To ensure that regular and GC64 jobs in packaging workflows don't
overwrite artifacts of each other, we need to use a different artifact
name per job.

NO_DOC=ci
NO_TEST=ci
NO_CHANGELOG=ci

In PR #7090 we forgot to update the path to gather failure artifacts for
many packaging workflows. Now it is fixed.

NO_DOC=ci
NO_TEST=ci
NO_CHANGELOG=ci

GitHub suggests to look on the document, when it is present. The banner
is shown for visitors of issues and pull requests pages. We can use this
feature to highlight typical stones on the contributor's road.

My goal is to leave suggestions for typical points of a confusion in a
short document (don't bury into details) and leave links for further
reading.

I hope we should move most of content of the 'open a pull request'
section to a wiki page and possibly merge it with [1] and [2]. It
requires more work and I would consider it as a future step toward good
developer guidelines.

[1]: https://github.com/tarantool/tarantool/wiki/Code-review-procedure
[2]: https://www.tarantool.io/en/doc/latest/dev_guide/developer_guidelines

NO_CHANGELOG=not a user visible change
NO_DOC=nothing requires an update on the website
NO_TEST=it is purely documentation change

Introduce helpers for each of our custom extension types.
The helpers will be used by mp_check() to validate extension contents
and make sure no malformed data is accepted by tarantool.

Closes #6857

NO_DOC=no user visible changes

The inject remote method is used in testing quite a lot. For example,
when someone needs to pass arbitrary MsgPack to Tarantool.

The current way to use it is:
conn:_request(netbox._method.inject, opts, nil, nil, custom_msgpack)

This is quite long and ugly, so let's add a helper:
conn:_inject(custom_msgpack, opts)

Also, while we're at it, lets fix the following issue.
The _inject remote method is barely usable without knowing which sync
net.box will use for the next request.

On the one hand, the user has to encode some custom sync to the request
he's injecting.
On the other hand, net.box doesn't parse the custom sync and always uses
its own pre-generated one to wait for the request response.

So the user has to pick the correct sync value, which net.box uses
internally.

Let's make life simpler and introduce _next_sync method, which returns
the next sync, which will be used by net.box.

In-scope-of #6857
Closes #7177

NO_DOC=internal change
NO_CHANGELOG=internal change
NO_TEST=tested implicitly in next commit

A Tarantool server is supposed to send a greeting message right after
accepting a new client so the first thing an applier does after
connecting to the master is reads the greeting. It does this without
timeouts. The problem is that if by mistake we connect to a wrong
instance, which doesn't send anything to clients, the applier will hang
forever (until the remote closes the socket), without logging any
errors.

This may happen even with a valid Tarantool instance - if SSL encryption
is enabled on the master, but not on the client, because the SSL
protocol assumes that the client initiates a connection by writing to
the socket first (before the server).

Let's add a timeout to the operation reading the greeting. The timeout
is set to replication_disconnect_timeout(), after which a connection is
broken if the master doesn't send heartbeats for that long. Note, we
don't add a timeout to other read/write operations issued to initiate a
replication connection, because if we received a greeting and it's
valid, then the master is likely to be fine.

Closes #7204

NO_DOC=bug

The upgrade test worked incorrectly: the instances WAL directory contained
xlog with the latest schema version and the snapshot contained an older
schema version (2.7.1) than required (2.9.1), whereas there only need to
be a snapshot with 2.9.1 schema version — fix this descrepancy.

Also, schema upgrade is performed automatically: remove the redundant
`box.schema.upgrade` call.

NO_CHANGELOG=test fix
NO_DOC=test fix
NO_DOC=test fix

Commit b18dd47f ("Introduce backtrace=true option to fiber.info()")
introduced a way to skip backtraces in fiber.info() calls.
Commit 9da7e03e ("fiber: introduce fiber_o:info() and fiber_o:csw()")
introduced `options` function for fiber object however it ignored
passed options.
This patch fixed it. Currently fiber:info({backtrace = false})
returns info without backtrace.

Closes #7210

NO_DOC=bugfix

- Ubuntu Groovy Gorilla (20.10)
- Ubuntu Hirsute Hippo (21.04)

See for details: https://wiki.ubuntu.com/Releases

NO_DOC=ci
NO_TEST=ci
NO_CHANGELOG=ci

- Fedora 30
- Fedora 31
- Fedora 32
- Fedora 33

See for details: https://docs.fedoraproject.org/en-US/releases/eol

NO_DOC=ci
NO_TEST=ci
NO_CHANGELOG=ci

FreeBSD Tarantool port has some problems if build uses cmake'
ExternalProject_Add which refers to the github repository.
Sumodule works better - so switching to using it.

NO_CHANGELOG=build
NO_DOC=build
NO_TEST=build

Previous version of tzcode/CMakeLists.txt leaves untracked files under
source directory src/lib/tzcode/tz/. We have changed slightly an approach
used to checkout and build IANA github sources under build/ directory,
not in ${PROJECT_BINARY_DIR} (which may be pointing not to out-of-source
build directory, but rather to the repository root itself, if insource
build used).

Closes #7173

NO_CHANGELOG=infra
NO_DOC=infra
NO_TEST=infra

By a mistake it was marked as API_EXPORT but was forgotten to
be added to exports.

Closes #7125

NO_DOC=bugfix

This client error assumes two arguments: space's name and index's name.
Since from the context it's clear that primary index is affected,
let's drop index name from error message. It's quite handy for space
upgrade case, since in such situation we have lack of information
concerning space.

NO_CHANGELOG=<Minor change>
NO_DOC=<No functional changes>

We are going to utilize primary key definition during space upgrade in
order to verify primary key invariant. So we need it to be stored in
space_upgrade metadata. Let's extend space_upgrade_new() signature and
pass there primary key definition. Also patch moves space_upgrade_new()
call a bit lower through code - now it's called after all indexes are
assigned (to simplify extraction of primary key definition).
Moreover, in order to print pretty error message during space upgrade
we should keep space in struct space_upgrade.

NO_DOC=ee
NO_TEST=ee
NO_CHANGELOG=ee

box.error() expects first argument to be numeric or table; passing
string to it is wrong. Accidentally in net.box module two places where
string is passed to box.error() as first argument: in :unprepare() and
:connect() methods. Let's fix them and pass error code ER_ILLEGAL_PARAMS.

NO_DOC=<bugfix>
NO_CHANGELOG=<Minor>

Fix wording, punctuation, and formatting.

NO_DOC=changelog
NO_TEST=changelog

Previously SQL didn't validate for all cases, that the amount of the
source and destination columns during insertion is equal. The problem
was detected when we insert an incorrect amount of values into the
table. For example, the query

insert into t(a) select a, b from t

produced an instance crash. Fixed.

Closes #7132

NO_DOC=bug fix

Co-authored-by: Mergen Imeev <imeevma@gmail.com>

We used to not implement comparison operators for interval
objects, thus any compare (even of equal) objects returned false:

	```
	tarantool> dt1 = datetime.new()
	---
	...

	tarantool> dt2 = datetime.now()
	---
	...

	tarantool> dt1 - dt1 == dt2 - dt2
	---
	- false
	...
	```

Now we implemented comparison for interval objects, where we compare
field by field starting from highest attribute (e.g. `year`) to
smallest attribute (e.g. `nsec`) and if any of comparison
returned non zero value then we return this result.

NO_DOC=bugfix
NO_CHANGELOG=bugfix

Use Olson/IANA tzdata for timezone handling in datetime string.

Closes of #6751

NO_TEST=doc

@TarantoolBot document
Title: Document timezone support

Timezones support
-----------------

Tarantool uses IANA tzdata aka Olson DB facilities for timezone
resolution at the moment of parsing of datetime literals or while
parsing `tz` attribute in constructor, or `:set{}` method call.

In addition to the `tzoffset` we provided before, we now define
`tzindex`, the unique index assigned by Tarantool to each known IANA
timezone.

```lua
date = require 'datetime'

tarantool> T = date.parse '2022-01-01T00:00 Europe/Moscow'

tarantool> T.tzindex
---
- 947
...

tarantool> T.tzoffset
---
- 180
...

tarantool> T.tz
---
- Europe/Moscow
...
```

Now `date.isdst` field (alone and as part of `:totable()` table) is
correctly calculated using `tzindex` and attributes of the selected
timezone in the Olson DB timezone.

```
tarantool> date.parse('2004-06-01T00:00 Europe/Moscow').isdst
---
- true
...

tarantool> date.parse('2004-12-01T00:00 Europe/Moscow'):totable()
---
- sec: 0
  min: 0
  yday: 336
  day: 1
  nsec: 0
  isdst: false
  wday: 4
  tzoffset: 180
  month: 12
  year: 2004
  hour: 0
...

```

All timezone names and abbreviations, known to the current Tarantool
version are available via `datetime.TZ` bidirectional array.

```lua
print(date.TZ['Europe/Moscow']) -- 947
print(date.TZ[947]) -- Europe/Moscow
```

Limitations
-----------

There were moments in past history, when local mean time in some
partcular zone used timezone offset not representable in whole
minutes, but rather in seconds, i.e. in Moscow before 1918 there
used to be offset +2 hours 31 minutes and 19 seconds. Please see
Olson dump for this period

```
$ ./src/lib/tzcode/install/usr/bin/zdump -c1880,1918 -i Europe/Moscow

TZ="Europe/Moscow"
-       -       +023017 MMT
1916-07-03      00:01:02        +023119 MMT
1917-07-02      00      +033119 MST     1
1917-12-27      23      +023119 MMT
```

Modern tzdata rules do not use such tiny fraction, and all timezones differ
to UTC in units measured in minutes, not seconds. Tarantool datetime uses
minutes internally as units for `tzoffset` so there is some loss of
precision if you try to operate with such ancient timestamps.

For performance reasons try to reuse previously
allocated structure across calls to tzalloc() if
zone appears to be the same.

NO_DOC=refactoring
NO_CHANGELOG=refactoring
NO_TEST=refactoring

Fixed incorrect interval initialization for case of `()` used
while it was correct for `{}` arguments.

NO_DOC=bugfix
NO_CHANGELOG=bugfix
NO_TEST=bugfix

```
tarantool> itv = require('datetime').interval.new{}
---
...

tarantool> itv
---
- 0 seconds
...

tarantool> itv = require('datetime').interval.new()
---
...

tarantool> itv
---
- +0.000000001 seconds
...
```

Extended gen-zone-abbrevs.pl script with capability to generate
timezones.lua in addition to timezones.h.

That's bidirectional Lua translation table for resolution of
timezone name by it's index and vice versa. Exported to Lua
world via `datetime.TZ`

NO_DOC=next commit

Properly calculate `isdst` field in datetime Lua object and
for attribute returned by `:totable()` function.

NO_DOC=next commit

* Modified localtime.c to use tnt_tm instead of tm structure;
* Use this structure for passing data to tnt_localtime_rz at
  the moment of time-zone resolution;
* As a side-effect of refactoring - simplified singleton we
  used to use for automagical sorting of array with timezone
  abbreviations

NO_DOC=internal
NO_CHANGELOG=internal

Using our own `tnt_tm` allows to properly pass `tm_tzindex`.

NO_DOC=refactoring
NO_CHANGELOG=refactoring
NO_TEST=refactoring

* localtime.c refactored to make tzalloc/tzfree public;
* Modified Perl script to generate primary Olson timezones and aliases.
  Get rid of NYI;

NO_DOC=refactoring
NO_CHANGELOG=refactoring
NO_TEST=refactoring

- localtime.c, tzfile.h added to build

NO_DOC=build
NO_CHANGELOG=build
NO_TEST=build

A certificate used for an encrypted replication connection may expire,
in which case the user will be forced to reconfigure the connection both
on the master and on the replica to use the new certificate. If the
replica completes the reconfiguration first, it will fail to connect to
the master and never retry, even though the master should complete its
reconfiguration in a moment.

To avoid that, we should make the applier retry to connect on SSLError,
just like it does on SocketError.

Needed for https://github.com/tarantool/tarantool-ee/issues/107

NO_DOC=ee
NO_TEST=ee
NO_CHANGELOG=ee

Applier uses box_error_code() to for better logging:
 - It remembers tha last raised error code and skips logging if the new
   error code is the same.
 - It logs "will retry every X seconds" only for retryable error codes
   (for example, ER_SYSTEM) while for non-retryable errors (for example,
   ER_PROC_LUA) the message isn't logged.

box_error_code() returns ER_PROC_LUA for SSLError, which is confusing
and would result in inconsistent logging in applier if we made SSLError
retryable. Let's add a separate error code for this error (ER_SSL) and
introduce a test case that checks that box_error_code() works as
expected for all kinds of errors.

Follow-up commit a7028dde ("Add SSL iostream stub").
Needed for https://github.com/tarantool/tarantool-ee/issues/107

NO_DOC=internal
NO_CHANGELOG=internal

LuaJIT submodule is bumped to introduce the following changes:
* sysprof: change C configuration API
* sysprof: enrich symtab on a new trace or a proto
* sysprof: fix SYSPROF_HANDLER_STACK_DEPTH
* sysprof: make internal API functions static
* sysprof: add LUAJIT_DISABLE_SYSPROF to Makefile
* symtab: check the _GNU_SOURCE definition

Within this changeset Tarantool-specific backtrace handler is introduced
and set to be used by sysprof machinery.

Besides, all new public Lua C API introduced within this changeset is
added to extra/exports.

Follows up #781

NO_DOC=LuaJIT submodule bump
NO_TEST=LuaJIT submodule bump
NO_CHANGELOG=LuaJIT submodule bump

This patch introduces operator [] that allows to get elements from MAP
and ARRAY values.

Closes #4762
Closes #4763
Part of #6251

@TarantoolBot document
Title: Operator [] in SQL

Operator `[]` allows to get an element of MAP and ARRAY values.
Examples:
```
tarantool> box.execute([[SELECT [1, 2, 3, 4, 5][3];]])
---
- metadata:
  - name: COLUMN_1
    type: any
  rows:
  - [3]
...

tarantool> box.execute([[SELECT {'a' : 123, 7: 'asd'}['a'];]])
---
- metadata:
  - name: COLUMN_1
    type: any
  rows:
  - [123]
...
```

The returned values is of type ANY.

If the operator is applied to a value that is not a MAP or ARRAY or is
NULL, an error is thrown.

Example:
```
tarantool> box.execute([[SELECT 1[1];]])
---
- null
- Selecting is only possible from map and array values
...
```

However, if there are two or more operators next to each other, the
second or following operators do not throw an error, but instead
return NULL.

Example:
```
tarantool> box.execute([[select [1][1][2][3][4];]])
---
- metadata:
  - name: COLUMN_1
    type: any
  rows:
  - [null]
...
```

NO_DOC=changelog update
NO_TEST=changelog update

Add the fedora_36.yml and fedora_36_aarch64.yml workflow files to build
Tarantool packages for x86_64 and aarch64 systems.

Closes tarantool/tarantool-qa#239

NO_DOC=ci
NO_TEST=ci

In Fedora 36 the default version of the `openssl-devel` package is 3.
But for now Tarantool doesn't work with this version. So this patch
adds a workaround with installation of the `openssl1.1-devel` package
to build rpms and exporting env variable `OPENSSL_CONF=/dev/null` to
pass tests.

See for details: https://github.com/tarantool/tarantool/issues/6477.

NO_DOC=ci
NO_TEST=ci
NO_CHANGELOG=ci

Apparently, the current implementation of `tree_iterator_start_raw` is
buggy for reverse iterators: instead of tracking gaps for successors of
keys, it tracks gaps for tuples shifted by one to the left of the
successor: reorder the code of `tree_iterator_start_raw` to get
the successor tuple prior to shifting done for reverse iterators and
simplify the implementation to make it more straightforward and thus
comprehensible.

Closes #7073
Closes #7113

NO_DOC=bugfix

Co-authored-by: Alexander Lyapunov <alyapunov@tarantool.org>

If the space is being upgraded, UPDATE/UPSERT requests matching the new
space format should work. However, currently they wouldn't work unless
the target tuple has already been upgraded, because we don't apply the
upgrade function to a tuple fetched from the space before applying
UPDATE operations. Fix this.

Follow-up commit 21e2def9 ("box: introduce result processor").
Needed for https://github.com/tarantool/tarantool-ee/issues/111

NO_DOC=ee
NO_TEST=ee
NO_CHANGELOG=ee

There's no need to apply a space upgrade function to the tuple returned
by INSERT, REPLACE, UPDATE, or UPSERT, because it must match the new
format anyway. We only need to process tuples returned by DELETE
requests. That being said, let's move the result processing from
box_process_rw() to memtx_space_execute_delete(). Note, we don't need to
patch vinyl_space_execute_delete(), because in Vinyl DELETE doesn't
return the deleted tuple.

Follow-up commit 21e2def9 ("box: introduce result processor").

NO_DOC=refactoring
NO_TEST=refactoring
NO_CHANGELOG=refactoring

We're going to use the 'result_process' name for a function that
processes a result without creating a result_processor struct.

Follow-up commit 21e2def9 ("box: introduce result processor").

NO_DOC=refactoring
NO_TEST=refactoring
NO_CHANGELOG=refactoring

Changelog: https://curl.se/changes.html#7_83_0

Curl replaced prefix CURL_ with CMAKE_ in private identifiers [1] since
the 'CMAKE_' prefix is reserved for CMake's own private use, see [2].

Patch adds a number of options defined in curl build infrastructure with
their default values used in 7.83.0.

1. https://github.com/curl/curl/commit/9108da2c26d18e927b91e33d3729d9cf0f3eb8fa
2. https://cmake.org/cmake/help/latest/manual/cmake-variables.7.html

NO_DOC=libcurl submodule bump
NO_TEST=libcurl submodule bump

Follows up #6029