Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>

This imports a simple unit test program to validate the coroutine
implementation.

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>

The 'struct ucontext_t' declared by system headers appears too small,
because getcontext() calls will scribble over memory after the end of
the struct. The libucontext impl is a suitable alternative, as is the
fallback gthread impl.

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>

libucontext is a lightweight implementation of ucontext for platforms
that do not have a built-in implementation. This allows us to use the
same code to support libucontext as ucontext.

This is a copy of the spice-gtk commit

  commit fa812c88492641005a768c72502a8953bd1223b2
  Author: osy <osy@turing.llc>
  Date:   Fri Mar 4 15:52:48 2022 -0800

    coroutine: add support for libucontext

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>

This will offer greater performance that the gthread impl

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>

Instead of using ucontext and setjmp together just use setjmp and
limit ucontext usage to initialise the initial jmp_buf context.
This simplifies the code and reduce number of context operations done.
In particular using _setjmp/_longjmp moving from a context to
another does not require any system call.
On x64 the continuation structure is reduced from 2176 to 248 bytes.

This is a copy of the spice-gtk commit

  commit 5989350429a79787f02e9c869c7769989781d320
  Author: Frediano Ziglio <freddy77@gmail.com>
  Date:   Fri Aug 14 05:26:35 2020 +0100

    continuation: Simplify implementation

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>

Changing stack is not usually instrumentation friendly.
Allows to enable the usage of Valgrind memcheck calling some
valgrind macros.

This is a copy of the spice-gtk commit:

  commit 942bd130ddd3c331045c1d3e4f580909901d285a
  Author: Frediano Ziglio <freddy77@gmail.com>
  Date:   Thu Aug 13 14:31:52 2020 +0100

      Allows to use Valgrind with ucontext coroutines

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>

This is a copy of spice-gtk commit:

  commit 9def63a7a402e491ba79a4aa98ef85f8142f11b4
  Author: Frediano Ziglio <freddy77@gmail.com>
  Date:   Tue Aug 11 21:24:26 2020 +0100

    coroutine_gthread: Fix recursive yield

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>

Not necessary from Glib 2.32 for static allocated conditions.

This is a copy of spice-gtk commit:

  commit 3b24b556fdef88e4582cffec01800d24bb65ae24
  Author: Frediano Ziglio <freddy77@gmail.com>
  Date:   Tue Aug 11 21:23:09 2020 +0100

      coroutine_gthread: Do not call g_cond_init

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>

This is a copy of spice-gtk commit

  commit 25875b2e26a54a08571376da31684eed316df6d5
  Author: Frediano Ziglio <freddy77@gmail.com>
  Date:   Tue Aug 11 17:51:18 2020 +0100

      coroutine: Initialize "current" at link time

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>

Not used and implemented in different ways by the various
coroutine implementations

This is a copy of spice-gtk commit:

  commit ca2e4ec3bb4f5ce590b8c35fe173247c71958f20
  Author: Frediano Ziglio <freddy77@gmail.com>
  Date:   Tue Aug 11 17:49:53 2020 +0100

      coroutine: Remove "release" field

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>

Only used by different coroutine implementations

This is a copy of the spice-gtk commit:

  commit 5f162843f96baf54c74cb9a5fe9b3b5801026d14
  Author: Frediano Ziglio <freddy77@gmail.com>
  Date:   Tue Aug 11 17:45:40 2020 +0100

      coroutine: Make coroutine_swap static

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>

This is a copy of the spice-gtk commit:

  commit 274ea2683bb1c9e88ce58e2640e84a47a8fcb939
  Author: Marc-André Lureau <marcandre.lureau@redhat.com>
  Date:   Fri Jan 4 17:38:28 2019 +0400

    coroutine-gthread: fix GThread deprecations

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>

Take an approach similar to gthreads, which are considered as low
level/must work features by glib, which aborts on failures.

This change is copied from the spice-gtk commit:

  commit 0508f586be974a49497a15d5ce39b2f94b68a4bf
  Author: Marc-André Lureau <marcandre.lureau@redhat.com>
  Date:   Tue Nov 19 15:29:28 2013 +0100

    coroutine: error out on OOM or impossible conditions

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>

The coroutine entry() return value must be passed via the caller->data,
since coroutine_swap() returns from->data. (this is needed because
coroutine_swap() argument is used both to return from entry() or to send
data to a running coroutine)

When leaving a coroutine, clear its caller.

Return correct coroutine_self() before other coroutine are initialized.

This is a copy of spice-gtk commit

  commit 0c120e0f6d94da5b7aa2009e8c07fc1290722c01
  Author: Marc-André Lureau <marcandre.lureau@redhat.com>
  Date:   Tue Nov 19 14:57:05 2013 +0100

      coroutine: make gthread coroutine pass tests

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>

When leaving a coroutine, it swaps back to where it came from, not to
the leader/main coroutine.

This is a copy of spice-gtk commit

  commit a9a2c162e0ee03f80e3e6f7450a002719a80d06a
  Author: Marc-André Lureau <marcandre.lureau@redhat.com>
  Date:   Fri Nov 15 22:03:40 2013 +0100

      coroutine: fix current coroutine

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>

This is a copy of spice-gtk commit

  commit baa51c5ee34ca6d0448d02901a0a61dc87c3dff8
  Author: Marc-André Lureau <marcandre.lureau@redhat.com>
  Date:   Fri Nov 15 22:03:39 2013 +0100

    coroutine: add some preconditions

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>

As described in http://www.1024cores.net/home/lock-free-algorithms/tricks/fibers



This is a copy from spice-gtk of

  commit 1cbfe2b8fccb86a20039867fc33bc475f358f8e4
  Author: Marc-André Lureau <marcandre.lureau@redhat.com>
  Date:   Fri Aug 12 16:32:20 2011 +0200

    use _setjmp/_longjmp to speed up coroutine switching

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>

The rules expect GNU make which is not the default.

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>

Many of the GNULIB rules are explicitly disabled, while others are
testing for scenarios that don't apply because we don't use GNULIB,
and prefer glib2 APIs over raw platform libc APIs.

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>

The syntax-check script was setting the wrong current directory, so make
was reporting nothing todo, making the test a no-op.

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>

It's in the public API.

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>

gtk-doc is being phased out, as unmaintained.

Despite some current limitations or annoyances, gi-docgen is what
Glib/Gtk+ is using these days, so follow along.

I didn't include gvncpulse, it should be trivial to add if necessary...

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>

Modernize a little bit the meson file, to avoid extra warnings when
running a recent meson versions.

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>

Fedora best practice is to now maintain native and mingw
builds together, to prevent mingw falling behind and reduce
the maint burden.

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>

The keycodemapdb Meson subproject provides a program and a source input.
Since it is a subproject, Meson wants to sandbox that and requires it to
be explicitly exported. But this never happened -- instead, we manually
poked at files using the actual string path "subprojects/......"

This was always a Meson sandbox violation, but Meson 0.63.0 started
noticing it and erroring out.

Instead, do the right thing. Update the subproject to a version that has
a meson.build with actually meaningful contents -- namely, a files
variable and a found program. Then use these in order to run the needed
custom_target.

In the process, it is also necessary to correct the argument ordering
when running keymap-gen.

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>

It has only existed as a virtual Provides line since Fedora 25

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>

This macro is a wrappar around g_debug and as such doesn't need to add
its own newline, unless trying to break a single message across lines.

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>

gcrypt refuses to encrypt DES data using an all-zeros key. While this
might seem reasonable, this is the key that you get when someone sets
a zero length password in a VNC server.

IOW, the gcrypt behaviour prevents gtk-vnc connecting to a server with
zero length password.

GNUTLS doesn't have this problem but also doesn't support DES-ECB,
so we have to fake it by calling DES-CBC for each block individually.
Conceptually this is horribly inefficient, but since the VNC challenge
is only 16 bytes long it isn't a problem in the real world.

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>