mutter/src/backends/native/meta-backend-native.c

729 lines
22 KiB
C
Raw Normal View History

/* -*- mode: C; c-file-style: "gnu"; indent-tabs-mode: nil; -*- */
/*
* Copyright (C) 2014 Red Hat
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License as
* published by the Free Software Foundation; either version 2 of the
* License, or (at your option) any later version.
*
* This program is distributed in the hope that it will be useful, but
* WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA
* 02111-1307, USA.
*
* Written by:
* Jasper St. Pierre <jstpierre@mecheye.net>
*/
2018-10-19 03:15:54 -04:00
/**
* SECTION:meta-backend-native
* @title: MetaBackendNative
* @short_description: A native (KMS/evdev) MetaBackend
*
* MetaBackendNative is an implementation of #MetaBackend that uses "native"
* technologies like DRM/KMS and libinput/evdev to perform the necessary
* functions.
*/
#include "config.h"
#include "backends/native/meta-backend-native.h"
#include "backends/native/meta-backend-native-private.h"
#include "backends/native/meta-input-thread.h"
#include <sched.h>
#include <stdlib.h>
#include "backends/meta-cursor-tracker-private.h"
#include "backends/meta-idle-monitor-private.h"
#include "backends/meta-keymap-utils.h"
#include "backends/meta-logical-monitor.h"
#include "backends/meta-monitor-manager-private.h"
#include "backends/meta-pointer-constraint.h"
#include "backends/meta-settings-private.h"
#include "backends/meta-stage-private.h"
#include "backends/native/meta-clutter-backend-native.h"
backends/native: Add basic KMS abstraction building blocks The intention with KMS abstraction is to hide away accessing the drm functions behind an API that allows us to have different kind of KMS implementations, including legacy non-atomic and atomic. The intention is also that the code interacting with the drm device should be able to be run in a different thread than the main thread. This means that we need to make sure that all drm*() API usage must only occur from within tasks that eventually can be run in the dedicated thread. The idea here is that MetaKms provides a outward facing API other places of mutter can use (e.g. MetaGpuKms and friends), while MetaKmsImpl is an internal implementation that only gets interacted with via "tasks" posted via the MetaKms object. These tasks will in the future potentially be run on the dedicated KMS thread. Initially, we don't create any new threads. Likewise, MetaKmsDevice is a outward facing representation of a KMS device, while MetaKmsImplDevice is the corresponding implementation, which only runs from within the MetaKmsImpl tasks. This commit only moves opening and closing the device to this new API, while leaking the fd outside of the impl enclosure, effectively making the isolation for drm*() calls pointless. This, however, is necessary to allow gradual porting of drm interaction, and eventually the file descriptor in MetaGpuKms will be removed. For now, it's harmless, since everything still run in the main thread. https://gitlab.gnome.org/GNOME/mutter/issues/548 https://gitlab.gnome.org/GNOME/mutter/merge_requests/525
2019-01-29 04:24:44 -05:00
#include "backends/native/meta-kms.h"
#include "backends/native/meta-kms-device.h"
#include "backends/native/meta-launcher.h"
#include "backends/native/meta-monitor-manager-native.h"
#include "backends/native/meta-renderer-native.h"
#include "backends/native/meta-seat-native.h"
Introduce regional stage rendering Add support for drawing a stage using multiple framebuffers each making up one part of the stage. This works by the stage backend (ClutterStageWindow) providing a list of views which will be for splitting up the stage in different regions. A view layout, for now, is a set of rectangles. The stage window (i.e. stage "backend" will use this information when drawing a frame, using one framebuffer for each view. The scene graph is adapted to explictly take a view when painting the stage. It will use this view, its assigned framebuffer and layout to offset and clip the drawing accordingly. This effectively removes any notion of "stage framebuffer", since each stage now may consist of multiple framebuffers. Therefore, API involving this has been deprecated and made no-ops; namely clutter_stage_ensure_context(). Callers are now assumed to either always use a framebuffer reference explicitly, or push/pop the framebuffer of a given view where the code has not yet changed to use the explicit-buffer-using cogl API. Currently only the nested X11 backend supports this mode fully, and the per view framebuffers are all offscreen. Upon frame completion, it'll blit each view's framebuffer onto the onscreen framebuffer before swapping. Other backends (X11 CM and native/KMS) are adapted to manage a full-stage view. The X11 CM backend will continue to use this method, while the native/KMS backend will be adopted to use multiple view drawing. https://bugzilla.gnome.org/show_bug.cgi?id=768976
2016-05-26 23:09:24 -04:00
#include "backends/native/meta-stage-native.h"
#include "cogl/cogl.h"
#include "core/meta-border.h"
#include "meta/main.h"
#ifdef HAVE_REMOTE_DESKTOP
#include "backends/meta-screen-cast.h"
#endif
enum
{
PROP_0,
PROP_HEADLESS,
N_PROPS
};
static GParamSpec *obj_props[N_PROPS];
struct _MetaBackendNative
{
MetaBackend parent;
MetaLauncher *launcher;
MetaUdev *udev;
backends/native: Add basic KMS abstraction building blocks The intention with KMS abstraction is to hide away accessing the drm functions behind an API that allows us to have different kind of KMS implementations, including legacy non-atomic and atomic. The intention is also that the code interacting with the drm device should be able to be run in a different thread than the main thread. This means that we need to make sure that all drm*() API usage must only occur from within tasks that eventually can be run in the dedicated thread. The idea here is that MetaKms provides a outward facing API other places of mutter can use (e.g. MetaGpuKms and friends), while MetaKmsImpl is an internal implementation that only gets interacted with via "tasks" posted via the MetaKms object. These tasks will in the future potentially be run on the dedicated KMS thread. Initially, we don't create any new threads. Likewise, MetaKmsDevice is a outward facing representation of a KMS device, while MetaKmsImplDevice is the corresponding implementation, which only runs from within the MetaKmsImpl tasks. This commit only moves opening and closing the device to this new API, while leaking the fd outside of the impl enclosure, effectively making the isolation for drm*() calls pointless. This, however, is necessary to allow gradual porting of drm interaction, and eventually the file descriptor in MetaGpuKms will be removed. For now, it's harmless, since everything still run in the main thread. https://gitlab.gnome.org/GNOME/mutter/issues/548 https://gitlab.gnome.org/GNOME/mutter/merge_requests/525
2019-01-29 04:24:44 -05:00
MetaKms *kms;
gboolean is_headless;
gulong udev_device_added_handler_id;
};
static GInitableIface *initable_parent_iface;
static void
initable_iface_init (GInitableIface *initable_iface);
G_DEFINE_TYPE_WITH_CODE (MetaBackendNative, meta_backend_native, META_TYPE_BACKEND,
G_IMPLEMENT_INTERFACE (G_TYPE_INITABLE,
initable_iface_init))
static void
disconnect_udev_device_added_handler (MetaBackendNative *native);
static void
meta_backend_native_dispose (GObject *object)
{
MetaBackendNative *native = META_BACKEND_NATIVE (object);
if (native->udev_device_added_handler_id)
{
disconnect_udev_device_added_handler (native);
native->udev_device_added_handler_id = 0;
}
backends/native: Add basic KMS abstraction building blocks The intention with KMS abstraction is to hide away accessing the drm functions behind an API that allows us to have different kind of KMS implementations, including legacy non-atomic and atomic. The intention is also that the code interacting with the drm device should be able to be run in a different thread than the main thread. This means that we need to make sure that all drm*() API usage must only occur from within tasks that eventually can be run in the dedicated thread. The idea here is that MetaKms provides a outward facing API other places of mutter can use (e.g. MetaGpuKms and friends), while MetaKmsImpl is an internal implementation that only gets interacted with via "tasks" posted via the MetaKms object. These tasks will in the future potentially be run on the dedicated KMS thread. Initially, we don't create any new threads. Likewise, MetaKmsDevice is a outward facing representation of a KMS device, while MetaKmsImplDevice is the corresponding implementation, which only runs from within the MetaKmsImpl tasks. This commit only moves opening and closing the device to this new API, while leaking the fd outside of the impl enclosure, effectively making the isolation for drm*() calls pointless. This, however, is necessary to allow gradual porting of drm interaction, and eventually the file descriptor in MetaGpuKms will be removed. For now, it's harmless, since everything still run in the main thread. https://gitlab.gnome.org/GNOME/mutter/issues/548 https://gitlab.gnome.org/GNOME/mutter/merge_requests/525
2019-01-29 04:24:44 -05:00
g_clear_object (&native->kms);
g_clear_object (&native->udev);
G_OBJECT_CLASS (meta_backend_native_parent_class)->dispose (object);
g_clear_pointer (&native->launcher, meta_launcher_free);
}
static ClutterBackend *
meta_backend_native_create_clutter_backend (MetaBackend *backend)
{
return g_object_new (META_TYPE_CLUTTER_BACKEND_NATIVE, NULL);
}
#ifdef HAVE_REMOTE_DESKTOP
static void
maybe_disable_screen_cast_dma_bufs (MetaBackendNative *native)
{
MetaBackend *backend = META_BACKEND (native);
MetaSettings *settings = meta_backend_get_settings (backend);
MetaRenderer *renderer = meta_backend_get_renderer (backend);
MetaRendererNative *renderer_native = META_RENDERER_NATIVE (renderer);
MetaScreenCast *screen_cast = meta_backend_get_screen_cast (backend);
MetaGpuKms *primary_gpu;
MetaKmsDevice *kms_device;
const char *driver_name;
static const char *enable_dma_buf_drivers[] = {
"i915",
NULL,
};
primary_gpu = meta_renderer_native_get_primary_gpu (renderer_native);
if (!primary_gpu)
{
g_message ("Disabling DMA buffer screen sharing (surfaceless)");
goto disable_dma_bufs;
}
kms_device = meta_gpu_kms_get_kms_device (primary_gpu);
driver_name = meta_kms_device_get_driver_name (kms_device);
if (g_strv_contains (enable_dma_buf_drivers, driver_name))
return;
if (meta_settings_is_experimental_feature_enabled (settings,
META_EXPERIMENTAL_FEATURE_DMA_BUF_SCREEN_SHARING))
return;
g_message ("Disabling DMA buffer screen sharing for driver '%s'.",
driver_name);
disable_dma_bufs:
meta_screen_cast_disable_dma_bufs (screen_cast);
}
#endif /* HAVE_REMOTE_DESKTOP */
static void
update_viewports (MetaBackend *backend)
{
MetaMonitorManager *monitor_manager =
meta_backend_get_monitor_manager (backend);
ClutterBackend *clutter_backend = meta_backend_get_clutter_backend (backend);
MetaSeatNative *seat =
META_SEAT_NATIVE (clutter_backend_get_default_seat (clutter_backend));
MetaViewportInfo *viewports;
viewports = meta_monitor_manager_get_viewports (monitor_manager);
meta_seat_native_set_viewports (seat, viewports);
g_object_unref (viewports);
}
static void
meta_backend_native_post_init (MetaBackend *backend)
{
MetaSettings *settings = meta_backend_get_settings (backend);
META_BACKEND_CLASS (meta_backend_native_parent_class)->post_init (backend);
if (meta_settings_is_experimental_feature_enabled (settings,
META_EXPERIMENTAL_FEATURE_RT_SCHEDULER))
{
int retval;
struct sched_param sp = {
.sched_priority = sched_get_priority_min (SCHED_RR)
};
retval = sched_setscheduler (0, SCHED_RR | SCHED_RESET_ON_FORK, &sp);
if (retval != 0)
g_warning ("Failed to set RT scheduler: %m");
}
#ifdef HAVE_REMOTE_DESKTOP
maybe_disable_screen_cast_dma_bufs (META_BACKEND_NATIVE (backend));
#endif
update_viewports (backend);
#ifdef HAVE_WAYLAND
meta_backend_init_wayland (backend);
#endif
}
static MetaMonitorManager *
meta_backend_native_create_monitor_manager (MetaBackend *backend,
GError **error)
{
MetaBackendNative *backend_native = META_BACKEND_NATIVE (backend);
MetaMonitorManager *manager;
manager = g_initable_new (META_TYPE_MONITOR_MANAGER_NATIVE, NULL, error,
"backend", backend,
"needs-outputs", !backend_native->is_headless,
NULL);
if (!manager)
return NULL;
g_signal_connect_swapped (manager, "monitors-changed-internal",
G_CALLBACK (update_viewports), backend);
return manager;
}
2014-04-22 15:15:11 -04:00
static MetaCursorRenderer *
meta_backend_native_get_cursor_renderer (MetaBackend *backend,
ClutterInputDevice *device)
2014-04-22 15:15:11 -04:00
{
ClutterBackend *clutter_backend = meta_backend_get_clutter_backend (backend);
MetaSeatNative *seat_native =
META_SEAT_NATIVE (clutter_backend_get_default_seat (clutter_backend));
return meta_seat_native_maybe_ensure_cursor_renderer (seat_native, device);
2014-04-22 15:15:11 -04:00
}
static MetaRenderer *
meta_backend_native_create_renderer (MetaBackend *backend,
GError **error)
{
MetaBackendNative *native = META_BACKEND_NATIVE (backend);
MetaRendererNative *renderer_native;
renderer_native = meta_renderer_native_new (native, error);
if (!renderer_native)
return NULL;
return META_RENDERER (renderer_native);
}
static MetaInputSettings *
meta_backend_native_get_input_settings (MetaBackend *backend)
{
ClutterBackend *clutter_backend = meta_backend_get_clutter_backend (backend);
MetaSeatNative *seat_native =
META_SEAT_NATIVE (clutter_backend_get_default_seat (clutter_backend));
return meta_seat_impl_get_input_settings (seat_native->impl);
}
static MetaLogicalMonitor *
meta_backend_native_get_current_logical_monitor (MetaBackend *backend)
{
MetaCursorTracker *cursor_tracker = meta_backend_get_cursor_tracker (backend);
MetaMonitorManager *monitor_manager =
meta_backend_get_monitor_manager (backend);
graphene_point_t point;
meta_cursor_tracker_get_pointer (cursor_tracker, &point, NULL);
return meta_monitor_manager_get_logical_monitor_at (monitor_manager,
point.x, point.y);
}
static void
meta_backend_native_set_keymap (MetaBackend *backend,
const char *layouts,
const char *variants,
const char *options)
{
ClutterBackend *clutter_backend = meta_backend_get_clutter_backend (backend);
ClutterSeat *seat;
seat = clutter_backend_get_default_seat (clutter_backend);
meta_seat_native_set_keyboard_map (META_SEAT_NATIVE (seat),
layouts, variants, options);
meta_backend_notify_keymap_changed (backend);
}
static struct xkb_keymap *
meta_backend_native_get_keymap (MetaBackend *backend)
{
ClutterBackend *clutter_backend = meta_backend_get_clutter_backend (backend);
ClutterSeat *seat;
seat = clutter_backend_get_default_seat (clutter_backend);
return meta_seat_native_get_keyboard_map (META_SEAT_NATIVE (seat));
}
static xkb_layout_index_t
meta_backend_native_get_keymap_layout_group (MetaBackend *backend)
{
ClutterBackend *clutter_backend = meta_backend_get_clutter_backend (backend);
ClutterSeat *seat;
seat = clutter_backend_get_default_seat (clutter_backend);
return meta_seat_native_get_keyboard_layout_index (META_SEAT_NATIVE (seat));
}
static void
meta_backend_native_lock_layout_group (MetaBackend *backend,
guint idx)
{
ClutterBackend *clutter_backend = meta_backend_get_clutter_backend (backend);
xkb_layout_index_t old_idx;
ClutterSeat *seat;
old_idx = meta_backend_native_get_keymap_layout_group (backend);
if (old_idx == idx)
return;
seat = clutter_backend_get_default_seat (clutter_backend);
meta_seat_native_set_keyboard_layout_index (META_SEAT_NATIVE (seat), idx);
meta_backend_notify_keymap_layout_group_changed (backend, idx);
}
const char *
meta_backend_native_get_seat_id (MetaBackendNative *backend_native)
{
if (backend_native->is_headless)
return "seat0";
else
return meta_launcher_get_seat_id (backend_native->launcher);
}
gboolean
meta_backend_native_is_headless (MetaBackendNative *backend_native)
{
return backend_native->is_headless;
}
static void
meta_backend_native_set_pointer_constraint (MetaBackend *backend,
MetaPointerConstraint *constraint)
{
ClutterBackend *clutter_backend = meta_backend_get_clutter_backend (backend);
ClutterSeat *seat = clutter_backend_get_default_seat (clutter_backend);
MetaPointerConstraintImpl *constraint_impl = NULL;
cairo_region_t *region;
if (constraint)
{
region = meta_pointer_constraint_get_region (constraint);
constraint_impl = meta_pointer_constraint_impl_native_new (constraint,
region);
}
meta_seat_native_set_pointer_constraint (META_SEAT_NATIVE (seat),
constraint_impl);
}
static void
meta_backend_native_update_screen_size (MetaBackend *backend,
int width, int height)
{
Introduce regional stage rendering Add support for drawing a stage using multiple framebuffers each making up one part of the stage. This works by the stage backend (ClutterStageWindow) providing a list of views which will be for splitting up the stage in different regions. A view layout, for now, is a set of rectangles. The stage window (i.e. stage "backend" will use this information when drawing a frame, using one framebuffer for each view. The scene graph is adapted to explictly take a view when painting the stage. It will use this view, its assigned framebuffer and layout to offset and clip the drawing accordingly. This effectively removes any notion of "stage framebuffer", since each stage now may consist of multiple framebuffers. Therefore, API involving this has been deprecated and made no-ops; namely clutter_stage_ensure_context(). Callers are now assumed to either always use a framebuffer reference explicitly, or push/pop the framebuffer of a given view where the code has not yet changed to use the explicit-buffer-using cogl API. Currently only the nested X11 backend supports this mode fully, and the per view framebuffers are all offscreen. Upon frame completion, it'll blit each view's framebuffer onto the onscreen framebuffer before swapping. Other backends (X11 CM and native/KMS) are adapted to manage a full-stage view. The X11 CM backend will continue to use this method, while the native/KMS backend will be adopted to use multiple view drawing. https://bugzilla.gnome.org/show_bug.cgi?id=768976
2016-05-26 23:09:24 -04:00
ClutterBackend *clutter_backend = meta_backend_get_clutter_backend (backend);
MetaStageNative *stage_native;
ClutterActor *stage = meta_backend_get_stage (backend);
Introduce regional stage rendering Add support for drawing a stage using multiple framebuffers each making up one part of the stage. This works by the stage backend (ClutterStageWindow) providing a list of views which will be for splitting up the stage in different regions. A view layout, for now, is a set of rectangles. The stage window (i.e. stage "backend" will use this information when drawing a frame, using one framebuffer for each view. The scene graph is adapted to explictly take a view when painting the stage. It will use this view, its assigned framebuffer and layout to offset and clip the drawing accordingly. This effectively removes any notion of "stage framebuffer", since each stage now may consist of multiple framebuffers. Therefore, API involving this has been deprecated and made no-ops; namely clutter_stage_ensure_context(). Callers are now assumed to either always use a framebuffer reference explicitly, or push/pop the framebuffer of a given view where the code has not yet changed to use the explicit-buffer-using cogl API. Currently only the nested X11 backend supports this mode fully, and the per view framebuffers are all offscreen. Upon frame completion, it'll blit each view's framebuffer onto the onscreen framebuffer before swapping. Other backends (X11 CM and native/KMS) are adapted to manage a full-stage view. The X11 CM backend will continue to use this method, while the native/KMS backend will be adopted to use multiple view drawing. https://bugzilla.gnome.org/show_bug.cgi?id=768976
2016-05-26 23:09:24 -04:00
stage_native = meta_clutter_backend_native_get_stage_native (clutter_backend);
meta_stage_native_rebuild_views (stage_native);
Introduce regional stage rendering Add support for drawing a stage using multiple framebuffers each making up one part of the stage. This works by the stage backend (ClutterStageWindow) providing a list of views which will be for splitting up the stage in different regions. A view layout, for now, is a set of rectangles. The stage window (i.e. stage "backend" will use this information when drawing a frame, using one framebuffer for each view. The scene graph is adapted to explictly take a view when painting the stage. It will use this view, its assigned framebuffer and layout to offset and clip the drawing accordingly. This effectively removes any notion of "stage framebuffer", since each stage now may consist of multiple framebuffers. Therefore, API involving this has been deprecated and made no-ops; namely clutter_stage_ensure_context(). Callers are now assumed to either always use a framebuffer reference explicitly, or push/pop the framebuffer of a given view where the code has not yet changed to use the explicit-buffer-using cogl API. Currently only the nested X11 backend supports this mode fully, and the per view framebuffers are all offscreen. Upon frame completion, it'll blit each view's framebuffer onto the onscreen framebuffer before swapping. Other backends (X11 CM and native/KMS) are adapted to manage a full-stage view. The X11 CM backend will continue to use this method, while the native/KMS backend will be adopted to use multiple view drawing. https://bugzilla.gnome.org/show_bug.cgi?id=768976
2016-05-26 23:09:24 -04:00
clutter_actor_set_size (stage, width, height);
}
static MetaGpuKms *
create_gpu_from_udev_device (MetaBackendNative *native,
GUdevDevice *device,
GError **error)
{
backends/native: Add basic KMS abstraction building blocks The intention with KMS abstraction is to hide away accessing the drm functions behind an API that allows us to have different kind of KMS implementations, including legacy non-atomic and atomic. The intention is also that the code interacting with the drm device should be able to be run in a different thread than the main thread. This means that we need to make sure that all drm*() API usage must only occur from within tasks that eventually can be run in the dedicated thread. The idea here is that MetaKms provides a outward facing API other places of mutter can use (e.g. MetaGpuKms and friends), while MetaKmsImpl is an internal implementation that only gets interacted with via "tasks" posted via the MetaKms object. These tasks will in the future potentially be run on the dedicated KMS thread. Initially, we don't create any new threads. Likewise, MetaKmsDevice is a outward facing representation of a KMS device, while MetaKmsImplDevice is the corresponding implementation, which only runs from within the MetaKmsImpl tasks. This commit only moves opening and closing the device to this new API, while leaking the fd outside of the impl enclosure, effectively making the isolation for drm*() calls pointless. This, however, is necessary to allow gradual porting of drm interaction, and eventually the file descriptor in MetaGpuKms will be removed. For now, it's harmless, since everything still run in the main thread. https://gitlab.gnome.org/GNOME/mutter/issues/548 https://gitlab.gnome.org/GNOME/mutter/merge_requests/525
2019-01-29 04:24:44 -05:00
MetaKmsDeviceFlag flags = META_KMS_DEVICE_FLAG_NONE;
const char *device_path;
backends/native: Add basic KMS abstraction building blocks The intention with KMS abstraction is to hide away accessing the drm functions behind an API that allows us to have different kind of KMS implementations, including legacy non-atomic and atomic. The intention is also that the code interacting with the drm device should be able to be run in a different thread than the main thread. This means that we need to make sure that all drm*() API usage must only occur from within tasks that eventually can be run in the dedicated thread. The idea here is that MetaKms provides a outward facing API other places of mutter can use (e.g. MetaGpuKms and friends), while MetaKmsImpl is an internal implementation that only gets interacted with via "tasks" posted via the MetaKms object. These tasks will in the future potentially be run on the dedicated KMS thread. Initially, we don't create any new threads. Likewise, MetaKmsDevice is a outward facing representation of a KMS device, while MetaKmsImplDevice is the corresponding implementation, which only runs from within the MetaKmsImpl tasks. This commit only moves opening and closing the device to this new API, while leaking the fd outside of the impl enclosure, effectively making the isolation for drm*() calls pointless. This, however, is necessary to allow gradual porting of drm interaction, and eventually the file descriptor in MetaGpuKms will be removed. For now, it's harmless, since everything still run in the main thread. https://gitlab.gnome.org/GNOME/mutter/issues/548 https://gitlab.gnome.org/GNOME/mutter/merge_requests/525
2019-01-29 04:24:44 -05:00
MetaKmsDevice *kms_device;
if (meta_is_udev_device_platform_device (device))
backends/native: Add basic KMS abstraction building blocks The intention with KMS abstraction is to hide away accessing the drm functions behind an API that allows us to have different kind of KMS implementations, including legacy non-atomic and atomic. The intention is also that the code interacting with the drm device should be able to be run in a different thread than the main thread. This means that we need to make sure that all drm*() API usage must only occur from within tasks that eventually can be run in the dedicated thread. The idea here is that MetaKms provides a outward facing API other places of mutter can use (e.g. MetaGpuKms and friends), while MetaKmsImpl is an internal implementation that only gets interacted with via "tasks" posted via the MetaKms object. These tasks will in the future potentially be run on the dedicated KMS thread. Initially, we don't create any new threads. Likewise, MetaKmsDevice is a outward facing representation of a KMS device, while MetaKmsImplDevice is the corresponding implementation, which only runs from within the MetaKmsImpl tasks. This commit only moves opening and closing the device to this new API, while leaking the fd outside of the impl enclosure, effectively making the isolation for drm*() calls pointless. This, however, is necessary to allow gradual porting of drm interaction, and eventually the file descriptor in MetaGpuKms will be removed. For now, it's harmless, since everything still run in the main thread. https://gitlab.gnome.org/GNOME/mutter/issues/548 https://gitlab.gnome.org/GNOME/mutter/merge_requests/525
2019-01-29 04:24:44 -05:00
flags |= META_KMS_DEVICE_FLAG_PLATFORM_DEVICE;
if (meta_is_udev_device_boot_vga (device))
backends/native: Add basic KMS abstraction building blocks The intention with KMS abstraction is to hide away accessing the drm functions behind an API that allows us to have different kind of KMS implementations, including legacy non-atomic and atomic. The intention is also that the code interacting with the drm device should be able to be run in a different thread than the main thread. This means that we need to make sure that all drm*() API usage must only occur from within tasks that eventually can be run in the dedicated thread. The idea here is that MetaKms provides a outward facing API other places of mutter can use (e.g. MetaGpuKms and friends), while MetaKmsImpl is an internal implementation that only gets interacted with via "tasks" posted via the MetaKms object. These tasks will in the future potentially be run on the dedicated KMS thread. Initially, we don't create any new threads. Likewise, MetaKmsDevice is a outward facing representation of a KMS device, while MetaKmsImplDevice is the corresponding implementation, which only runs from within the MetaKmsImpl tasks. This commit only moves opening and closing the device to this new API, while leaking the fd outside of the impl enclosure, effectively making the isolation for drm*() calls pointless. This, however, is necessary to allow gradual porting of drm interaction, and eventually the file descriptor in MetaGpuKms will be removed. For now, it's harmless, since everything still run in the main thread. https://gitlab.gnome.org/GNOME/mutter/issues/548 https://gitlab.gnome.org/GNOME/mutter/merge_requests/525
2019-01-29 04:24:44 -05:00
flags |= META_KMS_DEVICE_FLAG_BOOT_VGA;
if (meta_is_udev_device_requires_modifiers (device))
flags |= META_KMS_DEVICE_FLAG_REQUIRES_MODIFIERS;
if (meta_is_udev_device_preferred_primary (device))
flags |= META_KMS_DEVICE_FLAG_PREFERRED_PRIMARY;
device_path = g_udev_device_get_device_file (device);
backends/native: Add basic KMS abstraction building blocks The intention with KMS abstraction is to hide away accessing the drm functions behind an API that allows us to have different kind of KMS implementations, including legacy non-atomic and atomic. The intention is also that the code interacting with the drm device should be able to be run in a different thread than the main thread. This means that we need to make sure that all drm*() API usage must only occur from within tasks that eventually can be run in the dedicated thread. The idea here is that MetaKms provides a outward facing API other places of mutter can use (e.g. MetaGpuKms and friends), while MetaKmsImpl is an internal implementation that only gets interacted with via "tasks" posted via the MetaKms object. These tasks will in the future potentially be run on the dedicated KMS thread. Initially, we don't create any new threads. Likewise, MetaKmsDevice is a outward facing representation of a KMS device, while MetaKmsImplDevice is the corresponding implementation, which only runs from within the MetaKmsImpl tasks. This commit only moves opening and closing the device to this new API, while leaking the fd outside of the impl enclosure, effectively making the isolation for drm*() calls pointless. This, however, is necessary to allow gradual porting of drm interaction, and eventually the file descriptor in MetaGpuKms will be removed. For now, it's harmless, since everything still run in the main thread. https://gitlab.gnome.org/GNOME/mutter/issues/548 https://gitlab.gnome.org/GNOME/mutter/merge_requests/525
2019-01-29 04:24:44 -05:00
kms_device = meta_kms_create_device (native->kms, device_path, flags,
error);
if (!kms_device)
return NULL;
return meta_gpu_kms_new (native, kms_device, error);
}
static void
on_udev_device_added (MetaUdev *udev,
GUdevDevice *device,
MetaBackendNative *native)
{
MetaBackend *backend = META_BACKEND (native);
g_autoptr (GError) error = NULL;
const char *device_path;
MetaGpuKms *new_gpu_kms;
GList *gpus, *l;
if (!meta_udev_is_drm_device (udev, device))
return;
device_path = g_udev_device_get_device_file (device);
gpus = meta_backend_get_gpus (backend);
for (l = gpus; l; l = l->next)
{
MetaGpuKms *gpu_kms = l->data;
if (!g_strcmp0 (device_path, meta_gpu_kms_get_file_path (gpu_kms)))
{
g_warning ("Failed to hotplug secondary gpu '%s': %s",
device_path, "device already present");
return;
}
}
new_gpu_kms = create_gpu_from_udev_device (native, device, &error);
if (!new_gpu_kms)
{
g_warning ("Failed to hotplug secondary gpu '%s': %s",
device_path, error->message);
return;
}
meta_backend_add_gpu (backend, META_GPU (new_gpu_kms));
}
static void
connect_udev_device_added_handler (MetaBackendNative *native)
{
native->udev_device_added_handler_id =
g_signal_connect (native->udev, "device-added",
G_CALLBACK (on_udev_device_added), native);
}
static void
disconnect_udev_device_added_handler (MetaBackendNative *native)
{
g_clear_signal_handler (&native->udev_device_added_handler_id, native->udev);
}
static gboolean
init_gpus (MetaBackendNative *native,
GError **error)
{
MetaBackend *backend = META_BACKEND (native);
MetaUdev *udev = meta_backend_native_get_udev (native);
GList *devices;
GList *l;
devices = meta_udev_list_drm_devices (udev, error);
if (*error)
return FALSE;
for (l = devices; l; l = l->next)
{
GUdevDevice *device = l->data;
MetaGpuKms *gpu_kms;
GError *local_error = NULL;
gpu_kms = create_gpu_from_udev_device (native, device, &local_error);
if (!gpu_kms)
{
g_warning ("Failed to open gpu '%s': %s",
g_udev_device_get_device_file (device),
local_error->message);
g_clear_error (&local_error);
continue;
}
meta_backend_add_gpu (backend, META_GPU (gpu_kms));
}
g_list_free_full (devices, g_object_unref);
if (!native->is_headless &&
g_list_length (meta_backend_get_gpus (backend)) == 0)
{
g_set_error (error, G_IO_ERROR, G_IO_ERROR_NOT_FOUND,
"No GPUs found");
return FALSE;
}
connect_udev_device_added_handler (native);
return TRUE;
}
static gboolean
meta_backend_native_initable_init (GInitable *initable,
GCancellable *cancellable,
GError **error)
{
MetaBackendNative *native = META_BACKEND_NATIVE (initable);
MetaKmsFlags kms_flags;
if (!meta_is_stage_views_enabled ())
{
g_set_error (error, G_IO_ERROR, G_IO_ERROR_FAILED,
"The native backend requires stage views");
return FALSE;
}
if (!native->is_headless)
{
native->launcher = meta_launcher_new (error);
if (!native->launcher)
return FALSE;
}
#ifdef HAVE_WAYLAND
meta_backend_init_wayland_display (META_BACKEND (native));
#endif
native->udev = meta_udev_new (native);
kms_flags = META_KMS_FLAG_NONE;
if (native->is_headless)
kms_flags |= META_KMS_FLAG_NO_MODE_SETTING;
native->kms = meta_kms_new (META_BACKEND (native), kms_flags, error);
backends/native: Add basic KMS abstraction building blocks The intention with KMS abstraction is to hide away accessing the drm functions behind an API that allows us to have different kind of KMS implementations, including legacy non-atomic and atomic. The intention is also that the code interacting with the drm device should be able to be run in a different thread than the main thread. This means that we need to make sure that all drm*() API usage must only occur from within tasks that eventually can be run in the dedicated thread. The idea here is that MetaKms provides a outward facing API other places of mutter can use (e.g. MetaGpuKms and friends), while MetaKmsImpl is an internal implementation that only gets interacted with via "tasks" posted via the MetaKms object. These tasks will in the future potentially be run on the dedicated KMS thread. Initially, we don't create any new threads. Likewise, MetaKmsDevice is a outward facing representation of a KMS device, while MetaKmsImplDevice is the corresponding implementation, which only runs from within the MetaKmsImpl tasks. This commit only moves opening and closing the device to this new API, while leaking the fd outside of the impl enclosure, effectively making the isolation for drm*() calls pointless. This, however, is necessary to allow gradual porting of drm interaction, and eventually the file descriptor in MetaGpuKms will be removed. For now, it's harmless, since everything still run in the main thread. https://gitlab.gnome.org/GNOME/mutter/issues/548 https://gitlab.gnome.org/GNOME/mutter/merge_requests/525
2019-01-29 04:24:44 -05:00
if (!native->kms)
return FALSE;
if (!init_gpus (native, error))
return FALSE;
return initable_parent_iface->init (initable, cancellable, error);
}
static void
meta_backend_native_set_property (GObject *object,
guint prop_id,
const GValue *value,
GParamSpec *pspec)
{
MetaBackendNative *backend_native = META_BACKEND_NATIVE (object);
switch (prop_id)
{
case PROP_HEADLESS:
backend_native->is_headless = g_value_get_boolean (value);
break;
default:
G_OBJECT_WARN_INVALID_PROPERTY_ID (object, prop_id, pspec);
break;
}
}
static void
initable_iface_init (GInitableIface *initable_iface)
{
initable_parent_iface = g_type_interface_peek_parent (initable_iface);
initable_iface->init = meta_backend_native_initable_init;
}
static void
meta_backend_native_class_init (MetaBackendNativeClass *klass)
{
MetaBackendClass *backend_class = META_BACKEND_CLASS (klass);
GObjectClass *object_class = G_OBJECT_CLASS (klass);
object_class->set_property = meta_backend_native_set_property;
object_class->dispose = meta_backend_native_dispose;
backend_class->create_clutter_backend = meta_backend_native_create_clutter_backend;
backend_class->post_init = meta_backend_native_post_init;
backend_class->create_monitor_manager = meta_backend_native_create_monitor_manager;
backend_class->get_cursor_renderer = meta_backend_native_get_cursor_renderer;
backend_class->create_renderer = meta_backend_native_create_renderer;
backend_class->get_input_settings = meta_backend_native_get_input_settings;
backend_class->get_current_logical_monitor = meta_backend_native_get_current_logical_monitor;
backend_class->set_keymap = meta_backend_native_set_keymap;
backend_class->get_keymap = meta_backend_native_get_keymap;
backend_class->get_keymap_layout_group = meta_backend_native_get_keymap_layout_group;
backend_class->lock_layout_group = meta_backend_native_lock_layout_group;
backend_class->update_screen_size = meta_backend_native_update_screen_size;
backend_class->set_pointer_constraint = meta_backend_native_set_pointer_constraint;
obj_props[PROP_HEADLESS] =
g_param_spec_boolean ("headless",
"headless",
"Headless",
FALSE,
G_PARAM_WRITABLE |
G_PARAM_CONSTRUCT_ONLY |
G_PARAM_STATIC_STRINGS);
g_object_class_install_properties (object_class, N_PROPS, obj_props);
}
static void
meta_backend_native_init (MetaBackendNative *native)
{
}
MetaLauncher *
meta_backend_native_get_launcher (MetaBackendNative *native)
{
return native->launcher;
}
MetaUdev *
meta_backend_native_get_udev (MetaBackendNative *native)
{
return native->udev;
}
backends/native: Add basic KMS abstraction building blocks The intention with KMS abstraction is to hide away accessing the drm functions behind an API that allows us to have different kind of KMS implementations, including legacy non-atomic and atomic. The intention is also that the code interacting with the drm device should be able to be run in a different thread than the main thread. This means that we need to make sure that all drm*() API usage must only occur from within tasks that eventually can be run in the dedicated thread. The idea here is that MetaKms provides a outward facing API other places of mutter can use (e.g. MetaGpuKms and friends), while MetaKmsImpl is an internal implementation that only gets interacted with via "tasks" posted via the MetaKms object. These tasks will in the future potentially be run on the dedicated KMS thread. Initially, we don't create any new threads. Likewise, MetaKmsDevice is a outward facing representation of a KMS device, while MetaKmsImplDevice is the corresponding implementation, which only runs from within the MetaKmsImpl tasks. This commit only moves opening and closing the device to this new API, while leaking the fd outside of the impl enclosure, effectively making the isolation for drm*() calls pointless. This, however, is necessary to allow gradual porting of drm interaction, and eventually the file descriptor in MetaGpuKms will be removed. For now, it's harmless, since everything still run in the main thread. https://gitlab.gnome.org/GNOME/mutter/issues/548 https://gitlab.gnome.org/GNOME/mutter/merge_requests/525
2019-01-29 04:24:44 -05:00
MetaKms *
meta_backend_native_get_kms (MetaBackendNative *native)
{
return native->kms;
}
gboolean
meta_activate_vt (int vt, GError **error)
{
MetaBackend *backend = meta_get_backend ();
MetaBackendNative *native = META_BACKEND_NATIVE (backend);
MetaLauncher *launcher = meta_backend_native_get_launcher (native);
if (native->is_headless)
{
g_set_error (error, G_IO_ERROR, G_IO_ERROR_FAILED,
"Can't switch VT while headless");
return FALSE;
}
return meta_launcher_activate_vt (launcher, vt, error);
}
void
meta_backend_native_pause (MetaBackendNative *native)
{
MetaBackend *backend = META_BACKEND (native);
MetaMonitorManager *monitor_manager =
meta_backend_get_monitor_manager (backend);
MetaMonitorManagerNative *monitor_manager_native =
META_MONITOR_MANAGER_NATIVE (monitor_manager);
ClutterBackend *clutter_backend = meta_backend_get_clutter_backend (backend);
MetaSeatNative *seat =
META_SEAT_NATIVE (clutter_backend_get_default_seat (clutter_backend));
MetaRenderer *renderer = meta_backend_get_renderer (backend);
COGL_TRACE_BEGIN_SCOPED (MetaBackendNativePause,
"Backend (pause)");
meta_seat_native_release_devices (seat);
meta_renderer_pause (renderer);
disconnect_udev_device_added_handler (native);
meta_monitor_manager_native_pause (monitor_manager_native);
}
void meta_backend_native_resume (MetaBackendNative *native)
{
MetaBackend *backend = META_BACKEND (native);
ClutterStage *stage = CLUTTER_STAGE (meta_backend_get_stage (backend));
MetaMonitorManager *monitor_manager =
meta_backend_get_monitor_manager (backend);
MetaMonitorManagerNative *monitor_manager_native =
META_MONITOR_MANAGER_NATIVE (monitor_manager);
MetaIdleMonitor *idle_monitor;
ClutterBackend *clutter_backend = meta_backend_get_clutter_backend (backend);
MetaSeatNative *seat =
META_SEAT_NATIVE (clutter_backend_get_default_seat (clutter_backend));
MetaRenderer *renderer = meta_backend_get_renderer (backend);
MetaInputSettings *input_settings;
COGL_TRACE_BEGIN_SCOPED (MetaBackendNativeResume,
"Backend (resume)");
meta_monitor_manager_native_resume (monitor_manager_native);
meta_kms_resume (native->kms);
connect_udev_device_added_handler (native);
meta_seat_native_reclaim_devices (seat);
meta_renderer_resume (renderer);
clutter_actor_queue_redraw (CLUTTER_ACTOR (stage));
idle_monitor = meta_idle_monitor_get_core ();
meta_idle_monitor_reset_idletime (idle_monitor);
input_settings = meta_backend_get_input_settings (backend);
meta_input_settings_maybe_restore_numlock_state (input_settings);
clutter_seat_ensure_a11y_state (CLUTTER_SEAT (seat));
}