diff options
Diffstat (limited to 'runtime/rsyslog.c')
-rw-r--r-- | runtime/rsyslog.c | 237 |
1 files changed, 237 insertions, 0 deletions
diff --git a/runtime/rsyslog.c b/runtime/rsyslog.c new file mode 100644 index 00000000..54db12c2 --- /dev/null +++ b/runtime/rsyslog.c @@ -0,0 +1,237 @@ +/* rsyslog.c - the main entry point into rsyslog's runtime library (RTL) + * + * This module contains all function which work on a RTL global level. It's + * name is abbreviated to "rsrt" (rsyslog runtime). + * + * Please note that the runtime library tends to be plugin-safe. That is, it must be + * initialized by calling a global initialization function. However, that + * function checks if the library is already initialized and, if so, does + * nothing except incrementing a refeence count. Similarly, the deinit + * function does nothing as long as there are still other users (which + * is tracked via the refcount). As such, it is safe to call init and + * exit multiple times, as long as this are always matching calls. This + * capability is needed for a plugin system, where one plugin never + * knows what the other did. HOWEVER, as of this writing, not all runtime + * library objects may work cleanly without static global data (the + * debug system is a very good example of this). So while we aim at the + * ability to work well in a plugin environment, things may not really work + * out. If you intend to use the rsyslog runtime library inside plugins, + * you should investigate the situation in detail. Please note that the + * rsyslog project itself does not yet need this functionality - thus you + * can safely assume it is totally untested ;). + * + * rgerhards, 2008-04-17: I have now once again checked on the plugin-safety. + * Unfortunately, there is currently no hook at all with which we could + * abstract a global data instance class. As such, we can NOT make the + * runtime plugin-safe in the above-described sense. As the rsyslog + * project itself does not need this functionality (and it is quesationable + * if someone else ever will), we do currently do not make an effort to + * support it. So if you intend to use rsyslog runtime inside a non-rsyslog + * plugin system, be careful! + * + * The rsyslog runtime library is in general reentrant and thread-safe. There + * are some intentional exceptions (e.g. inside the msg object). These are + * documented. Any other threading and reentrency issue can be considered a bug. + * + * Module begun 2008-04-16 by Rainer Gerhards + * + * Copyright 2008 Rainer Gerhards and Adiscon GmbH. + * + * This file is part of the rsyslog runtime library. + * + * The rsyslog runtime library is free software: you can redistribute it and/or modify + * it under the terms of the GNU Lesser General Public License as published by + * the Free Software Foundation, either version 3 of the License, or + * (at your option) any later version. + * + * The rsyslog runtime library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with the rsyslog runtime library. If not, see <http://www.gnu.org/licenses/>. + * + * A copy of the GPL can be found in the file "COPYING" in this distribution. + * A copy of the LGPL can be found in the file "COPYING.LESSER" in this distribution. + */ +#include "config.h" +#include <stdio.h> +#include <stdlib.h> +#include <assert.h> + +#include "rsyslog.h" +#include "obj.h" +#include "vm.h" +#include "sysvar.h" +#include "stringbuf.h" +#include "wti.h" +#include "wtp.h" +#include "expr.h" +#include "ctok.h" +#include "vmop.h" +#include "vmstk.h" +#include "vmprg.h" +#include "datetime.h" +#include "queue.h" +#include "conf.h" +#include "glbl.h" +#include "errmsg.h" + +/* forward definitions */ +static rsRetVal dfltErrLogger(int, uchar *errMsg); + +/* globally visible static data - see comment in rsyslog.h for details */ +uchar *glblModPath; /* module load path */ +rsRetVal (*glblErrLogger)(int, uchar*) = dfltErrLogger; /* the error logger to use by the errmsg module */ + +/* static data */ +static int iRefCount = 0; /* our refcount - it MUST exist only once inside a process (not thread) + thus it is perfectly OK to use a static. MUST be initialized to 0! */ + +/* This is the default instance of the error logger. It simply writes the message + * to stderr. It is expected that this is replaced by the runtime user very early + * during startup (at least if the default is unsuitable). However, we provide a + * default so that we can log errors during the intial phase, most importantly + * during initialization. -- rgerhards. 2008-04-17 + */ +static rsRetVal dfltErrLogger(int iErr, uchar *errMsg) +{ + DEFiRet; + fprintf(stderr, "rsyslog runtime error(%d): %s\n", iErr, errMsg); + RETiRet; +} + + +/* set the error log function + * rgerhards, 2008-04-18 + */ +rsRetVal +rsrtSetErrLogger(rsRetVal (*errLogger)(int, uchar*)) +{ + DEFiRet; + assert(errLogger != NULL); + glblErrLogger = errLogger; + RETiRet; +} + + +/* globally initialze the runtime system + * NOTE: this is NOT thread safe and must not be called concurrently. If that + * ever poses a problem, we may use proper mutex calls - not considered needed yet. + * If ppErrObj is provided, it receives a char pointer to the name of the object that + * caused the problem (if one occured). The caller must never free this pointer. If + * ppErrObj is NULL, no such information will be provided. pObjIF is the pointer to + * the "obj" object interface, which may be used to query any other rsyslog objects. + * rgerhards, 2008-04-16 + */ +rsRetVal +rsrtInit(char **ppErrObj, obj_if_t *pObjIF) +{ + DEFiRet; + + if(iRefCount == 0) { + /* init runtime only if not yet done */ + if(ppErrObj != NULL) *ppErrObj = "obj"; + CHKiRet(objClassInit(NULL)); /* *THIS* *MUST* always be the first class initilizer being called! */ + CHKiRet(objGetObjInterface(pObjIF)); /* this provides the root pointer for all other queries */ + + /* initialize core classes. We must be very careful with the order of events. Some + * classes use others and if we do not initialize them in the right order, we may end + * up with an invalid call. The most important thing that can happen is that an error + * is detected and needs to be logged, wich in turn requires a broader number of classes + * to be available. The solution is that we take care in the order of calls AND use a + * class immediately after it is initialized. And, of course, we load those classes + * first that we use ourselfs... -- rgerhards, 2008-03-07 + */ + if(ppErrObj != NULL) *ppErrObj = "glbl"; + CHKiRet(glblClassInit(NULL)); + if(ppErrObj != NULL) *ppErrObj = "datetime"; + CHKiRet(datetimeClassInit(NULL)); + if(ppErrObj != NULL) *ppErrObj = "msg"; + CHKiRet(msgClassInit(NULL)); + if(ppErrObj != NULL) *ppErrObj = "str,"; + CHKiRet(strmClassInit(NULL)); + if(ppErrObj != NULL) *ppErrObj = "wti"; + CHKiRet(wtiClassInit(NULL)); + if(ppErrObj != NULL) *ppErrObj = "wtp"; + CHKiRet(wtpClassInit(NULL)); + if(ppErrObj != NULL) *ppErrObj = "queue"; + CHKiRet(queueClassInit(NULL)); + if(ppErrObj != NULL) *ppErrObj = "vmstk"; + CHKiRet(vmstkClassInit(NULL)); + if(ppErrObj != NULL) *ppErrObj = "sysvar"; + CHKiRet(sysvarClassInit(NULL)); + if(ppErrObj != NULL) *ppErrObj = "vm"; + CHKiRet(vmClassInit(NULL)); + if(ppErrObj != NULL) *ppErrObj = "vmop"; + CHKiRet(vmopClassInit(NULL)); + if(ppErrObj != NULL) *ppErrObj = "vmprg"; + CHKiRet(vmprgClassInit(NULL)); + if(ppErrObj != NULL) *ppErrObj = "ctok_token"; + CHKiRet(ctok_tokenClassInit(NULL)); + if(ppErrObj != NULL) *ppErrObj = "ctok"; + CHKiRet(ctokClassInit(NULL)); + if(ppErrObj != NULL) *ppErrObj = "expr"; + CHKiRet(exprClassInit(NULL)); + if(ppErrObj != NULL) *ppErrObj = "conf"; + CHKiRet(confClassInit(NULL)); + + /* dummy "classes" */ + if(ppErrObj != NULL) *ppErrObj = "str"; + CHKiRet(strInit()); + } + + ++iRefCount; + dbgprintf("rsyslog runtime initialized, version %s, current users %d\n", VERSION, iRefCount); + +finalize_it: + RETiRet; +} + + +/* globally de-initialze the runtime system + * NOTE: this is NOT thread safe and must not be called concurrently. If that + * ever poses a problem, we may use proper mutex calls - not considered needed yet. + * This function must be provided with the caller's obj object pointer. This is + * automatically deinitialized by the runtime system. + * rgerhards, 2008-04-16 + */ +rsRetVal +rsrtExit(void) +{ + DEFiRet; + + if(iRefCount == 1) { + /* do actual de-init only if we are the last runtime user */ + confClassExit(); + glblClassExit(); + objClassExit(); /* *THIS* *MUST/SHOULD?* always be the first class initilizer being called (except debug)! */ + } + + --iRefCount; + /* TODO we must deinit this pointer! pObjIF = NULL; / * no longer exists for this caller */ + + dbgprintf("rsyslog runtime de-initialized, current users %d\n", iRefCount); + + RETiRet; +} + + +/* returns 0 if the rsyslog runtime is not initialized and another value + * if it is. This function is primarily meant to be used by runtime functions + * itself. However, it is safe to call it before initializing the runtime. + * Plugins should NOT rely on this function. The reason is that another caller + * may have already initialized it but deinits it before this plugin is done. + * So for plugins and like architectures, the right course of action is to + * call rsrtInit() and rsrtExit(), which can be called by multiple callers. + * rgerhards, 2008-04-16 + */ +int rsrtIsInit(void) +{ + return iRefCount; +} + + +/* vim:set ai: + */ |