summaryrefslogtreecommitdiff
path: root/src/mongo/db/exec/collection_scan.cpp
blob: e53810c75386bff3c431538b578afd2d41e21c00 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
/**
 *    Copyright (C) 2013 10gen Inc.
 *
 *    This program is free software: you can redistribute it and/or  modify
 *    it under the terms of the GNU Affero General Public License, version 3,
 *    as published by the Free Software Foundation.
 *
 *    This program 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 Affero General Public License for more details.
 *
 *    You should have received a copy of the GNU Affero General Public License
 *    along with this program.  If not, see <http://www.gnu.org/licenses/>.
 *
 *    As a special exception, the copyright holders give permission to link the
 *    code of portions of this program with the OpenSSL library under certain
 *    conditions as described in each individual source file and distribute
 *    linked combinations including the program with the OpenSSL library. You
 *    must comply with the GNU Affero General Public License in all respects for
 *    all of the code used other than as permitted herein. If you modify file(s)
 *    with this exception, you may extend this exception to your version of the
 *    file(s), but you are not obligated to do so. If you do not wish to do so,
 *    delete this exception statement from your version. If you delete this
 *    exception statement from all source files in the program, then also delete
 *    it in the license file.
 */

#include "mongo/db/exec/collection_scan.h"

#include "mongo/db/catalog/database.h"
#include "mongo/db/exec/collection_scan_common.h"
#include "mongo/db/exec/filter.h"
#include "mongo/db/exec/working_set.h"
#include "mongo/db/catalog/collection.h"
#include "mongo/util/fail_point_service.h"

#include "mongo/db/client.h" // XXX-ERH

namespace mongo {

    // Some fail points for testing.
    MONGO_FP_DECLARE(collscanInMemoryFail);
    MONGO_FP_DECLARE(collscanInMemorySucceed);

    // static
    bool CollectionScan::diskLocInMemory(DiskLoc loc) {
        if (MONGO_FAIL_POINT(collscanInMemoryFail)) {
            return false;
        }

        if (MONGO_FAIL_POINT(collscanInMemorySucceed)) {
            return true;
        }

        return _iter->recordFor(loc)->likelyInPhysicalMemory();
    }

    CollectionScan::CollectionScan(const CollectionScanParams& params,
                                   WorkingSet* workingSet,
                                   const MatchExpression* filter)
        : _workingSet(workingSet),
          _filter(filter),
          _params(params),
          _nsDropped(false) {

        // We pre-allocate a WSID and use it to pass up fetch requests.  It is only
        // used to pass up fetch requests and we should never use it for anything else.
        _wsidForFetch = _workingSet->allocate();
        WorkingSetMember* member = _workingSet->get(_wsidForFetch);
        // Kind of a lie since the obj isn't pointing to the data at loc. but the obj
        // won't be used.
        member->state = WorkingSetMember::LOC_AND_UNOWNED_OBJ;
    }

    PlanStage::StageState CollectionScan::work(WorkingSetID* out) {
        ++_commonStats.works;
        if (_nsDropped) { return PlanStage::DEAD; }

        // Do some init if we haven't already.
        if (NULL == _iter) {
            if ( _params.collection == NULL ) {
                _nsDropped = true;
                return PlanStage::DEAD;
            }

            _iter.reset( _params.collection->getIterator( _params.start,
                                                          _params.tailable,
                                                          _params.direction ) );

            ++_commonStats.needTime;
            return PlanStage::NEED_TIME;
        }

        // See if the record we're about to access is in memory.  If it's not, pass a fetch
        // request up.
        if (!isEOF()) {
            DiskLoc curr = _iter->curr();
            if (!curr.isNull() && !diskLocInMemory(curr)) {
                WorkingSetMember* member = _workingSet->get(_wsidForFetch);
                member->loc = curr;
                *out = _wsidForFetch;
                return PlanStage::NEED_FETCH;
            }
        }

        // What we'll return to the user.
        DiskLoc nextLoc;

        // Should we try getNext() on the underlying _iter if we're EOF?  Yes, if we're tailable.
        if (isEOF()) {
            if (!_params.tailable) {
                return PlanStage::IS_EOF;
            }
            else {
                // See if _iter gives us anything new.
                nextLoc = _iter->getNext();
                if (nextLoc.isNull()) {
                    // Nope, still EOF.
                    return PlanStage::IS_EOF;
                }
            }
        }
        else {
            nextLoc = _iter->getNext();
        }

        WorkingSetID id = _workingSet->allocate();
        WorkingSetMember* member = _workingSet->get(id);
        member->loc = nextLoc;
        member->obj = _params.collection->docFor(member->loc);
        member->state = WorkingSetMember::LOC_AND_UNOWNED_OBJ;

        ++_specificStats.docsTested;

        if (Filter::passes(member, _filter)) {
            *out = id;
            ++_commonStats.advanced;
            return PlanStage::ADVANCED;
        }
        else {
            _workingSet->free(id);
            ++_commonStats.needTime;
            return PlanStage::NEED_TIME;
        }
    }

    bool CollectionScan::isEOF() {
        if ((0 != _params.maxScan) && (_specificStats.docsTested >= _params.maxScan)) {
            return true;
        }
        if (_nsDropped) { return true; }
        if (NULL == _iter) { return false; }
        return _iter->isEOF();
    }

    void CollectionScan::invalidate(const DiskLoc& dl, InvalidationType type) {
        ++_commonStats.invalidates;

        // We don't care about mutations since we apply any filters to the result when we (possibly)
        // return it.
        if (INVALIDATION_DELETION != type) {
            return;
        }

        // If we're here, 'dl' is being deleted.

        // Deletions can harm the underlying RecordIterator so we must pass them down.
        if (NULL != _iter) {
            _iter->invalidate(dl);
        }

        // We might have 'dl' inside of the WSM that _wsidForFetch references.  This is OK because
        // the runner who handles the fetch request does so before releasing any locks (and allowing
        // the DiskLoc to be deleted).  We also don't use any data in the WSM referenced by
        // _wsidForFetch so it's OK to leave the DiskLoc there.
    }

    void CollectionScan::prepareToYield() {
        ++_commonStats.yields;
        if (NULL != _iter) {
            _iter->prepareToYield();
        }
    }

    void CollectionScan::recoverFromYield() {
        ++_commonStats.unyields;
        if (NULL != _iter) {
            if (!_iter->recoverFromYield()) {
                warning() << "Collection dropped or state deleted during yield of CollectionScan";
                _nsDropped = true;
            }
        }
    }

    PlanStageStats* CollectionScan::getStats() {
        _commonStats.isEOF = isEOF();
        auto_ptr<PlanStageStats> ret(new PlanStageStats(_commonStats, STAGE_COLLSCAN));
        ret->specific.reset(new CollectionScanStats(_specificStats));
        return ret.release();
    }

}  // namespace mongo