mirror/mariadb
1
0
Fork 0
mirror of https://github.com/MariaDB/server.git synced 2025-02-02 03:51:50 +01:00
Code Issues Projects Releases Packages Wiki Activity Actions
mariadb/storage/tokudb/PerconaFT/ft/ft-verify.cc
Sergei Golubchik 2c8c652978 5.6.26-74.0
2015-10-26 12:57:57 +01:00

523 lines
23 KiB
C++
Raw Blame History

/* -*- mode: C++; c-basic-offset: 4; indent-tabs-mode: nil -*- */
// vim: ft=cpp:expandtab:ts=8:sw=4:softtabstop=4:
#ident "$Id$"
/*======
This file is part of PerconaFT.
Copyright (c) 2006, 2015, Percona and/or its affiliates. All rights reserved.
PerconaFT is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License, version 2,
as published by the Free Software Foundation.
PerconaFT 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 PerconaFT. If not, see <http://www.gnu.org/licenses/>.
----------------------------------------
PerconaFT 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.
PerconaFT 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 PerconaFT. If not, see <http://www.gnu.org/licenses/>.
======= */
#ident "Copyright (c) 2006, 2015, Percona and/or its affiliates. All rights reserved."
/* Verify an FT. */
/* Check:
* The tree is of uniform depth (and the height is correct at every node)
* For each pivot key: the max of the stuff to the left is <= the pivot key < the min of the stuff to the right.
* For each leaf node: All the keys are in strictly increasing order.
* For each nonleaf node: All the messages have keys that are between the associated pivot keys ( left_pivot_key < message <= right_pivot_key)
*/
#include "ft/serialize/block_table.h"
#include "ft/ft.h"
#include "ft/ft-cachetable-wrappers.h"
#include "ft/ft-internal.h"
#include "ft/node.h"
static int
compare_pairs (FT_HANDLE ft_handle, const DBT *a, const DBT *b) {
return ft_handle->ft->cmp(a, b);
}
static int
compare_pair_to_key (FT_HANDLE ft_handle, const DBT *a, const void *key, uint32_t keylen) {
DBT y;
return ft_handle->ft->cmp(a, toku_fill_dbt(&y, key, keylen));
}
static int
verify_msg_in_child_buffer(FT_HANDLE ft_handle, enum ft_msg_type type, MSN msn, const void *key, uint32_t keylen, const void *UU(data), uint32_t UU(datalen), XIDS UU(xids), const DBT *lesser_pivot, const DBT *greatereq_pivot)
__attribute__((warn_unused_result));
UU()
static int
verify_msg_in_child_buffer(FT_HANDLE ft_handle, enum ft_msg_type type, MSN msn, const void *key, uint32_t keylen, const void *UU(data), uint32_t UU(datalen), XIDS UU(xids), const DBT *lesser_pivot, const DBT *greatereq_pivot) {
int result = 0;
if (msn.msn == ZERO_MSN.msn)
result = EINVAL;
switch (type) {
default:
break;
case FT_INSERT:
case FT_INSERT_NO_OVERWRITE:
case FT_DELETE_ANY:
case FT_ABORT_ANY:
case FT_COMMIT_ANY:
// verify key in bounds
if (lesser_pivot) {
int compare = compare_pair_to_key(ft_handle, lesser_pivot, key, keylen);
if (compare >= 0)
result = EINVAL;
}
if (result == 0 && greatereq_pivot) {
int compare = compare_pair_to_key(ft_handle, greatereq_pivot, key, keylen);
if (compare < 0)
result = EINVAL;
}
break;
}
return result;
}
static DBT
get_ith_key_dbt (BASEMENTNODE bn, int i) {
DBT kdbt;
int r = bn->data_buffer.fetch_key_and_len(i, &kdbt.size, &kdbt.data);
invariant_zero(r); // this is a bad failure if it happens.
return kdbt;
}
#define VERIFY_ASSERTION(predicate, i, string) ({ \
if(!(predicate)) { \
fprintf(stderr, "%s:%d: Looking at child %d of block %" PRId64 ": %s\n", __FILE__, __LINE__, i, blocknum.b, string); \
result = TOKUDB_NEEDS_REPAIR; \
if (!keep_going_on_failure) goto done; \
}})
#define VERIFY_ASSERTION_BASEMENT(predicate, bn, entry, string) ({ \
if(!(predicate)) { \
fprintf(stderr, "%s:%d: Looking at block %" PRId64 " bn %d entry %d: %s\n", __FILE__, __LINE__, blocknum.b, bn, entry, string); \
result = TOKUDB_NEEDS_REPAIR; \
if (!keep_going_on_failure) goto done; \
}})
struct count_msgs_extra {
int count;
MSN msn;
message_buffer *msg_buffer;
};
// template-only function, but must be extern
int count_msgs(const int32_t &offset, const uint32_t UU(idx), struct count_msgs_extra *const e)
__attribute__((nonnull(3)));
int count_msgs(const int32_t &offset, const uint32_t UU(idx), struct count_msgs_extra *const e)
{
MSN msn;
e->msg_buffer->get_message_key_msn(offset, nullptr, &msn);
if (msn.msn == e->msn.msn) {
e->count++;
}
return 0;
}
struct verify_message_tree_extra {
message_buffer *msg_buffer;
bool broadcast;
bool is_fresh;
int i;
int verbose;
BLOCKNUM blocknum;
int keep_going_on_failure;
bool messages_have_been_moved;
};
int verify_message_tree(const int32_t &offset, const uint32_t UU(idx), struct verify_message_tree_extra *const e) __attribute__((nonnull(3)));
int verify_message_tree(const int32_t &offset, const uint32_t UU(idx), struct verify_message_tree_extra *const e)
{
BLOCKNUM blocknum = e->blocknum;
int keep_going_on_failure = e->keep_going_on_failure;
int result = 0;
DBT k, v;
ft_msg msg = e->msg_buffer->get_message(offset, &k, &v);
bool is_fresh = e->msg_buffer->get_freshness(offset);
if (e->broadcast) {
VERIFY_ASSERTION(ft_msg_type_applies_all((enum ft_msg_type) msg.type()) || ft_msg_type_does_nothing((enum ft_msg_type) msg.type()),
e->i, "message found in broadcast list that is not a broadcast");
} else {
VERIFY_ASSERTION(ft_msg_type_applies_once((enum ft_msg_type) msg.type()),
e->i, "message found in fresh or stale message tree that does not apply once");
if (e->is_fresh) {
if (e->messages_have_been_moved) {
VERIFY_ASSERTION(is_fresh,
e->i, "message found in fresh message tree that is not fresh");
}
} else {
VERIFY_ASSERTION(!is_fresh,
e->i, "message found in stale message tree that is fresh");
}
}
done:
return result;
}
int error_on_iter(const int32_t &UU(offset), const uint32_t UU(idx), void *UU(e));
int error_on_iter(const int32_t &UU(offset), const uint32_t UU(idx), void *UU(e)) {
return TOKUDB_NEEDS_REPAIR;
}
int verify_marked_messages(const int32_t &offset, const uint32_t UU(idx), struct verify_message_tree_extra *const e) __attribute__((nonnull(3)));
int verify_marked_messages(const int32_t &offset, const uint32_t UU(idx), struct verify_message_tree_extra *const e)
{
BLOCKNUM blocknum = e->blocknum;
int keep_going_on_failure = e->keep_going_on_failure;
int result = 0;
bool is_fresh = e->msg_buffer->get_freshness(offset);
VERIFY_ASSERTION(!is_fresh, e->i, "marked message found in the fresh message tree that is fresh");
done:
return result;
}
template<typename verify_omt_t>
static int
verify_sorted_by_key_msn(FT_HANDLE ft_handle, message_buffer *msg_buffer, const verify_omt_t &mt) {
int result = 0;
size_t last_offset = 0;
for (uint32_t i = 0; i < mt.size(); i++) {
int32_t offset;
int r = mt.fetch(i, &offset);
assert_zero(r);
if (i > 0) {
struct toku_msg_buffer_key_msn_cmp_extra extra(ft_handle->ft->cmp, msg_buffer);
if (toku_msg_buffer_key_msn_cmp(extra, last_offset, offset) >= 0) {
result = TOKUDB_NEEDS_REPAIR;
break;
}
}
last_offset = offset;
}
return result;
}
template<typename count_omt_t>
static int
count_eq_key_msn(FT_HANDLE ft_handle, message_buffer *msg_buffer, const count_omt_t &mt, const DBT *key, MSN msn) {
struct toku_msg_buffer_key_msn_heaviside_extra extra(ft_handle->ft->cmp, msg_buffer, key, msn);
int r = mt.template find_zero<struct toku_msg_buffer_key_msn_heaviside_extra, toku_msg_buffer_key_msn_heaviside>(extra, nullptr, nullptr);
int count;
if (r == 0) {
count = 1;
} else {
assert(r == DB_NOTFOUND);
count = 0;
}
return count;
}
void
toku_get_node_for_verify(
BLOCKNUM blocknum,
FT_HANDLE ft_handle,
FTNODE* nodep
)
{
uint32_t fullhash = toku_cachetable_hash(ft_handle->ft->cf, blocknum);
ftnode_fetch_extra bfe;
bfe.create_for_full_read(ft_handle->ft);
toku_pin_ftnode(
ft_handle->ft,
blocknum,
fullhash,
&bfe,
PL_WRITE_EXPENSIVE, // may_modify_node
nodep,
false
);
}
struct verify_msg_fn {
FT_HANDLE ft_handle;
NONLEAF_CHILDINFO bnc;
const DBT *curr_less_pivot;
const DBT *curr_geq_pivot;
BLOCKNUM blocknum;
MSN this_msn;
int verbose;
int keep_going_on_failure;
bool messages_have_been_moved;
MSN last_msn;
int msg_i;
int result = 0; // needed by VERIFY_ASSERTION
verify_msg_fn(FT_HANDLE handle, NONLEAF_CHILDINFO nl, const DBT *less, const DBT *geq,
BLOCKNUM b, MSN tmsn, int v, int k, bool m) :
ft_handle(handle), bnc(nl), curr_less_pivot(less), curr_geq_pivot(geq),
blocknum(b), this_msn(tmsn), verbose(v), keep_going_on_failure(k), messages_have_been_moved(m), last_msn(ZERO_MSN), msg_i(0) {
}
int operator()(const ft_msg &msg, bool is_fresh) {
enum ft_msg_type type = (enum ft_msg_type) msg.type();
MSN msn = msg.msn();
XIDS xid = msg.xids();
const void *key = msg.kdbt()->data;
const void *data = msg.vdbt()->data;
uint32_t keylen = msg.kdbt()->size;
uint32_t datalen = msg.vdbt()->size;
int r = verify_msg_in_child_buffer(ft_handle, type, msn, key, keylen, data, datalen, xid,
curr_less_pivot,
curr_geq_pivot);
VERIFY_ASSERTION(r == 0, msg_i, "A message in the buffer is out of place");
VERIFY_ASSERTION((msn.msn > last_msn.msn), msg_i, "msn per msg must be monotonically increasing toward newer messages in buffer");
VERIFY_ASSERTION((msn.msn <= this_msn.msn), msg_i, "all messages must have msn within limit of this node's max_msn_applied_to_node_in_memory");
if (ft_msg_type_applies_once(type)) {
int count;
DBT keydbt;
toku_fill_dbt(&keydbt, key, keylen);
int total_count = 0;
count = count_eq_key_msn(ft_handle, &bnc->msg_buffer, bnc->fresh_message_tree, toku_fill_dbt(&keydbt, key, keylen), msn);
total_count += count;
if (is_fresh) {
VERIFY_ASSERTION(count == 1, msg_i, "a fresh message was not found in the fresh message tree");
} else if (messages_have_been_moved) {
VERIFY_ASSERTION(count == 0, msg_i, "a stale message was found in the fresh message tree");
}
VERIFY_ASSERTION(count <= 1, msg_i, "a message was found multiple times in the fresh message tree");
count = count_eq_key_msn(ft_handle, &bnc->msg_buffer, bnc->stale_message_tree, &keydbt, msn);
total_count += count;
if (is_fresh) {
VERIFY_ASSERTION(count == 0, msg_i, "a fresh message was found in the stale message tree");
} else if (messages_have_been_moved) {
VERIFY_ASSERTION(count == 1, msg_i, "a stale message was not found in the stale message tree");
}
VERIFY_ASSERTION(count <= 1, msg_i, "a message was found multiple times in the stale message tree");
VERIFY_ASSERTION(total_count <= 1, msg_i, "a message was found in both message trees (or more than once in a single tree)");
VERIFY_ASSERTION(total_count >= 1, msg_i, "a message was not found in either message tree");
} else {
VERIFY_ASSERTION(ft_msg_type_applies_all(type) || ft_msg_type_does_nothing(type), msg_i, "a message was found that does not apply either to all or to only one key");
struct count_msgs_extra extra = { .count = 0, .msn = msn, .msg_buffer = &bnc->msg_buffer };
bnc->broadcast_list.iterate<struct count_msgs_extra, count_msgs>(&extra);
VERIFY_ASSERTION(extra.count == 1, msg_i, "a broadcast message was not found in the broadcast list");
}
last_msn = msn;
msg_i++;
done:
return result;
}
};
static int
toku_verify_ftnode_internal(FT_HANDLE ft_handle,
MSN rootmsn, MSN parentmsn_with_messages, bool messages_exist_above,
FTNODE node, int height,
const DBT *lesser_pivot, // Everything in the subtree should be > lesser_pivot. (lesser_pivot==NULL if there is no lesser pivot.)
const DBT *greatereq_pivot, // Everything in the subtree should be <= lesser_pivot. (lesser_pivot==NULL if there is no lesser pivot.)
int verbose, int keep_going_on_failure, bool messages_have_been_moved)
{
int result=0;
MSN this_msn;
BLOCKNUM blocknum = node->blocknum;
//printf("%s:%d pin %p\n", __FILE__, __LINE__, node_v);
toku_ftnode_assert_fully_in_memory(node);
this_msn = node->max_msn_applied_to_node_on_disk;
if (height >= 0) {
invariant(height == node->height); // this is a bad failure if wrong
}
if (node->height > 0 && messages_exist_above) {
VERIFY_ASSERTION((parentmsn_with_messages.msn >= this_msn.msn), 0, "node msn must be descending down tree, newest messages at top");
}
// Verify that all the pivot keys are in order.
for (int i = 0; i < node->n_children-2; i++) {
DBT x, y;
int compare = compare_pairs(ft_handle, node->pivotkeys.fill_pivot(i, &x), node->pivotkeys.fill_pivot(i + 1, &y));
VERIFY_ASSERTION(compare < 0, i, "Value is >= the next value");
}
// Verify that all the pivot keys are lesser_pivot < pivot <= greatereq_pivot
for (int i = 0; i < node->n_children-1; i++) {
DBT x;
if (lesser_pivot) {
int compare = compare_pairs(ft_handle, lesser_pivot, node->pivotkeys.fill_pivot(i, &x));
VERIFY_ASSERTION(compare < 0, i, "Pivot is >= the lower-bound pivot");
}
if (greatereq_pivot) {
int compare = compare_pairs(ft_handle, greatereq_pivot, node->pivotkeys.fill_pivot(i, &x));
VERIFY_ASSERTION(compare >= 0, i, "Pivot is < the upper-bound pivot");
}
}
for (int i = 0; i < node->n_children; i++) {
DBT x, y;
const DBT *curr_less_pivot = (i==0) ? lesser_pivot : node->pivotkeys.fill_pivot(i - 1, &x);
const DBT *curr_geq_pivot = (i==node->n_children-1) ? greatereq_pivot : node->pivotkeys.fill_pivot(i, &y);
if (node->height > 0) {
NONLEAF_CHILDINFO bnc = BNC(node, i);
// Verify that messages in the buffers are in the right place.
VERIFY_ASSERTION(verify_sorted_by_key_msn(ft_handle, &bnc->msg_buffer, bnc->fresh_message_tree) == 0, i, "fresh_message_tree");
VERIFY_ASSERTION(verify_sorted_by_key_msn(ft_handle, &bnc->msg_buffer, bnc->stale_message_tree) == 0, i, "stale_message_tree");
verify_msg_fn verify_msg(ft_handle, bnc, curr_less_pivot, curr_geq_pivot,
blocknum, this_msn, verbose, keep_going_on_failure, messages_have_been_moved);
int r = bnc->msg_buffer.iterate(verify_msg);
if (r != 0) { result = r; goto done; }
struct verify_message_tree_extra extra = { .msg_buffer = &bnc->msg_buffer, .broadcast = false, .is_fresh = true, .i = i, .verbose = verbose, .blocknum = node->blocknum, .keep_going_on_failure = keep_going_on_failure, .messages_have_been_moved = messages_have_been_moved };
r = bnc->fresh_message_tree.iterate<struct verify_message_tree_extra, verify_message_tree>(&extra);
if (r != 0) { result = r; goto done; }
extra.is_fresh = false;
r = bnc->stale_message_tree.iterate<struct verify_message_tree_extra, verify_message_tree>(&extra);
if (r != 0) { result = r; goto done; }
bnc->fresh_message_tree.verify_marks_consistent();
if (messages_have_been_moved) {
VERIFY_ASSERTION(!bnc->fresh_message_tree.has_marks(), i, "fresh message tree still has marks after moving messages");
r = bnc->fresh_message_tree.iterate_over_marked<void, error_on_iter>(nullptr);
if (r != 0) { result = r; goto done; }
}
else {
r = bnc->fresh_message_tree.iterate_over_marked<struct verify_message_tree_extra, verify_marked_messages>(&extra);
if (r != 0) { result = r; goto done; }
}
extra.broadcast = true;
r = bnc->broadcast_list.iterate<struct verify_message_tree_extra, verify_message_tree>(&extra);
if (r != 0) { result = r; goto done; }
}
else {
BASEMENTNODE bn = BLB(node, i);
for (uint32_t j = 0; j < bn->data_buffer.num_klpairs(); j++) {
VERIFY_ASSERTION((rootmsn.msn >= this_msn.msn), 0, "leaf may have latest msn, but cannot be greater than root msn");
DBT kdbt = get_ith_key_dbt(bn, j);
if (curr_less_pivot) {
int compare = compare_pairs(ft_handle, curr_less_pivot, &kdbt);
VERIFY_ASSERTION_BASEMENT(compare < 0, i, j, "The leafentry is >= the lower-bound pivot");
}
if (curr_geq_pivot) {
int compare = compare_pairs(ft_handle, curr_geq_pivot, &kdbt);
VERIFY_ASSERTION_BASEMENT(compare >= 0, i, j, "The leafentry is < the upper-bound pivot");
}
if (0 < j) {
DBT prev_key_dbt = get_ith_key_dbt(bn, j-1);
int compare = compare_pairs(ft_handle, &prev_key_dbt, &kdbt);
VERIFY_ASSERTION_BASEMENT(compare < 0, i, j, "Adjacent leafentries are out of order");
}
}
}
}
done:
return result;
}
// input is a pinned node, on exit, node is unpinned
int
toku_verify_ftnode (FT_HANDLE ft_handle,
MSN rootmsn, MSN parentmsn_with_messages, bool messages_exist_above,
FTNODE node, int height,
const DBT *lesser_pivot, // Everything in the subtree should be > lesser_pivot. (lesser_pivot==NULL if there is no lesser pivot.)
const DBT *greatereq_pivot, // Everything in the subtree should be <= lesser_pivot. (lesser_pivot==NULL if there is no lesser pivot.)
int (*progress_callback)(void *extra, float progress), void *progress_extra,
int recurse, int verbose, int keep_going_on_failure)
{
MSN this_msn;
//printf("%s:%d pin %p\n", __FILE__, __LINE__, node_v);
toku_ftnode_assert_fully_in_memory(node);
this_msn = node->max_msn_applied_to_node_on_disk;
int result = 0;
int result2 = 0;
if (node->height > 0) {
// Otherwise we'll just do the next call
result = toku_verify_ftnode_internal(
ft_handle, rootmsn, parentmsn_with_messages, messages_exist_above, node, height, lesser_pivot, greatereq_pivot,
verbose, keep_going_on_failure, false);
if (result != 0 && (!keep_going_on_failure || result != TOKUDB_NEEDS_REPAIR)) goto done;
}
if (node->height > 0) {
toku_move_ftnode_messages_to_stale(ft_handle->ft, node);
}
result2 = toku_verify_ftnode_internal(
ft_handle, rootmsn, parentmsn_with_messages, messages_exist_above, node, height, lesser_pivot, greatereq_pivot,
verbose, keep_going_on_failure, true);
if (result == 0) {
result = result2;
if (result != 0 && (!keep_going_on_failure || result != TOKUDB_NEEDS_REPAIR)) goto done;
}
// Verify that the subtrees have the right properties.
if (recurse && node->height > 0) {
for (int i = 0; i < node->n_children; i++) {
FTNODE child_node;
toku_get_node_for_verify(BP_BLOCKNUM(node, i), ft_handle, &child_node);
DBT x, y;
int r = toku_verify_ftnode(ft_handle, rootmsn,
(toku_bnc_n_entries(BNC(node, i)) > 0
? this_msn
: parentmsn_with_messages),
messages_exist_above || toku_bnc_n_entries(BNC(node, i)) > 0,
child_node, node->height-1,
(i==0) ? lesser_pivot : node->pivotkeys.fill_pivot(i - 1, &x),
(i==node->n_children-1) ? greatereq_pivot : node->pivotkeys.fill_pivot(i, &y),
progress_callback, progress_extra,
recurse, verbose, keep_going_on_failure);
if (r) {
result = r;
if (!keep_going_on_failure || result != TOKUDB_NEEDS_REPAIR) goto done;
}
}
}
done:
toku_unpin_ftnode(ft_handle->ft, node);
if (result == 0 && progress_callback)
result = progress_callback(progress_extra, 0.0);
return result;
}
int
toku_verify_ft_with_progress (FT_HANDLE ft_handle, int (*progress_callback)(void *extra, float progress), void *progress_extra, int verbose, int keep_on_going) {
assert(ft_handle->ft);
FTNODE root_node = NULL;
{
uint32_t root_hash;
CACHEKEY root_key;
toku_calculate_root_offset_pointer(ft_handle->ft, &root_key, &root_hash);
toku_get_node_for_verify(root_key, ft_handle, &root_node);
}
int r = toku_verify_ftnode(ft_handle, ft_handle->ft->h->max_msn_in_ft, ft_handle->ft->h->max_msn_in_ft, false, root_node, -1, NULL, NULL, progress_callback, progress_extra, 1, verbose, keep_on_going);
if (r == 0) {
toku_ft_lock(ft_handle->ft);
ft_handle->ft->h->time_of_last_verification = time(NULL);
ft_handle->ft->h->dirty = 1;
toku_ft_unlock(ft_handle->ft);
}
return r;
}
int
toku_verify_ft (FT_HANDLE ft_handle) {
return toku_verify_ft_with_progress(ft_handle, NULL, NULL, 0, 0);
}
Reference in a new issue View git blame Copy permalink