The Ctrl-Alt-u keyboard shortcut restores the screen to its original
size. In the SDL2 UI this is done by destroying the window and
creating a new one. The old window emits SDL_WINDOWEVENT_HIDDEN when
it's destroyed, but trying to call SDL_GetWindowFromID() from that
event's window ID returns a null pointer. handle_windowevent() assumes
that the pointer is never null so it results in a crash.

Cc: qemu-stable@nongnu.org
Signed-off-by: NAlberto Garcia <berto@igalia.com>
Signed-off-by: NGerd Hoffmann <kraxel@redhat.com>

Add new sdl2-gl.c file, with display
rendering functions using opengl.
Signed-off-by: NGerd Hoffmann <kraxel@redhat.com>
Reviewed-by: NMax Reitz <mreitz@redhat.com>

Signed-off-by: NGerd Hoffmann <kraxel@redhat.com>
Reviewed-by: NMax Reitz <mreitz@redhat.com>

Signed-off-by: NGerd Hoffmann <kraxel@redhat.com>

Apparently it is possible for X to send an event to a hidden SDL2
window, leading to SDL2 believing it is now shown. SDL2 will pass the
SDL_WINDOWEVENT_SHOWN message to the application without actually
showing the window; the problem is that the next SDL_ShowWindow() will
be a no-op because SDL2 assumes the window is already shown.

The correct way to react to SDL_WINDOWEVENT_SHOWN would be to clear
scon->hidden (analogous for SDL_WINDOWEVENT_HIDDEN). However, due to the
window not actually being shown, this will somehow not be correct after
all.

Therefore, just hide the window on SDL_WINDOWEVENT_SHOWN if it is
supposed to be hidden (and analogous for SDL_WINDOWEVENT_HIDDEN).
Signed-off-by: NMax Reitz <mreitz@redhat.com>
Signed-off-by: NGerd Hoffmann <kraxel@redhat.com>

SDL_PollEvent() polls events for all windows; therefore,
sdl2_poll_events() will poll the events for all windows and not only for
the one identified by the given sdl2_console.

This should be considered in handle_windowevent(): The window affected
by the event is not necessarily the one identified by the sdl2_console
object given to sdl2_poll_events(), but the one identified by
ev->window.windowID.
Signed-off-by: NMax Reitz <mreitz@redhat.com>
Signed-off-by: NGerd Hoffmann <kraxel@redhat.com>

Now that common event handling code is split off, we can move
over sdl_refresh to sdl2-2d.c, and rename it to sdl2_2d_refresh.
Signed-off-by: NGerd Hoffmann <kraxel@redhat.com>
Reviewed-by: NMax Reitz <mreitz@redhat.com>

Create a new function to poll and handle sdl2 events,
which is then just called from the refresh timer.
Signed-off-by: NGerd Hoffmann <kraxel@redhat.com>
Reviewed-by: NMax Reitz <mreitz@redhat.com>

Add a new sdl2_2d_redraw function for a complete screen refresh,
so we can stop using graphic_hw_invalidate for that.  There is
no need to bother console / gfx emulation code if we are just
going to re-blit the screen after window resizes.
Signed-off-by: NGerd Hoffmann <kraxel@redhat.com>
Reviewed-by: NMax Reitz <mreitz@redhat.com>

Move sdl_switch to sdl2-2d.c file, rename to sdl2_2d_switch.
Signed-off-by: NGerd Hoffmann <kraxel@redhat.com>
Reviewed-by: NMax Reitz <mreitz@redhat.com>

Split do_sdl_resize function (which does alot more than just resizing)
into three: sdl2_window_{create,destroy,resize}.

Fix SDL_Renderer handling: must be guest display size not host window
size, and SDL2 will magically handle all scaling for us.

Make fullscreen actually enter fullscreen mode and simplify the code.
There is no need to store the original window size, the window manager
will do that for us.
Signed-off-by: NGerd Hoffmann <kraxel@redhat.com>
Reviewed-by: NMax Reitz <mreitz@redhat.com>

Create new sdl2-2d file for 2d display rendering.
Move over sdl_update code, and rename to sdl2_2d_update.
Signed-off-by: NGerd Hoffmann <kraxel@redhat.com>
Reviewed-by: NMax Reitz <mreitz@redhat.com>

Makes quite some keys actually go to the guest instead of
being captured by the host window manager.
Signed-off-by: NGerd Hoffmann <kraxel@redhat.com>
Reviewed-by: NMax Reitz <mreitz@redhat.com>

Signed-off-by: NGerd Hoffmann <kraxel@redhat.com>
Reviewed-by: NMax Reitz <mreitz@redhat.com>

Create sdl2.h header file, in preparation for sdl2 code splitup.
Populate it with sdl2_console struct (renamed from sdl2_state).
Signed-off-by: NGerd Hoffmann <kraxel@redhat.com>
Reviewed-by: NMax Reitz <mreitz@redhat.com>

Compile sdl.c / sdl2.c depending on CONFIG_SDLABI instead of
compiling both and have version #ifdefs in the source code.
Signed-off-by: NGerd Hoffmann <kraxel@redhat.com>
Reviewed-by: NMax Reitz <mreitz@redhat.com>

It isn't used.
Signed-off-by: NGerd Hoffmann <kraxel@redhat.com>

Signed-off-by: NGerd Hoffmann <kraxel@redhat.com>

Signed-off-by: NGerd Hoffmann <kraxel@redhat.com>

So the input layer knows where the input is coming from
and input routing works correctly.
Signed-off-by: NGerd Hoffmann <kraxel@redhat.com>

In SDL2, wheel movement is its own event, not a button event. Wire
it up similar to gtk.c
Signed-off-by: NCole Robinson <crobinso@redhat.com>
Signed-off-by: NGerd Hoffmann <kraxel@redhat.com>

Allows you to resize the sdl2 window and have the guest notice.

[ kraxel: zero-initialize QemuUIInfo ]
Signed-off-by: NDave Airlie <airlied@redhat.com>
Signed-off-by: NGerd Hoffmann <kraxel@redhat.com>

Right now relative mode accelerates too fast, and has the 'invisible wall'
problem. SDL2 added an explicit API to handle this use case, so let's use
it.
Signed-off-by: NCole Robinson <crobinso@redhat.com>
Signed-off-by: NGerd Hoffmann <kraxel@redhat.com>

Unbreaks relative mouse mode with sdl2, just like was done with sdl.c
in c3aa84b6.
Signed-off-by: NCole Robinson <crobinso@redhat.com>
Signed-off-by: NGerd Hoffmann <kraxel@redhat.com>

I've ported the SDL1.2 code over, and rewritten it to use the SDL2 interface.

The biggest changes were in the input handling, where SDL2 has done a major
overhaul, and I've had to include a generated translation file to get from
SDL2 codes back to qemu compatible ones. I'm still not sure how the keyboard
layout code works in qemu, so there may be further work if someone can point
me a test case that works with SDL1.2 and doesn't with SDL2.

Some SDL env vars we used to set are no longer used by SDL2,
Windows, OSX support is untested,

I don't think we can link to SDL1.2 and SDL2 at the same time, so I felt
using --with-sdlabi=2.0 to select the new code should be fine, like how
gtk does it.

v1.1: fix keys in text console
v1.2: fix shutdown, cleanups a bit of code, support ARGB cursor

v2.0: merge the SDL multihead patch into this, g_new the number of consoles
needed, wrap DCL inside per-console structure.
Signed-off-by: NDave Airlie <airlied@redhat.com>

Fixes & improvements by kraxel:
 * baum build fix
 * remove text console logic
 * adapt to new input core
 * codestyle fixups
Signed-off-by: NGerd Hoffmann <kraxel@redhat.com>