rsyslog/plugins/imtemplate/imtemplate.c
Rainer Gerhards 02bb3c901d consolidated time calls during msg object creation
...this improves performance and consistency and also fixes
a bug where subsecond time properties generated by imfile, imklog and
internal messages could be slightly inconsistent.
2008-10-02 15:52:24 +02:00

459 lines
19 KiB
C

/* imtemplate.c
*
* This is NOT a real input module but a (copy)-template to create one. Please
* do NOT edit this file directly. Rather, copy it, together with the rest of
* the directory, to a new location ./plugins/im<yourname>, then replace
* all references to imtemplate in Makefile.am to im<yourname>. Be sure to
* fix the copyright notices to gain proper credit ;) Any derived version,
* however, needs to be placed under GPLv3 (see GPLv3 for details). If you
* do not like that policy, do not use this template or any of the header
* files. The rsyslog project greatly appreciates module contributions, so
* please consider contributing your work - even if you may think it only
* server a single very special purpose. It has turned out that at least some
* folks have similiar special purposes ;)
*
* IMPORTANT
* The comments in this file are actually the interface specification. I decided
* not to put it into a separate file as it is much simpler to keep it up to
* date when it is part of the actual template module.
*
* NAMING
* All input modules shall be named im<something>. While this is not a hard
* requirement, it helps keeping track of things.
*
* Global variables and functions should have a prefix - use as somewhat
* longer one to prevent conflicts with rsyslog itself and other modules
* (OK, hopefully I'll have some more precise advise in the future...).
*
* INCLUDE MODULE IN THE MAIN MAKE SCRIPT
* If the module shall be provided as part of rsyslog (or simply as a build aid,
* you need to add it to the main autoconf files). To do so, you need to edit
* Makefile.am and configure.ac in the main directory. Search for imtemplate
* and copy/modify the relevant code for your plugin.
*
* DEBUGGING
* While you develop your code, you may want to add
* --enable-debug --enable-rtinst
* to your ./configure settings. These enable extra run-time checks, which cost
* a lot of performance but can help detect some of the most frequently made
* bugs. These settings will also provide you with a nice stack dump if something
* goes really wrong.
*
* MORE SAMPLES
* Remember that rsyslog ships with a number of input modules (./plugins/im*). It
* is always a good idea to have a look at them before starting your own. imudp
* may be a good, relatively trivial, sample.
*
* --------------------------------------------------------------------------------
*
* This template was cretead on 2008-02-01 by Rainer Gerhards.
*
* Copyright 2008 Rainer Gerhards and Adiscon GmbH.
*
* This file is part of rsyslog.
*
* Rsyslog is free software: you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation, either version 3 of the License, or
* (at your option) any later version.
*
* Rsyslog 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 General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with Rsyslog. If not, see <http://www.gnu.org/licenses/>.
*
* A copy of the GPL can be found in the file "COPYING" in this distribution.
*/
#include "config.h" /* this is for autotools and always must be the first include */
#include <stdlib.h>
#include <assert.h>
#include <string.h>
#include <errno.h>
#include <pthread.h> /* do NOT remove: will soon be done by the module generation macros */
#include "rsyslog.h" /* error codes etc... */
#include "cfsysline.h" /* access to config file objects */
#include "module-template.h" /* generic module interface code - very important, read it! */
#include "srUtils.h" /* some utility functions */
MODULE_TYPE_INPUT /* must be present for input modules, do not remove */
/* defines */
/* Module static data */
DEF_IMOD_STATIC_DATA /* must be present, starts static data */
/* Here, define whatever static data is needed. Is it suggested that static variables only are
* used (not externally visible). If you need externally visible variables, make sure you use a
* prefix in order not to conflict with other modules or rsyslogd itself (also see comment
* at file header).
*/
/* static int imtemplateWhateverVar = 0; */
/* config settings */
/* You may add any functions that you feel are useful for your needs. No specific restrictions
* apply, but we suggest that you use the "iRet" call order, which enables you to use debug
* support for your own functions and which also makes it easy to communicate exceptions back
* to the upstream caller (rsyslog framework, for example.
*
* The function below is a sample of how one of your functions may look like. Again, the sample
* below is *not* needed to be present in order to meet the interface requirements.
*
* Be sure to use static functions (suggested) or prefixes to prevent name conflicts -- see file
* header for more information.
*/
static rsRetVal /* rsRetVal is our generic error-reporting return type */
imtemplateMyFunc(int iMyParam)
{
DEFiRet; /* define iRet, the return code and other plumbing */
/* define your local variables here */
/* code whatever you need to code here. The "iRet" system can be helpful:
*
* CHKiRet(function(param1, param2, ...));
* calls a function and checks if it returns RS_RET_OK. If so, work
* proceeds. If some other code is returned, the function is aborted
* and control transferred to finalize_it (which you need to define)
*
* CHKiRet_Hdlr(function(param1, param2, ...))
* much like CHKiRet, but allows you to specify your own code that is
* executed if the function does not return RS_RET_OK, e.g.:
* CHKiRet_Hdlr(function(a, b)) {
* ... some error handling here ...
* }
* control is not transferred to finalize_it, except if you use one
* of the relevant macros (described below)
*
* FINALIZE
* immediately transfers control to finalize_it, using the current
* value of iRet, e.g.
* if(bDone)
* FINALIZE;
*
* ABORT_FINALIZE(retcode)
* just like FINALIZE, except that iRet is set to the provided error
* code before control is transferred, e.g.
* if((ptr = malloc(20)) == NULL)
* ABORT_FINALIZE(RS_RET_OUT_OF_MEMORY);
*
* In order for all this to work, you need to define finalize_it, e.g.
*
* finalize_it:
* RETiRet;
*
* RETiRet does some housekeeping and then does a "return iRet" to transfer
* control back to the caller. There shall only be one function exit and
* it shall be via RETiRet, preferrably at the end of the function code.
*
*/
finalize_it:
/* clean up anything that needs to be cleaned up if processing did not
* go well, for example:
*/
if(iRet != RS_RET_OK) {
/* cleanup, e.g.
* free(somePtr);
*/
}
RETiRet;
}
/* This function is the cancel cleanup handler. It is called when rsyslog decides the
* module must be stopped, what most probably happens during shutdown of rsyslogd. When
* this function is called, the runInput() function (below) is already terminated - somewhere
* in the middle of what it was doing. The cancel cleanup handler below should take
* care of any locked mutexes and such, things that really need to be cleaned up
* before processing continues. In general, many plugins do not need to provide
* any code at all here.
*
* IMPORTANT: the calling interface of this function can NOT be modified. It actually is
* called by pthreads. The provided argument is currently not being used.
*/
/* ------------------------------------------------------------------------------------------ *
* DO NOT TOUCH the following code - it will soon be part of the module generation macros! */
static void
inputModuleCleanup(void *arg)
{
BEGINfunc
/* END no-touch zone *
* ------------------------------------------------------------------------------------------ */
/* your code here */
/* ------------------------------------------------------------------------------------------ *
* DO NOT TOUCH the following code - it will soon be part of the module generation macros! */
ENDfunc
}
/* END no-touch zone *
* ------------------------------------------------------------------------------------------ */
/* This function is called by the framework to gather the input. The module stays
* most of its lifetime inside this function. It MUST NEVER exit this function. Doing
* so would end module processing and rsyslog would NOT reschedule the module. If
* you exit from this function, you violate the interface specification!
*
* So how is it terminated? When it is time to terminate, rsyslog actually cancels
* the threads. This may sound scary, but is not. There is a cancel cleanup handler
* defined (the function directly above). See comments there for specifics.
*
* runInput is always called on a single thread. If the module neees multiple threads,
* it is free to create them. HOWEVER, it must make sure that any threads created
* are killed and joined in the cancel cleanup handler.
*/
BEGINrunInput
/* define any local variables you need here */
CODESTARTrunInput
/* ------------------------------------------------------------------------------------------ *
* DO NOT TOUCH the following code - it will soon be part of the module generation macros! */
pthread_cleanup_push(inputModuleCleanup, NULL);
while(1) { /* endless loop - do NOT break; out of it! */
/* END no-touch zone *
* ------------------------------------------------------------------------------------------ */
/* your code here */
/* All rsyslog objects (see other modules, e.g. msg.c) are available
* to your here. Some useful things are:
*
* errmsg.LogError(NO_ERRCODE, format-string, ... params ...);
* logs an error message as syslogd, just as printf, e.g.
* errmsg.LogError(NO_ERRCODE, "Error %d occured during %s", 1, "test");
*
* There are several ways how a message can be enqueued. This part of the
* interface is currently underspecified. Have a look at the function definitions
* in syslogd.c (sorry, folks...).
*
* If you received a full syslog message that must be decoded by a message
* parser, parseAndSubmitMessage() is the way to go. It's not just a funny name
* but also a quite some legacy. Consequently, its interface is, ummm, not
* well designed.
* parseAndSubmitMessage((char*)fromHost, (char*) pRcvBuf, lenRcvd, bParseHost);
* fromHost
* is the host that we received the message from (a string)
* pRcvBuf
* is the received (to-be-decoded) message
* lenRcvd
* is the length of the received message. Please note that pRcvBuf is
* NOT a standard C-string. Most importantly it is NOT expected to be
* \0-terminated. Thus the lenght is vitally imporant (if it is wrong,
* rsyslogd will probably segfault).
* bParseHost
* is a boolean (0-no, 1-yes). It tells the parser whether or not
* a hostname should be parsed from the message. This is important
* for sources that are known not to provide a hostname.
* Use define MSG_PARSE_HOSTNAME and MSG_DONT_PARSE_HOSTNAME
*
* Another, more elaborate, way is to create the message object ourselves and
* pass it to the rule engine. That way is more appropriate if the message
* does not need to be parsed, for example when reading text (log) files. In that way,
* we can set the message properties as of our liking. This is how it works:
*
msg_t *pMsg;
CHKiRet(msgConstruct(&pMsg));
MsgSetUxTradMsg(pMsg, msg);
MsgSetRawMsg(pMsg, msg);
MsgSetHOSTNAME(pMsg, LocalHostName);
MsgSetTAG(pMsg, "rsyslogd:");
pMsg->iFacility = LOG_FAC(pri);
pMsg->iSeverity = LOG_PRI(pri);
pMsg->bParseHOSTNAME = 0;
flags |= INTERNAL_MSG;
logmsg(pMsg, flags); / * some time, CHKiRet() will work here, too [today NOT!] * /
*
* Note that UxTradMsg is a wild construct. For the time being, set it to
* the raw message text. I am hard thinking at dropping that beast at all...
*
* This example probably does not set all message properties (but the ones
* that are of practical importance). If you need all, check msg.h. Use
* method access functions whereever possible, unfortunately not all structure
* members are currently exposed in that clean way - so you sometimes need
* to access them directly (it goes without saying that we will fix that
* over time ;)).
*/
/* ------------------------------------------------------------------------------------------ *
* DO NOT TOUCH the following code - it will soon be part of the module generation macros! */
}
/*NOTREACHED*/
pthread_cleanup_pop(0); /* just for completeness, but never called... */
RETiRet; /* use it to make sure the housekeeping is done! */
ENDrunInput
/* END no-touch zone *
* ------------------------------------------------------------------------------------------ */
/* The function is called by rsyslog before runInput() is called. It is a last chance
* to set up anything specific. Most importantly, it can be used to tell rsyslog if the
* input shall run or not. The idea is that if some config settings (or similiar things)
* are not OK, the input can tell rsyslog it will not execute. To do so, return
* RS_RET_NO_RUN or a specific error code. If RS_RET_OK is returned, rsyslog will
* proceed and call the runInput() entry point. If you do not return anything
* specific, RS_RET_OK is automatically returned (as in all functions).
*/
BEGINwillRun
/* place any variables needed here */
CODESTARTwillRun
/* ... your code here ... */
/* Just to give you an idea, here are some samples (from the actual imudp module:
*
if(udpLstnSocks == NULL)
ABORT_FINALIZE(RS_RET_NO_RUN);
if((pRcvBuf = malloc(glbl.GetMaxLine * sizeof(char))) == NULL) {
ABORT_FINALIZE(RS_RET_OUT_OF_MEMORY);
}
*
*/
finalize_it:
ENDwillRun
/* This function is called by the framework after runInput() has been terminated. It
* shall free any resources and prepare the module for unload.
*
* So it is important that runInput() keeps track of what needs to be cleaned up.
* Objects to think about are files (must be closed), network connections, threads (must
* be stopped and joined) and memory (must be freed). Of course, there are a myriad
* of other things, so use your own judgement what you need to do.
*
* Another important chore of this function is to persist whatever state the module
* needs to persist. Unfortunately, there is currently no standard way of doing that.
* Future version of the module interface will probably support it, but that doesn't
* help you right at the moment. In general, it is suggested that anything that needs
* to be persisted is saved in a file, whose name and location is passed in by a
* module-specific config directive.
*/
BEGINafterRun
/* place any variables needed here */
CODESTARTafterRun
/* ... do cleanup here ... */
/* if you have a string config variable, remember to free its content:
*
if(pszStr != NULL) {
free(pszStr);
pszStr = NULL;
}
*/
ENDafterRun
/* The following entry points are defined in module-template.h.
* In general, they need to be present, but you do NOT need to provide
* any code here.
*/
BEGINmodExit
CODESTARTmodExit
ENDmodExit
BEGINqueryEtryPt
CODESTARTqueryEtryPt
CODEqueryEtryPt_STD_IMOD_QUERIES
ENDqueryEtryPt
/* The following function shall reset all configuration variables to their
* default values. The code provided in modInit() below registers it to be
* called on "$ResetConfigVariables". You may also call it from other places,
* but in general this is not necessary. Once runInput() has been called, this
* function here is never again called.
*/
static rsRetVal resetConfigVariables(uchar __attribute__((unused)) *pp, void __attribute__((unused)) *pVal)
{
DEFiRet;
/* if you have string variables in you config settings, you need to do this:
if(pszStr != NULL) {
free(pszStr);
pszStr = NULL;
}
* Note that it is vitally important that the pointer is set to NULL, because
* otherwise the framework handler will try to free it a second time when
* a new value is set!
*/
/* ... your code here ... */
RETiRet;
}
/* modInit() is called once the module is loaded. It must perform all module-wide
* initialization tasks. There are also a number of housekeeping tasks that the
* framework requires. These are handled by the macros. Please note that the
* complexity of processing is depending on the actual module. However, only
* thing absolutely necessary should be done here. Actual app-level processing
* is to be performed in runInput(). A good sample of what to do here may be to
* set some variable defaults. The most important thing probably is registration
* of config command handlers.
*/
BEGINmodInit()
CODESTARTmodInit
*ipIFVersProvided = 1; /* interface spec version this module is written to (currently always 1) */
CODEmodInit_QueryRegCFSLineHdlr
/* register config file handlers
* For details, see cfsysline.c/.h. The config file is automatically handled. In general,
* a pointer to a variable receiving the value and the config directive is to be supplied.
* A custom function pointer can only be provided, which then is called when the config
* directive appears. Limit this to cases where it is absolutely necessary. The
* STD_LOADABLE_MODULE_ID is a value that identifies the module. It is use to automatically
* unregister the module's config file handlers upon module unload. Do NOT use any other
* value for this parameter! Available Syntaxes (supported types) can be seen in cfsysline.h,
* the ecslCmdHdrlType enum has all that are currently defined.
*
* Config file directives should always be along the lines of
*
* $Input<moduleobject>ObjObjName
*
* An example would be $InputImtemplateRetriesMax. This is currently not enforced,
* but when we get to our new config file format and reader, this becomes quite
* important.
*
* Please note that config directives must be provided in lower case. The engine
* makes the mapping (what currently means case-insensitive comparison). The dollar
* sign is NOT part of the directive and thus not specified.
*
* Some samples:
*
* A hypothetical integer variable:
* CHKiRet(omsdRegCFSLineHdlr((uchar *)"inputimtemplatemessagenumber", 0, eCmdHdlrInt,
NULL, &intVariable, STD_LOADABLE_MODULE_ID));
*
* and a hypothetical string variable:
* CHKiRet(omsdRegCFSLineHdlr((uchar *)"inputimtemplatemessagetext", 0, eCmdHdlrGetWord,
* NULL, &pszBindAddr, STD_LOADABLE_MODULE_ID));
*/
/* whenever config variables exist, they should be resettable via $ResetConfigVariables.
* The following line adds our handler for that. Note that if you do not have any config
* variables at all (unlikely, I think...), you can remove this handler.
*/
CHKiRet(omsdRegCFSLineHdlr((uchar *)"resetconfigvariables", 1, eCmdHdlrCustomHandler,
resetConfigVariables, NULL, STD_LOADABLE_MODULE_ID));
/* ... do whatever else you need to do, but keep it brief ... */
ENDmodInit
/*
* vim:set ai:
*/