We use own IO callbacks from server code anyway.

- allow incremental linking for Debug build
- do not allow EDITANDCONTINUE (/ZI)
- better clang-cl warnings flags

Many InnoDB internal variables and counters were only exposed
in an unstructured fashion via SHOW ENGINE INNODB STATUS.

Expose more variables via SHOW STATUS. Many of these were
exported in XtraDB.

Also, introduce SHOW_SIZE_T and use the proper size for
exporting the InnoDB variables.

Remove some unnecessary indirection via export_vars, and
bind some variables directly.

dict_sys_t::rough_size(): Replaces dict_sys_get_size()
and includes the hash table sizes.

This is based on a contribution by Tony Liu from ServiceNow.

Shorten some VARCHAR attributes to a more reasonable length.

INNODB_METRICS: Rename the column STATUS to ENABLED, and make it Boolean.

Replace with INT(1) many Boolean attributes that were declared as VARCHAR
containing 'NO','YES','disabled','enabled','Uninitialized','Initialized'.

Replace some VARCHAR attributes with ENUM.

Replace some BIGINT with INT when 32 bits are sufficient.

Remove INNODB_SYS_TABLESPACES.SPACE_TYPE. The type of a tablespace
can be derived from the tablespace ID. A fixed number is used for
the system tablespace and the temporary tablespace. All other tablespaces
are single-table or single-partition tablespaces.

i_s_locks_row_t::lock_type, lock_get_type_str(): Remove.
This is a redundant field. Table and record locks can be
distinguished by whether i_s_locks_row_t::lock_index is NULL.

fill_trx_row(): Do not unnecessarily copy the constant strings that
trx->op_info is pointing to.

i_s_locks_row_t::lock_mode: Replace string with integer.

lock_get_mode_str(), lock_get_trx_id(), lock_get_trx(): Remove.

field_store_ulint(): Remove.

Fix the build on clang. GCC only checks that a member function with
the C++11 override keyword is actually overriding a virtual member
function in a base class. In clang, you will get a warning for
overridden member functions that lack the keyword, if any overridden
member function of the class is declared with the keyword.

Sources did not compile in some builds because of undeclared
ER_BLOB_KEY_WITHOUT_LENGTH. Moving the implementations of
Key_part_spec::check_key_length_for_blob() from sql_class.h to sql_class.cc

Uniformly use the `.git` suffix format

Create "Sql_cmd_show_slave_status" class for "SHOW SLAVE STATUS" command.

The gcc version in CentOS 7 is 4.8.2 (released in Mar 2013).
But Arm added the support for 'arm_acle.h' in gcc 5.1 (released in Jun 2014).
And there are also some bugs in Arm gcc 4.8.
The patch is to detect GCC(greater than 5.1) version for Arm crc32c.

Change-Id: I718b9b89f952fed72c8795a176ebbe5395c8dd5e
Signed-off-by: Yuqi Gu <yuqi.gu@arm.com>

Fix "consistent override", and "unused private field" warnings
in popular header files.

Remove this flag only from domdoc.cpp, which uses #import directive.

If it is not enabled, build wsrep as static "stub" library from wsrep_dummy.cc
´

Allow static plugins to export symbols (on Unix)
wsrep_info relies on wsrep defined symbols (e.g  LOCK_wsrep_config_state)
exported from mysqld

Not much use of this feature so far.

`cmake . -DWITH_EXTRA_CHARSETS=complex` did not compile.
Fixing forgotten "utf8" to "utf8mb3" in definitions of
CHARSET and CHARSETS_AVAILABLE in cmake/character_sets.cmake

installers

Rely on Universal CRT to be on every computer we install.

Improve previous patch so we do not depend on (tiny) compiler-version
dependent libraries, i.e vcruntime140.dll, and msvcp140.dll

The remaining dependency is Universal CRT, is part of Win10+ OS,
and should already be installed by Windows update on most downlevel
systems.

Do not load GetSystemTimePreciseAsFileTime() dynamically - it is available
since Windows 8. Win7 is going out of support before 10.5 reaches GA.

MDEV-19585  Assertion with S3 table and flush_tables

There was a bug in the page cache that didn't take into account that
another thread could be waiting for a page to be read by read_big_block().
Fixed by releasing all waiters in read_big_block()

There are two options when coping S3 tables with mysqldump
(there is startup option --copy_s3_tables, boolean, default no)
1) Ignore all tables with engine S3, as the data is already safe in S3 and any
computer where you restore the backup will automatically discover the S3 table.
2) Copy the table as a normal table with the following 2 changes:
- Change ENGINE=S3 to ENGINE=ARIA;
- After copy add to log 'ALTER TABLE table_name ENGINE=S3'

MDEV-19585  Assertion with S3 table and flush_tables

The limit has to be increased so that MariaDB can create system tables.
It should not have any notable impact on performance.

There should not be any notable performance differences between 1K and 4K,
especially for temporary tables. In most cases using bigger blocks is also
faster (with the possible exception of doing key reads of not fixed length
keys).