Home - Waterfall Grid T-Grid Console Builders Recent Builds Buildslaves Changesources - JSON API - About

Console View


Categories: connectors experimental galera main
Legend:   Passed Failed Warnings Failed Again Running Exception Offline No data

connectors experimental galera main
Oleksandr Byelkin
Merge branch 'bb-10.11-release' into bb-11.4-release
Oleksandr Byelkin
Merge branch 'bb-11.4-release' into bb-11.8-release
Sergei Golubchik
MDEV-39363 Logical Bug in `NOT ( ... XOR ... )` Evaluation with Implicit Type Conversion

for 0.1f val_int() is 0, val_bool() is 1.
Oleksandr Byelkin
Merge branch 'bb-10.11-release' into bb-11.4-release
Sergei Golubchik
MDEV-39481 ASAN error on malformed WKB polygon

let's make is difficult for wkb and len to desync
Marko Mäkelä
MDEV-32115: Log checkpoint race with wsrep_sst_method=rsync

Galera snapshot transfer (SST) using the default wsrep_sst_method=rsync
is prone to creating corrupted snapshots. The probability for this is
rather low and might only affect installations that include
ENGINE=InnoDB tables that contain FULLTEXT INDEX.

The function sst_disable_innodb_writes() aims to disable all InnoDB writes
during the time a snapshot transfer (SST) is in progress using the
default wsrep_sst_method=rsync.

The logic based on invoking log_make_checkpoint() almost works, except
for two things: We failed to ensure that fts_optimize_callback() has
stopped executing, and we did not block updates of the log checkpoint
header.

log_checkpoint_low(): Assert that writes to the log are allowed.

buf_flush_page_cleaner(): Do not try to advance the checkpoint while
wsrep_sst_method=rsync is in progress. This prevents the assertion
in log_checkpoint_low() from failing.

fts_optimize_pause(), fts_optimize_resume(): Pause and resume the
fts_optimize_callback().

sst_disable_innodb_writes(): Disable all background writers
before initiating the log checkpoint.

fts_optimize_callback(): Assert that wsrep_sst_method=rsync is not
active, and remove the previous incorrect attempt at fixing this race.
Oleksandr Byelkin
Merge branch 'bb-11.4-release' into bb-11.8-release
Oleksandr Byelkin
Merge branch '10.6' into bb-10.11-release
Dave Gosselin
MDEV-37020: Derived table merge optimization does not work for delete and update

A derived table in a multitable DELETE or UPDATE was materialized
while a derived table in an equivalent query using a VIEW was merged.

The cause was a blanket guard in TABLE_LIST::init_derived added by
commit fe89df42686f.  That commit fixed a ROWNUM crash on VIEWs with
ORDER BY, but its derived table guard was wider than it needed to be.

Narrow that guard to the case when the derived table lives inside a
VIEW's body (the case when belong_to_view is set).  A derived table at
the top level of a multitable query will be merged, while a derived
nested table within a VIEW will be materialized.

Narrowing that guard exposes a separate latent bug.  The access check
in multi_update_check_table_access has a branch for VIEWs and another
for 'not VIEWs' which dereferences table->table->map.  A merged
derived table that is not a VIEW fits neither condition.  As is the
case in main.lock_multi_bug38499, when concurrent ALTER on the target
forces the prepared statement to be prepared again, table->table on
the merged derived table might be NULL and this leads to a crash.
Privileges for the underlying base tables are already checked by
multi_update_precheck, so multi_update_check_table_access now returns
early in its else branch when the input is a merged derived table.
Lawrin Novitsky
ODBC-492 SQLGetTypeInfo: Wrong COLUMN_SIZE for SQL_TYPE_TYPESTAMP

Specs say that column size for temporal types has to be length in chars for representation with
max possible precision. i.e. in case of MariaDB this is 26
Also other types have been corrected:
- SQL_TYPE_TIME for the same logic has to be 17. Event though ODBC type
  isn't supposed to have fractional part and hours value beyond 0-23
- SET and ENUM effective max length is ~65216
- TYPE_(W)VARCHAR is 65532

There were also some discrepences between values for ODBCv3 and ODBCv2
Enhanced existing test to check all these values.
  • codbc-alma8-aarch64: build linux-connector_odbc failed -  stdiowarnings (1)
  • codbc-alma9-aarch64: build linux-connector_odbc failed -  stdio
  • codbc-alma9-amd64: build linux-connector_odbc failed -  stdio
  • codbc-alma84-amd64: build linux-connector_odbc failed -  stdiowarnings (1)
  • codbc-bookworm-aarch64: build linux-connector_odbc failed -  stdio
  • codbc-bookworm-amd64: build linux-connector_odbc failed -  stdio
  • codbc-bullseye-aarch64: build linux-connector_odbc failed -  stdio
  • codbc-bullseye-amd64: build linux-connector_odbc failed -  stdio
  • codbc-jammy-amd64: build linux-connector_odbc failed -  stdio
  • codbc-noble-amd64-memcheck: build linux-connector_odbc failed -  stdiowarnings (4)
  • codbc-rhel8-aarch64: build linux-connector_odbc failed -  stdiowarnings (1)
  • codbc-rhel8-amd64: build linux-connector_odbc failed -  stdiowarnings (1)
  • codbc-rhel9-aarch64: build linux-connector_odbc failed -  stdiowarnings (1)
  • codbc-rhel9-amd64: build linux-connector_odbc failed -  stdiowarnings (1)
  • codbc-rocky8-aarch64: build linux-connector_odbc failed -  stdio
  • codbc-sles15-amd64: build linux-connector_odbc failed -  stdiowarnings (5)
  • codbc-windows: 'dojob SET TEST_DSN=odbc-3.1 SET TEST_DRIVER=odbc-3.1 SET TEST_PORT=3306 SET TEST_SCHEMA=odbcodbc-3.1 cd win32/test ctest --output-on-failure' failed -  stdio
Yuchen Pei
MDEV-39535 Check for signs of EOF in ha_partition::handle_unordered_prev

The fields m_part_spec.start_part and m_part_spec.end_part defines the
range of partitions for index scans.

Of these two fields, m_part_spec.start_part is also commonly assigned
NO_CURRENT_PART_ID to indicate uninitialisation or failure mode.

MDEV-20195 shows problems when "unordered" scan fails to detect such a
condition for HANDLER ... NEXT statements.

MDEV-37330 allows "unordered" scan to handle index_prev. So we have a
similar problem for HANDLER ... PREV statements, and fix it by
checking for m_part_spec.start_part == NO_CURRENT_PART_ID in
ha_partition::handle_unordered_prev.

Furthermore, "ordered" index scans, i.e. index scans utilising a
priority queue checks m_top_entry against NO_CURRENT_PART_ID to detect
similar issues. Unlike m_part_spec.start_part, m_top_entry is set by
"ordered" scans only. This is why we have to do similar checks on
m_part_spec.start_part in "unordered" scans regardless of the
direction of the scan. Also because of this, in this patch we now set
m_part_spec.start_part only to NO_CURRENT_PART_ID in
ha_partition::handle_unordered_scan_next_partition for EOF.

Also renamed ha_partition::m_unordered_reverse_index to
ha_partition::m_unordered_reverse_scan because it is always false when
there is only one partition to scan.

Two of the testcases (the non-original ones) don't fail before the
patch, but they take the same path i.e. enter
ha_partition::handle_unordered_prev with m_part_spec.start_part ==
NO_CURRENT_PART_ID.
Rucha Deodhar
MDEV-39449: Memory corruption (heap-buffer-overflow) in uint4korr and
Gcalc_function::count_internal, apparent partial stack looping in
Gcalc_function::count_internal and Assertion `(0)' failed in
Item_func_spatial_precise_rel::val_bool

Analysis:
ST_COVEREDBY() was linked with Item_func_spatial_precise_rel with
SP_COVEREDBY_FUNC, but the case was missing in val_bool() method.
This resulted in fall through switch leading to an assert/ASAN failure during

Fix:
Added the missing case and implemented it similar to WITHIN
(with MBR precheck + buffer). This fixes the crash and makes
ST_COVEREDBY() behave correctly.
Rucha Deodhar
MDEV-39213: json range syntax crash

Analysis:
When json is being parsed, the step decreases without a out-of-bound check
resulting in failure.
Fix:
Before decreasing the step, check if it will result into out of bound.
Oleksandr Byelkin
Merge branch '10.6' into bb-10.11-release
Yuchen Pei
MDEV-20195 Check for signs of EOF in partition "unordered scan"

When running HANDLER commands, there is no optimization to detect
const tables of empty rows and the handler states are not reset
between HANDLER NEXT statements, either in ha_partition::reset or in
ha_partition::partition_scan_set_up (for m_part_spec.start_part,
relevant to this task)

When doing a partition "unordered scan" (i.e. an ordered scan without
using priority queues), with the "first scan" in
ha_partition::handle_unordered_scan_next_partition, EOF causes
NO_CURRENT_PART_ID to be assigned to m_part_spec.start_part. In the
"subsequent scan" ha_partition::handle_unordered_next should check
this condition and return EOF directly, just like the "ordered scan"
ha_partition::handle_ordered_next.
Oleksandr Byelkin
fix
Oleksandr Byelkin
Merge branch '10.11' into bb-10.11-release
Oleksandr Byelkin
Merge branch 'bb-11.8-release' into bb-12.3-release
Oleksandr Byelkin
Merge branch 'bb-10.11-release' into bb-11.4-release
Oleksandr Byelkin
fix
Oleksandr Byelkin
Merge branch 'bb-11.4-release' into bb-11.8-release
Oleksandr Byelkin
Merge branch 'bb-11.4-release' into bb-11.8-release
ParadoxV5
MDEV-39418: Revert most of MDEV-39240 for MDEV-32188

MDEV-39240 fixed how servers before 11.5/11.4-enterprise
accepted timestamps beyond Year 2038 from row-based replication,
which were invalid until 11.5/11.4-enterprise’s MDEV-32188.
MDEV-39240 does not apply after MDEV-32188 extended the valid range,
so those versions should exclude this fix,
as if MDEV-32188 already covers it.

This commit reverts commits 3234045953 and most of f9c34a1442, keeping
only the tweak to the MTR script `include/check_type` for consistency.
Oleksandr Byelkin
fix
Oleksandr Byelkin
Merge branch 'bb-11.8-release' into bb-12.3-release
Rucha Deodhar
Fixes json test failures in func_json.test and mdev_32854.test in bb-12.3-release
Sergei Golubchik
MDEV-39540 crash due to narrowing cast in update_ref_and_keys()
Sergei Golubchik
MDEV-39516 s3 curl_easy_setopt requires long values otherwise compile failure

update submodule to compile on fc44
Rucha Deodhar
Potential fix for pull request finding

Co-authored-by: Copilot Autofix powered by AI <[email protected]>
Jan Lindström
MDEV-39413 wsrep unsafe handling of parameters

When server is started with unsafe parameter --exec could
return also error 134. Furthermore, --exed pkill could
return error 15. Added these error codes as accepted.