diff options
author | Ulrich Drepper <drepper@redhat.com> | 1997-08-27 20:26:10 +0000 |
---|---|---|
committer | Ulrich Drepper <drepper@redhat.com> | 1997-08-27 20:26:10 +0000 |
commit | 92f1da4da04a7a86ddee91be5eaf0b10c333ac64 (patch) | |
tree | 2a10ce9e4e407e7e5b5ca092ca0947d234b5ff60 /db2/hash/hash.c | |
parent | 22be878ecbc66606371bd33258f56e6711e6ba7a (diff) | |
download | glibc-92f1da4da04a7a86ddee91be5eaf0b10c333ac64.tar.gz glibc-92f1da4da04a7a86ddee91be5eaf0b10c333ac64.tar.xz glibc-92f1da4da04a7a86ddee91be5eaf0b10c333ac64.zip |
Update. cvs/libc-ud-970827
1997-08-10 19:17 Philip Blundell <Philip.Blundell@pobox.com> * nss/nss_db/db-XXX.c: Include <db_185.h> not <db.h>. Somebody should update this to use the new db API. * nss/nss_db/db-netgrp.c: Likewise. * nss/nss_db/db-alias.c: Likewise. * db2/Makefile: Makefile for db-2.x in glibc. 1997-08-27 21:20 Ulrich Drepper <drepper@cygnus.com> * csu/Makefile (before-compile): New goal. Make sure abi-tag.h is generated. [$(elf)=yes] (asm-CPPFLAGS): Make sure abi-tag.h file can be found. * Makeconfig [$(build-omitfp)=yes] (CFLAGS-.o): Add -D__USE_STRING_INLINES. * string/string.f: Move strnlen optimization after inclusion of <bits/string.h>. Include <bits/string.h> only if __USE_STRING_INLINES is defined. * sysdeps/generic/memcpy.c: Undef memcpy to allow macro of this name in <bits/string.h>. * sysdeps/generic/memset.c: Likewise. * sysdeps/i386/string.h: i386 optimized string functions. * sysdeps/i386/i486string.h: i486+ optimized string functions. * Makefile (subdirs): Change db to db2. * shlib-versions: Bump libdb verion number to 3. * include/db.h: Include from db2 directory. * include/db_185.h: New file. * sysdeps/i386/Makefile [$(subdirs)=db2] (CPPFLAGS): Add macros to provide spinlock information for db2. * sysdeps/m68k/m68020/Makefile: New file. Likewise. * sysdeps/sparc/Makefile: New file. Likewise. * sysdeps/unix/sysv/linux/Makefile [$(subdirs)=db2] (CPPFLAGS): Add -DHAVE_LLSEEK. * db2/config.h: Hand-edited config file for db2 in glibc. * db2/compat.h: New file from db-2.3.4. * db2/db.h: Likewise. * db2/db_185.h: Likewise. * db2/db_int.h: Likewise. * db2/makedb.c: Likewise. * db2/btree/bt_close.c: Likewise. * db2/btree/bt_compare.c: Likewise. * db2/btree/bt_conv.c: Likewise. * db2/btree/bt_cursor.c: Likewise. * db2/btree/bt_delete.c: Likewise. * db2/btree/bt_open.c: Likewise. * db2/btree/bt_page.c: Likewise. * db2/btree/bt_put.c: Likewise. * db2/btree/bt_rec.c: Likewise. * db2/btree/bt_recno.c: Likewise. * db2/btree/btree_auto.c: Likewise. * db2/btree/bt_rsearch.c: Likewise. * db2/btree/bt_search.c: Likewise. * db2/btree/bt_split.c: Likewise. * db2/btree/bt_stat.c: Likewise. * db2/btree/btree.src: Likewise. * db2/common/db_appinit.c: Likewise. * db2/common/db_err.c: Likewise. * db2/common/db_byteorder.c: Likewise. * db2/common/db_apprec.c: Likewise. * db2/common/db_salloc.c: Likewise. * db2/common/db_log2.c: Likewise. * db2/common/db_region.c: Likewise. * db2/common/db_shash.c: Likewise. * db2/db/db.c: Likewise. * db2/db/db.src: Likewise. * db2/db/db_conv.c: Likewise. * db2/db/db_dispatch.c: Likewise. * db2/db/db_dup.c: Likewise. * db2/db/db_overflow.c: Likewise. * db2/db/db_pr.c: Likewise. * db2/db/db_rec.c: Likewise. * db2/db/db_ret.c: Likewise. * db2/db/db_thread.c: Likewise. * db2/db/db_auto.c: Likewise. * db2/db185/db185.c: Likewise. * db2/db185/db185_int.h: Likewise. * db2/dbm/dbm.c: Likewise. * db2/hash/hash.c: Likewise. * db2/hash/hash.src: Likewise. * db2/hash/hash_page.c: Likewise. * db2/hash/hash_conv.c: Likewise. * db2/hash/hash_debug.c: Likewise. * db2/hash/hash_stat.c: Likewise. * db2/hash/hash_rec.c: Likewise. * db2/hash/hash_dup.c: Likewise. * db2/hash/hash_func.c: Likewise. * db2/hash/hash_auto.c: Likewise. * db2/include/mp.h: Likewise. * db2/include/btree.h: Likewise. * db2/include/db.h.src: Likewise. * db2/include/db_int.h.src: Likewise. * db2/include/db_shash.h: Likewise. * db2/include/db_swap.h: Likewise. * db2/include/db_185.h.src: Likewise. * db2/include/txn.h: Likewise. * db2/include/db_am.h: Likewise. * db2/include/shqueue.h: Likewise. * db2/include/hash.h: Likewise. * db2/include/db_dispatch.h: Likewise. * db2/include/lock.h: Likewise. * db2/include/db_page.h: Likewise. * db2/include/log.h: Likewise. * db2/include/db_auto.h: Likewise. * db2/include/btree_auto.h: Likewise. * db2/include/hash_auto.h: Likewise. * db2/include/log_auto.h: Likewise. * db2/include/txn_auto.h: Likewise. * db2/include/db_ext.h: Likewise. * db2/include/btree_ext.h: Likewise. * db2/include/clib_ext.h: Likewise. * db2/include/common_ext.h: Likewise. * db2/include/hash_ext.h: Likewise. * db2/include/lock_ext.h: Likewise. * db2/include/log_ext.h: Likewise. * db2/include/mp_ext.h: Likewise. * db2/include/mutex_ext.h: Likewise. * db2/include/os_ext.h: Likewise. * db2/include/txn_ext.h: Likewise. * db2/include/cxx_int.h: Likewise. * db2/include/db_cxx.h: Likewise. * db2/include/queue.h: Likewise. * db2/lock/lock.c: Likewise. * db2/lock/lock_conflict.c: Likewise. * db2/lock/lock_util.c: Likewise. * db2/lock/lock_deadlock.c: Likewise. * db2/log/log.c: Likewise. * db2/log/log_get.c: Likewise. * db2/log/log.src: Likewise. * db2/log/log_compare.c: Likewise. * db2/log/log_put.c: Likewise. * db2/log/log_rec.c: Likewise. * db2/log/log_archive.c: Likewise. * db2/log/log_register.c: Likewise. * db2/log/log_auto.c: Likewise. * db2/log/log_findckp.c: Likewise. * db2/mp/mp_bh.c: Likewise. * db2/mp/mp_fget.c: Likewise. * db2/mp/mp_fopen.c: Likewise. * db2/mp/mp_fput.c: Likewise. * db2/mp/mp_fset.c: Likewise. * db2/mp/mp_open.c: Likewise. * db2/mp/mp_region.c: Likewise. * db2/mp/mp_pr.c: Likewise. * db2/mp/mp_sync.c: Likewise. * db2/mutex/68020.gcc: Likewise. * db2/mutex/mutex.c: Likewise. * db2/mutex/README: Likewise. * db2/mutex/x86.gcc: Likewise. * db2/mutex/sparc.gcc: Likewise. * db2/mutex/uts4.cc.s: Likewise. * db2/mutex/alpha.dec: Likewise. * db2/mutex/alpha.gcc: Likewise. * db2/mutex/parisc.gcc: Likewise. * db2/mutex/parisc.hp: Likewise. * db2/os/db_os_abs.c: Likewise. * db2/os/db_os_dir.c: Likewise. * db2/os/db_os_fid.c: Likewise. * db2/os/db_os_lseek.c: Likewise. * db2/os/db_os_mmap.c: Likewise. * db2/os/db_os_open.c: Likewise. * db2/os/db_os_rw.c: Likewise. * db2/os/db_os_sleep.c: Likewise. * db2/os/db_os_stat.c: Likewise. * db2/os/db_os_unlink.c: Likewise. * db2/txn/txn.c: Likewise. * db2/txn/txn.src: Likewise. * db2/txn/txn_rec.c: Likewise. * db2/txn/txn_auto.c: Likewise. * db2/clib/getlong.c: Likewise. * db2/progs/db_archive/db_archive.c: Likewise. * db2/progs/db_checkpoint/db_checkpoint.c: Likewise. * db2/progs/db_deadlock/db_deadlock.c: Likewise. * db2/progs/db_dump/db_dump.c: Likewise. * db2/progs/db_dump185/db_dump185.c: Likewise. * db2/progs/db_load/db_load.c: Likewise. * db2/progs/db_printlog/db_printlog.c: Likewise. * db2/progs/db_recover/db_recover.c: Likewise. * db2/progs/db_stat/db_stat.c: Likewise. * libio/stdio.h [__cplusplus] (__STDIO_INLINE): Define as inline. * po/de.po, po/sv.po: Update from 2.0.5 translations. * sysdeps/unix/sysv/linux/netinet/tcp.h: Pretty print. * sunrpc/rpc/xdr.h (XDR): Don't define argument of x_destroy callback as const. * sunrpc/xdr_mem.c (xdrmem_destroy): Don't define argument as const. * sunrpx/xdr_rec.c (xdrrec_destroy): Likewise. * sunrpx/xdr_stdio.c (xdrstdio_destroy): Likewise. 1997-08-27 18:47 Ulrich Drepper <drepper@cygnus.com> * sysdeps/unix/sysv/linux/if_index.c: Include <errno.h>. Reported by Benjamin Kosnik <bkoz@cygnus.com>. 1997-08-27 02:27 Roland McGrath <roland@baalperazim.frob.com> * abi-tags: New file. * csu/Makefile (distribute): Remove abi-tag.h. ($(objpfx)abi-tag.h): New target. * Makefile (distribute): Add abi-tags. * sysdeps/unix/sysv/linux/abi-tag.h: File removed. * sysdeps/mach/hurd/abi-tag.h: File removed. * sysdeps/stub/abi-tag.h: File removed. 1997-08-25 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/unix/make-syscalls.sh: Change output so that it generates compilation rules only for the currently selected object suffixes. 1997-08-25 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/m68k/dl-machine.h (RTLD_START): Switch back to previous section to avoid confusing the compiler. * sysdeps/alpha/dl-machine.h (RTLD_START): Likewise. * sysdeps/i386/dl-machine.h (RTLD_START): Likewise. * sysdeps/mips/dl-machine.h (RTLD_START): Likewise. * sysdeps/mips/mips64/dl-machine.h (RTLD_START): Likewise. * sysdeps/sparc/sparc32/dl-machine.h (RTLD_START): Likewise. * sysdeps/m68k/dl-machine.h (elf_machine_load_address): Use a GOT relocation instead of a constant to avoid text relocation. (ELF_MACHINE_BEFORE_RTLD_RELOC): Removed. (RTLD_START): Declare global labels as functions and add size directive. 1997-08-25 17:01 Ulrich Drepper <drepper@cygnus.com> * sysdeps/i386/bits/select.h: Correct assembler versions to work even for descriptors >= 32. * stdlib/alloca.h: Don't define alloca to __alloca since if gcc is used __alloca is not defined to __builtin_alloca and so might not be available. Reported by Uwe Ohse <uwe@ohse.de>. * sysdeps/unix/sysv/linux/sys/sysmacros.h: Define macros in a special way if gcc is not used and so dev_t is an array. Reported by Uwe Ohse <uwe@ohse.de>. 1997-08-23 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * manual/libc.texinfo: Reorder chapters to match logical order. 1997-08-25 12:22 Ulrich Drepper <drepper@cygnus.com> * sunrpc/rpc/xdr.h: Change name of parameters in prototypes of xdr_reference, xdrmem_create, and xdrstdio_create because of clash with g++ internal symbols. Patch by Sudish Joseph <sj@eng.mindspring.net>. * elf/dl-deps.c: Implement handling of DT_FILTER.
Diffstat (limited to 'db2/hash/hash.c')
-rw-r--r-- | db2/hash/hash.c | 1440 |
1 files changed, 1440 insertions, 0 deletions
diff --git a/db2/hash/hash.c b/db2/hash/hash.c new file mode 100644 index 0000000000..6d8c40057d --- /dev/null +++ b/db2/hash/hash.c @@ -0,0 +1,1440 @@ +/*- + * See the file LICENSE for redistribution information. + * + * Copyright (c) 1996, 1997 + * Sleepycat Software. All rights reserved. + */ +/* + * Copyright (c) 1990, 1993, 1994 + * Margo Seltzer. All rights reserved. + */ +/* + * Copyright (c) 1990, 1993, 1994 + * The Regents of the University of California. All rights reserved. + * + * This code is derived from software contributed to Berkeley by + * Margo Seltzer. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * 3. All advertising materials mentioning features or use of this software + * must display the following acknowledgement: + * This product includes software developed by the University of + * California, Berkeley and its contributors. + * 4. Neither the name of the University nor the names of its contributors + * may be used to endorse or promote products derived from this software + * without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + */ + +#include "config.h" + +#ifndef lint +static const char sccsid[] = "@(#)hash.c 10.25 (Sleepycat) 8/24/97"; +#endif /* not lint */ + +#ifndef NO_SYSTEM_INCLUDES +#include <sys/types.h> +#include <sys/stat.h> + +#include <errno.h> +#include <fcntl.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <unistd.h> +#endif + +#include "shqueue.h" +#include "db_int.h" +#include "db_page.h" +#include "db_am.h" +#include "db_ext.h" +#include "hash.h" +#include "log.h" + +static int __ham_c_close __P((DBC *)); +static int __ham_c_del __P((DBC *, int)); +static int __ham_c_get __P((DBC *, DBT *, DBT *, int)); +static int __ham_c_put __P((DBC *, DBT *, DBT *, int)); +static int __ham_c_init __P((DB *, DB_TXN *, DBC **)); +static int __ham_cursor __P((DB *, DB_TXN *, DBC **)); +static int __ham_delete __P((DB *, DB_TXN *, DBT *, int)); +static int __ham_dup_return __P((HTAB *, HASH_CURSOR *, DBT *, int)); +static int __ham_get __P((DB *, DB_TXN *, DBT *, DBT *, int)); +static void __ham_init_htab __P((HTAB *)); +static int __ham_lookup __P((HTAB *, + HASH_CURSOR *, const DBT *, u_int32_t, db_lockmode_t)); +static int __ham_overwrite __P((HTAB *, HASH_CURSOR *, DBT *)); +static int __ham_put __P((DB *, DB_TXN *, DBT *, DBT *, int)); +static int __ham_sync __P((DB *, int)); + +/************************** INTERFACE ROUTINES ***************************/ +/* OPEN/CLOSE */ + +/* + * __ham_open -- + * + * PUBLIC: int __ham_open __P((DB *, DB_INFO *)); + */ +int +__ham_open(dbp, dbinfo) + DB *dbp; + DB_INFO *dbinfo; +{ + DB_ENV *dbenv; + DBC *curs; + HTAB *hashp; + int file_existed, ret; + + dbenv = dbp->dbenv; + + if ((hashp = (HTAB *)calloc(1, sizeof(HTAB))) == NULL) + return (ENOMEM); + hashp->dbp = dbp; + + /* Set the hash function if specified by the user. */ + if (dbinfo != NULL && dbinfo->h_hash != NULL) + hashp->hash = dbinfo->h_hash; + + /* + * Initialize the remaining fields of the dbp. The type, close and + * fd functions are all set in db_open. + */ + dbp->internal = hashp; + dbp->cursor = __ham_cursor; + dbp->del = __ham_delete; + dbp->get = __ham_get; + dbp->put = __ham_put; + dbp->sync = __ham_sync; + + /* If locking is turned on, lock the meta data page. */ + if (F_ISSET(dbp, DB_AM_LOCKING)) { + dbp->lock.pgno = BUCKET_INVALID; + if ((ret = lock_get(dbenv->lk_info, dbp->locker, + 0, &dbp->lock_dbt, DB_LOCK_READ, &hashp->hlock)) != 0) { + if (ret < 0) + ret = EAGAIN; + goto out; + } + } + + /* + * Now, we can try to read the meta-data page and figure out + * if we set up locking and get the meta-data page properly. + * If this is a new file, initialize it, and put it back dirty. + */ + if ((ret = __ham_get_page(hashp->dbp, 0, (PAGE **)&hashp->hdr)) != 0) + goto out; + + /* Initialize the hashp structure */ + if (hashp->hdr->magic == DB_HASHMAGIC) { + file_existed = 1; + /* File exists, verify the data in the header. */ + if (hashp->hash == NULL) + hashp->hash = + hashp->hdr->version < 5 ? __ham_func4 : __ham_func5; + if (hashp->hash(CHARKEY, sizeof(CHARKEY)) != + hashp->hdr->h_charkey) { + __db_err(hashp->dbp->dbenv, + "hash: incompatible hash function"); + ret = EINVAL; + goto out; + } + if (F_ISSET(hashp->hdr, DB_HASH_DUP)) + F_SET(dbp, DB_AM_DUP); + } else { + /* + * File does not exist, we must initialize the header. If + * locking is enabled that means getting a write lock first. + */ + file_existed = 0; + if (F_ISSET(dbp, DB_AM_LOCKING) && + ((ret = lock_put(dbenv->lk_info, hashp->hlock)) != 0 || + (ret = lock_get(dbenv->lk_info, dbp->locker, 0, + &dbp->lock_dbt, DB_LOCK_WRITE, &hashp->hlock)) != 0)) { + if (ret < 0) + ret = EAGAIN; + goto out; + } + + hashp->hdr->nelem = dbinfo != NULL ? dbinfo->h_nelem : 0; + hashp->hdr->ffactor = + dbinfo != NULL && dbinfo->h_ffactor ? dbinfo->h_ffactor : 0; + __ham_init_htab(hashp); + if (F_ISSET(dbp, DB_AM_DUP)) + F_SET(hashp->hdr, DB_HASH_DUP); + if ((ret = __ham_dirty_page(hashp, (PAGE *)hashp->hdr)) != 0) + goto out; + } + + /* Initialize the default cursor. */ + __ham_c_init(dbp, NULL, &curs); + TAILQ_INSERT_TAIL(&dbp->curs_queue, curs, links); + + /* Allocate memory for our split buffer. */ + if ((hashp->split_buf = (PAGE *)malloc(dbp->pgsize)) == NULL) { + ret = ENOMEM; + goto out; + } + +#ifdef NO_STATISTICS_FOR_DB_ERR + __db_err(dbp->dbenv, + "%s%lx\n%s%ld\n%s%ld\n%s%ld\n%s%ld\n%s0x%lx\n%s0x%lx\n%s%ld\n%s%ld\n%s0x%lx", + "TABLE POINTER ", (long)hashp, + "BUCKET SIZE ", (long)hashp->hdr->pagesize, + "FILL FACTOR ", (long)hashp->hdr->ffactor, + "MAX BUCKET ", (long)hashp->hdr->max_bucket, + "OVFL POINT ", (long)hashp->hdr->ovfl_point, + "LAST FREED ", (long)hashp->hdr->last_freed, + "HIGH MASK ", (long)hashp->hdr->high_mask, + "LOW MASK ", (long)hashp->hdr->low_mask, + "NELEM ", (long)hashp->hdr->nelem, + "FLAGS ", (long)hashp->hdr->flags); +#endif + + /* Release the meta data page */ + (void)__ham_put_page(hashp->dbp, (PAGE *)hashp->hdr, 0); + if (F_ISSET(dbp, DB_AM_LOCKING) && + (ret = lock_put(dbenv->lk_info, hashp->hlock)) != 0) { + if (ret < 0) + ret = EAGAIN; + goto out; + } + + hashp->hlock = 0; + hashp->hdr = NULL; + /* Sync the file so that we know that the meta data goes to disk. */ + if (!file_existed && (ret = dbp->sync(dbp, 0)) != 0) + goto out; + return (0); + +out: (void)__ham_close(dbp); + return (ret); +} + +/* + * PUBLIC: int __ham_close __P((DB *)); + */ +int +__ham_close(dbp) + DB *dbp; +{ + HTAB *hashp; + int ret, t_ret; + + DEBUG_LWRITE(dbp, NULL, "ham_close", NULL, NULL, 0); + hashp = (HTAB *)dbp->internal; + ret = 0; + + /* Free the split page. */ + if (hashp->split_buf) + FREE(hashp->split_buf, dbp->pgsize); + + if (hashp->hdr && (t_ret = __ham_put_page(hashp->dbp, + (PAGE *)hashp->hdr, 0)) != 0 && ret == 0) + ret = t_ret; + if (hashp->hlock && (t_ret = lock_put(hashp->dbp->dbenv->lk_info, + hashp->hlock)) != 0 && ret == 0) + ret = t_ret; + + FREE(hashp, sizeof(HTAB)); + dbp->internal = NULL; + return (ret); +} + +/************************** LOCAL CREATION ROUTINES **********************/ +/* + * Returns 0 on No Error + */ +static void +__ham_init_htab(hashp) + HTAB *hashp; +{ + u_int32_t nelem; + int32_t l2, nbuckets; + + nelem = hashp->hdr->nelem; + hashp->hdr->pagesize = hashp->dbp->pgsize; + ZERO_LSN(hashp->hdr->lsn); + hashp->hdr->magic = DB_HASHMAGIC; + hashp->hdr->version = DB_HASHVERSION; + if (hashp->hash == NULL) + hashp->hash = + hashp->hdr->version < 5 ? __ham_func4 : __ham_func5; + hashp->hdr->h_charkey = hashp->hash(CHARKEY, sizeof(CHARKEY)); + if (nelem != 0 && hashp->hdr->ffactor != 0) { + nelem = (nelem - 1) / hashp->hdr->ffactor + 1; + l2 = __db_log2(nelem > 2 ? nelem : 2); + } else + l2 = 2; + + nbuckets = 1 << l2; + + hashp->hdr->spares[l2] = 0; + hashp->hdr->spares[l2 + 1] = 0; + hashp->hdr->ovfl_point = l2; + hashp->hdr->last_freed = PGNO_INVALID; + + hashp->hdr->max_bucket = hashp->hdr->high_mask = nbuckets - 1; + hashp->hdr->low_mask = (nbuckets >> 1) - 1; + memcpy(hashp->hdr->uid, hashp->dbp->lock.fileid, DB_FILE_ID_LEN); +} + +/********************** DESTROY/CLOSE ROUTINES ************************/ + + +/* + * Write modified pages to disk + * + * Returns: + * 0 == OK + * -1 ERROR + */ +static int +__ham_sync(dbp, flags) + DB *dbp; + int flags; +{ + int ret; + + DEBUG_LWRITE(dbp, NULL, "ham_sync", NULL, NULL, flags); + if ((ret = __db_syncchk(dbp, flags)) != 0) + return (ret); + if (F_ISSET(dbp, DB_AM_RDONLY)) + return (0); + + if ((ret = memp_fsync(dbp->mpf)) == DB_INCOMPLETE) + ret = 0; + + return (ret); +} + +/*******************************SEARCH ROUTINES *****************************/ +/* + * All the access routines return + * + * Returns: + * 0 on SUCCESS + * 1 to indicate an external ERROR (i.e. key not found, etc) + * -1 to indicate an internal ERROR (i.e. out of memory, etc) + */ + +static int +__ham_get(dbp, txn, key, data, flags) + DB *dbp; + DB_TXN *txn; + DBT *key; + DBT *data; + int flags; +{ + DB *ldbp; + DBC *cp; + HTAB *hashp; + HASH_CURSOR *hcp; + int ret, t_ret; + + DEBUG_LREAD(dbp, txn, "ham_get", key, NULL, flags); + if ((ret = __db_getchk(dbp, key, data, flags)) != 0) + return (ret); + + ldbp = dbp; + if (F_ISSET(dbp, DB_AM_THREAD) && + (ret = __db_gethandle(dbp, __ham_hdup, &ldbp)) != 0) + return (ret); + + hashp = (HTAB *)ldbp->internal; + SET_LOCKER(ldbp, txn); + GET_META(ldbp, hashp); + cp = TAILQ_FIRST(&ldbp->curs_queue); + + hashp->hash_accesses++; + hcp = (HASH_CURSOR *)TAILQ_FIRST(&ldbp->curs_queue)->internal; + if ((ret = __ham_lookup(hashp, hcp, key, 0, DB_LOCK_READ)) == 0) + if (F_ISSET(hcp, H_OK)) + ret = __ham_dup_return(hashp, hcp, data, DB_FIRST); + else /* Key was not found */ + ret = DB_NOTFOUND; + + if ((t_ret = __ham_item_done(hashp, hcp, 0)) != 0 && ret == 0) + ret = t_ret; + RELEASE_META(ldbp, hashp); + if (F_ISSET(dbp, DB_AM_THREAD)) + __db_puthandle(ldbp); + return (ret); +} + +static int +__ham_put(dbp, txn, key, data, flags) + DB *dbp; + DB_TXN *txn; + DBT *key; + DBT *data; + int flags; +{ + DB *ldbp; + HTAB *hashp; + HASH_CURSOR *hcp; + DBT tmp_val, *myval; + int ret, t_ret; + u_int32_t nbytes; + + DEBUG_LWRITE(dbp, txn, "ham_put", key, data, flags); + if ((ret = __db_putchk(dbp, key, data, + flags, F_ISSET(dbp, DB_AM_RDONLY), F_ISSET(dbp, DB_AM_DUP))) != 0) + return (ret); + + ldbp = dbp; + if (F_ISSET(dbp, DB_AM_THREAD) && + (ret = __db_gethandle(dbp, __ham_hdup, &ldbp)) != 0) + return (ret); + + hashp = (HTAB *)ldbp->internal; + SET_LOCKER(ldbp, txn); + GET_META(ldbp, hashp); + hcp = TAILQ_FIRST(&ldbp->curs_queue)->internal; + + nbytes = (ISBIG(hashp, key->size) ? HOFFPAGE_PSIZE : + HKEYDATA_PSIZE(key->size)) + + (ISBIG(hashp, data->size) ? HOFFPAGE_PSIZE : + HKEYDATA_PSIZE(data->size)); + + hashp->hash_accesses++; + ret = __ham_lookup(hashp, hcp, key, nbytes, DB_LOCK_WRITE); + + if (ret == DB_NOTFOUND) { + ret = 0; + if (hcp->seek_found_page != PGNO_INVALID && + hcp->seek_found_page != hcp->pgno) { + if ((ret = __ham_item_done(hashp, hcp, 0)) != 0) + goto out; + hcp->pgno = hcp->seek_found_page; + hcp->bndx = NDX_INVALID; + } + + if (F_ISSET(data, DB_DBT_PARTIAL) && data->doff != 0) { + /* + * Doing a partial put, but the key does not exist + * and we are not beginning the write at 0. We + * must create a data item padded up to doff and + * then write the new bytes represented by val. + */ + ret = __ham_init_dbt(&tmp_val, data->size + data->doff, + &hcp->big_data, &hcp->big_datalen); + if (ret == 0) { + memset(tmp_val.data, 0, data->doff); + memcpy((u_int8_t *)tmp_val.data + data->doff, + data->data, data->size); + myval = &tmp_val; + } + } else + myval = (DBT *)data; + + if (ret == 0) + ret = __ham_add_el(hashp, hcp, key, myval, H_KEYDATA); + } else if (ret == 0 && F_ISSET(hcp, H_OK)) { + if (flags == DB_NOOVERWRITE) + ret = DB_KEYEXIST; + else if (F_ISSET(ldbp, DB_AM_DUP)) + ret = __ham_add_dup(hashp, hcp, data, DB_KEYLAST); + else + ret = __ham_overwrite(hashp, hcp, data); + } + + /* Free up all the cursor pages. */ + if ((t_ret = __ham_item_done(hashp, hcp, ret == 0)) != 0 && ret == 0) + ret = t_ret; + /* Now check if we have to grow. */ +out: if (ret == 0 && F_ISSET(hcp, H_EXPAND)) { + ret = __ham_expand_table(hashp); + F_CLR(hcp, H_EXPAND); + } + + if ((t_ret = __ham_item_done(hashp, hcp, ret == 0)) != 0 && ret == 0) + ret = t_ret; + RELEASE_META(ldbp, hashp); + if (F_ISSET(dbp, DB_AM_THREAD)) + __db_puthandle(ldbp); + return (ret); +} + +static int +__ham_cursor(dbp, txnid, dbcp) + DB *dbp; + DB_TXN *txnid; + DBC **dbcp; +{ + int ret; + + DEBUG_LWRITE(dbp, txnid, "ham_cursor", NULL, NULL, 0); + if ((ret = __ham_c_init(dbp, txnid, dbcp)) != 0) + return (ret); + + DB_THREAD_LOCK(dbp); + TAILQ_INSERT_TAIL(&dbp->curs_queue, *dbcp, links); + DB_THREAD_UNLOCK(dbp); + return (ret); +} + +static int +__ham_c_init(dbp, txnid, dbcp) + DB *dbp; + DB_TXN *txnid; + DBC **dbcp; +{ + DBC *db_curs; + HASH_CURSOR *new_curs; + + if ((db_curs = (DBC *)calloc(sizeof(DBC), 1)) == NULL) + return (ENOMEM); + + if ((new_curs = + (HASH_CURSOR *)calloc(sizeof(struct cursor_t), 1)) == NULL) { + FREE(db_curs, sizeof(DBC)); + return (ENOMEM); + } + + db_curs->internal = new_curs; + db_curs->c_close = __ham_c_close; + db_curs->c_del = __ham_c_del; + db_curs->c_get = __ham_c_get; + db_curs->c_put = __ham_c_put; + db_curs->txn = txnid; + db_curs->dbp = dbp; + + new_curs->db_cursor = db_curs; + __ham_item_init(new_curs); + + if (dbcp != NULL) + *dbcp = db_curs; + return (0); +} + +static int +__ham_delete(dbp, txn, key, flags) + DB *dbp; + DB_TXN *txn; + DBT *key; + int flags; +{ + DB *ldbp; + HTAB *hashp; + HASH_CURSOR *hcp; + int ret, t_ret; + + DEBUG_LWRITE(dbp, txn, "ham_delete", key, NULL, flags); + if ((ret = __db_delchk(dbp, flags, F_ISSET(dbp, DB_AM_RDONLY))) != 0) + return (ret); + + ldbp = dbp; + if (F_ISSET(dbp, DB_AM_THREAD) && + (ret = __db_gethandle(dbp, __ham_hdup, &ldbp)) != 0) + return (ret); + hashp = (HTAB *)ldbp->internal; + SET_LOCKER(ldbp, txn); + GET_META(ldbp, hashp); + hcp = TAILQ_FIRST(&ldbp->curs_queue)->internal; + + hashp->hash_accesses++; + if ((ret = __ham_lookup(hashp, hcp, key, 0, DB_LOCK_WRITE)) == 0) + if (F_ISSET(hcp, H_OK)) + ret = __ham_del_pair(hashp, hcp); + else + ret = DB_NOTFOUND; + + if ((t_ret = __ham_item_done(hashp, hcp, ret == 0)) != 0 && ret == 0) + ret = t_ret; + RELEASE_META(ldbp, hashp); + if (F_ISSET(dbp, DB_AM_THREAD)) + __db_puthandle(ldbp); + return (ret); +} + +/* ****************** CURSORS ********************************** */ +static int +__ham_c_close(cursor) + DBC *cursor; +{ + DB *ldbp; + HTAB *hashp; + HASH_CURSOR *hcp; + int ret; + + DEBUG_LWRITE(cursor->dbp, cursor->txn, "ham_c_close", NULL, NULL, 0); + /* + * If the pagep, dpagep, and lock fields of the cursor are all NULL, + * then there really isn't a need to get a handle here. However, + * the normal case is that at least one of those fields is non-NULL, + * and putting those checks in here would couple the ham_item_done + * functionality with cursor close which would be pretty disgusting. + * Instead, we pay the overhead here of always getting the handle. + */ + ldbp = cursor->dbp; + if (F_ISSET(cursor->dbp, DB_AM_THREAD) && + (ret = __db_gethandle(cursor->dbp, __ham_hdup, &ldbp)) != 0) + return (ret); + hashp = (HTAB *)ldbp->internal; + hcp = (HASH_CURSOR *)cursor->internal; + ret = __ham_item_done(hashp, hcp, 0); + + if (hcp->big_key) + FREE(hcp->big_key, hcp->big_keylen); + if (hcp->big_data) + FREE(hcp->big_data, hcp->big_datalen); + + /* + * All cursors (except the default ones) are linked off the master. + * Therefore, when we close the cursor, we have to remove it from + * the master, not the local one. When we are closing the file in + * its entirety, then we clear the THREAD bit and the master and + * local are identical, so we remove the correct one. + */ + DB_THREAD_LOCK(cursor->dbp); + TAILQ_REMOVE(&cursor->dbp->curs_queue, cursor, links); + DB_THREAD_UNLOCK(cursor->dbp); + + if (F_ISSET(cursor->dbp, DB_AM_THREAD)) + __db_puthandle(ldbp); + + FREE(hcp, sizeof(HASH_CURSOR)); + FREE(cursor, sizeof(DBC)); + return (ret); +} + +static int +__ham_c_del(cursor, flags) + DBC *cursor; + int flags; +{ + DB *ldbp; + HTAB *hashp; + HASH_CURSOR *hcp; + HASH_CURSOR save_curs; + db_pgno_t ppgno, chg_pgno; + int ret, t_ret; + + DEBUG_LWRITE(cursor->dbp, cursor->txn, "ham_c_del", NULL, NULL, flags); + ldbp = cursor->dbp; + if (F_ISSET(cursor->dbp, DB_AM_THREAD) && + (ret = __db_gethandle(cursor->dbp, __ham_hdup, &ldbp)) != 0) + return (ret); + hashp = (HTAB *)ldbp->internal; + hcp = (HASH_CURSOR *)cursor->internal; + save_curs = *hcp; + if ((ret = __db_cdelchk(ldbp, flags, + F_ISSET(ldbp, DB_AM_RDONLY), IS_VALID(hcp))) != 0) + return (ret); + if (F_ISSET(hcp, H_DELETED)) + return (DB_NOTFOUND); + + SET_LOCKER(hashp->dbp, cursor->txn); + GET_META(hashp->dbp, hashp); + hashp->hash_accesses++; + if ((ret = __ham_get_cpage(hashp, hcp, DB_LOCK_WRITE)) != 0) + goto out; + if (F_ISSET(hcp, H_ISDUP) && hcp->dpgno != PGNO_INVALID) { + ppgno = PREV_PGNO(hcp->dpagep); + + /* Remove item from duplicate page. */ + chg_pgno = hcp->dpgno; + if ((ret = __db_drem(hashp->dbp, + &hcp->dpagep, hcp->dndx, __ham_del_page)) != 0) + goto out; + + /* + * There are 4 cases. + * 1. We removed an item on a page, but nothing else changed. + * 2. We removed the last item on a page, but there is a + * following page of duplicates. + * 3. We removed the last item on a page, this page was the + * last page in a duplicate set, but there were dups before + * it. + * 4. We removed the last item on a page, removing the last + * duplicate. + * In case 1 hcp->dpagep is unchanged. + * In case 2 hcp->dpagep comes back pointing to the next dup + * page. + * In case 3 hcp->dpagep comes back NULL. + * In case 4 hcp->dpagep comes back NULL. + */ + if (hcp->dpagep == NULL) { + if (ppgno != PGNO_INVALID) { /* Case 3 */ + hcp->dpgno = ppgno; + if ((ret = __ham_get_cpage(hashp, hcp, + DB_LOCK_READ)) != 0) + goto out; + hcp->dndx = NUM_ENT(hcp->dpagep); + F_SET(hcp, H_DELETED); + } else { /* Case 4 */ + ret = __ham_del_pair(hashp, hcp); + hcp->dpgno = PGNO_INVALID; + /* + * Delpair updated the cursor queue, so we + * don't have to do that here. + */ + chg_pgno = PGNO_INVALID; + } + } else if (PGNO(hcp->dpagep) != hcp->dpgno) { + hcp->dndx = 0; /* Case 2 */ + hcp->dpgno = PGNO(hcp->dpagep); + if (ppgno == PGNO_INVALID) + memcpy(P_ENTRY(hcp->pagep, + H_DATAINDEX(hcp->bndx)) + + SSZ(HOFFDUP, pgno), &hcp->dpgno, + sizeof(db_pgno_t)); + F_SET(hcp, H_DELETED); + } else /* Case 1 */ + F_SET(hcp, H_DELETED); + if (chg_pgno != PGNO_INVALID) + __ham_c_update(hashp, hcp, chg_pgno, 0, 0, 1); + } else if (F_ISSET(hcp, H_ISDUP)) { /* on page */ + if (hcp->dup_off == 0 && DUP_SIZE(hcp->dup_len) == + LEN_HDATA(hcp->pagep, hashp->hdr->pagesize, hcp->bndx)) + ret = __ham_del_pair(hashp, hcp); + else { + DBT repldbt; + + repldbt.flags = 0; + F_SET(&repldbt, DB_DBT_PARTIAL); + repldbt.doff = hcp->dup_off; + repldbt.dlen = DUP_SIZE(hcp->dup_len); + repldbt.size = 0; + ret = __ham_replpair(hashp, hcp, &repldbt, 0); + hcp->dup_tlen -= DUP_SIZE(hcp->dup_len); + __ham_c_update(hashp, hcp, hcp->pgno, + DUP_SIZE(hcp->dup_len), 0, 1); + F_SET(hcp, H_DELETED); + } + + } else + /* Not a duplicate */ + ret = __ham_del_pair(hashp, hcp); + +out: if ((t_ret = __ham_item_done(hashp, hcp, ret == 0)) != 0 && ret == 0) + t_ret = ret; + if (ret != 0) + *hcp = save_curs; + RELEASE_META(hashp->dbp, hashp); + if (F_ISSET(cursor->dbp, DB_AM_THREAD)) + __db_puthandle(ldbp); + return (ret); +} + +static int +__ham_c_get(cursor, key, data, flags) + DBC *cursor; + DBT *key; + DBT *data; + int flags; +{ + DB *ldbp; + HTAB *hashp; + HASH_CURSOR *hcp, save_curs; + int get_key, ret, t_ret; + + DEBUG_LREAD(cursor->dbp, cursor->txn, "ham_c_get", + flags == DB_SET || flags == DB_SET_RANGE ? key : NULL, + NULL, flags); + ldbp = cursor->dbp; + if (F_ISSET(cursor->dbp, DB_AM_THREAD) && + (ret = __db_gethandle(cursor->dbp, __ham_hdup, &ldbp)) != 0) + return (ret); + hashp = (HTAB *)(ldbp->internal); + hcp = (HASH_CURSOR *)cursor->internal; + save_curs = *hcp; + if ((ret = + __db_cgetchk(hashp->dbp, key, data, flags, IS_VALID(hcp))) != 0) + return (ret); + + SET_LOCKER(hashp->dbp, cursor->txn); + GET_META(hashp->dbp, hashp); + hashp->hash_accesses++; + + hcp->seek_size = 0; + + ret = 0; + get_key = 1; + switch (flags) { + case DB_PREV: + if (hcp->bucket != BUCKET_INVALID) { + ret = __ham_item_prev(hashp, hcp, DB_LOCK_READ); + break; + } + /* FALL THROUGH */ + case DB_LAST: + ret = __ham_item_last(hashp, hcp, DB_LOCK_READ); + break; + case DB_FIRST: + ret = __ham_item_first(hashp, hcp, DB_LOCK_READ); + break; + case DB_NEXT: + if (hcp->bucket == BUCKET_INVALID) + hcp->bucket = 0; + ret = __ham_item_next(hashp, hcp, DB_LOCK_READ); + break; + case DB_SET: + case DB_SET_RANGE: + ret = __ham_lookup(hashp, hcp, key, 0, DB_LOCK_READ); + get_key = 0; + break; + case DB_CURRENT: + if (F_ISSET(hcp, H_DELETED)) { + ret = DB_KEYEMPTY; + goto out; + } + + ret = __ham_item(hashp, hcp, DB_LOCK_READ); + break; + } + + /* + * Must always enter this loop to do error handling and + * check for big key/data pair. + */ + while (1) { + if (ret != 0 && ret != DB_NOTFOUND) + goto out1; + else if (F_ISSET(hcp, H_OK)) { + /* Get the key. */ + if (get_key && (ret = __db_ret(hashp->dbp, hcp->pagep, + H_KEYINDEX(hcp->bndx), key, &hcp->big_key, + &hcp->big_keylen)) != 0) + goto out1; + + ret = __ham_dup_return(hashp, hcp, data, flags); + break; + } else if (!F_ISSET(hcp, H_NOMORE)) { + abort(); + break; + } + + /* + * Ran out of entries in a bucket; change buckets. + */ + switch (flags) { + case DB_LAST: + case DB_PREV: + ret = __ham_item_done(hashp, hcp, 0); + if (hcp->bucket == 0) { + ret = DB_NOTFOUND; + goto out1; + } + hcp->bucket--; + hcp->bndx = NDX_INVALID; + if (ret == 0) + ret = __ham_item_prev(hashp, + hcp, DB_LOCK_READ); + break; + case DB_FIRST: + case DB_NEXT: + ret = __ham_item_done(hashp, hcp, 0); + hcp->bndx = NDX_INVALID; + hcp->bucket++; + hcp->pgno = PGNO_INVALID; + hcp->pagep = NULL; + if (hcp->bucket > hashp->hdr->max_bucket) { + ret = DB_NOTFOUND; + goto out1; + } + if (ret == 0) + ret = __ham_item_next(hashp, + hcp, DB_LOCK_READ); + break; + case DB_SET: + case DB_SET_RANGE: + /* Key not found. */ + ret = DB_NOTFOUND; + goto out1; + } + } +out1: if ((t_ret = __ham_item_done(hashp, hcp, 0)) != 0 && ret == 0) + t_ret = ret; +out: if (ret) + *hcp = save_curs; + RELEASE_META(hashp->dbp, hashp); + if (F_ISSET(cursor->dbp, DB_AM_THREAD)) + __db_puthandle(ldbp); + return (ret); +} + +static int +__ham_c_put(cursor, key, data, flags) + DBC *cursor; + DBT *key; + DBT *data; + int flags; +{ + DB *ldbp; + HTAB *hashp; + HASH_CURSOR *hcp, save_curs; + int ret, t_ret; + u_int32_t nbytes; + + DEBUG_LWRITE(cursor->dbp, cursor->txn, "ham_c_put", + flags == DB_KEYFIRST || flags == DB_KEYLAST ? key : NULL, + NULL, flags); + ldbp = cursor->dbp; + if (F_ISSET(cursor->dbp, DB_AM_THREAD) && + (ret = __db_gethandle(cursor->dbp, __ham_hdup, &ldbp)) != 0) + return (ret); + hashp = (HTAB *)(ldbp->internal); + hcp = (HASH_CURSOR *)cursor->internal; + save_curs = *hcp; + + if ((ret = __db_cputchk(hashp->dbp, key, data, flags, + F_ISSET(ldbp, DB_AM_RDONLY), IS_VALID(hcp))) != 0) + return (ret); + if (F_ISSET(hcp, H_DELETED)) + return (DB_NOTFOUND); + + SET_LOCKER(hashp->dbp, cursor->txn); + GET_META(hashp->dbp, hashp); + ret = 0; + + switch (flags) { + case DB_KEYLAST: + case DB_KEYFIRST: + nbytes = (ISBIG(hashp, key->size) ? HOFFPAGE_PSIZE : + HKEYDATA_PSIZE(key->size)) + + (ISBIG(hashp, data->size) ? HOFFPAGE_PSIZE : + HKEYDATA_PSIZE(data->size)); + ret = __ham_lookup(hashp, hcp, key, nbytes, DB_LOCK_WRITE); + break; + case DB_BEFORE: + case DB_AFTER: + case DB_CURRENT: + ret = __ham_item(hashp, hcp, DB_LOCK_WRITE); + break; + } + + if (ret == 0) { + if (flags == DB_CURRENT && !F_ISSET(ldbp, DB_AM_DUP)) + ret = __ham_overwrite(hashp, hcp, data); + else + ret = __ham_add_dup(hashp, hcp, data, flags); + } + + if (ret == 0 && F_ISSET(hcp, H_EXPAND)) { + ret = __ham_expand_table(hashp); + F_CLR(hcp, H_EXPAND); + } + + if ((t_ret = __ham_item_done(hashp, hcp, ret == 0)) != 0 && ret == 0) + ret = t_ret; + if (ret != 0) + *hcp = save_curs; + RELEASE_META(hashp->dbp, hashp); + if (F_ISSET(cursor->dbp, DB_AM_THREAD)) + __db_puthandle(ldbp); + return (ret); +} + +/********************************* UTILITIES ************************/ + +/* + * __ham_expand_table -- + * + * PUBLIC: int __ham_expand_table __P((HTAB *)); + */ +int +__ham_expand_table(hashp) + HTAB *hashp; +{ + u_int32_t old_bucket, new_bucket; + u_int32_t spare_ndx; + int ret; + + ret = 0; + DIRTY_META(hashp, ret); + if (ret) + return (ret); + + if (DB_LOGGING(hashp->dbp)) { + DB_LSN new_lsn; + + if ((ret = __ham_splitmeta_log(hashp->dbp->dbenv->lg_info, + (DB_TXN *)hashp->dbp->txn, &new_lsn, 0, + hashp->dbp->log_fileid, + hashp->hdr->max_bucket, hashp->hdr->ovfl_point, + hashp->hdr->spares[hashp->hdr->ovfl_point], + &hashp->hdr->lsn)) != 0) + return (ret); + + hashp->hdr->lsn = new_lsn; + } + + hashp->hash_expansions++; + new_bucket = ++hashp->hdr->max_bucket; + old_bucket = (hashp->hdr->max_bucket & hashp->hdr->low_mask); + + /* + * If the split point is increasing (hdr.max_bucket's log base 2 + * increases), max sure that we have enough extra pages, then + * copy the current contents of the spare split bucket to the + * next bucket. + */ + spare_ndx = __db_log2(hashp->hdr->max_bucket + 1); + if (spare_ndx > hashp->hdr->ovfl_point) { + /* + * We are about to shift the split point. Make sure that + * if the next doubling is going to be big (more than 8 + * pages), we have some extra pages around. + */ + if (hashp->hdr->spares[hashp->hdr->ovfl_point] == 0 && + new_bucket >= 8) + __ham_init_ovflpages(hashp); + + hashp->hdr->spares[spare_ndx] = + hashp->hdr->spares[hashp->hdr->ovfl_point]; + hashp->hdr->ovfl_point = spare_ndx; + } + + if (new_bucket > hashp->hdr->high_mask) { + /* Starting a new doubling */ + hashp->hdr->low_mask = hashp->hdr->high_mask; + hashp->hdr->high_mask = new_bucket | hashp->hdr->low_mask; + } + + if (BUCKET_TO_PAGE(hashp, new_bucket) > MAX_PAGES(hashp)) { + __db_err(hashp->dbp->dbenv, + "hash: Cannot allocate new bucket. Pages exhausted."); + return (ENOSPC); + } + + /* Relocate records to the new bucket */ + return (__ham_split_page(hashp, old_bucket, new_bucket)); +} + +/* + * PUBLIC: u_int32_t __ham_call_hash __P((HTAB *, u_int8_t *, int32_t)); + */ +u_int32_t +__ham_call_hash(hashp, k, len) + HTAB *hashp; + u_int8_t *k; + int32_t len; +{ + u_int32_t n, bucket; + + n = (u_int32_t)hashp->hash(k, len); + bucket = n & hashp->hdr->high_mask; + if (bucket > hashp->hdr->max_bucket) + bucket = bucket & hashp->hdr->low_mask; + return (bucket); +} + +/* + * Check for duplicates, and call __db_ret appropriately. Release + * everything held by the cursor. + */ +static int +__ham_dup_return(hashp, hcp, val, flags) + HTAB *hashp; + HASH_CURSOR *hcp; + DBT *val; + int flags; +{ + HKEYDATA *hk; + PAGE *pp; + DBT *myval, tmp_val; + db_indx_t ndx; + db_pgno_t pgno; + u_int8_t type; + int indx, ret; + db_indx_t len; + + /* Check for duplicate and return the first one. */ + ndx = H_DATAINDEX(hcp->bndx); + type = GET_HKEYDATA(hcp->pagep, ndx)->type; + pp = hcp->pagep; + myval = val; + + /* + * There are 3 cases: + * 1. We are not in duplicate, simply call db_ret. + * 2. We are looking at keys and stumbled onto a duplicate. + * 3. We are in the middle of a duplicate set. (ISDUP set) + */ + + /* + * Here we check for the case where we just stumbled onto a + * duplicate. In this case, we do initialization and then + * let the normal duplicate code handle it. + */ + if (!F_ISSET(hcp, H_ISDUP)) + if (type == H_DUPLICATE) { + F_SET(hcp, H_ISDUP); + hcp->dup_tlen = LEN_HDATA(hcp->pagep, + hashp->hdr->pagesize, hcp->bndx); + hk = H_PAIRDATA(hcp->pagep, hcp->bndx); + if (flags == DB_LAST || flags == DB_PREV) { + hcp->dndx = 0; + hcp->dup_off = 0; + do { + memcpy(&len, hk->data + hcp->dup_off, + sizeof(db_indx_t)); + hcp->dup_off += DUP_SIZE(len); + hcp->dndx++; + } while (hcp->dup_off < hcp->dup_tlen); + hcp->dup_off -= DUP_SIZE(len); + hcp->dndx--; + } else { + memcpy(&len, hk->data, sizeof(db_indx_t)); + hcp->dup_off = 0; + hcp->dndx = 0; + } + hcp->dup_len = len; + } else if (type == H_OFFDUP) { + F_SET(hcp, H_ISDUP); + memcpy(&pgno, + P_ENTRY(hcp->pagep, ndx) + SSZ(HOFFDUP, pgno), + sizeof(db_pgno_t)); + if (flags == DB_LAST || flags == DB_PREV) { + indx = (int)hcp->dndx; + if ((ret = __db_dend(hashp->dbp, + pgno, &hcp->dpagep)) != 0) + return (ret); + hcp->dpgno = PGNO(hcp->dpagep); + hcp->dndx = NUM_ENT(hcp->dpagep) - 1; + } else if ((ret = __ham_next_cpage(hashp, + hcp, pgno, 0, H_ISDUP)) != 0) + return (ret); + } + + + /* + * Now, everything is initialized, grab a duplicate if + * necessary. + */ + if (F_ISSET(hcp, H_ISDUP)) + if (hcp->dpgno != PGNO_INVALID) { + pp = hcp->dpagep; + ndx = hcp->dndx; + } else { + /* + * Copy the DBT in case we are retrieving into + * user memory and we need the parameters for + * it. + */ + memcpy(&tmp_val, val, sizeof(*val)); + F_SET(&tmp_val, DB_DBT_PARTIAL); + tmp_val.dlen = hcp->dup_len; + tmp_val.doff = hcp->dup_off + sizeof(db_indx_t); + myval = &tmp_val; + } + + + /* + * Finally, if we had a duplicate, pp, ndx, and myval should be + * set appropriately. + */ + if ((ret = __db_ret(hashp->dbp, pp, ndx, myval, &hcp->big_data, + &hcp->big_datalen)) != 0) + return (ret); + + /* + * In case we sent a temporary off to db_ret, set the real + * return values. + */ + val->data = myval->data; + val->size = myval->size; + + return (0); +} + +static int +__ham_overwrite(hashp, hcp, nval) + HTAB *hashp; + HASH_CURSOR *hcp; + DBT *nval; +{ + DBT *myval, tmp_val; + HKEYDATA *hk; + + if (F_ISSET(hashp->dbp, DB_AM_DUP)) + return (__ham_add_dup(hashp, hcp, nval, DB_KEYLAST)); + else if (!F_ISSET(nval, DB_DBT_PARTIAL)) { + /* Put/overwrite */ + memcpy(&tmp_val, nval, sizeof(*nval)); + F_SET(&tmp_val, DB_DBT_PARTIAL); + tmp_val.doff = 0; + hk = H_PAIRDATA(hcp->pagep, hcp->bndx); + if (hk->type == H_OFFPAGE) + memcpy(&tmp_val.dlen, + (u_int8_t *)hk + SSZ(HOFFPAGE, tlen), + sizeof(u_int32_t)); + else + tmp_val.dlen = LEN_HDATA(hcp->pagep, + hashp->hdr->pagesize,hcp->bndx); + myval = &tmp_val; + } else /* Regular partial put */ + myval = nval; + + return (__ham_replpair(hashp, hcp, myval, 0)); +} + +/* + * Given a key and a cursor, sets the cursor to the page/ndx on which + * the key resides. If the key is found, the cursor H_OK flag is set + * and the pagep, bndx, pgno (dpagep, dndx, dpgno) fields are set. + * If the key is not found, the H_OK flag is not set. If the sought + * field is non-0, the pagep, bndx, pgno (dpagep, dndx, dpgno) fields + * are set indicating where an add might take place. If it is 0, + * non of the cursor pointer field are valid. + */ +static int +__ham_lookup(hashp, hcp, key, sought, mode) + HTAB *hashp; + HASH_CURSOR *hcp; + const DBT *key; + u_int32_t sought; + db_lockmode_t mode; +{ + HKEYDATA *hk; + db_pgno_t pgno; + u_int32_t tlen; + int match, ret, t_ret; + + /* + * Set up cursor so that we're looking for space to add an item + * as we cycle through the pages looking for the key. + */ + if ((ret = __ham_item_reset(hashp, hcp)) != 0) + return (ret); + hcp->seek_size = sought; + + hcp->bucket = __ham_call_hash(hashp, (u_int8_t *)key->data, key->size); + while (1) { + if ((ret = __ham_item_next(hashp, hcp, mode)) != 0) + return (ret); + + if (F_ISSET(hcp, H_NOMORE)) + break; + + hk = H_PAIRKEY(hcp->pagep, hcp->bndx); + switch (hk->type) { + case H_OFFPAGE: + memcpy(&tlen, (u_int8_t *)hk + SSZ(HOFFPAGE, tlen), + sizeof(u_int32_t)); + if (tlen == key->size) { + memcpy(&pgno, + (u_int8_t *)hk + SSZ(HOFFPAGE, pgno), + sizeof(db_pgno_t)); + match = __db_moff(hashp->dbp, key, pgno); + if (match == 0) { + F_SET(hcp, H_OK); + return (0); + } + } + break; + case H_KEYDATA: + if (key->size == LEN_HKEY(hcp->pagep, + hashp->hdr->pagesize, hcp->bndx) && + memcmp(key->data, hk->data, key->size) == 0) { + F_SET(hcp, H_OK); + return (0); + } + break; + case H_DUPLICATE: + case H_OFFDUP: + /* + * These are errors because keys are never + * duplicated, only data items are. + */ + return (__db_pgfmt(hashp->dbp, PGNO(hcp->pagep))); + } + hashp->hash_collisions++; + } + + /* + * Item was not found, adjust cursor properly. + */ + + if (sought != 0) + return (ret); + + if ((t_ret = __ham_item_done(hashp, hcp, 0)) != 0 && ret == 0) + ret = t_ret; + return (ret); +} + +/* + * Initialize a dbt using some possibly already allocated storage + * for items. + * PUBLIC: int __ham_init_dbt __P((DBT *, u_int32_t, void **, u_int32_t *)); + */ +int +__ham_init_dbt(dbt, size, bufp, sizep) + DBT *dbt; + u_int32_t size; + void **bufp; + u_int32_t *sizep; +{ + memset(dbt, 0, sizeof(*dbt)); + if (*sizep < size) { + if ((*bufp = (void *)(*bufp == NULL ? + malloc(size) : realloc(*bufp, size))) == NULL) { + *sizep = 0; + return (ENOMEM); + } + *sizep = size; + } + dbt->data = *bufp; + dbt->size = size; + return (0); +} + +/* + * Adjust the cursor after an insert or delete. The cursor passed is + * the one that was operated upon; we just need to check any of the + * others. + * + * len indicates the length of the item added/deleted + * add indicates if the item indicated by the cursor has just been + * added (add == 1) or deleted (add == 0). + * dup indicates if the addition occurred into a duplicate set. + * + * PUBLIC: void __ham_c_update __P((HTAB *, + * PUBLIC: HASH_CURSOR *, db_pgno_t, u_int32_t, int, int)); + */ +void +__ham_c_update(hashp, hcp, chg_pgno, len, add, dup) + HTAB *hashp; + HASH_CURSOR *hcp; + db_pgno_t chg_pgno; + u_int32_t len; + int add; + int dup; +{ + DBC *cp; + HTAB *hp; + HASH_CURSOR *lcp; + int page_deleted; + + /* + * Regular adds are always at the end of a given page, + * so we never have to adjust anyone's cursor after + * a regular add. + */ + if (!dup && add) + return; + + page_deleted = chg_pgno != PGNO_INVALID && + ((!dup && chg_pgno != hcp->pgno) || + (dup && chg_pgno != hcp->dpgno)); + + hp = hcp->db_cursor->dbp->master->internal; + DB_THREAD_LOCK(hp->dbp); + + for (cp = TAILQ_FIRST(&hp->dbp->curs_queue); cp != NULL; + cp = TAILQ_NEXT(cp, links)) { + if (cp->internal == hcp) + continue; + + lcp = (HASH_CURSOR *)cp->internal; + + if (!dup && lcp->pgno != chg_pgno) + continue; + + if (dup && F_ISSET(hcp, H_DELETED) && lcp->pgno != chg_pgno) + continue; + + if (dup && !F_ISSET(hcp, H_DELETED) && lcp->dpgno != chg_pgno) + continue; + + if (page_deleted) { + if (dup) { + lcp->dpgno = hcp->dpgno; + lcp->dndx = hcp->dndx; + } else { + lcp->pgno = hcp->pgno; + lcp->bndx = hcp->bndx; + lcp->bucket = hcp->bucket; + } + F_CLR(lcp, H_ISDUP); + continue; + } + + if (!dup && lcp->bndx > hcp->bndx) + lcp->bndx--; + else if (!dup && lcp->bndx == hcp->bndx) + F_SET(lcp, H_DELETED); + else if (dup && lcp->bndx == hcp->bndx) { + /* Assign dpgno in case there was page conversion. */ + lcp->dpgno = hcp->dpgno; + if (add && lcp->dndx >= hcp->dndx ) + lcp->dndx++; + else if (!add && lcp->dndx > hcp->dndx) + lcp->dndx--; + else if (!add && lcp->dndx == hcp->dndx) + F_SET(lcp, H_DELETED); + + /* Now adjust on-page information. */ + if (lcp->dpgno == PGNO_INVALID) + if (add) { + lcp->dup_tlen += len; + if (lcp->dndx > hcp->dndx) + lcp->dup_off += len; + } else { + lcp->dup_tlen -= len; + if (lcp->dndx > hcp->dndx) + lcp->dup_off -= len; + } + } + } + DB_THREAD_UNLOCK(hp->dbp); +} + +/* + * __ham_hdup -- + * This function gets called when we create a duplicate handle for a + * threaded DB. It should create the private part of the DB structure. + * PUBLIC: int __ham_hdup __P((DB *, DB *)); + */ +int +__ham_hdup(orig, new) + DB *orig, *new; +{ + HTAB *hashp; + DBC *curs; + int ret; + + if ((hashp = (HTAB *)malloc(sizeof(HTAB))) == NULL) + return (ENOMEM); + + new->internal = hashp; + + hashp->dbp = new; + hashp->hlock = 0; + hashp->hdr = NULL; + hashp->hash = ((HTAB *)orig->internal)->hash; + if ((hashp->split_buf = (PAGE *)malloc(orig->pgsize)) == NULL) + return (ENOMEM); + hashp->local_errno = 0; + hashp->hash_accesses = 0; + hashp->hash_collisions = 0; + hashp->hash_expansions = 0; + hashp->hash_overflows = 0; + hashp->hash_bigpages = 0; + /* Initialize the cursor queue. */ + ret = __ham_c_init(new, NULL, &curs); + TAILQ_INSERT_TAIL(&new->curs_queue, curs, links); + return (ret); +} |