/*- * See the file LICENSE for redistribution information. * * Copyright (c) 1996, 1997, 1998, 1999 * Sleepycat Software. All rights reserved. */ #include "db_config.h" #ifndef lint static const char sccsid[] = "@(#)mp_bh.c 11.5 (Sleepycat) 9/21/99"; #endif /* not lint */ #ifndef NO_SYSTEM_INCLUDES #include #include #include #ifndef _MSC_VER /* _WIN32 */ #include #endif #endif #include "db_int.h" #include "db_shash.h" #include "mp.h" #ifdef DEBUG #include "WordMonitor.h" #include "db_page.h" #endif /* DEBUG */ static int CDB___memp_upgrade __P((DB_MPOOL *, DB_MPOOLFILE *, MPOOLFILE *)); /* * CDB___memp_bhwrite -- * Write the page associated with a given bucket header. * * PUBLIC: int CDB___memp_bhwrite * PUBLIC: __P((DB_MPOOL *, MPOOLFILE *, BH *, int *, int *)); */ int CDB___memp_bhwrite(dbmp, mfp, bhp, restartp, wrotep) DB_MPOOL *dbmp; MPOOLFILE *mfp; BH *bhp; int *restartp, *wrotep; { DB_MPOOLFILE *dbmfp; DB_MPREG *mpreg; int incremented, ret; if (restartp != NULL) *restartp = 0; if (wrotep != NULL) *wrotep = 0; incremented = 0; /* * Walk the process' DB_MPOOLFILE list and find a file descriptor for * the file. We also check that the descriptor is open for writing. * If we find a descriptor on the file that's not open for writing, we * try and upgrade it to make it writeable. If that fails, we're done. */ MUTEX_THREAD_LOCK(dbmp->mutexp); for (dbmfp = TAILQ_FIRST(&dbmp->dbmfq); dbmfp != NULL; dbmfp = TAILQ_NEXT(dbmfp, q)) if (dbmfp->mfp == mfp) { if (F_ISSET(dbmfp, MP_READONLY) && CDB___memp_upgrade(dbmp, dbmfp, mfp)) { MUTEX_THREAD_UNLOCK(dbmp->mutexp); return (0); } /* * Increment the reference count -- see the comment in * CDB_memp_fclose(). */ ++dbmfp->ref; incremented = 1; break; } MUTEX_THREAD_UNLOCK(dbmp->mutexp); if (dbmfp != NULL) goto found; /* * !!! * Don't try to attach to temporary files. There are two problems in * trying to do that. First, if we have different privileges than the * process that "owns" the temporary file, we might create the backing * disk file such that the owning process couldn't read/write its own * buffers, e.g., CDB_memp_trickle() running as root creating a file owned * as root, mode 600. Second, if the temporary file has already been * created, we don't have any way of finding out what its real name is, * and, even if we did, it was already unlinked (so that it won't be * left if the process dies horribly). This decision causes a problem, * however: if the temporary file consumes the entire buffer cache, * and the owner doesn't flush the buffers to disk, we could end up * with resource starvation, and the CDB_memp_trickle() thread couldn't do * anything about it. That's a pretty unlikely scenario, though. */ if (F_ISSET(mfp, MP_TEMP)) return (0); /* * It's possible the file has been removed. Drop into the page-write * function, which has to handle the fact that we don't have any real * file descriptor information. */ if (F_ISSET(mfp, MP_REMOVED)) { dbmfp = NULL; goto found; } /* * It's not a page from a file we've opened. If the file requires * input/output processing, see if this process has ever registered * information as to how to write this type of file. If not, there's * nothing we can do. */ if (mfp->ftype != 0) { MUTEX_THREAD_LOCK(dbmp->mutexp); for (mpreg = LIST_FIRST(&dbmp->dbregq); mpreg != NULL; mpreg = LIST_NEXT(mpreg, q)) if (mpreg->ftype == mfp->ftype) break; MUTEX_THREAD_UNLOCK(dbmp->mutexp); if (mpreg == NULL) return (0); } /* * Try and open the file, attaching to the underlying shared area. * Ignore any error, assume it's a permissions problem. * * XXX * There's no negative cache, so we may repeatedly try and open files * that we have previously tried (and failed) to open. */ if (CDB___memp_fopen(dbmp, mfp, R_ADDR(&dbmp->reginfo, mfp->path_off), 0, 0, mfp->stat.st_pagesize, 0, NULL, &dbmfp) != 0) return (0); found: ret = CDB___memp_pgwrite(dbmp, dbmfp, bhp, restartp, wrotep); if (incremented) { MUTEX_THREAD_LOCK(dbmp->mutexp); --dbmfp->ref; MUTEX_THREAD_UNLOCK(dbmp->mutexp); } return (ret); } /* * CDB___memp_pgread -- * Read a page from a file. * * PUBLIC: int CDB___memp_pgread __P((DB_MPOOLFILE *, BH *, int)); */ int CDB___memp_pgread(dbmfp, bhp, can_create) DB_MPOOLFILE *dbmfp; BH *bhp; int can_create; { DB_IO db_io; DB_ENV *dbenv; DB_MPOOL *dbmp; MPOOLFILE *mfp; size_t len, pagesize; ssize_t nr; int created, ret; dbmp = dbmfp->dbmp; dbenv = dbmp->dbenv; mfp = dbmfp->mfp; pagesize = mfp->stat.st_pagesize; F_SET(bhp, BH_LOCKED | BH_TRASH); MUTEX_LOCK(&bhp->mutex, dbenv->lockfhp); R_UNLOCK(dbenv, &dbmp->reginfo); /* * Temporary files may not yet have been created. We don't create * them now, we create them when the pages have to be flushed. */ nr = 0; if (F_ISSET(&dbmfp->fh, DB_FH_VALID)) { /* * Ignore read errors if we have permission to create the page. * Assume that the page doesn't exist, and that we'll create it * when we write it out. * * XXX * Theoretically, we could overwrite a page of data if it were * possible for a file to be successfully opened for reading * and then for the read to fail. Shouldn't ever happen, but * it might be worth checking to see if the offset is past the * known end-of-file. */ db_io.fhp = &dbmfp->fh; db_io.mutexp = dbmfp->mutexp; db_io.pagesize = db_io.bytes = pagesize; db_io.pgno = bhp->pgno; db_io.buf = bhp->buf; #ifdef DEBUG /* * Prevent signal to occur during IO */ word_monitor_click(); #endif /* DEBUG */ if(F_ISSET(dbmfp, MP_CMPR)) { ret = CDB___memp_cmpr(dbmfp, bhp, &db_io, DB_IO_READ, &nr); } else { ret = CDB___os_io(&db_io, DB_IO_READ, &nr); } #ifdef DEBUG if(ret == 0) { PAGE* pp = (PAGE*)db_io.buf; word_monitor_add(WORD_MONITOR_READ, 1); switch(TYPE(pp)) { case P_IBTREE: word_monitor_add(WORD_MONITOR_PAGE_IBTREE, 1); break; case P_LBTREE: word_monitor_add(WORD_MONITOR_PAGE_LBTREE, 1); break; default: word_monitor_add(WORD_MONITOR_PAGE_UNKNOWN, 1); break; } } #endif /* DEBUG */ } else ret = 0; created = 0; if (nr < (ssize_t)pagesize) { if (can_create) created = 1; else { /* * If we had a short read, ret may be 0. This may not * be an error -- in particular DB recovery processing * may request pages that have never been written to * disk, in which case we won't find the page. So, the * caller must know how to handle the error. */ if (ret == 0) ret = EIO; goto err; } } /* * Clear any bytes we didn't read that need to be cleared. If we're * running in diagnostic mode, smash any bytes on the page that are * unknown quantities for the caller. */ if (nr != (ssize_t)pagesize) { len = mfp->clear_len == 0 ? pagesize : mfp->clear_len; if (nr < (ssize_t)len) memset(bhp->buf + nr, 0, len - nr); #ifdef DIAGNOSTIC if (nr > (ssize_t)len) len = nr; if (len < pagesize) memset(bhp->buf + len, CLEAR_BYTE, pagesize - len); #endif } /* Call any pgin function. */ ret = mfp->ftype == 0 ? 0 : CDB___memp_pg(dbmfp, bhp, 1); /* Unlock the buffer and reacquire the region lock. */ err: MUTEX_UNLOCK(&bhp->mutex); R_LOCK(dbenv, &dbmp->reginfo); /* * If no errors occurred, the data is now valid, clear the BH_TRASH * flag; regardless, clear the lock bit and let other threads proceed. */ F_CLR(bhp, BH_LOCKED); if (ret == 0) { F_CLR(bhp, BH_TRASH); /* Update the statistics. */ if (created) ++mfp->stat.st_page_create; else ++mfp->stat.st_page_in; } return (ret); } /* * CDB___memp_pgwrite -- * Write a page to a file. * * PUBLIC: int CDB___memp_pgwrite * PUBLIC: __P((DB_MPOOL *, DB_MPOOLFILE *, BH *, int *, int *)); */ int CDB___memp_pgwrite(dbmp, dbmfp, bhp, restartp, wrotep) DB_MPOOL *dbmp; DB_MPOOLFILE *dbmfp; BH *bhp; int *restartp, *wrotep; { DB_ENV *dbenv; DB_IO db_io; DB_LSN lsn; MCACHE *mc; MPOOL *mp; MPOOLFILE *mfp; ssize_t nw; int callpgin, dosync, ret, syncfail; const char *fail; dbenv = dbmp->dbenv; mp = dbmp->reginfo.primary; mfp = dbmfp == NULL ? NULL : dbmfp->mfp; if (restartp != NULL) *restartp = 0; if (wrotep != NULL) *wrotep = 0; callpgin = 0; /* * Check the dirty bit -- this buffer may have been written since we * decided to write it. */ if (!F_ISSET(bhp, BH_DIRTY)) { if (wrotep != NULL) *wrotep = 1; return (0); } MUTEX_LOCK(&bhp->mutex, dbenv->lockfhp); /* * If there were two writers, we may have just been waiting while the * other writer completed I/O on this buffer. Check the dirty bit one * more time. */ if (!F_ISSET(bhp, BH_DIRTY)) { MUTEX_UNLOCK(&bhp->mutex); if (wrotep != NULL) *wrotep = 1; return (0); } F_SET(bhp, BH_LOCKED); R_UNLOCK(dbenv, &dbmp->reginfo); if (restartp != NULL) *restartp = 1; /* * It's possible that the underlying file was removed. Currently, * this is only used by the filesystem operation recovery routines, * but it's possible that it may be more widely used in the future. * * !!! * Once we pass this point, we know that dbmfp and mfp aren't NULL, * and that we have a valid file reference. */ if (mfp == NULL || F_ISSET(mfp, MP_REMOVED)) goto file_removed; /* Copy the LSN off the page if we're going to need it. */ if (F_ISSET(dbenv, DB_ENV_LOGGING) || F_ISSET(bhp, BH_WRITE)) memcpy(&lsn, bhp->buf + mfp->lsn_off, sizeof(DB_LSN)); /* Ensure the appropriate log records are on disk. */ if (F_ISSET(dbenv, DB_ENV_LOGGING) && (ret = CDB_log_flush(dbenv, &lsn)) != 0) goto err; /* * Call any pgout function. We set the callpgin flag so that we flag * that the contents of the buffer will need to be passed through pgin * before they are reused. */ if (mfp->ftype == 0) ret = 0; else { callpgin = 1; if ((ret = CDB___memp_pg(dbmfp, bhp, 0)) != 0) goto err; } /* Temporary files may not yet have been created. */ if (!F_ISSET(&dbmfp->fh, DB_FH_VALID)) { MUTEX_THREAD_LOCK(dbmp->mutexp); if (!F_ISSET(&dbmfp->fh, DB_FH_VALID) && ((ret = CDB___db_appname(dbenv, DB_APP_TMP, NULL, NULL, DB_OSO_CREATE | DB_OSO_EXCL | DB_OSO_TEMP, &dbmfp->fh, NULL)) != 0 || !F_ISSET(&dbmfp->fh, DB_FH_VALID))) { MUTEX_THREAD_UNLOCK(dbmp->mutexp); CDB___db_err(dbenv, "unable to create temporary backing file"); goto err; } MUTEX_THREAD_UNLOCK(dbmp->mutexp); } /* Write the page. */ db_io.fhp = &dbmfp->fh; db_io.mutexp = dbmfp->mutexp; db_io.pagesize = db_io.bytes = mfp->stat.st_pagesize; db_io.pgno = bhp->pgno; db_io.buf = bhp->buf; #ifdef DEBUG { PAGE* pp = (PAGE*)db_io.buf; word_monitor_add(WORD_MONITOR_WRITE, 1); switch(TYPE(pp)) { case P_IBTREE: word_monitor_add(WORD_MONITOR_PAGE_IBTREE, 1); break; case P_LBTREE: word_monitor_add(WORD_MONITOR_PAGE_LBTREE, 1); break; default: word_monitor_add(WORD_MONITOR_PAGE_UNKNOWN, 1); break; } } /* * Prevent signal to occur during IO */ word_monitor_click(); #endif /* DEBUG */ if(F_ISSET(dbmfp, MP_CMPR)) { ret = CDB___memp_cmpr(dbmfp, bhp, &db_io, DB_IO_WRITE, &nw); } else { ret = CDB___os_io(&db_io, DB_IO_WRITE, &nw); } if (ret != 0) { CDB___db_panic(dbenv, ret); fail = "write"; goto syserr; } if (nw != (ssize_t)mfp->stat.st_pagesize) { ret = EIO; fail = "write"; goto syserr; } file_removed: /* * !!! * Once we pass this point, dbmfp and mfp may be NULL, we may not have * a valid file reference. * * Unlock the buffer and reacquire the region lock. */ MUTEX_UNLOCK(&bhp->mutex); R_LOCK(dbenv, &dbmp->reginfo); /* * Clean up the flags based on a successful write. * * If we rewrote the page, it will need processing by the pgin * routine before reuse. */ if (callpgin) F_SET(bhp, BH_CALLPGIN); F_CLR(bhp, BH_DIRTY | BH_LOCKED); /* * If we write a buffer for which a checkpoint is waiting, update * the count of pending buffers (both in the mpool as a whole and * for this file). If the count for this file goes to zero, set a * flag so we flush the writes. */ dosync = 0; if (F_ISSET(bhp, BH_WRITE)) { F_CLR(bhp, BH_WRITE); --mp->lsn_cnt; if (mfp != NULL) dosync = --mfp->lsn_cnt == 0 ? 1 : 0; } /* Update the page clean/dirty statistics. */ mc = BH_TO_CACHE(dbmp, bhp); ++mc->stat.st_page_clean; --mc->stat.st_page_dirty; /* Update I/O statistics. */ if (mfp != NULL) ++mfp->stat.st_page_out; /* * Do the sync after everything else has been updated, so any incoming * checkpoint doesn't see inconsistent information. * * XXX: * Don't lock the region around the sync, fsync(2) has no atomicity * issues. * * XXX: * We ignore errors from the sync -- it makes no sense to return an * error to the calling process, so set a flag causing the checkpoint * to be retried later. There is a possibility, of course, that a * subsequent checkpoint was started and that we're going to force it * to fail. That should be unlikely, and fixing it would be difficult. */ if (dosync) { R_UNLOCK(dbenv, &dbmp->reginfo); syncfail = CDB___os_fsync(&dbmfp->fh) != 0; R_LOCK(dbenv, &dbmp->reginfo); if (syncfail) F_SET(mp, MP_LSN_RETRY); } if (wrotep != NULL) *wrotep = 1; return (0); syserr: CDB___db_err(dbenv, "%s: %s failed for page %lu", CDB___memp_fn(dbmfp), fail, (u_long)bhp->pgno); err: /* Unlock the buffer and reacquire the region lock. */ MUTEX_UNLOCK(&bhp->mutex); R_LOCK(dbenv, &dbmp->reginfo); /* * Clean up the flags based on a failure. * * The page remains dirty but we remove our lock. If we rewrote the * page, it will need processing by the pgin routine before reuse. */ if (callpgin) F_SET(bhp, BH_CALLPGIN); F_CLR(bhp, BH_LOCKED); return (ret); } /* * CDB___memp_pg -- * Call the pgin/pgout routine. * * PUBLIC: int CDB___memp_pg __P((DB_MPOOLFILE *, BH *, int)); */ int CDB___memp_pg(dbmfp, bhp, is_pgin) DB_MPOOLFILE *dbmfp; BH *bhp; int is_pgin; { DBT dbt, *dbtp; DB_MPOOL *dbmp; DB_MPREG *mpreg; MPOOLFILE *mfp; int ftype, ret; dbmp = dbmfp->dbmp; mfp = dbmfp->mfp; MUTEX_THREAD_LOCK(dbmp->mutexp); ftype = mfp->ftype; for (mpreg = LIST_FIRST(&dbmp->dbregq); mpreg != NULL; mpreg = LIST_NEXT(mpreg, q)) { if (ftype != mpreg->ftype) continue; if (mfp->pgcookie_len == 0) dbtp = NULL; else { dbt.size = mfp->pgcookie_len; dbt.data = R_ADDR(&dbmp->reginfo, mfp->pgcookie_off); dbtp = &dbt; } MUTEX_THREAD_UNLOCK(dbmp->mutexp); if (is_pgin) { if (mpreg->pgin != NULL && (ret = mpreg->pgin(bhp->pgno, bhp->buf, dbtp)) != 0) goto err; } else if (mpreg->pgout != NULL && (ret = mpreg->pgout(bhp->pgno, bhp->buf, dbtp)) != 0) goto err; break; } if (mpreg == NULL) MUTEX_THREAD_UNLOCK(dbmp->mutexp); return (0); err: MUTEX_THREAD_UNLOCK(dbmp->mutexp); CDB___db_err(dbmp->dbenv, "%s: %s failed for page %lu", CDB___memp_fn(dbmfp), is_pgin ? "pgin" : "pgout", (u_long)bhp->pgno); return (ret); } /* * CDB___memp_bhfree -- * Free a bucket header and its referenced data. * * PUBLIC: void CDB___memp_bhfree __P((DB_MPOOL *, BH *, int)); */ void CDB___memp_bhfree(dbmp, bhp, free_mem) DB_MPOOL *dbmp; BH *bhp; int free_mem; { DB_HASHTAB *dbht; MCACHE *mc; MPOOL *mp; int n_bucket, n_cache; mp = dbmp->reginfo.primary; mc = BH_TO_CACHE(dbmp, bhp); n_cache = NCACHE(mp, bhp->pgno); n_bucket = NBUCKET(mc, bhp->mf_offset, bhp->pgno); dbht = R_ADDR(&dbmp->c_reginfo[n_cache], mc->htab); /* Delete the buffer header from the hash bucket queue. */ SH_TAILQ_REMOVE(&dbht[n_bucket], bhp, hq, __bh); /* Delete the buffer header from the LRU queue. */ SH_TAILQ_REMOVE(&mc->bhq, bhp, q, __bh); DB_ASSERT(mc->stat.st_page_clean != 0); --mc->stat.st_page_clean; /* * If we're not reusing it immediately, free the buffer header * and data for real. */ CDB___memp_cmpr_free_chain(dbmp, bhp); if (free_mem) { CDB___db_shalloc_free(dbmp->c_reginfo[n_cache].addr, bhp); } } /* * CDB___memp_upgrade -- * Upgrade a file descriptor from readonly to readwrite. */ static int CDB___memp_upgrade(dbmp, dbmfp, mfp) DB_MPOOL *dbmp; DB_MPOOLFILE *dbmfp; MPOOLFILE *mfp; { DB_FH fh; int ret; char *rpath; /* * !!! * We expect the handle to already be locked. */ /* Check to see if we've already upgraded. */ if (F_ISSET(dbmfp, MP_UPGRADE)) return (0); /* Check to see if we've already failed. */ if (F_ISSET(dbmfp, MP_UPGRADE_FAIL)) return (1); /* * Calculate the real name for this file and try to open it read/write. * We know we have a valid pathname for the file because it's the only * way we could have gotten a file descriptor of any kind. */ if ((ret = CDB___db_appname(dbmp->dbenv, DB_APP_DATA, NULL, R_ADDR(&dbmp->reginfo, mfp->path_off), 0, NULL, &rpath)) != 0) return (ret); if (CDB___os_open(rpath, 0, 0, &fh) != 0) { F_SET(dbmfp, MP_UPGRADE_FAIL); ret = 1; } else { /* Swap the descriptors and set the upgrade flag. */ (void)CDB___os_closehandle(&dbmfp->fh); dbmfp->fh = fh; F_SET(dbmfp, MP_UPGRADE); ret = 0; } CDB___os_freestr(rpath); return (ret); }