache_ptr->resize_ctl).decr_mode == H5C_decr__age_out ) || ( ( (cache_ptr->resize_ctl).decr_mode == H5C_decr__age_out_with_threshold ) && ( hit_rate >= (cache_ptr->resize_ctl).upper_hr_threshold ) ) ) { if ( cache_ptr->max_cache_size > (cache_ptr->resize_ctl).min_size ){ /* evict aged out cache entries if appropriate... */ if(H5C__autoadjust__ageout__evict_aged_out_entries(f, primary_dxpl_id, secondary_dxpl_id, write_permitted, first_flush_ptr) < 0) HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "error flushing aged out entries.") /* ... and then reduce cache size if appropriate */ if ( cache_ptr->index_size < cache_ptr->max_cache_size ) { if ( (cache_ptr->resize_ctl).apply_empty_reserve ) { test_size = (size_t)(((double)cache_ptr->index_size) / (1 - (cache_ptr->resize_ctl).empty_reserve)); if ( test_size < cache_ptr->max_cache_size ) { *status_ptr = decrease; *new_max_cache_size_ptr = test_size; } } else { *status_ptr = decrease; *new_max_cache_size_ptr = cache_ptr->index_size; } if ( *status_ptr == decrease ) { /* clip to min size if necessary */ if ( *new_max_cache_size_ptr < (cache_ptr->resize_ctl).min_size ) { *new_max_cache_size_ptr = (cache_ptr->resize_ctl).min_size; } /* clip to max decrement if necessary */ if ( ( (cache_ptr->resize_ctl).apply_max_decrement ) && ( ((cache_ptr->resize_ctl).max_decrement + *new_max_cache_size_ptr) < cache_ptr->max_cache_size ) ) { *new_max_cache_size_ptr = cache_ptr->max_cache_size - (cache_ptr->resize_ctl).max_decrement; } } } } else { *status_ptr = at_min_size; } } done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C__autoadjust__ageout() */ /*------------------------------------------------------------------------- * * Function: H5C__autoadjust__ageout__cycle_epoch_marker * * Purpose: Remove the oldest epoch marker from the LRU list, * and reinsert it at the head of the LRU list. Also * remove the epoch marker's index from the head of the * ring buffer, and re-insert it at the tail of the ring * buffer. * * Return: SUCCEED on success/FAIL on failure. * * Programmer: John Mainzer, 11/22/04 * * Modifications: * * None. * *------------------------------------------------------------------------- */ static herr_t H5C__autoadjust__ageout__cycle_epoch_marker(H5C_t * cache_ptr) { herr_t ret_value = SUCCEED; /* Return value */ int i; FUNC_ENTER_NOAPI_NOINIT(H5C__autoadjust__ageout__cycle_epoch_marker) HDassert( cache_ptr ); HDassert( cache_ptr->magic == H5C__H5C_T_MAGIC ); if ( cache_ptr->epoch_markers_active <= 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "No active epoch markers on entry?!?!?.") } /* remove the last marker from both the ring buffer and the LRU list */ i = cache_ptr->epoch_marker_ringbuf[cache_ptr->epoch_marker_ringbuf_first]; cache_ptr->epoch_marker_ringbuf_first = (cache_ptr->epoch_marker_ringbuf_first + 1) % (H5C__MAX_EPOCH_MARKERS + 1); cache_ptr->epoch_marker_ringbuf_size -= 1; if ( cache_ptr->epoch_marker_ringbuf_size < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "ring buffer underflow.") } if ( (cache_ptr->epoch_marker_active)[i] != TRUE ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "unused marker in LRU?!?") } H5C__DLL_REMOVE((&((cache_ptr->epoch_markers)[i])), \ (cache_ptr)->LRU_head_ptr, \ (cache_ptr)->LRU_tail_ptr, \ (cache_ptr)->LRU_list_len, \ (cache_ptr)->LRU_list_size, \ (FAIL)) /* now, re-insert it at the head of the LRU list, and at the tail of * the ring buffer. */ HDassert( ((cache_ptr->epoch_markers)[i]).addr == (haddr_t)i ); HDassert( ((cache_ptr->epoch_markers)[i]).next == NULL ); HDassert( ((cache_ptr->epoch_markers)[i]).prev == NULL ); cache_ptr->epoch_marker_ringbuf_last = (cache_ptr->epoch_marker_ringbuf_last + 1) % (H5C__MAX_EPOCH_MARKERS + 1); (cache_ptr->epoch_marker_ringbuf)[cache_ptr->epoch_marker_ringbuf_last] = i; cache_ptr->epoch_marker_ringbuf_size += 1; if ( cache_ptr->epoch_marker_ringbuf_size > H5C__MAX_EPOCH_MARKERS ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "ring buffer overflow.") } H5C__DLL_PREPEND((&((cache_ptr->epoch_markers)[i])), \ (cache_ptr)->LRU_head_ptr, \ (cache_ptr)->LRU_tail_ptr, \ (cache_ptr)->LRU_list_len, \ (cache_ptr)->LRU_list_size, \ (FAIL)) done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C__autoadjust__ageout__cycle_epoch_marker() */ /*------------------------------------------------------------------------- * * Function: H5C__autoadjust__ageout__evict_aged_out_entries * * Purpose: Evict clean entries in the cache that haven't * been accessed for at least * (cache_ptr->resize_ctl).epochs_before_eviction epochs, * and flush dirty entries that haven't been accessed for * that amount of time. * * Depending on configuration, the function will either * flush or evict all such entries, or all such entries it * encounters until it has freed the maximum amount of space * allowed under the maximum decrement. * * If we are running in parallel mode, writes may not be * permitted. If so, the function simply skips any dirty * entries it may encounter. * * The function makes no attempt to maintain the minimum * clean size, as there is no guarantee that the cache size * will be changed. * * If there is no cache size change, the minimum clean size * constraint will be met through a combination of clean * entries and free space in the cache. * * If there is a cache size reduction, the minimum clean size * will be re-calculated, and will be enforced the next time * we have to make space in the cache. * * The primary_dxpl_id and secondary_dxpl_id parameters * specify the dxpl_ids used depending on the value of * *first_flush_ptr. The idea is to use the primary_dxpl_id * on the first write in a sequence of writes, and to use * the secondary_dxpl_id on all subsequent writes. * * This is useful in the metadata cache, but may not be * needed elsewhere. If so, just use the same dxpl_id for * both parameters. * * Observe that this function cannot occasion a read. * * Return: Non-negative on success/Negative on failure. * * Programmer: John Mainzer, 11/22/04 * * Modifications: * * JRM -- 10/13/07 * Added code to detect and manage the case in which a * flush callback changes the LRU-list out from under * the function. The only way I can think of in which this * can happen is if a flush function loads an entry * into the cache that isn't there already. Quincey tells * me that this will never happen, but I'm not sure I * believe him. * * Note that this is a pretty bad scenario if it ever * happens. The code I have added should allow us to * handle the situation under all but the worst conditions, * but one can argue that I should just scream and die if I * ever detect the condidtion. * *------------------------------------------------------------------------- */ static herr_t H5C__autoadjust__ageout__evict_aged_out_entries(H5F_t * f, hid_t primary_dxpl_id, hid_t secondary_dxpl_id, hbool_t write_permitted, hbool_t * first_flush_ptr) { H5C_t * cache_ptr = f->shared->cache; herr_t result; size_t eviction_size_limit; size_t bytes_evicted = 0; hbool_t prev_is_dirty = FALSE; H5C_cache_entry_t * entry_ptr; H5C_cache_entry_t * next_ptr; H5C_cache_entry_t * prev_ptr; herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI_NOINIT(H5C__autoadjust__ageout__evict_aged_out_entries) HDassert( f ); HDassert( cache_ptr ); HDassert( cache_ptr->magic == H5C__H5C_T_MAGIC ); /* if there is a limit on the amount that the cache size can be decrease * in any one round of the cache size reduction algorithm, load that * limit into eviction_size_limit. Otherwise, set eviction_size_limit * to the equivalent of infinity. The current size of the index will * do nicely. */ if ( (cache_ptr->resize_ctl).apply_max_decrement ) { eviction_size_limit = (cache_ptr->resize_ctl).max_decrement; } else { eviction_size_limit = cache_ptr->index_size; /* i.e. infinity */ } if ( write_permitted ) { entry_ptr = cache_ptr->LRU_tail_ptr; while ( ( entry_ptr != NULL ) && ( (entry_ptr->type)->id != H5C__EPOCH_MARKER_TYPE ) && ( bytes_evicted < eviction_size_limit ) ) { HDassert( ! (entry_ptr->is_protected) ); next_ptr = entry_ptr->next; prev_ptr = entry_ptr->prev; if ( prev_ptr != NULL ) { prev_is_dirty = prev_ptr->is_dirty; } if ( entry_ptr->is_dirty ) { result = H5C_flush_single_entry(f, primary_dxpl_id, secondary_dxpl_id, entry_ptr->type, entry_ptr->addr, H5C__NO_FLAGS_SET, first_flush_ptr, FALSE); } else { bytes_evicted += entry_ptr->size; result = H5C_flush_single_entry(f, primary_dxpl_id, secondary_dxpl_id, entry_ptr->type, entry_ptr->addr, H5C__FLUSH_INVALIDATE_FLAG, first_flush_ptr, TRUE); } if ( result < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTFLUSH, FAIL, \ "unable to flush entry") } if ( prev_ptr != NULL ) { #ifndef NDEBUG if ( prev_ptr->magic != H5C__H5C_CACHE_ENTRY_T_MAGIC ) { /* something horrible has happened to *prev_ptr -- * scream and die. */ HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "*prev_ptr corrupt") } else #endif /* NDEBUG */ if ( ( prev_ptr->is_dirty != prev_is_dirty ) || ( prev_ptr->next != next_ptr ) || ( prev_ptr->is_protected ) || ( prev_ptr->is_pinned ) ) { /* something has happened to the LRU -- start over * from the tail. */ entry_ptr = cache_ptr->LRU_tail_ptr; } else { entry_ptr = prev_ptr; } } else { entry_ptr = NULL; } } /* end while */ /* for now at least, don't bother to maintain the minimum clean size, * as the cache should now be less than its maximum size. Due to * the vaguries of the cache size reduction algorthim, we may not * reduce the size of the cache. * * If we do, we will calculate a new minimum clean size, which will * be enforced the next time we try to make space in the cache. * * If we don't, no action is necessary, as we have just evicted and/or * or flushed a bunch of entries and therefore the sum of the clean * and free space in the cache must be greater than or equal to the * min clean space requirement (assuming that requirement was met on * entry). */ } else /* ! write_permitted */ { /* since we are not allowed to write, all we can do is evict * any clean entries that we may encounter before we either * hit the eviction size limit, or encounter the epoch marker. * * If we are operating read only, this isn't an issue, as there * will not be any dirty entries. * * If we are operating in R/W mode, all the dirty entries we * skip will be flushed the next time we attempt to make space * when writes are permitted. This may have some local * performance implications, but it shouldn't cause any net * slowdown. */ HDassert( H5C_MAINTAIN_CLEAN_AND_DIRTY_LRU_LISTS ); entry_ptr = cache_ptr->LRU_tail_ptr; while ( ( entry_ptr != NULL ) && ( (entry_ptr->type)->id != H5C__EPOCH_MARKER_TYPE ) && ( bytes_evicted < eviction_size_limit ) ) { HDassert( ! (entry_ptr->is_protected) ); prev_ptr = entry_ptr->prev; if ( ! (entry_ptr->is_dirty) ) { result = H5C_flush_single_entry(f, primary_dxpl_id, secondary_dxpl_id, entry_ptr->type, entry_ptr->addr, H5C__FLUSH_INVALIDATE_FLAG, first_flush_ptr, TRUE); if ( result < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTFLUSH, FAIL, \ "unable to flush clean entry") } } /* just skip the entry if it is dirty, as we can't do * anything with it now since we can't write. */ entry_ptr = prev_ptr; } /* end while */ } if ( cache_ptr->index_size < cache_ptr->max_cache_size ) { cache_ptr->cache_full = FALSE; } done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C__autoadjust__ageout__evict_aged_out_entries() */ /*------------------------------------------------------------------------- * * Function: H5C__autoadjust__ageout__insert_new_marker * * Purpose: Find an unused marker cache entry, mark it as used, and * insert it at the head of the LRU list. Also add the * marker's index in the epoch_markers array. * * Return: SUCCEED on success/FAIL on failure. * * Programmer: John Mainzer, 11/19/04 * * Modifications: * * None. * *------------------------------------------------------------------------- */ static herr_t H5C__autoadjust__ageout__insert_new_marker(H5C_t * cache_ptr) { herr_t ret_value = SUCCEED; /* Return value */ int i; FUNC_ENTER_NOAPI_NOINIT(H5C__autoadjust__ageout__insert_new_marker) HDassert( cache_ptr ); HDassert( cache_ptr->magic == H5C__H5C_T_MAGIC ); if ( cache_ptr->epoch_markers_active >= (cache_ptr->resize_ctl).epochs_before_eviction ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "Already have a full complement of markers.") } /* find an unused marker */ i = 0; while ( ( (cache_ptr->epoch_marker_active)[i] ) && ( i < H5C__MAX_EPOCH_MARKERS ) ) { i++; } if(i >= H5C__MAX_EPOCH_MARKERS) HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Can't find unused marker.") HDassert( ((cache_ptr->epoch_markers)[i]).addr == (haddr_t)i ); HDassert( ((cache_ptr->epoch_markers)[i]).next == NULL ); HDassert( ((cache_ptr->epoch_markers)[i]).prev == NULL ); (cache_ptr->epoch_marker_active)[i] = TRUE; cache_ptr->epoch_marker_ringbuf_last = (cache_ptr->epoch_marker_ringbuf_last + 1) % (H5C__MAX_EPOCH_MARKERS + 1); (cache_ptr->epoch_marker_ringbuf)[cache_ptr->epoch_marker_ringbuf_last] = i; cache_ptr->epoch_marker_ringbuf_size += 1; if ( cache_ptr->epoch_marker_ringbuf_size > H5C__MAX_EPOCH_MARKERS ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "ring buffer overflow.") } H5C__DLL_PREPEND((&((cache_ptr->epoch_markers)[i])), \ (cache_ptr)->LRU_head_ptr, \ (cache_ptr)->LRU_tail_ptr, \ (cache_ptr)->LRU_list_len, \ (cache_ptr)->LRU_list_size, \ (FAIL)) cache_ptr->epoch_markers_active += 1; done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C__autoadjust__ageout__insert_new_marker() */ /*------------------------------------------------------------------------- * * Function: H5C__autoadjust__ageout__remove_all_markers * * Purpose: Remove all epoch markers from the LRU list and mark them * as inactive. * * Return: SUCCEED on success/FAIL on failure. * * Programmer: John Mainzer, 11/22/04 * * Modifications: * * None. * *------------------------------------------------------------------------- */ static herr_t H5C__autoadjust__ageout__remove_all_markers(H5C_t * cache_ptr) { herr_t ret_value = SUCCEED; /* Return value */ int i; int ring_buf_index; FUNC_ENTER_NOAPI_NOINIT(H5C__autoadjust__ageout__remove_all_markers) HDassert( cache_ptr ); HDassert( cache_ptr->magic == H5C__H5C_T_MAGIC ); while ( cache_ptr->epoch_markers_active > 0 ) { /* get the index of the last epoch marker in the LRU list * and remove it from the ring buffer. */ ring_buf_index = cache_ptr->epoch_marker_ringbuf_first; i = (cache_ptr->epoch_marker_ringbuf)[ring_buf_index]; cache_ptr->epoch_marker_ringbuf_first = (cache_ptr->epoch_marker_ringbuf_first + 1) % (H5C__MAX_EPOCH_MARKERS + 1); cache_ptr->epoch_marker_ringbuf_size -= 1; if ( cache_ptr->epoch_marker_ringbuf_size < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "ring buffer underflow.") } if ( (cache_ptr->epoch_marker_active)[i] != TRUE ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "unused marker in LRU?!?") } /* remove the epoch marker from the LRU list */ H5C__DLL_REMOVE((&((cache_ptr->epoch_markers)[i])), \ (cache_ptr)->LRU_head_ptr, \ (cache_ptr)->LRU_tail_ptr, \ (cache_ptr)->LRU_list_len, \ (cache_ptr)->LRU_list_size, \ (FAIL)) /* mark the epoch marker as unused. */ (cache_ptr->epoch_marker_active)[i] = FALSE; HDassert( ((cache_ptr->epoch_markers)[i]).addr == (haddr_t)i ); HDassert( ((cache_ptr->epoch_markers)[i]).next == NULL ); HDassert( ((cache_ptr->epoch_markers)[i]).prev == NULL ); /* decrement the number of active epoch markers */ cache_ptr->epoch_markers_active -= 1; HDassert( cache_ptr->epoch_markers_active == \ cache_ptr->epoch_marker_ringbuf_size ); } done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C__autoadjust__ageout__remove_all_markers() */ /*------------------------------------------------------------------------- * * Function: H5C__autoadjust__ageout__remove_excess_markers * * Purpose: Remove epoch markers from the end of the LRU list and * mark them as inactive until the number of active markers * equals the the current value of * (cache_ptr->resize_ctl).epochs_before_eviction. * * Return: SUCCEED on success/FAIL on failure. * * Programmer: John Mainzer, 11/19/04 * * Modifications: * * None. * *------------------------------------------------------------------------- */ static herr_t H5C__autoadjust__ageout__remove_excess_markers(H5C_t * cache_ptr) { herr_t ret_value = SUCCEED; /* Return value */ int i; int ring_buf_index; FUNC_ENTER_NOAPI_NOINIT(H5C__autoadjust__ageout__remove_excess_markers) HDassert( cache_ptr ); HDassert( cache_ptr->magic == H5C__H5C_T_MAGIC ); if ( cache_ptr->epoch_markers_active <= (cache_ptr->resize_ctl).epochs_before_eviction ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "no excess markers on entry.") } while ( cache_ptr->epoch_markers_active > (cache_ptr->resize_ctl).epochs_before_eviction ) { /* get the index of the last epoch marker in the LRU list * and remove it from the ring buffer. */ ring_buf_index = cache_ptr->epoch_marker_ringbuf_first; i = (cache_ptr->epoch_marker_ringbuf)[ring_buf_index]; cache_ptr->epoch_marker_ringbuf_first = (cache_ptr->epoch_marker_ringbuf_first + 1) % (H5C__MAX_EPOCH_MARKERS + 1); cache_ptr->epoch_marker_ringbuf_size -= 1; if ( cache_ptr->epoch_marker_ringbuf_size < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "ring buffer underflow.") } if ( (cache_ptr->epoch_marker_active)[i] != TRUE ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "unused marker in LRU?!?") } /* remove the epoch marker from the LRU list */ H5C__DLL_REMOVE((&((cache_ptr->epoch_markers)[i])), \ (cache_ptr)->LRU_head_ptr, \ (cache_ptr)->LRU_tail_ptr, \ (cache_ptr)->LRU_list_len, \ (cache_ptr)->LRU_list_size, \ (FAIL)) /* mark the epoch marker as unused. */ (cache_ptr->epoch_marker_active)[i] = FALSE; HDassert( ((cache_ptr->epoch_markers)[i]).addr == (haddr_t)i ); HDassert( ((cache_ptr->epoch_markers)[i]).next == NULL ); HDassert( ((cache_ptr->epoch_markers)[i]).prev == NULL ); /* decrement the number of active epoch markers */ cache_ptr->epoch_markers_active -= 1; HDassert( cache_ptr->epoch_markers_active == \ cache_ptr->epoch_marker_ringbuf_size ); } done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C__autoadjust__ageout__remove_excess_markers() */ /*------------------------------------------------------------------------- * * Function: H5C__flash_increase_cache_size * * Purpose: If there is not at least new_entry_size - old_entry_size * bytes of free space in the cache and the current * max_cache_size is less than (cache_ptr->resize_ctl).max_size, * perform a flash increase in the cache size and then reset * the full cache hit rate statistics, and exit. * * Return: Non-negative on success/Negative on failure. * * Programmer: John Mainzer, 12/31/07 * * Modifications: * * None. * *------------------------------------------------------------------------- */ static herr_t H5C__flash_increase_cache_size(H5C_t * cache_ptr, size_t old_entry_size, size_t new_entry_size) { size_t new_max_cache_size = 0; size_t old_max_cache_size = 0; size_t new_min_clean_size = 0; size_t old_min_clean_size = 0; size_t space_needed; enum H5C_resize_status status = flash_increase; /* may change */ double hit_rate; herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI_NOINIT(H5C__flash_increase_cache_size) HDassert( cache_ptr ); HDassert( cache_ptr->magic == H5C__H5C_T_MAGIC ); HDassert( cache_ptr->flash_size_increase_possible ); HDassert( new_entry_size > cache_ptr->flash_size_increase_threshold ); HDassert( old_entry_size < new_entry_size ); if ( old_entry_size >= new_entry_size ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "old_entry_size >= new_entry_size") } space_needed = new_entry_size - old_entry_size; if ( ( (cache_ptr->index_size + space_needed) > cache_ptr->max_cache_size ) && ( cache_ptr->max_cache_size < (cache_ptr->resize_ctl).max_size ) ) { /* we have work to do */ switch ( (cache_ptr->resize_ctl).flash_incr_mode ) { case H5C_flash_incr__off: HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "flash_size_increase_possible but H5C_flash_incr__off?!") break; case H5C_flash_incr__add_space: if ( cache_ptr->index_size < cache_ptr->max_cache_size ) { HDassert( (cache_ptr->max_cache_size - cache_ptr->index_size) < space_needed ); space_needed -= cache_ptr->max_cache_size - cache_ptr->index_size; } space_needed = (size_t)(((double)space_needed) * (cache_ptr->resize_ctl).flash_multiple); new_max_cache_size = cache_ptr->max_cache_size + space_needed; break; default: /* should be unreachable */ HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "Unknown flash_incr_mode?!?!?.") break; } if ( new_max_cache_size > (cache_ptr->resize_ctl).max_size ) { new_max_cache_size = (cache_ptr->resize_ctl).max_size; } HDassert( new_max_cache_size > cache_ptr->max_cache_size ); new_min_clean_size = (size_t) ((double)new_max_cache_size * ((cache_ptr->resize_ctl).min_clean_fraction)); HDassert( new_min_clean_size <= new_max_cache_size ); old_max_cache_size = cache_ptr->max_cache_size; old_min_clean_size = cache_ptr->min_clean_size; cache_ptr->max_cache_size = new_max_cache_size; cache_ptr->min_clean_size = new_min_clean_size; /* update flash cache size increase fields as appropriate */ HDassert ( cache_ptr->flash_size_increase_possible ); switch ( (cache_ptr->resize_ctl).flash_incr_mode ) { case H5C_flash_incr__off: HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "flash_size_increase_possible but H5C_flash_incr__off?!") break; case H5C_flash_incr__add_space: cache_ptr->flash_size_increase_threshold = (size_t) (((double)(cache_ptr->max_cache_size)) * ((cache_ptr->resize_ctl).flash_threshold)); break; default: /* should be unreachable */ HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "Unknown flash_incr_mode?!?!?.") break; } /* note that we don't cycle the epoch markers. We can * argue either way as to whether we should, but for now * we don't. */ if ( (cache_ptr->resize_ctl).rpt_fcn != NULL ) { /* get the hit rate for the reporting function. Should still * be good as we havent reset the hit rate statistics. */ if ( H5C_get_cache_hit_rate(cache_ptr, &hit_rate) != SUCCEED ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Can't get hit rate.") } (*((cache_ptr->resize_ctl).rpt_fcn)) (cache_ptr, H5C__CURR_AUTO_RESIZE_RPT_FCN_VER, hit_rate, status, old_max_cache_size, new_max_cache_size, old_min_clean_size, new_min_clean_size); } if ( H5C_reset_cache_hit_rate_stats(cache_ptr) != SUCCEED ) { /* this should be impossible... */ HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "H5C_reset_cache_hit_rate_stats failed.") } } done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C__flash_increase_cache_size() */ /*------------------------------------------------------------------------- * Function: H5C_flush_invalidate_cache * * Purpose: Flush and destroy the entries contained in the target * cache. * * If the cache contains protected entries, the function will * fail, as protected entries cannot be either flushed or * destroyed. However all unprotected entries should be * flushed and destroyed before the function returns failure. * * While pinned entries can usually be flushed, they cannot * be destroyed. However, they should be unpinned when all * the entries that reference them have been destroyed (thus * reduding the pinned entry's reference count to 0, allowing * it to be unpinned). * * If pinned entries are present, the function makes repeated * passes through the cache, flushing all dirty entries * (including the pinned dirty entries where permitted) and * destroying all unpinned entries. This process is repeated * until either the cache is empty, or the number of pinned * entries stops decreasing on each pass. * * The primary_dxpl_id and secondary_dxpl_id parameters * specify the dxpl_ids used on the first write occasioned * by the flush (primary_dxpl_id), and on all subsequent * writes (secondary_dxpl_id). * * Return: Non-negative on success/Negative on failure or if there was * a request to flush all items and something was protected. * * Programmer: John Mainzer * 3/24/065 * * Modifications: * * To support the fractal heap, the cache must now deal with * entries being dirtied, resized, and/or moved inside * flush callbacks. Updated function to support this. * * -- JRM 8/27/06 * * Added code to detect and manage the case in which a * flush callback changes the s-list out from under * the function. The only way I can think of in which this * can happen is if a flush function loads an entry * into the cache that isn't there already. Quincey tells * me that this will never happen, but I'm not sure I * believe him. * * Note that this is a pretty bad scenario if it ever * happens. The code I have added should allow us to * handle the situation under all but the worst conditions, * but one can argue that I should just scream and die if I * ever detect the condidtion. * * -- JRM 10/13/07 * *------------------------------------------------------------------------- */ static herr_t H5C_flush_invalidate_cache(H5F_t * f, hid_t primary_dxpl_id, hid_t secondary_dxpl_id, unsigned flags) { H5C_t * cache_ptr = f->shared->cache; herr_t status; hbool_t first_flush = TRUE; int32_t protected_entries = 0; int32_t i; int32_t cur_pel_len; int32_t old_pel_len; int32_t passes = 0; unsigned cooked_flags; H5SL_node_t * node_ptr = NULL; H5C_cache_entry_t * entry_ptr = NULL; H5C_cache_entry_t * next_entry_ptr = NULL; #if H5C_DO_SANITY_CHECKS int64_t actual_slist_len = 0; int64_t initial_slist_len = 0; size_t actual_slist_size = 0; size_t initial_slist_size = 0; #endif /* H5C_DO_SANITY_CHECKS */ herr_t ret_value = SUCCEED; FUNC_ENTER_NOAPI(H5C_flush_invalidate_cache, FAIL) HDassert( f ); HDassert( cache_ptr ); HDassert( cache_ptr->magic == H5C__H5C_T_MAGIC ); HDassert( cache_ptr->skip_file_checks || f ); HDassert( cache_ptr->slist_ptr ); /* Filter out the flags that are not relevant to the flush/invalidate. * At present, only the H5C__FLUSH_CLEAR_ONLY_FLAG is kept. */ cooked_flags = flags & H5C__FLUSH_CLEAR_ONLY_FLAG; /* remove ageout markers if present */ if ( cache_ptr->epoch_markers_active > 0 ) { status = H5C__autoadjust__ageout__remove_all_markers(cache_ptr); if ( status != SUCCEED ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "error removing all epoch markers.") } } /* The flush proceedure here is a bit strange. * * In the outer while loop we make at least one pass through the * cache, and then repeat until either all the pinned entries * unpin themselves, or until the number of pinned entries stops * declining. In this later case, we scream and die. * * Since the fractal heap can dirty, resize, and/or move entries * in is flush callback, it is possible that the cache will still * contain dirty entries at this point. If so, we must make up to * H5C__MAX_PASSES_ON_FLUSH more passes through the skip list * to allow it to empty. If is is not empty at this point, we again * scream and die. * * Further, since clean entries can be dirtied, resized, and/or moved * as the result of a flush call back (either the entries own, or that * for some other cache entry), we can no longer promise to flush * the cache entries in increasing address order. * * Instead, we just do the best we can -- making a pass through * the skip list, and then a pass through the "clean" entries, and * then repeating as needed. Thus it is quite possible that an * entry will be evicted from the cache only to be re-loaded later * in the flush process (From what Quincey tells me, the pin * mechanism makes this impossible, but even it it is true now, * we shouldn't count on it in the future.) * * The bottom line is that entries will probably be flushed in close * to increasing address order, but there are no guarantees. */ cur_pel_len = cache_ptr->pel_len; old_pel_len = cache_ptr->pel_len; while ( cache_ptr->index_len > 0 ) { unsigned curr_flush_dep_height = 0; unsigned flush_dep_passes = 0; /* Loop over all flush dependency heights of entries */ while((curr_flush_dep_height <= H5C__NUM_FLUSH_DEP_HEIGHTS) && (cache_ptr->index_len > 0 ) && (flush_dep_passes < H5C__MAX_PASSES_ON_FLUSH) ) { hbool_t flushed_during_dep_loop = FALSE; /* first, try to flush-destroy any dirty entries. Do this by * making a scan through the slist. Note that new dirty entries * may be created by the flush call backs. Thus it is possible * that the slist will not be empty after we finish the scan. */ if ( cache_ptr->slist_len == 0 ) { node_ptr = NULL; HDassert( cache_ptr->slist_size == 0 ); } else { /* Start at beginning of skip list each time */ node_ptr = H5SL_first(cache_ptr->slist_ptr); HDassert( node_ptr != NULL ); /* Get cache entry for this node */ next_entry_ptr = (H5C_cache_entry_t *)H5SL_item(node_ptr); if ( NULL == next_entry_ptr ) HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "next_entry_ptr == NULL ?!?!") HDassert( next_entry_ptr->magic == H5C__H5C_CACHE_ENTRY_T_MAGIC ); HDassert( next_entry_ptr->is_dirty ); HDassert( next_entry_ptr->in_slist ); } #if H5C_DO_SANITY_CHECKS /* Depending on circumstances, H5C_flush_single_entry() will * remove dirty entries from the slist as it flushes them. * Thus for sanity checks we must make note of the initial * slist length and size before we do any flushes. */ initial_slist_len = cache_ptr->slist_len; initial_slist_size = cache_ptr->slist_size; /* There is also the possibility that entries will be * dirtied, resized, and/or moved as the result of * calls to the flush callbacks. We use the slist_len_increase * and slist_size_increase increase fields in struct H5C_t * to track these changes for purpose of sanity checking. * To this end, we must zero these fields before we start * the pass through the slist. */ cache_ptr->slist_len_increase = 0; cache_ptr->slist_size_increase = 0; /* Finally, reset the actual_slist_len and actual_slist_size * fields to zero, as these fields are used to accumulate * the slist lenght and size that we see as we scan through * the slist. */ actual_slist_len = 0; actual_slist_size = 0; #endif /* H5C_DO_SANITY_CHECKS */ while ( node_ptr != NULL ) { entry_ptr = next_entry_ptr; /* With the advent of the fractal heap, it is possible * that the flush callback will dirty and/or resize * other entries in the cache. In particular, while * Quincey has promised me that this will never happen, * it is possible that the flush callback for an * entry may protect an entry that is not in the cache, * perhaps causing the cache to flush and possibly * evict the entry associated with node_ptr to make * space for the new entry. * * Thus we do a bit of extra sanity checking on entry_ptr, * and break out of this scan of the skip list if we * detect major problems. We have a bit of leaway on the * number of passes though the skip list, so this shouldn't * be an issue in the flush in and of itself, as it should * be all but impossible for this to happen more than once * in any flush. * * Observe that that breaking out of the scan early * shouldn't break the sanity checks just after the end * of this while loop. * * If an entry has merely been marked clean and removed from * the s-list, we simply break out of the scan. * * If the entry has been evicted, we flag an error and * exit. */ #ifndef NDEBUG if ( entry_ptr->magic != H5C__H5C_CACHE_ENTRY_T_MAGIC ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "entry_ptr->magic is invalid ?!?!") } else #endif /* NDEBUG */ if ( ( ! entry_ptr->is_dirty ) || ( ! entry_ptr->in_slist ) ) { /* the s-list has been modified out from under us. * break out of the loop. */ goto end_of_inner_loop;; } /* increment node pointer now, before we delete its target * from the slist. */ node_ptr = H5SL_next(node_ptr); if ( node_ptr != NULL ) { next_entry_ptr = (H5C_cache_entry_t *)H5SL_item(node_ptr); if ( NULL == next_entry_ptr ) HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "next_entry_ptr == NULL ?!?!") HDassert( next_entry_ptr->magic == H5C__H5C_CACHE_ENTRY_T_MAGIC ); HDassert( next_entry_ptr->is_dirty ); HDassert( next_entry_ptr->in_slist ); } else { next_entry_ptr = NULL; } /* Note that we now remove nodes from the slist as we flush * the associated entries, instead of leaving them there * until we are done, and then destroying all nodes in * the slist. * * While this optimization used to be easy, with the possibility * of new entries being added to the slist in the midst of the * flush, we must keep the slist in cannonical form at all * times. */ HDassert( entry_ptr != NULL ); HDassert( entry_ptr->in_slist ); #if H5C_DO_SANITY_CHECKS /* update actual_slist_len & actual_slist_size before * the flush. Note that the entry will be removed * from the slist after the flush, and thus may be * resized by the flush callback. This is OK, as * we will catch the size delta in * cache_ptr->slist_size_increase. * * Note that we include pinned entries in this count, even * though we will not actually flush them. */ actual_slist_len++; actual_slist_size += entry_ptr->size; #endif /* H5C_DO_SANITY_CHECKS */ if ( entry_ptr->is_protected ) { /* we have major problems -- but lets flush * everything we can before we flag an error. */ protected_entries++; } else if ( entry_ptr->is_pinned ) { /* Test to see if we are can flush the entry now. * If we can, go ahead and flush, but don't tell * H5C_flush_single_entry() to destroy the entry * as pinned entries can't be evicted. */ if(entry_ptr->flush_dep_height == curr_flush_dep_height ) { status = H5C_flush_single_entry(f, primary_dxpl_id, secondary_dxpl_id, NULL, entry_ptr->addr, H5C__NO_FLAGS_SET, &first_flush, FALSE); if ( status < 0 ) { /* This shouldn't happen -- if it does, we are toast * so just scream and die. */ HGOTO_ERROR(H5E_CACHE, H5E_CANTFLUSH, FAIL, \ "dirty pinned entry flush failed.") } /* end if */ flushed_during_dep_loop = TRUE; } /* end if */ else if(entry_ptr->flush_dep_height < curr_flush_dep_height) /* This shouldn't happen -- if it does, just scream and die. */ HGOTO_ERROR(H5E_CACHE, H5E_CANTFLUSH, FAIL, "dirty entry below current flush dep. height.") } /* end if */ else { if(entry_ptr->flush_dep_height == curr_flush_dep_height ){ status = H5C_flush_single_entry(f, primary_dxpl_id, secondary_dxpl_id, NULL, entry_ptr->addr, (cooked_flags | H5C__FLUSH_INVALIDATE_FLAG), &first_flush, TRUE); if ( status < 0 ) { /* This shouldn't happen -- if it does, we are toast so * just scream and die. */ HGOTO_ERROR(H5E_CACHE, H5E_CANTFLUSH, FAIL, \ "dirty entry flush destroy failed.") } /* end if */ flushed_during_dep_loop = TRUE; } /* end if */ else if(entry_ptr->flush_dep_height < curr_flush_dep_height) /* This shouldn't happen -- if it does, just scream and die. */ HGOTO_ERROR(H5E_CACHE, H5E_CANTFLUSH, FAIL, "dirty entry below current flush dep. height.") } /* end else */ } /* end while loop scanning skip list */ #if H5C_DO_SANITY_CHECKS /* It is possible that entries were added to the slist during * the scan, either before or after scan pointer. The following * asserts take this into account. * * Don't bother with the sanity checks if node_ptr != NULL, as * in this case we broke out of the loop because it got changed * out from under us. */ if ( node_ptr == NULL ) { HDassert( (actual_slist_len + cache_ptr->slist_len) == (initial_slist_len + cache_ptr->slist_len_increase) ); HDassert( (actual_slist_size + cache_ptr->slist_size) == (initial_slist_size + cache_ptr->slist_size_increase) ); } #endif /* H5C_DO_SANITY_CHECKS */ /* Since we are doing a destroy, we must make a pass through * the hash table and try to flush - destroy all entries that * remain. * * It used to be that all entries remaining in the cache at * this point had to be clean, but with the fractal heap mods * this may not be the case. If so, we will flush entries out * of increasing address order. * * Writes to disk are possible here. */ for ( i = 0; i < H5C__HASH_TABLE_LEN; i++ ) { next_entry_ptr = cache_ptr->index[i]; while ( next_entry_ptr != NULL ) { entry_ptr = next_entry_ptr; #ifndef NDEBUG HDassert( entry_ptr->magic == H5C__H5C_CACHE_ENTRY_T_MAGIC ); #endif /* NDEBUG */ next_entry_ptr = entry_ptr->ht_next; HDassert ( ( next_entry_ptr == NULL ) || ( next_entry_ptr->magic == H5C__H5C_CACHE_ENTRY_T_MAGIC ) ); if ( entry_ptr->is_protected ) { /* we have major problems -- but lets flush and destroy * everything we can before we flag an error. */ protected_entries++; if ( ! entry_ptr->in_slist ) { HDassert( !(entry_ptr->is_dirty) ); } } else if ( ! ( entry_ptr->is_pinned ) ) { /* Test to see if we are can flush the entry now. * If we can, go ahead and flush. */ if(entry_ptr->flush_dep_height == curr_flush_dep_height ){ status = H5C_flush_single_entry(f, primary_dxpl_id, secondary_dxpl_id, NULL, entry_ptr->addr, (cooked_flags | H5C__FLUSH_INVALIDATE_FLAG), &first_flush, TRUE); if ( status < 0 ) { /* This shouldn't happen -- if it does, we are toast so * just scream and die. */ HGOTO_ERROR(H5E_CACHE, H5E_CANTFLUSH, FAIL, \ "Entry flush destroy failed.") } flushed_during_dep_loop = TRUE; } /* end if */ else if(entry_ptr->flush_dep_height < curr_flush_dep_height) /* This shouldn't happen -- if it does, just scream and die. */ HGOTO_ERROR(H5E_CACHE, H5E_CANTFLUSH, FAIL, "dirty entry below current flush dep. height.") } /* end if */ /* We can't do anything if the entry is pinned. The * hope is that the entry will be unpinned as the * result of destroys of entries that reference it. * * We detect this by noting the change in the number * of pinned entries from pass to pass. If it stops * shrinking before it hits zero, we scream and die. */ /* if the flush function on the entry we last evicted * loaded an entry into cache (as Quincey has promised me * it never will), and if the cache was full, it is * possible that *next_entry_ptr was flushed or evicted. * * Test to see if this happened here. Note that if this * test is triggred, we are accessing a deallocated piece * of dynamically allocated memory, so we just scream and * die. */ #ifndef NDEBUG if ( ( next_entry_ptr != NULL ) && ( next_entry_ptr->magic != H5C__H5C_CACHE_ENTRY_T_MAGIC ) ) { /* Something horrible has happened to * *next_entry_ptr -- scream and die. */ HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "next_entry_ptr->magic is invalid?!?!?.") } #endif /* NDEBUG */ } /* end while loop scanning hash table bin */ } /* end for loop scanning hash table */ /* Check for incrementing flush dependency height */ if(flushed_during_dep_loop) { /* If we flushed an entry at this flush dependency height * start over at the bottom level of the flush dependencies */ curr_flush_dep_height = 0; /* Make certain we don't get stuck in an infinite loop */ flush_dep_passes++; } /* end if */ else curr_flush_dep_height++; } /* end while loop over flush dependency heights */ end_of_inner_loop: old_pel_len = cur_pel_len; cur_pel_len = cache_ptr->pel_len; if ( ( cur_pel_len > 0 ) && ( cur_pel_len >= old_pel_len ) ) { /* The number of pinned entries is positive, and it is not * declining. Scream and die. */ HGOTO_ERROR(H5E_CACHE, H5E_CANTFLUSH, FAIL, \ "Pinned entry count not decreasing.") } else if ( ( cur_pel_len == 0 ) && ( old_pel_len == 0 ) ) { /* increment the pass count */ passes++; } if ( passes >= H5C__MAX_PASSES_ON_FLUSH ) { /* we have exceeded the maximum number of passes through the * cache to flush and destroy all entries. Scream and die. */ HGOTO_ERROR(H5E_CACHE, H5E_CANTFLUSH, FAIL, \ "Maximum passes on flush exceeded.") } } /* main while loop */ /* Invariants, after destroying all entries in the hash table */ HDassert( cache_ptr->index_size == 0 ); HDassert( cache_ptr->clean_index_size == 0 ); HDassert( cache_ptr->dirty_index_size == 0 ); HDassert( cache_ptr->slist_len == 0 ); HDassert( cache_ptr->slist_size == 0 ); HDassert( cache_ptr->pel_len == 0 ); HDassert( cache_ptr->pel_size == 0 ); HDassert( cache_ptr->pl_len == 0 ); HDassert( cache_ptr->pl_size == 0 ); HDassert( cache_ptr->LRU_list_len == 0 ); HDassert( cache_ptr->LRU_list_size == 0 ); HDassert( protected_entries <= cache_ptr->pl_len ); if ( protected_entries > 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTFLUSH, FAIL, \ "Cache has protected entries.") } else if ( cur_pel_len > 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTFLUSH, FAIL, \ "Can't unpin all pinned entries.") } done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C_flush_invalidate_cache() */ /*------------------------------------------------------------------------- * * Function: H5C_flush_single_entry * * Purpose: Flush or clear (and evict if requested) the cache entry * with the specified address and type. If the type is NULL, * any unprotected entry at the specified address will be * flushed (and possibly evicted). * * Attempts to flush a protected entry will result in an * error. * * *first_flush_ptr should be true if only one * flush is contemplated before the next load, or if this * is the first of a sequence of flushes that will be * completed before the next load. *first_flush_ptr is set * to false if a flush actually takes place, and should be * left false until the end of the sequence. * * The primary_dxpl_id is used if *first_flush_ptr is TRUE * on entry, and a flush actually takes place. The * secondary_dxpl_id is used in any subsequent flush where * *first_flush_ptr is FALSE on entry. * * If the H5C__FLUSH_INVALIDATE_FLAG flag is set, the entry will * be cleared and not flushed -- in the case *first_flush_ptr, * primary_dxpl_id, and secondary_dxpl_id are all irrelevent, * and the call can't be part of a sequence of flushes. * * If the caller knows the address of the TBBT node at * which the target entry resides, it can avoid a lookup * by supplying that address in the tgt_node_ptr parameter. * If this parameter is NULL, the function will do a TBBT * search for the entry instead. * * The function does nothing silently if there is no entry * at the supplied address, or if the entry found has the * wrong type. * * Return: Non-negative on success/Negative on failure or if there was * an attempt to flush a protected item. * * Programmer: John Mainzer, 5/5/04 * * Modifications: * * JRM -- 7/21/04 * Updated function for the addition of the hash table. * * QAK -- 11/26/04 * Updated function for the switch from TBBTs to skip lists. * * JRM -- 1/6/05 * Updated function to reset the flush_marker field. * Also replace references to H5F_FLUSH_INVALIDATE and * H5F_FLUSH_CLEAR_ONLY with references to * H5C__FLUSH_INVALIDATE_FLAG and H5C__FLUSH_CLEAR_ONLY_FLAG * respectively. * * JRM -- 6/24/05 * Added code to remove dirty entries from the slist after * they have been flushed. Also added a sanity check that * will scream if we attempt a write when writes are * completely disabled. * * JRM -- 7/5/05 * Added code to call the new log_flush callback whenever * a dirty entry is written to disk. Note that the callback * is not called if the H5C__FLUSH_CLEAR_ONLY_FLAG is set, * as there is no write to file in this case. * * JRM -- 8/21/06 * Added code maintaining the flush_in_progress and * destroy_in_progress fields in H5C_cache_entry_t. * * Also added flush_flags parameter to the call to * type_ptr->flush() so that the flush routine can report * whether the entry has been resized or moved. Added * code using the flush_flags variable to detect the case * in which the target entry is resized during flush, and * update the caches data structures accordingly. * * JRM -- 3/29/07 * Added sanity checks on the new is_read_only and * ro_ref_count fields. * * QAK -- 2/07/08 * Separated "destroy entry" concept from "remove entry from * cache" concept, by adding the 'take_ownership' flag and * the "destroy_entry" variable. * * JRM -- 11/5/08 * Added call to H5C__UPDATE_INDEX_FOR_ENTRY_CLEAN() to * maintain the new clean_index_size and clean_index_size * fields of H5C_t. * *------------------------------------------------------------------------- */ static herr_t H5C_flush_single_entry(H5F_t * f, hid_t primary_dxpl_id, hid_t secondary_dxpl_id, const H5C_class_t * type_ptr, haddr_t addr, unsigned flags, hbool_t * first_flush_ptr, hbool_t del_entry_from_slist_on_destroy) { H5C_t * cache_ptr = f->shared->cache; hbool_t destroy; hbool_t clear_only; hbool_t take_ownership; hbool_t was_dirty; hbool_t destroy_entry; herr_t status; int type_id; unsigned flush_flags = H5C_CALLBACK__NO_FLAGS_SET; H5C_cache_entry_t * entry_ptr = NULL; herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI_NOINIT(H5C_flush_single_entry) HDassert( f ); HDassert( cache_ptr ); HDassert( cache_ptr->magic == H5C__H5C_T_MAGIC ); HDassert( cache_ptr->skip_file_checks || f ); HDassert( H5F_addr_defined(addr) ); HDassert( first_flush_ptr ); destroy = ( (flags & H5C__FLUSH_INVALIDATE_FLAG) != 0 ); clear_only = ( (flags & H5C__FLUSH_CLEAR_ONLY_FLAG) != 0); take_ownership = ( (flags & H5C__TAKE_OWNERSHIP_FLAG) != 0); /* Set the flag for destroying the entry, based on the 'take ownership' * and 'destroy' flags */ if(take_ownership) destroy_entry = FALSE; else destroy_entry = destroy; /* attempt to find the target entry in the hash table */ H5C__SEARCH_INDEX(cache_ptr, addr, entry_ptr, FAIL) #if H5C_DO_SANITY_CHECKS if ( entry_ptr != NULL ) { HDassert( ! ( ( destroy ) && ( entry_ptr->is_pinned ) ) ); if ( entry_ptr->in_slist ) { if ( ( ( entry_ptr->flush_marker ) && ( ! entry_ptr->is_dirty ) ) || ( entry_ptr->addr != addr ) ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "entry in slist failed sanity checks.") } } else { if ( ( entry_ptr->is_dirty ) || ( entry_ptr->flush_marker ) || ( entry_ptr->addr != addr ) ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "entry failed sanity checks.") } } } #if 0 /* this should be useful for debugging from time to time. * lets leave it in for now. -- JRM 12/15/04 */ else { HDfprintf(stdout, "H5C_flush_single_entry(): non-existant entry. addr = %a\n", addr); HDfflush(stdout); } #endif #endif /* H5C_DO_SANITY_CHECKS */ if ( ( entry_ptr != NULL ) && ( entry_ptr->is_protected ) ) { /* Attempt to flush a protected entry -- scream and die. */ HGOTO_ERROR(H5E_CACHE, H5E_PROTECT, FAIL, \ "Attempt to flush a protected entry.") } if ( ( entry_ptr != NULL ) && ( ( type_ptr == NULL ) || ( type_ptr->id == entry_ptr->type->id ) ) ) { /* we have work to do */ /* We will set flush_in_progress back to FALSE at the end if the * entry still exists at that point. */ entry_ptr->flush_in_progress = TRUE; #ifdef H5_HAVE_PARALLEL #ifndef NDEBUG /* If MPI based VFD is used, do special parallel I/O sanity checks. * Note that we only do these sanity checks when the clear_only flag * is not set, and the entry to be flushed is dirty. Don't bother * otherwise as no file I/O can result. * * There are also cases (testing for instance) where it is convenient * to pass in dummy dxpl_ids. Since we don't use the dxpl_ids directly, * this isn't a problem -- but we do have to turn off sanity checks * involving them. We use cache_ptr->skip_dxpl_id_checks to do this. */ if ( ( ! cache_ptr->skip_dxpl_id_checks ) && ( ! clear_only ) && ( entry_ptr->is_dirty ) && ( IS_H5FD_MPI(f) ) ) { H5P_genplist_t *dxpl; /* Dataset transfer property list */ H5FD_mpio_xfer_t xfer_mode; /* I/O xfer mode property value */ /* Get the dataset transfer property list */ if ( NULL == (dxpl = H5I_object(primary_dxpl_id)) ) { HGOTO_ERROR(H5E_ARGS, H5E_BADTYPE, FAIL, \ "not a dataset transfer property list") } /* Get the transfer mode property */ if( H5P_get(dxpl, H5D_XFER_IO_XFER_MODE_NAME, &xfer_mode) < 0 ) { HGOTO_ERROR(H5E_PLIST, H5E_CANTGET, FAIL, \ "can't retrieve xfer mode") } /* Sanity check transfer mode */ HDassert( xfer_mode == H5FD_MPIO_COLLECTIVE ); } #endif /* NDEBUG */ #endif /* H5_HAVE_PARALLEL */ was_dirty = entry_ptr->is_dirty; type_id = entry_ptr->type->id; entry_ptr->flush_marker = FALSE; if ( clear_only ) { H5C__UPDATE_STATS_FOR_CLEAR(cache_ptr, entry_ptr) } else { H5C__UPDATE_STATS_FOR_FLUSH(cache_ptr, entry_ptr) } if ( destroy ) { H5C__UPDATE_STATS_FOR_EVICTION(cache_ptr, entry_ptr) } /* If the entry's type has a 'notify' callback and the entry is about * to be removed from the cache, send a 'before eviction' notice while * the entry is still fully integrated in the cache. */ if(destroy) { if(entry_ptr->type->notify && (entry_ptr->type->notify)(H5C_NOTIFY_ACTION_BEFORE_EVICT, entry_ptr) < 0) HGOTO_ERROR(H5E_CACHE, H5E_CANTNOTIFY, FAIL, "can't notify client about entry to evict") } /* end if */ /* Always remove the entry from the hash table on a destroy. On a * flush with destroy, it is cheaper to discard the skip list all at * once rather than remove the entries one by one, so we only delete * from the slist only if requested. * * We must do deletions now as the callback routines will free the * entry if destroy is true. * * Note that it is possible that the entry will be moved during * its call to flush. This will upset H5C_move_entry() if we * don't tell it that it doesn't have to worry about updating the * index and SLIST. Use the destroy_in_progress field for this * purpose. */ if ( destroy ) { entry_ptr->destroy_in_progress = TRUE; H5C__DELETE_FROM_INDEX(cache_ptr, entry_ptr) if ( ( entry_ptr->in_slist ) && ( del_entry_from_slist_on_destroy ) ) { H5C__REMOVE_ENTRY_FROM_SLIST(cache_ptr, entry_ptr) } } /* Update the replacement policy for the flush or eviction. * Again, do this now so we don't have to reference freed * memory in the destroy case. */ if ( destroy ) { /* AKA eviction */ #if 0 /* JRM */ /* This test code may come in handy -- lets keep it for a while. * * Note that it will cause spurious errors in the serial case * unless we are maintaining the clean and dirty LRU lists. */ { if ( entry_ptr->is_dirty ) { if ( cache_ptr->dLRU_head_ptr == NULL ) HDfprintf(stdout, "%s: cache_ptr->dLRU_head_ptr == NULL.\n", FUNC); if ( cache_ptr->dLRU_tail_ptr == NULL ) HDfprintf(stdout, "%s: cache_ptr->dLRU_tail_ptr == NULL.\n", FUNC); if ( cache_ptr->dLRU_list_len <= 0 ) HDfprintf(stdout, "%s: cache_ptr->dLRU_list_len <= 0.\n", FUNC); if ( cache_ptr->dLRU_list_size <= 0 ) HDfprintf(stdout, "%s: cache_ptr->dLRU_list_size <= 0.\n", FUNC); if ( cache_ptr->dLRU_list_size < entry_ptr->size ) HDfprintf(stdout, "%s: cache_ptr->dLRU_list_size < entry_ptr->size.\n", FUNC); if ( ( (cache_ptr->dLRU_list_size) == entry_ptr->size ) && ( ! ( (cache_ptr->dLRU_list_len) == 1 ) ) ) HDfprintf(stdout, "%s: dLRU_list_size == size && dLRU_list_len != 1\n", FUNC); if ( ( entry_ptr->aux_prev == NULL ) && ( cache_ptr->dLRU_head_ptr != entry_ptr ) ) HDfprintf(stdout, "%s: entry_ptr->aux_prev == NULL && dLRU_head_ptr != entry_ptr\n", FUNC); if ( ( entry_ptr->aux_next == NULL ) && ( cache_ptr->dLRU_tail_ptr != entry_ptr ) ) HDfprintf(stdout, "%s: entry_ptr->aux_next == NULL && dLRU_tail_ptr != entry_ptr\n", FUNC); if ( ( cache_ptr->dLRU_list_len == 1 ) && ( ! ( ( cache_ptr->dLRU_head_ptr == entry_ptr ) && ( cache_ptr->dLRU_tail_ptr == entry_ptr ) && ( entry_ptr->aux_next == NULL ) && ( entry_ptr->aux_prev == NULL ) && ( cache_ptr->dLRU_list_size == entry_ptr->size ) ) ) ) { HDfprintf(stdout, "%s: single entry dlru sanity check fails\n", FUNC); } } else { if ( cache_ptr->cLRU_head_ptr == NULL ) HDfprintf(stdout, "%s: cache_ptr->cLRU_head_ptr == NULL.\n", FUNC); if ( cache_ptr->cLRU_tail_ptr == NULL ) HDfprintf(stdout, "%s: cache_ptr->cLRU_tail_ptr == NULL.\n", FUNC); if ( cache_ptr->cLRU_list_len <= 0 ) HDfprintf(stdout, "%s: cache_ptr->cLRU_list_len <= 0.\n", FUNC); if ( cache_ptr->cLRU_list_size <= 0 ) HDfprintf(stdout, "%s: cache_ptr->cLRU_list_size <= 0.\n", FUNC); if ( cache_ptr->cLRU_list_size < entry_ptr->size ) HDfprintf(stdout, "%s: cache_ptr->cLRU_list_size < entry_ptr->size.\n", FUNC); if ( ( (cache_ptr->cLRU_list_size) == entry_ptr->size ) && ( ! ( (cache_ptr->cLRU_list_len) == 1 ) ) ) HDfprintf(stdout, "%s: cLRU_list_size == size && cLRU_list_len != 1\n", FUNC); if ( ( entry_ptr->aux_prev == NULL ) && ( cache_ptr->cLRU_head_ptr != entry_ptr ) ) HDfprintf(stdout, "%s: entry_ptr->aux_prev == NULL && cLRU_head_ptr != entry_ptr\n", FUNC); if ( ( entry_ptr->aux_next == NULL ) && ( cache_ptr->cLRU_tail_ptr != entry_ptr ) ) HDfprintf(stdout, "%s: entry_ptr->aux_next == NULL && cLRU_tail_ptr != entry_ptr\n", FUNC); if ( ( cache_ptr->cLRU_list_len == 1 ) && ( ! ( ( cache_ptr->cLRU_head_ptr == entry_ptr ) && ( cache_ptr->cLRU_tail_ptr == entry_ptr ) && ( entry_ptr->aux_next == NULL ) && ( entry_ptr->aux_prev == NULL ) && ( cache_ptr->cLRU_list_size == entry_ptr->size ) ) ) ) { HDfprintf(stdout, "%s: single entry clru sanity check fails\n", FUNC); } } } #endif /* JRM */ H5C__UPDATE_RP_FOR_EVICTION(cache_ptr, entry_ptr, FAIL) } else { H5C__UPDATE_RP_FOR_FLUSH(cache_ptr, entry_ptr, FAIL) } /* Clear the dirty flag only, if requested */ if ( clear_only ) { if ( destroy ) { #ifndef NDEBUG /* we are about to call the clear callback with the * destroy flag set -- this will result in *entry_ptr * being freed. Set the magic field to bad magic * so we can detect a freed cache entry if we see * one. */ entry_ptr->magic = H5C__H5C_CACHE_ENTRY_T_BAD_MAGIC; #endif /* NDEBUG */ entry_ptr->cache_ptr = NULL; } /* Call the callback routine to clear all dirty flags for object */ if ( (entry_ptr->type->clear)(f, entry_ptr, destroy_entry) < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTFLUSH, FAIL, "can't clear entry") } } else { #if H5C_DO_SANITY_CHECKS if ( ( entry_ptr->is_dirty ) && ( cache_ptr->check_write_permitted == NULL ) && ( ! (cache_ptr->write_permitted) ) ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "Write when writes are always forbidden!?!?!") } #endif /* H5C_DO_SANITY_CHECKS */ if ( destroy ) { #ifndef NDEBUG /* we are about to call the flush callback with the * destroy flag set -- this will result in *entry_ptr * being freed. Set the magic field to bad magic * so we can detect a freed cache entry if we see * one. */ entry_ptr->magic = H5C__H5C_CACHE_ENTRY_T_BAD_MAGIC; #endif /* NDEBUG */ entry_ptr->cache_ptr = NULL; } /* Only block for all the processes on the first piece of metadata */ if ( *first_flush_ptr && entry_ptr->is_dirty ) { status = (entry_ptr->type->flush)(f, primary_dxpl_id, destroy_entry, entry_ptr->addr, entry_ptr, &flush_flags); *first_flush_ptr = FALSE; } else { status = (entry_ptr->type->flush)(f, secondary_dxpl_id, destroy_entry, entry_ptr->addr, entry_ptr, &flush_flags); } if ( status < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTFLUSH, FAIL, \ "unable to flush entry") } #ifdef H5_HAVE_PARALLEL if ( flush_flags != H5C_CALLBACK__NO_FLAGS_SET ) { /* In the parallel case, flush operations can * cause problems. If they occur, scream and * die. * * At present, in the parallel case, the aux_ptr * will only be set if there is more than one * process. Thus we can use this to detect * the parallel case. * * This works for now, but if we start using the * aux_ptr for other purposes, we will have to * change this test accordingly. * * NB: While this test detects entryies that attempt * to resize or move themselves during a flush * in the parallel case, it will not detect an * entry that dirties, resizes, and/or moves * other entries during its flush. * * From what Quincey tells me, this test is * sufficient for now, as any flush routine that * does the latter will also do the former. * * If that ceases to be the case, further * tests will be necessary. */ if ( cache_ptr->aux_ptr != NULL ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "resize/move in serialize occured in parallel case.") } } #endif /* H5_HAVE_PARALLEL */ } if ( ( ! destroy ) && ( entry_ptr->in_slist ) ) { H5C__REMOVE_ENTRY_FROM_SLIST(cache_ptr, entry_ptr) } if ( ( ! destroy ) && ( was_dirty ) ) { H5C__UPDATE_INDEX_FOR_ENTRY_CLEAN(cache_ptr, entry_ptr); } if ( ! destroy ) { /* i.e. if the entry still exists */ HDassert( !(entry_ptr->is_dirty) ); HDassert( !(entry_ptr->flush_marker) ); HDassert( !(entry_ptr->in_slist) ); HDassert( !(entry_ptr->is_protected) ); HDassert( !(entry_ptr->is_read_only) ); HDassert( (entry_ptr->ro_ref_count) == 0 ); if ( (flush_flags & H5C_CALLBACK__SIZE_CHANGED_FLAG) != 0 ) { /* The entry size changed as a result of the flush. * * Most likely, the entry was compressed, and the * new version is of a different size than the old. * * In any case, we must update entry and cache size * accordingly. */ size_t new_size; if ( (entry_ptr->type->size)(f, (void *)entry_ptr, &new_size) < 0 ) { HGOTO_ERROR(H5E_RESOURCE, H5E_CANTGETSIZE, FAIL, \ "Can't get entry size after flush") } if ( new_size != entry_ptr->size ) { HDassert( entry_ptr->size < H5C_MAX_ENTRY_SIZE ); /* update the hash table for the size change * We pass TRUE as the was_clean parameter, as we * have already updated the clean and dirty index * size fields for the fact that the entry has * been flushed. (See above call to * H5C__UPDATE_INDEX_FOR_ENTRY_CLEAN()). */ H5C__UPDATE_INDEX_FOR_SIZE_CHANGE((cache_ptr), \ (entry_ptr->size),\ (new_size), \ (entry_ptr), \ (TRUE)); /* The entry can't be protected since we just flushed it. * Thus we must update the replacement policy data * structures for the size change. The macro deals * with the pinned case. */ H5C__UPDATE_RP_FOR_SIZE_CHANGE(cache_ptr, entry_ptr, \ new_size) /* The entry can't be in the slist, so no need to update * the slist for the size change. */ /* update stats for the size change */ H5C__UPDATE_STATS_FOR_ENTRY_SIZE_CHANGE(cache_ptr, \ entry_ptr, \ new_size) /* finally, update the entry size proper */ entry_ptr->size = new_size; } } if ( (flush_flags & H5C_CALLBACK__MOVED_FLAG) != 0 ) { /* The entry was moved as the result of the flush. * * Most likely, the entry was compressed, and the * new version is larger than the old and thus had * to be relocated. * * At preset, all processing for this case is * handled elsewhere. But lets keep the if statement * around just in case. */ } entry_ptr->flush_in_progress = FALSE; } if ( cache_ptr->log_flush ) { status = (cache_ptr->log_flush)(cache_ptr, addr, was_dirty, flags, type_id); if ( status < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTFLUSH, FAIL, \ "log_flush callback failed.") } } } done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C_flush_single_entry() */ /*------------------------------------------------------------------------- * * Function: H5C_load_entry * * Purpose: Attempt to load the entry at the specified disk address * and with the specified type into memory. If successful. * return the in memory address of the entry. Return NULL * on failure. * * Note that this function simply loads the entry into * core. It does not insert it into the cache. * * Return: Non-NULL on success / NULL on failure. * * Programmer: John Mainzer, 5/18/04 * * Modifications: * * JRM - 7/21/04 * Updated function for the addition of the hash table. * * JRM - 6/23/06 * Deleted assertion that verified that a newly loaded * entry is clean. Due to a bug fix, this need not be * the case, as our code will attempt to repair errors * on load. * * JRM - 8/21/06 * Added initialization for the new flush_in_progress and * destroy in progress fields. * * JRM - 3/29/07 * Added initialization for the new is_read_only and * ro_ref_count fields. * * QAK -- 1/31/08 * Added initialization for the new free_file_space_on_destroy * field. * *------------------------------------------------------------------------- */ static void * H5C_load_entry(H5F_t * f, hid_t dxpl_id, const H5C_class_t * type, haddr_t addr, void * udata, #ifndef NDEBUG hbool_t skip_file_checks) #else /* NDEBUG */ hbool_t UNUSED skip_file_checks) #endif /* NDEBUG */ { void * thing = NULL; H5C_cache_entry_t * entry_ptr = NULL; unsigned u; /* Local index variable */ void * ret_value = NULL; /* Return value */ FUNC_ENTER_NOAPI_NOINIT(H5C_load_entry) HDassert( f ); HDassert( f->shared ); HDassert( f->shared->cache ); HDassert( skip_file_checks || f ); HDassert( type ); HDassert( type->load ); HDassert( type->size ); HDassert( H5F_addr_defined(addr) ); if ( NULL == (thing = (type->load)(f, dxpl_id, addr, udata)) ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTLOAD, NULL, "unable to load entry") } entry_ptr = (H5C_cache_entry_t *)thing; /* In general, an entry should be clean just after it is loaded. * * However, when this code is used in the metadata cache, it is * possible that object headers will be dirty at this point, as * the load function will alter object headers if necessary to * fix an old bug. * * To support this bug fix, I have replace the old assert: * * HDassert( entry_ptr->is_dirty == FALSE ); * * with: * * HDassert( ( entry_ptr->is_dirty == FALSE ) || ( type->id == 5 ) ); * * Note that type id 5 is associated with object headers in the metadata * cache. * * When we get to using H5C for other purposes, we may wish to * tighten up the assert so that the loophole only applies to the * metadata cache. */ HDassert( ( entry_ptr->is_dirty == FALSE ) || ( type->id == 5 ) ); #ifndef NDEBUG entry_ptr->magic = H5C__H5C_CACHE_ENTRY_T_MAGIC; #endif /* NDEBUG */ entry_ptr->cache_ptr = f->shared->cache; entry_ptr->addr = addr; entry_ptr->type = type; entry_ptr->is_protected = FALSE; entry_ptr->is_read_only = FALSE; entry_ptr->ro_ref_count = 0; entry_ptr->in_slist = FALSE; entry_ptr->flush_marker = FALSE; #ifdef H5_HAVE_PARALLEL entry_ptr->clear_on_unprotect = FALSE; #endif /* H5_HAVE_PARALLEL */ entry_ptr->flush_in_progress = FALSE; entry_ptr->destroy_in_progress = FALSE; entry_ptr->free_file_space_on_destroy = FALSE; if ( (type->size)(f, thing, &(entry_ptr->size)) < 0 ) { HGOTO_ERROR(H5E_RESOURCE, H5E_CANTGETSIZE, NULL, \ "Can't get size of thing") } HDassert( entry_ptr->size < H5C_MAX_ENTRY_SIZE ); /* Initialize flush dependency height fields */ entry_ptr->flush_dep_parent = NULL; for(u = 0; u < H5C__NUM_FLUSH_DEP_HEIGHTS; u++) entry_ptr->child_flush_dep_height_rc[u] = 0; entry_ptr->flush_dep_height = 0; entry_ptr->ht_next = NULL; entry_ptr->ht_prev = NULL; entry_ptr->next = NULL; entry_ptr->prev = NULL; entry_ptr->aux_next = NULL; entry_ptr->aux_prev = NULL; H5C__RESET_CACHE_ENTRY_STATS(entry_ptr); ret_value = thing; done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C_load_entry() */ /*------------------------------------------------------------------------- * * Function: H5C_make_space_in_cache * * Purpose: Attempt to evict cache entries until the index_size * is at least needed_space below max_cache_size. * * In passing, also attempt to bring cLRU_list_size to a * value greater than min_clean_size. * * Depending on circumstances, both of these goals may * be impossible, as in parallel mode, we must avoid generating * a write as part of a read (to avoid deadlock in collective * I/O), and in all cases, it is possible (though hopefully * highly unlikely) that the protected list may exceed the * maximum size of the cache. * * Thus the function simply does its best, returning success * unless an error is encountered. * * The primary_dxpl_id and secondary_dxpl_id parameters * specify the dxpl_ids used on the first write occasioned * by the call (primary_dxpl_id), and on all subsequent * writes (secondary_dxpl_id). This is useful in the metadata * cache, but may not be needed elsewhere. If so, just use the * same dxpl_id for both parameters. * * Observe that this function cannot occasion a read. * * Return: Non-negative on success/Negative on failure. * * Programmer: John Mainzer, 5/14/04 * * Modifications: * * JRM --7/21/04 * Minor modifications in support of the addition of a hash * table to facilitate lookups. * * JRM -- 11/22/04 * Added the first_flush_ptr parameter, which replaces the * old first_flush local variable. This allows the function * to coordinate on the first flush issue with other functions. * * JRM -- 12/13/04 * Added code to skip over epoch markers if present. * * JRM -- 1/3/06 * Modified function to work correctly when the the cache * is not full. This case occurs when we need to flush to * min clean size before the cache has filled. * * JRM -- 3/29/07 * Added sanity checks using the new is_read_only and * ro_ref_count fields. * * JRM -- 10/13/07 * Added code to detect and manage the case in which a * flush callback changes the LRU-list out from under * the function. The only way I can think of in which this * can happen is if a flush function loads an entry * into the cache that isn't there already. Quincey tells * me that this will never happen, but I'm not sure I * believe him. * * Note that this is a pretty bad scenario if it ever * happens. The code I have added should allow us to * handle the situation under all but the worst conditions, * but one can argue that I should just scream and die if I * ever detect the condidtion. * * JRM -- 11/13/08 * Modified function to always observe the min_clean_size * whether we are maintaining the clean and dirt LRU lists * or not. To do this, we had to add the new clean_index_size * and dirty_index_size fields to H5C_t, and supporting code * as needed throughout the cache. * * The purpose of this modification is to avoid "metadata * blizzards" in the write only case. In such instances, * the cache was allowed to fill with dirty metadata. When * we finally needed to evict an entry to make space, we had * to flush out a whole cache full of metadata -- which has * interesting performance effects. We hope to avoid (or * perhaps more accurately hide) this effect by maintaining * the min_clean_size, which should force us to start flushing * entries long before we actually have to evict something * to make space. * * MAM -- 01/06/09 * Added code to maintain clean_entries_skipped and total_entries * scanned statistics. *------------------------------------------------------------------------- */ static herr_t H5C_make_space_in_cache(H5F_t * f, hid_t primary_dxpl_id, hid_t secondary_dxpl_id, size_t space_needed, hbool_t write_permitted, hbool_t * first_flush_ptr) { H5C_t * cache_ptr = f->shared->cache; herr_t result; #if H5C_COLLECT_CACHE_STATS int32_t clean_entries_skipped = 0; int32_t total_entries_scanned = 0; #endif /* H5C_COLLECT_CACHE_STATS */ int32_t entries_examined = 0; int32_t initial_list_len; size_t empty_space; hbool_t prev_is_dirty = FALSE; hbool_t didnt_flush_entry = FALSE; H5C_cache_entry_t * entry_ptr; H5C_cache_entry_t * prev_ptr; H5C_cache_entry_t * next_ptr; herr_t ret_value = SUCCEED; /* Return value */ FUNC_ENTER_NOAPI_NOINIT(H5C_make_space_in_cache) HDassert( f ); HDassert( cache_ptr ); HDassert( cache_ptr->magic == H5C__H5C_T_MAGIC ); HDassert( first_flush_ptr != NULL ); HDassert( ( *first_flush_ptr == TRUE ) || ( *first_flush_ptr == FALSE ) ); HDassert( cache_ptr->index_size == (cache_ptr->clean_index_size + cache_ptr->dirty_index_size) ); if ( write_permitted ) { initial_list_len = cache_ptr->LRU_list_len; entry_ptr = cache_ptr->LRU_tail_ptr; if ( cache_ptr->index_size >= cache_ptr->max_cache_size ) { empty_space = 0; } else { empty_space = cache_ptr->max_cache_size - cache_ptr->index_size; } while ( ( ( (cache_ptr->index_size + space_needed) > cache_ptr->max_cache_size ) || ( ( empty_space + cache_ptr->clean_index_size ) < ( cache_ptr->min_clean_size ) ) ) && ( entries_examined <= (2 * initial_list_len) ) && ( entry_ptr != NULL ) ) { HDassert( ! (entry_ptr->is_protected) ); HDassert( ! (entry_ptr->is_read_only) ); HDassert( (entry_ptr->ro_ref_count) == 0 ); next_ptr = entry_ptr->next; prev_ptr = entry_ptr->prev; if ( prev_ptr != NULL ) { prev_is_dirty = prev_ptr->is_dirty; } if ( (entry_ptr->type)->id != H5C__EPOCH_MARKER_TYPE ) { didnt_flush_entry = FALSE; if ( entry_ptr->is_dirty ) { #if H5C_COLLECT_CACHE_STATS if ( (cache_ptr->index_size + space_needed) > cache_ptr->max_cache_size ) { cache_ptr->entries_scanned_to_make_space++; } #endif /* H5C_COLLECT_CACHE_STATS */ result = H5C_flush_single_entry(f, primary_dxpl_id, secondary_dxpl_id, entry_ptr->type, entry_ptr->addr, H5C__NO_FLAGS_SET, first_flush_ptr, FALSE); } else if ( (cache_ptr->index_size + space_needed) > cache_ptr->max_cache_size ) { #if H5C_COLLECT_CACHE_STATS cache_ptr->entries_scanned_to_make_space++; #endif /* H5C_COLLECT_CACHE_STATS */ result = H5C_flush_single_entry(f, primary_dxpl_id, secondary_dxpl_id, entry_ptr->type, entry_ptr->addr, H5C__FLUSH_INVALIDATE_FLAG, first_flush_ptr, TRUE); } else { /* We have enough space so don't flush clean entry. * Set result to SUCCEED to avoid triggering the error * code below. */ #if H5C_COLLECT_CACHE_STATS clean_entries_skipped++; #endif /* H5C_COLLECT_CACHE_STATS */ didnt_flush_entry = TRUE; result = SUCCEED; } #if H5C_COLLECT_CACHE_STATS total_entries_scanned++; #endif /* H5C_COLLECT_CACHE_STATS */ } else { /* Skip epoch markers. Set result to SUCCEED to avoid * triggering the error code below. */ didnt_flush_entry = TRUE; result = SUCCEED; } if ( result < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTFLUSH, FAIL, \ "unable to flush entry") } if ( prev_ptr != NULL ) { #ifndef NDEBUG if ( prev_ptr->magic != H5C__H5C_CACHE_ENTRY_T_MAGIC ) { /* something horrible has happened to *prev_ptr -- * scream and die. */ HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "*prev_ptr corrupt 1") } #endif /* NDEBUG */ if ( didnt_flush_entry ) { /* epoch markers don't get flushed, so the sanity checks * on normal entries will fail -- thus just set entry_ptr * to prev_ptr and go on. */ entry_ptr = prev_ptr; } else if ( ( prev_ptr->is_dirty != prev_is_dirty ) || ( prev_ptr->next != next_ptr ) || ( prev_ptr->is_protected ) || ( prev_ptr->is_pinned ) ) { /* something has happened to the LRU -- start over * from the tail. */ entry_ptr = cache_ptr->LRU_tail_ptr; } else { entry_ptr = prev_ptr; } } else { entry_ptr = NULL; } entries_examined++; if ( cache_ptr->index_size >= cache_ptr->max_cache_size ) { empty_space = 0; } else { empty_space = cache_ptr->max_cache_size - cache_ptr->index_size; } HDassert( cache_ptr->index_size == (cache_ptr->clean_index_size + cache_ptr->dirty_index_size) ); } #if H5C_COLLECT_CACHE_STATS cache_ptr->calls_to_msic++; cache_ptr->total_entries_skipped_in_msic += clean_entries_skipped; cache_ptr->total_entries_scanned_in_msic += total_entries_scanned; if ( clean_entries_skipped > cache_ptr->max_entries_skipped_in_msic ) { cache_ptr->max_entries_skipped_in_msic = clean_entries_skipped; } if ( total_entries_scanned > cache_ptr->max_entries_scanned_in_msic ) { cache_ptr->max_entries_scanned_in_msic = total_entries_scanned; } #endif /* H5C_COLLECT_CACHE_STATS */ HDassert( ( entries_examined > (2 * initial_list_len) ) || ( (cache_ptr->pl_size + cache_ptr->min_clean_size) > cache_ptr->max_cache_size ) || ( ( cache_ptr->clean_index_size + empty_space ) >= cache_ptr->min_clean_size ) ); #if H5C_MAINTAIN_CLEAN_AND_DIRTY_LRU_LISTS HDassert( ( entries_examined > (2 * initial_list_len) ) || ( cache_ptr->cLRU_list_size <= cache_ptr->clean_index_size ) ); HDassert( ( entries_examined > (2 * initial_list_len) ) || ( cache_ptr->dLRU_list_size <= cache_ptr->dirty_index_size ) ); #endif /* H5C_MAINTAIN_CLEAN_AND_DIRTY_LRU_LISTS */ } else { HDassert( H5C_MAINTAIN_CLEAN_AND_DIRTY_LRU_LISTS ); initial_list_len = cache_ptr->cLRU_list_len; entry_ptr = cache_ptr->cLRU_tail_ptr; while ( ( (cache_ptr->index_size + space_needed) > cache_ptr->max_cache_size ) && ( entries_examined <= initial_list_len ) && ( entry_ptr != NULL ) ) { HDassert( ! (entry_ptr->is_protected) ); HDassert( ! (entry_ptr->is_read_only) ); HDassert( (entry_ptr->ro_ref_count) == 0 ); HDassert( ! (entry_ptr->is_dirty) ); prev_ptr = entry_ptr->aux_prev; result = H5C_flush_single_entry(f, primary_dxpl_id, secondary_dxpl_id, entry_ptr->type, entry_ptr->addr, H5C__FLUSH_INVALIDATE_FLAG, first_flush_ptr, TRUE); if ( result < 0 ) { HGOTO_ERROR(H5E_CACHE, H5E_CANTFLUSH, FAIL, \ "unable to flush entry") } entry_ptr = prev_ptr; entries_examined++; } } done: FUNC_LEAVE_NOAPI(ret_value) } /* H5C_make_space_in_cache() */ /*------------------------------------------------------------------------- * * Function: H5C_validate_lru_list * * Purpose: Debugging function that scans the LRU list for errors. * * If an error is detected, the function generates a * diagnostic and returns FAIL. If no error is detected, * the function returns SUCCEED. * * Return: FAIL if error is detected, SUCCEED otherwise. * * Programmer: John Mainzer, 7/14/05 * * Modifications: * *------------------------------------------------------------------------- */ #if H5C_DO_EXTREME_SANITY_CHECKS static herr_t H5C_validate_lru_list(H5C_t * cache_ptr) { herr_t ret_value = SUCCEED; /* Return value */ int32_t len = 0; size_t size = 0; H5C_cache_entry_t * entry_ptr = NULL; FUNC_ENTER_NOAPI_NOINIT(H5C_validate_lru_list) HDassert( cache_ptr ); HDassert( cache_ptr->magic == H5C__H5C_T_MAGIC ); if ( ( ( cache_ptr->LRU_head_ptr == NULL ) || ( cache_ptr->LRU_tail_ptr == NULL ) ) && ( cache_ptr->LRU_head_ptr != cache_ptr->LRU_tail_ptr ) ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Check 1 failed") } if ( ( cache_ptr->LRU_list_len < 0 ) || ( cache_ptr->LRU_list_size < 0 ) ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Check 2 failed") } if ( ( cache_ptr->LRU_list_len == 1 ) && ( ( cache_ptr->LRU_head_ptr != cache_ptr->LRU_tail_ptr ) || ( cache_ptr->LRU_head_ptr == NULL ) || ( cache_ptr->LRU_head_ptr->size != cache_ptr->LRU_list_size ) ) ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Check 3 failed") } if ( ( cache_ptr->LRU_list_len >= 1 ) && ( ( cache_ptr->LRU_head_ptr == NULL ) || ( cache_ptr->LRU_head_ptr->prev != NULL ) || ( cache_ptr->LRU_tail_ptr == NULL ) || ( cache_ptr->LRU_tail_ptr->next != NULL ) ) ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Check 4 failed") } entry_ptr = cache_ptr->LRU_head_ptr; while ( entry_ptr != NULL ) { if ( ( entry_ptr != cache_ptr->LRU_head_ptr ) && ( ( entry_ptr->prev == NULL ) || ( entry_ptr->prev->next != entry_ptr ) ) ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Check 5 failed") } if ( ( entry_ptr != cache_ptr->LRU_tail_ptr ) && ( ( entry_ptr->next == NULL ) || ( entry_ptr->next->prev != entry_ptr ) ) ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Check 6 failed") } len++; size += entry_ptr->size; entry_ptr = entry_ptr->next; } if ( ( cache_ptr->LRU_list_len != len ) || ( cache_ptr->LRU_list_size != size ) ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, "Check 7 failed") } done: if ( ret_value != SUCCEED ) { HDassert(0); } FUNC_LEAVE_NOAPI(ret_value) } /* H5C_validate_lru_list() */ #endif /* H5C_DO_EXTREME_SANITY_CHECKS */ /*------------------------------------------------------------------------- * * Function: H5C_verify_not_in_index * * Purpose: Debugging function that scans the hash table to verify * that the specified instance of H5C_cache_entry_t is not * present. * * If an error is detected, the function generates a * diagnostic and returns FAIL. If no error is detected, * the function returns SUCCEED. * * Return: FAIL if error is detected, SUCCEED otherwise. * * Programmer: John Mainzer, 7/14/05 * * Modifications: * *------------------------------------------------------------------------- */ #if H5C_DO_EXTREME_SANITY_CHECKS static herr_t H5C_verify_not_in_index(H5C_t * cache_ptr, H5C_cache_entry_t * entry_ptr) { herr_t ret_value = SUCCEED; /* Return value */ int32_t i; int32_t depth; H5C_cache_entry_t * scan_ptr = NULL; FUNC_ENTER_NOAPI_NOINIT(H5C_verify_not_in_index) HDassert( cache_ptr != NULL ); HDassert( cache_ptr->magic == H5C__H5C_T_MAGIC ); HDassert( entry_ptr != NULL ); for ( i = 0; i < H5C__HASH_TABLE_LEN; i++ ) { depth = 0; scan_ptr = cache_ptr->index[i]; while ( scan_ptr != NULL ) { if ( scan_ptr == entry_ptr ) { HGOTO_ERROR(H5E_CACHE, H5E_SYSTEM, FAIL, \ "Entry already in index.") } depth++; scan_ptr = scan_ptr->ht_next; } } done: if ( ret_value != SUCCEED ) { HDassert(0); } FUNC_LEAVE_NOAPI(ret_value) } /* H5C_verify_not_in_index() */ #endif /* H5C_DO_EXTREME_SANITY_CHECKS */