Lines Matching defs:records

21  * records are stored in an array in the order they arrive. Additionally, a reference to each
24 * records is 1-based so that record number 0 can be used to indicate an unused hash slot.
26 * Deleted records are marked with a flag rather than actually removed to simplify hash table
29 * power of two that is greater than the number of records to be indexed, guaranteeing that hash
30 * insertion cannot fail, and that there are sufficient flags for all records.
32 * Once any open chapter zone fills its available space, the chapter is closed. The records from
34 * Empty or deleted records are replaced by copies of a valid record so that the record pages only
35 * contain valid records. The chapter then constructs a delta index which maps each record name to
39 * When the index is saved, the open chapter records are saved in a single array, once again
41 * different number of zones than previously, so the records must be parcelled out to their new
43 * records than it has space. In this case, the latest records for that zone will be discarded.
80 &open_chapter->records);
95 memset(open_chapter->records, 0, records_size(open_chapter));
122 record = &open_chapter->records[record_number];
148 *metadata = open_chapter->records[record_number].data;
172 record = &open_chapter->records[record_number];
197 vdo_free(open_chapter->records);
212 unsigned int records = 0;
220 * the chapter zone if it was closed early, and also to replace any deleted records. The
228 fill_record = &zone->records[zone->size];
236 for (records = 0; records < records_per_chapter; records++) {
237 struct uds_volume_record *record = &collated_records[records];
241 record_index = 1 + (records / zone_count);
242 page_number = records / records_per_page;
243 open_chapter = chapter_zones[records % zone_count];
252 *record = open_chapter->records[record_index];
333 record = &open_chapter->records[record_index];
364 * Track which zones cannot accept any more records. If the open chapter had a different
365 * number of zones previously, some new zones may have more records than they have space
366 * for. These overflow records will be discarded.