runtime/wti.c - main - githubmirror/rsyslog

runtime/wti.c
602 строки21 KB
Rainer Gerhards
queue: separate DA idle and worker timeouts
14 июл 2026, 21:57
14 июл 2026, 21:575ff8895
/* wti.c
 *
 * This file implements the worker thread instance (wti) class.
 *
 * File begun on 2008-01-20 by RGerhards based on functions from the
 * previous queue object class (the wti functions have been extracted)
 *
 * There is some in-depth documentation available in doc/dev_queue.html
 * (and in the web doc set on https://www.rsyslog.com/doc/). Be sure to read it
 * if you are getting acquainted to the object.
 *
 * Copyright 2008-2026 Adiscon GmbH.
 *
 * This file is part of the rsyslog runtime library.
 *
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at
 *
 *       http://www.apache.org/licenses/LICENSE-2.0
 *       -or-
 *       see COPYING.ASL20 in the source distribution
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */
#include "config.h"

#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <assert.h>
#include <signal.h>
#include <pthread.h>
#include <errno.h>
#include <stdint.h>

#include "rsyslog.h"
#include "stringbuf.h"
#include "srUtils.h"
#include "errmsg.h"
#include "wtp.h"
#include "wti.h"
#include "obj.h"
#include "glbl.h"
#include "action.h"
#include "atomic.h"
#include "rsconf.h"

/* static data */
DEFobjStaticHelpers;
DEFobjCurrIf(glbl)

    pthread_key_t thrd_wti_key;


/* methods */

/* get the header for debug messages
 * The caller must NOT free or otherwise modify the returned string!
 */
uchar *ATTR_NONNULL() wtiGetDbgHdr(const wti_t *const pThis) {
    ISOBJ_TYPE_assert(pThis, wti);

    if (pThis->pszDbgHdr == NULL)
        return (uchar *)"wti"; /* should not normally happen */
    else
        return pThis->pszDbgHdr;
}


/* return the current worker processing state. For the sake of
 * simplicity, we do not use the iRet interface. -- rgerhards, 2009-07-17
 */
int ATTR_NONNULL() wtiGetState(wti_t *pThis) {
    return ATOMIC_LOAD_32BIT(&pThis->bIsRunning, &pThis->mutIsRunning);
}

/* join terminated worker thread
 * This may be called in any thread state, it will be a NOP if the
 * thread is not to join.
 */
void ATTR_NONNULL() wtiJoinThrd(wti_t *const pThis) {
    int r;
    ISOBJ_TYPE_assert(pThis, wti);
    if (wtiGetState(pThis) == WRKTHRD_WAIT_JOIN) {
        DBGPRINTF("%s: joining terminated worker\n", wtiGetDbgHdr(pThis));
        if ((r = pthread_join(pThis->thrdID, NULL)) != 0) {
            LogMsg(r, RS_RET_INTERNAL_ERROR, LOG_WARNING, "rsyslog bug? wti cannot join terminated wrkr");
            return;
        }
        DBGPRINTF("%s: worker fully terminated\n", wtiGetDbgHdr(pThis));
        wtiSetState(pThis, WRKTHRD_STOPPED);
        if (dbgTimeoutToStderr) {
            fprintf(stderr, "rsyslog debug: %s: thread joined\n", wtiGetDbgHdr(pThis));
        }
    }
}

/* Set this thread to "always running" state (can not be unset)
 * rgerhards, 2009-07-20
 */
rsRetVal ATTR_NONNULL() wtiSetAlwaysRunning(wti_t *pThis) {
    ISOBJ_TYPE_assert(pThis, wti);
    pThis->bAlwaysRunning = RSTRUE;
    return RS_RET_OK;
}

/* Set status (thread is running or not), actually an property of
 * use for wtp, but we need to have it per thread instance (thus it
 * is inside wti). -- rgerhards, 2009-07-17
 */
rsRetVal ATTR_NONNULL() wtiSetState(wti_t *pThis, const int newVal) {
    ISOBJ_TYPE_assert(pThis, wti);
    if (newVal == WRKTHRD_STOPPED) {
        ATOMIC_STORE_32BIT(&pThis->bIsRunning, &pThis->mutIsRunning, 0);
    } else {
        ATOMIC_OR_INT_TO_INT(&pThis->bIsRunning, &pThis->mutIsRunning, newVal);
    }
    return RS_RET_OK;
}

/* Compare and Set state (atomic).
 */
rsRetVal ATTR_NONNULL() wtiCASState(wti_t *pThis, const int oldVal, const int newVal) {
    ISOBJ_TYPE_assert(pThis, wti);
    if (ATOMIC_CAS(&pThis->bIsRunning, oldVal, newVal, &pThis->mutIsRunning)) {
        return RS_RET_OK;
    } else {
        return RS_RET_FALSE;
    }
}


/* advise all workers to start by interrupting them. That should unblock all srSleep()
 * calls.
 */
rsRetVal wtiWakeupThrd(wti_t *pThis) {
    DEFiRet;

    ISOBJ_TYPE_assert(pThis, wti);

    if (wtiGetState(pThis)) {
        /* we first try the cooperative "cancel" interface */
        pthread_kill(pThis->thrdID, SIGTTIN);
        DBGPRINTF("sent SIGTTIN to worker thread %p\n", (void *)pThis->thrdID);
    }

    RETiRet;
}


/* Cancel the thread. If the thread is not running. But it is save and legal to
 * call wtiCancelThrd() in such situations. This function only returns when the
 * thread has terminated. Else we may get race conditions all over the code...
 * Note that when waiting for the thread to terminate, we do a busy wait, checking
 * progress every 10ms. It is very unlikely that we will ever cancel a thread
 * and, if so, it will only happen at the end of the rsyslog run. So doing this
 * kind of non-optimal wait is considered preferable over using condition variables.
 * rgerhards, 2008-02-26
 */
rsRetVal ATTR_NONNULL() wtiCancelThrd(wti_t *pThis, const uchar *const cancelobj) {
    DEFiRet;
    int state;
    wtp_t *pWtp;
    pthread_t thrdID;

    ISOBJ_TYPE_assert(pThis, wti);
    pWtp = pThis->pWtp;
    ISOBJ_TYPE_assert(pWtp, wtp);

    wtiJoinThrd(pThis);

    /* Lock wtp mutex to safely check state and access thrdID atomically.
     * This prevents race where thread terminates and thrdID becomes invalid
     * between state check and pthread_kill/pthread_cancel calls.
     */
    d_pthread_mutex_lock(&pWtp->mutWtp);
    state = wtiGetState(pThis);
    if (state == WRKTHRD_RUNNING || state == WRKTHRD_INITIALIZING) {
        thrdID = pThis->thrdID; /* Safe copy under mutex */
        d_pthread_mutex_unlock(&pWtp->mutWtp);

        LogMsg(0, RS_RET_ERR, LOG_WARNING,
               "%s: need to do cooperative cancellation "
               "- some data may be lost, increase timeout?",
               cancelobj);
        pthread_kill(thrdID, SIGTTIN);
        DBGPRINTF("sent SIGTTIN to worker thread %p, giving it a chance to terminate\n", (void *)thrdID);
        srSleep(0, 50000);
        wtiJoinThrd(pThis);
    } else {
        d_pthread_mutex_unlock(&pWtp->mutWtp);
        if (state == WRKTHRD_STOPPED) {
            FINALIZE; /* Already stopped, nothing to do */
        }
    }

    /* Second attempt with hard cancellation if needed */
    d_pthread_mutex_lock(&pWtp->mutWtp);
    state = wtiGetState(pThis);
    if (state == WRKTHRD_RUNNING || state == WRKTHRD_INITIALIZING) {
        thrdID = pThis->thrdID; /* Safe copy under mutex */
        d_pthread_mutex_unlock(&pWtp->mutWtp);

        LogMsg(0, RS_RET_ERR, LOG_WARNING, "%s: need to do hard cancellation", cancelobj);
        if (dbgTimeoutToStderr) {
            fprintf(stderr, "rsyslog debug: %s: need to do hard cancellation\n", cancelobj);
        }
        pthread_cancel(thrdID);
        pthread_kill(thrdID, SIGTTIN);
        DBGPRINTF("cooperative worker termination failed, using cancellation...\n");
        DBGOPRINT((obj_t *)pThis, "canceling worker thread\n");

        /* Wait for thread to terminate */
        while ((state = wtiGetState(pThis)) != WRKTHRD_STOPPED && state != WRKTHRD_WAIT_JOIN) {
            DBGOPRINT((obj_t *)pThis, "waiting on termination, state %d\n", state);
            srSleep(0, 10000);
        }
    } else {
        d_pthread_mutex_unlock(&pWtp->mutWtp);
    }

    wtiJoinThrd(pThis);

finalize_it:
    RETiRet;
}

/* note: this function is only called once in action.c */
rsRetVal wtiNewIParam(wti_t *const pWti, action_t *const pAction, actWrkrIParams_t **piparams) {
    actWrkrInfo_t *const wrkrInfo = &(pWti->actWrkrInfo[pAction->iActionNbr]);
    actWrkrIParams_t *iparams;
    size_t allocCount;
    size_t newMax;
    size_t startOffset;
    size_t zeroCount;
    DEFiRet;

    if (wrkrInfo->p.tx.currIParam == wrkrInfo->p.tx.maxIParams) {
        /* we need to extend */
        newMax = (wrkrInfo->p.tx.maxIParams == 0) ? CONF_IPARAMS_BUFSIZE : 2u * (size_t)wrkrInfo->p.tx.maxIParams;
        if ((size_t)pAction->iNumTpls > SIZE_MAX / newMax) {
            ABORT_FINALIZE(RS_RET_OUT_OF_MEMORY);
        }
        allocCount = (size_t)pAction->iNumTpls * newMax;
        if (allocCount > SIZE_MAX / sizeof(actWrkrIParams_t)) {
            ABORT_FINALIZE(RS_RET_OUT_OF_MEMORY);
        }
        CHKmalloc(iparams = realloc(wrkrInfo->p.tx.iparams, sizeof(actWrkrIParams_t) * allocCount));
        startOffset = (size_t)wrkrInfo->p.tx.currIParam * (size_t)pAction->iNumTpls;
        zeroCount = ((size_t)newMax - (size_t)wrkrInfo->p.tx.maxIParams) * (size_t)pAction->iNumTpls;
        memset(iparams + startOffset, 0, sizeof(actWrkrIParams_t) * zeroCount);
        wrkrInfo->p.tx.iparams = iparams;
        wrkrInfo->p.tx.maxIParams = (int)newMax;
    }
    *piparams = wrkrInfo->p.tx.iparams + wrkrInfo->p.tx.currIParam * pAction->iNumTpls;
    ++wrkrInfo->p.tx.currIParam;

finalize_it:
    RETiRet;
}


/* Destructor */
BEGINobjDestruct(wti) /* be sure to specify the object type also in END and CODESTART macros! */
    CODESTARTobjDestruct(wti);
    if (wtiGetState(pThis) != WRKTHRD_STOPPED) {
        DBGPRINTF("%s: rsyslog bug: worker not stopped during shutdown\n", wtiGetDbgHdr(pThis));
        if (dbgTimeoutToStderr) {
            fprintf(stderr, "RSYSLOG BUG: %s: worker not stopped during shutdown\n", wtiGetDbgHdr(pThis));
        } else {
            assert(wtiGetState(pThis) == WRKTHRD_STOPPED);
        }
    }
    /* actual destruction */
    batchFree(&pThis->batch);
    free(pThis->actWrkrInfo);
    pthread_cond_destroy(&pThis->pcondBusy);
    DESTROY_ATOMIC_HELPER_MUT(pThis->mutIsRunning);
    free(pThis->pszDbgHdr);
ENDobjDestruct(wti)


/* Standard-Constructor for the wti object
 */
BEGINobjConstruct(wti) /* be sure to specify the object type also in END macro! */
    INIT_ATOMIC_HELPER_MUT(pThis->mutIsRunning);
    pthread_cond_init(&pThis->pcondBusy, NULL);
ENDobjConstruct(wti)


/* Construction finalizer
 * rgerhards, 2008-01-17
 */
rsRetVal wtiConstructFinalize(wti_t *pThis) {
    DEFiRet;
    int iDeqBatchSize;

    ISOBJ_TYPE_assert(pThis, wti);

    DBGPRINTF("%s: finalizing construction of worker instance data (for %d actions)\n", wtiGetDbgHdr(pThis),
              runConf->actions.iActionNbr);

    /* initialize our thread instance descriptor (no concurrency here) */
    pThis->bIsRunning = WRKTHRD_STOPPED;

    /* must use calloc as we need zero-init */
    CHKmalloc(pThis->actWrkrInfo = calloc(runConf->actions.iActionNbr, sizeof(actWrkrInfo_t)));

    if (pThis->pWtp == NULL) {
        dbgprintf("wtiConstructFinalize: pWtp not set, this may be intentional\n");
        FINALIZE;
    }

    /* we now alloc the array for user pointers. We obtain the max from the queue itself. */
    CHKiRet(pThis->pWtp->pfGetDeqBatchSize(pThis->pWtp->pUsr, &iDeqBatchSize));
    CHKiRet(batchInit(&pThis->batch, iDeqBatchSize));

finalize_it:
    RETiRet;
}


/* cancellation cleanup handler for queueWorker ()
 * Most importantly, it must bring back the batch into a consistent state.
 * Keep in mind that cancellation is disabled if we run into
 * the cancel cleanup handler (and have been cancelled).
 * rgerhards, 2008-01-16
 */
static void wtiWorkerCancelCleanup(void *arg) {
    wti_t *pThis = (wti_t *)arg;
    wtp_t *pWtp;

    ISOBJ_TYPE_assert(pThis, wti);
    pWtp = pThis->pWtp;
    ISOBJ_TYPE_assert(pWtp, wtp);

    DBGPRINTF("%s: cancellation cleanup handler called.\n", wtiGetDbgHdr(pThis));
    pWtp->pfObjProcessed(pWtp->pUsr, pThis);
    DBGPRINTF("%s: done cancellation cleanup handler.\n", wtiGetDbgHdr(pThis));
}


/* wait for queue to become non-empty or timeout
 * this is introduced as helper to support queue minimum batch sizes, but may
 * also be used for other cases. This function waits until the queue is non-empty
 * or a timeout occurs. The timeout must be passed in as absolute value.
 * @returns 0 if timeout occurs (queue still empty), something else otherwise
 */
int ATTR_NONNULL() wtiWaitNonEmpty(wti_t *const pThis, const struct timespec timeout) {
    wtp_t *__restrict__ const pWtp = pThis->pWtp;
    int r;

    DBGOPRINT((obj_t *)pThis, "waiting on queue to become non-empty\n");
    if (d_pthread_cond_timedwait(&pThis->pcondBusy, pWtp->pmutUsr, &timeout) != 0) {
        r = 0;
    } else {
        r = 1;
    }
    DBGOPRINT((obj_t *)pThis, "waited on queue to become non-empty, result %d\n", r);
    return r;
}


/* wait for queue to become non-empty or timeout
 * helper to wtiWorker. Note the the predicate is
 * re-tested by the caller, so it is OK to NOT do it here.
 * rgerhards, 2009-05-20
 */
static void ATTR_NONNULL() doIdleProcessing(wti_t *const pThis, wtp_t *const pWtp, int *const pbInactivityTOOccurred) {
    struct timespec t;

    DBGPRINTF("%s: worker IDLE, waiting for work.\n", wtiGetDbgHdr(pThis));

    if (pThis->bAlwaysRunning) {
        /* never shut down any started worker */
        d_pthread_cond_wait(&pThis->pcondBusy, pWtp->pmutUsr);
    } else {
        const int timeout =
            pThis->workerIndex == 0 && pWtp->toFirstWrkShutdown != -2 ? pWtp->toFirstWrkShutdown : pWtp->toWrkShutdown;
        timeoutComp(&t, timeout); /* get absolute timeout */
        if (d_pthread_cond_timedwait(&pThis->pcondBusy, pWtp->pmutUsr, &t) != 0) {
            DBGPRINTF("%s: inactivity timeout, worker terminating...\n", wtiGetDbgHdr(pThis));
            *pbInactivityTOOccurred = 1; /* indicate we had a timeout */
        }
    }
    DBGOPRINT((obj_t *)pThis, "worker awoke from idle processing\n");
}


/* generic worker thread framework. Note that we prohibit cancellation
 * during almost all times, because it can have very undesired side effects.
 * However, we may need to cancel a thread if the consumer blocks for too
 * long (during shutdown). So what we do is block cancellation, and every
 * consumer must enable it during the periods where it is safe.
 */
PRAGMA_DIAGNOSTIC_PUSH
PRAGMA_IGNORE_Wempty_body rsRetVal wtiWorker(wti_t *__restrict__ const pThis) {
    wtp_t *__restrict__ const pWtp = pThis->pWtp; /* our worker thread pool -- shortcut */
    action_t *__restrict__ pAction;
    rsRetVal localRet;
    rsRetVal terminateRet;
    actWrkrInfo_t *__restrict__ wrkrInfo;
    int iCancelStateSave;
    int i, j, k;
    DEFiRet;

    dbgSetThrdName(pThis->pszDbgHdr);
    pthread_cleanup_push(wtiWorkerCancelCleanup, pThis);
    int bInactivityTOOccurred = 0;
    /* Cancellation contract:
     * - default state is DISABLE in worker context
     * - queue consumers enable it only around explicitly cancel-safe work
     * - critical cleanup/state transitions run with cancellation disabled
     */
    pthread_setcancelstate(PTHREAD_CANCEL_DISABLE, &iCancelStateSave);
    DBGPRINTF("wti %p: worker starting\n", pThis);
    /* now we have our identity, on to real processing */

    /* note: in this loop, the mutex is "never" unlocked. Of course,
     * this is not true: it actually is unlocked when the actual processing
     * is done, as part of pWtp->pfDoWork() processing. Note that this
     * function is required to re-lock it when done. We cannot do the
     * lock/unlock here ourselves, as pfDoWork() needs to access queue
     * structures itself.
     * The same goes for pfRateLimiter(). While we could unlock/lock when
     * we call it, in practice the function is often called without any
     * ratelimiting actually done. Only the rate limiter itself knows
     * that. As such, it needs to bear the burden of doing the locking
     * when required. -- rgerhards, 2013-11-20
     */
    d_pthread_mutex_lock(pWtp->pmutUsr);
    while (1) { /* loop will be broken below */
        if (pWtp->pfRateLimiter != NULL) { /* call rate-limiter, if defined */
            pWtp->pfRateLimiter(pWtp->pUsr);
        }

        /* first check if we are in shutdown process (but evaluate a bit later) */
        terminateRet = wtpChkStopWrkr(pWtp, MUTEX_ALREADY_LOCKED);
        if (terminateRet == RS_RET_TERMINATE_NOW) {
            /* we now need to free the old batch */
            localRet = pWtp->pfObjProcessed(pWtp->pUsr, pThis);
            DBGOPRINT((obj_t *)pThis,
                      "terminating worker because of "
                      "TERMINATE_NOW mode, del iRet %d\n",
                      localRet);
            break;
        }

        /* try to execute and process whatever we have */
        localRet = pWtp->pfDoWork(pWtp->pUsr, pThis);

        if (localRet == RS_RET_ERR_QUEUE_EMERGENCY) {
            break; /* end of loop */
        } else if (localRet == RS_RET_IDLE) {
            if (terminateRet == RS_RET_TERMINATE_WHEN_IDLE || bInactivityTOOccurred) {
                if (bInactivityTOOccurred && pWtp->pfIdleTimeout != NULL && pWtp->pWrkr[0] == pThis &&
                    pWtp->pfIdleTimeout(pWtp->pUsr) == RS_RET_RETRY) {
                    bInactivityTOOccurred = 0;
                    continue;
                }
                DBGOPRINT((obj_t *)pThis,
                          "terminating worker terminateRet=%d, "
                          "bInactivityTOOccurred=%d\n",
                          terminateRet, bInactivityTOOccurred);
                break; /* end of loop */
            }
            doIdleProcessing(pThis, pWtp, &bInactivityTOOccurred);
            continue; /* request next iteration */
        }

        bInactivityTOOccurred = 0; /* reset for next run */
    }

    d_pthread_mutex_unlock(pWtp->pmutUsr);

    DBGPRINTF("DDDD: wti %p: worker cleanup action instances\n", pThis);
    for (i = 0; i < runConf->actions.iActionNbr; ++i) {
        wrkrInfo = &(pThis->actWrkrInfo[i]);
        dbgprintf("wti %p, action %d, ptr %p\n", pThis, i, wrkrInfo->actWrkrData);
        if (wrkrInfo->actWrkrData != NULL) {
            pAction = wrkrInfo->pAction;
            actionRemoveWorker(pAction, wrkrInfo->actWrkrData);
            pAction->pMod->mod.om.freeWrkrInstance(wrkrInfo->actWrkrData);
            if (pAction->isTransactional) {
                /* free iparam "cache" - we need to go through to max! */
                for (j = 0; j < wrkrInfo->p.tx.maxIParams; ++j) {
                    for (k = 0; k < pAction->iNumTpls; ++k) {
                        free(actParam(wrkrInfo->p.tx.iparams, pAction->iNumTpls, j, k).param);
                    }
                }
                free(wrkrInfo->p.tx.iparams);
                wrkrInfo->p.tx.iparams = NULL;
                wrkrInfo->p.tx.currIParam = 0;
                wrkrInfo->p.tx.maxIParams = 0;
            } else {
                releaseDoActionParams(pAction, pThis, 1);
            }
            wrkrInfo->actWrkrData = NULL; /* re-init for next activation */
        }
    }

    /* indicate termination */
    pthread_cleanup_pop(0); /* remove cleanup handler */
    pthread_setcancelstate(iCancelStateSave, NULL);
    dbgprintf("wti %p: exiting\n", pThis);

    RETiRet;
}
PRAGMA_DIAGNOSTIC_POP


/* some simple object access methods */
rsRetVal wtiSetpWtp(wti_t *pThis, wtp_t *pVal) {
    pThis->pWtp = pVal;
    return RS_RET_OK;
}

/* set the debug header message
 * The passed-in string is duplicated. So if the caller does not need
 * it any longer, it must free it. Must be called only before object is finalized.
 * rgerhards, 2008-01-09
 */
rsRetVal wtiSetDbgHdr(wti_t *pThis, uchar *pszMsg, const size_t lenMsg) {
    DEFiRet;

    ISOBJ_TYPE_assert(pThis, wti);
    assert(pszMsg != NULL);

    if (lenMsg < 1) ABORT_FINALIZE(RS_RET_PARAM_ERROR);

    if (pThis->pszDbgHdr != NULL) {
        free(pThis->pszDbgHdr);
    }

    if ((pThis->pszDbgHdr = malloc(lenMsg + 1)) == NULL) ABORT_FINALIZE(RS_RET_OUT_OF_MEMORY);

    memcpy(pThis->pszDbgHdr, pszMsg, lenMsg + 1); /* always think about the \0! */

finalize_it:
    RETiRet;
}


/* This function returns (and creates if necessary) a dummy wti suitable
 * for use by the rule engine. It is intended to be used for direct-mode
 * main queues (folks, don't do that!). Once created, data is stored in
 * thread-specific storage.
 * Note: we do NOT do error checking -- if this functions fails, all the
 * rest will fail as well... (also, it will only fail under OOM, so...).
 * Memleak: we leak pWti's when run in direct mode. However, this is only
 * a cosmetic leak, as we need them until all inputs are terminated,
 * what means essentially until rsyslog itself is terminated. So we
 * don't care -- it's just not nice in valgrind, but that's it.
 */
wti_t *wtiGetDummy(void) {
    wti_t *pWti;

    pWti = (wti_t *)pthread_getspecific(thrd_wti_key);
    if (pWti == NULL) {
        wtiConstruct(&pWti);
        if (pWti != NULL) wtiConstructFinalize(pWti);
        if (pthread_setspecific(thrd_wti_key, pWti) != 0) {
            DBGPRINTF("wtiGetDummy: error setspecific thrd_wti_key\n");
        }
    }
    return pWti;
}

/* dummy */
static rsRetVal wtiQueryInterface(interface_t __attribute__((unused)) * i) {
    return RS_RET_NOT_IMPLEMENTED;
}

/* exit our class
 */
BEGINObjClassExit(wti, OBJ_IS_CORE_MODULE) /* CHANGE class also in END MACRO! */
    CODESTARTObjClassExit(wti);
    /* release objects we no longer need */
    objRelease(glbl, CORE_COMPONENT);
    pthread_key_delete(thrd_wti_key);
ENDObjClassExit(wti)


/* Initialize the wti class. Must be called as the very first method
 * before anything else is called inside this class.
 * rgerhards, 2008-01-09
 */
BEGINObjClassInit(wti, 1, OBJ_IS_CORE_MODULE) /* one is the object version (most important for persisting) */
    int r;
    /* request objects we use */
    CHKiRet(objUse(glbl, CORE_COMPONENT));
    r = pthread_key_create(&thrd_wti_key, NULL);
    if (r != 0) {
        dbgprintf("wti.c: pthread_key_create failed\n");
        ABORT_FINALIZE(RS_RET_ERR);
    }
ENDObjClassInit(wti)