* test: make sysprof tests more deterministic
* test: increase sampling interval in sysprof tests
* LJ_GC64: Fix IR_VARG offset for fixed number of results.
* sysprof: implement stack sandwich support
* symtab: fix .symtab section dump of the executable
* sysprof: disable proto and trace dumps in default
* ci: introduce GitHub action for environment setup
* build: configure parallel jobs
* ci: replace hw.ncpu with hw.logicalcpu for macOS
* ci: add Tarantool integration testing
* ci: remove GC64 matrix entries for ARM64 workflows
* ci: fix --parallel argument for MacOS runners

Closes #7172
Closes #7244
Closes #7264
Part of #7230
Needed for #7472

NO_DOC=LuaJIT submodule bump
NO_TEST=LuaJIT submodule bump

The problem is that even if we block all signals on all
threads except the main thread, the signals still can be
delivered to other threads (#7206). And another problem
is that user can spawn his own thread and not block
signals.

That is why the patch introduces tt_sigaction function that
guarantees that all signals will be handled only by the main
thread. We use this helper in clock_lowres module.
This is supposed to solve the problem, described in #7408.

NO_CHANGELOG=internal
NO_DOC=internal

Currently, tarantool uses fiber() macro in crash_collect()
to collect backtraces, which is redundant and leads to
NULL dereferencing if crash signal callback is executed
on thread with no initialized cord.

The patch makes it possible not to use fiber module for
collecting backtraces and gets rid of fiber() macro
in crash_collect().

NO_CHANGELOG=internal
NO_TEST=internal
NO_DOC=internal

Added a function (see the API in the documentation request below), which
reflects the `tuple[json_path]` Lua API (see #1285).

Part of #7228

@TarantoolBot document
Title: tuple: access a field using JSON path via module API

The following function is added into the module API:

```c
/**
 * Return a raw tuple field in the MsgPack format pointed by
 * a JSON path.
 *
 * The JSON path includes the outmost field. For example, "c" in
 * ["a", ["b", "c"], "d"] can be accessed using "[2][2]" path (if
 * index_base is 1, as in Lua). If index_base is set to 0, the
 * same field will be pointed by the "[1][1]" path.
 *
 * The first JSON path token may be a field name if the tuple
 * has associated format with named fields. A field of a nested
 * map can be accessed in the same way: "foo.bar" or ".foo.bar".
 *
 * The return value is valid until the tuple is destroyed, see
 * box_tuple_ref().
 *
 * Return NULL if the field does not exist or if the JSON path is
 * malformed or invalid. Multikey JSON path token [*] is treated
 * as invalid in this context.
 *
 * \param tuple a tuple
 * \param path a JSON path
 * \param path_len a length of @a path
 * \param index_base 0 if array element indexes in @a path are
 *        zero-based (like in C) or 1 if they're one-based (like
 *        in Lua)
 * \retval a pointer to a field data if the field exists or NULL
 */
API_EXPORT const char *
box_tuple_field_by_path(box_tuple_t *tuple, const char *path,
			uint32_t path_len, int index_base);
```

It is preparatory change to expose the `box_tuple_field_by_path()`
function into the module API. A user should have ability to choose
zero-based (as in C) or one-based (as in Lua) array element selectors.
This commit adds the `index_base` parameter into several internal
functions, which access a field using a JSON path.

Part of #7228

NO_DOC=refactoring, no user-visible changes
NO_TEST=will be tested in a next commit implicitly by
        box_tuple_field_by_path() test cases
NO_CHANGELOG=refactoring, no user-visible changes

Ubuntu Impish Indri (21.10) has reached end of life
(https://wiki.ubuntu.com/Releases). There will be
no further releases of Tarantool for Ubuntu 21.10,
hence there's no need to run tests and check
packaging.

NO_TEST=ci
NO_DOC=ci

Resolves tarantool/tarantool-qa#257

In Lua field's numbers are counted from base 1, however currently
space:format() and space.foreign_key return zero-based foreign key
fields, which leads to an error on space:format(space:format()).

Closes #7350

NO_DOC=bugfix

Currently a foreign_key field in the `format` argument, passed to
normalize_format, can be changed inside normalize_foreign_key_one.
Fix this by using a local copy of def.field.

NO_DOC=bugfix
NO_CHANGELOG=minor bug

The packpack-dont-decline-custom-version.patch file was introduced in
9707a7bd (github-ci: support the new version format) and tried to resolve
the issue that actually did not exist. PackPack supports accepting a
custom version (pay attention to `?=`) [1] and the patch is not needed.

[1] https://github.com/packpack/packpack/blob/194f898750cbbee6dd35255fe8aa7ec878eb3836/pack/config.mk#L45

NO_DOC=ci
NO_TEST=ci
NO_CHANGELOG=ci

The 2.8 branch is not longer supported. So there is no sense to keep
the logic for it.

NO_DOC=ci
NO_TEST=ci
NO_CHANGELOG=ci

Now it's time to drop the `.gitlab.mk` makefile because `.pack.mk` has
fully replaced it.

NO_DOC=ci
NO_TEST=ci
NO_CHANGELOG=ci

NO_DOC=ci
NO_TEST=ci
NO_CHANGELOG=ci

This patch introduces the new test makefile `.pack.mk` instead of the
ancient `.gitlab.mk`. The `.pack.mk` file is a refactored version of
`.gitlab.mk` and contains only those targets that are indeed used.

NO_DOC=ci
NO_TEST=ci
NO_CHANGELOG=ci

It is inconvenient to create self-referencing FK constraints, as the
space ID will only be generated during space creation. This is
especially useful for SQL, since the format for the space is created
at compile time, and the space ID is only obtained at run time.

Closes #7200

@TarantoolBot document
Title: Describe foreign keys referring to the same space
Since: 2.11
Root document: https://www.tarantool.io/en/doc/latest/book/box/data_model/#foreign-keys

It is possible to create a foreign key that refers to the same space (a
child space equals to the parent space).
To do that, omit `space` in the `foreign_key` parameter, or set it to
the id or to the name of the current space.

According to man 3 exec: "The first argument, by convention, should point
to the filename associated with the file being executed.". Using empty
string as the first argument while calling `true` program, provided by
coreutils led to error message being printed to stderr, which failed the
test.

This patch passes 'true' as the first argument.

Closes #7452.

NO_DOC=test
NO_CHANGELOG=test

Currently net.box's methods doesn't check types of passed options.
This can lead to Lua's internal errors, which are not self-explaining.

Let's add this functionality and raise errors with meaningful messages
in case of incorrect options.

Closes #6063
Closes #6530

NO_DOC=Not a visible change

Test is needed to be updated every time net.box's internals are patched
(e.g. in case of adding new functions of deleting old ones), as the order
of suggested autocomplete options changes.

Let's replace `assert_equals` with `assert_items_equals`, which is not
order dependent.

NO_DOC=test
NO_CHANGELOG=test

In order to support parallel cmake build in the `test-jepsen` target
we need to run Jepsen testing in a `debian-buster` container instead
of `debian-stretch` due to the old `cmake` version in it. The `cmake`
utility started to support `--parallel` option since 3.12 version [1].

In Debian Stretch `cmake` has version 3.7.
In Debian Buster `cmake` has version 3.13.

[1] https://cmake.org/cmake/help/latest/release/3.12.html#command-line

NO_DOC=ci
NO_TEST=ci
NO_CHANGELOG=ci

A note for the `.gitlab.mk` makefile:

Mentions about `.travis.mk` in `.gitlab.mk` file will be removed in
the scope of refactoring it later.

NO_DOC=ci
NO_TEST=ci
NO_CHANGELOG=ci

Now it's time to drop the `.travis.mk` makefile because `.test.mk` has
fully replaced it.

NO_DOC=ci
NO_TEST=ci
NO_CHANGELOG=ci

A note for arm64 OSX workflows:

The `arch -arm64` command prior to calling the makefile was omitted
because now the GitHub Actions agent has a build for arm64 machines.
Hence the agent is able to run without Rosetta 2 binary translator.

NO_DOC=ci
NO_TEST=ci
NO_CHANGELOG=ci

This patch adds actions with installing deps for Debian, FreeBSD, and
OSX operation systems. Also, it adds an action with installing deps for
Jepsen testing. Debian, FreeBSD, and OSX actions should be used before
running tests on the corresponding operation systems. Jepsen action
should be used before running Jepsen tests.

Reasoning:

1. The makefile w/o targets with deps installation is OS-independent.
   It should work on all systems.

2. Deps installation is more about env preparation rather than building
   and testing. Moreover, it is OS-dependent. So it's logical to take it
   outside the makefile into a GitHub action.

3. If we need testing on other operation systems, for example, CentOS,
   we will just create a special action with deps installation for that.
   No changes for the makefile will be needed. All should work out of
   the box.

NO_DOC=ci
NO_TEST=ci
NO_CHANGELOG=ci

This patch introduces the new test makefile `.test.mk` instead of the
ancient `.travis.mk`. The `.test.mk` file is a fully reworked version of
`.travis.mk` with the huge refactoring and contains only those targets
that are indeed used.

NO_DOC=ci
NO_TEST=ci
NO_CHANGELOG=ci

Created simplistic unit test for checking boundary conditions
for `datetime_unpack` function.

NO_DOC=internal
NO_CHANGELOG=internal

Introduced fuzzing testing for datetime decoder.

NO_DOC=internal
NO_CHANGELOG=internal

Introduced few checks of validity of incoming data
at the moment messagepack is converted to datetime.

- we check that size is either big enough for single
  epoch or whole record;
- and we check range of data in decoded fields.

Fixes #6723

NO_DOC=bugfix
NO_TEST=see the other commit for fuzzer

NO_DOC=refactoring
NO_CHANGELOG=refactoring

This patch fixes the ctrl+c behavior in multiline mode.
The old behavior was ignoring sigint in this mode.
The new behavior is quitting this mode and printing the new
default prompt.
Added a new test for covering the mentioned case.

Follows up #2717
Closes #7109

NO_DOC=BUGFIX

This commit replaces enter/leave_delayed_free_mode() MemtxAllocator
methods with open/close_read_view(). The open_read_view() method
returns a pointer to the new read view, which is supposed to be pssed
to close_read_view() when the read view is no longer needed. The new
API will allow us to optimize garbage collection, see #7185.

Currently, open_read_view() always returns nullptr and both ReadView
and memtx_read_view_opts are empty structs. We will extend them in
the future.

Closes #7364

NO_DOC=refactoring
NO_TEST=refactoring
NO_CHANGELOG=refactoring

std::get by type is unavailable in C++11. Implement our own version.
Let's name it util::get and put it in trivia/tuple.h.

NO_DOC=internal
NO_TEST=internal
NO_CHANGELOG=internal

To immediately delete tuples that were allocated before a read view was
created, but are not visible from the read view (see #7185), we will
need to store them in per-read view garbage collection lists. We will
also need to merge those lists. With lifo, merging is impossible. Let's
switch to stailq, which allows this operation (see stailq_concat).

Note, since the min value of box.cfg.slab_alloc_granularity is 4, we
align memtx_tuple by 4 bytes.

Needed for #7185

NO_DOC=refactoring
NO_TEST=refactoring
NO_CHANGELOG=refactoring

box.cfg.slab_alloc_granularity sets the min alignment used for struct
memtx_tuple. By default, it's set to the machine word size (8 bytes
typically), but it can be set to 4 bytes to reduce memory usage.

The problem is that we add memtx tuples to a garbage collection list
using the intrusive list pattern (basically, store a pointer to the next
entry in the first 8 bytes of the `memtx_tuple` struct). If
`box.cfg.slab_alloc_factor` is set to 4, the pointer will be unaligned
while the compiler will generate the machine code assuming it is
naturally aligned. For some architectures (x86), this works fine, but
for others (ARM), this may result in a runtime failure.

To fix this issue, we need to instruct the compiler that the list
pointer stored in `memtx_tuple` may be unaligned.

For now, let's silently ignore granularity < the machine word size,
because we're planning to switch from lifo to stailq for memtx garbage
collection lists, which would result in explicit compiler warnings.

See #7422
Needed for #7185

NO_DOC=invisible to the user
NO_CHANGELOG=invisible to the user
NO_TEST=will be added later when the bug is fixed

It isn't needed - we can check if delay_free_mode > 0 to figure our if
we are in the delayed free mode.

NO_DOC=refactoring
NO_TEST=refactoring
NO_CHANGELOG=refactoring

Checking #7185 with functional tests is complicated. Let's add a unit
test for this.

Needed for #7185

NO_DOC=test
NO_CHANGELOG=test

Just regular update to bring openssl security fixes into tarantool.

Added a patch for Mac OS to fix a build failure, see
https://github.com/openssl/openssl/issues/18720.

Changelog: https://www.openssl.org/news/cl111.txt
Vulnerabilities: https://www.openssl.org/news/vulnerabilities.html

NO_TEST=security update of a dependency
NO_DOC=security update of a dependency

Now we can make interactive transactions like in case of
local or txt remote consoles if peer supports streams.

Closes #7413

NO_DOC=minor change

The issue is if in binary remote console a error is thrown in expression
like "box.begin() error('something') box.commit()" then it is
overwritten by iproto check for active transactions at the end of eval.
The solution is to rollback active transaction in this case in console
code before iproto check.

Let's also assert in tests a behaviour that after successful transaction
begin it stays active in next evaluations (successful or not) until
explicit rollback.

Closes #7288

NO_DOC=minor change

Starting server is failed after chdir(3) for example if we call
box.cfg{} with work_dir set.

NO_DOC=fix testing harness
NO_TEST=fix testing harness
NO_CHANGELOG=fix testing harness

We keep per server directories there too, not only socket files.

NO_DOC=refactoring
NO_TEST=refactoring
NO_CHANGELOG=refactoring

For the client_print there are also other checks and actions but they
are unnecesessary.

NO_DOC=refactoring
NO_TEST=refactoring
NO_CHANGELOG=refactoring