Add a patch fixing issues when using cards which default to the radeon
kms driver with the amdgpu kms driver (rhbz#1490490) Extend default DeviceTimeout to 8 seconds (rhbz#1737221)
This commit is contained in:
Hans de Goede
parent
3796291739
commit
2e875c999d
@ -670,3 +670,201 @@ index 795bded..2e596d5 100644
|
|||||||
--
|
--
|
||||||
2.21.0
|
2.21.0
|
||||||
|
|
||||||
|
From 4b3be154867ddfc3315a39e57d23a374142c7d2f Mon Sep 17 00:00:00 2001
|
||||||
|
From: Hans de Goede <hdegoede@redhat.com>
|
||||||
|
Date: Sun, 25 Aug 2019 18:01:59 +0200
|
||||||
|
Subject: [PATCH 1/3] two-step: bgrt: Add workaround for desktops which do not
|
||||||
|
use the golden ratio
|
||||||
|
|
||||||
|
On desktops (no panel) we normally do not use the BGRT provided
|
||||||
|
xoffset and yoffset because the resolution they are intended for
|
||||||
|
may be differtent then the resolution of the current display.
|
||||||
|
|
||||||
|
On some desktops (no panel) the image gets centered not only
|
||||||
|
horizontally, but also vertically. In this case our default of using
|
||||||
|
the golden ratio for the vertical position causes the BGRT image
|
||||||
|
to jump.
|
||||||
|
|
||||||
|
To avoid this this commits adds an extra check to see if the provided
|
||||||
|
xoffset and yoffset perfectly center the image and in that case uses them.
|
||||||
|
|
||||||
|
An example of a system needing this workaround is the Minix Neo Z83-4.
|
||||||
|
|
||||||
|
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
|
||||||
|
---
|
||||||
|
src/plugins/splash/two-step/plugin.c | 18 ++++++++++++++++++
|
||||||
|
1 file changed, 18 insertions(+)
|
||||||
|
|
||||||
|
diff --git a/src/plugins/splash/two-step/plugin.c b/src/plugins/splash/two-step/plugin.c
|
||||||
|
index 2e596d5..b5ef926 100644
|
||||||
|
--- a/src/plugins/splash/two-step/plugin.c
|
||||||
|
+++ b/src/plugins/splash/two-step/plugin.c
|
||||||
|
@@ -506,6 +506,24 @@ view_set_bgrt_background (view_t *view)
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
|
+ /*
|
||||||
|
+ * On desktops (no panel) we normally do not use the BGRT provided
|
||||||
|
+ * xoffset and yoffset because the resolution they are intended for
|
||||||
|
+ * may be differtent then the resolution of the current display.
|
||||||
|
+ *
|
||||||
|
+ * On some desktops (no panel) the image gets centered not only
|
||||||
|
+ * horizontally, but also vertically. In this case our default of using
|
||||||
|
+ * the golden ratio for the vertical position causes the BGRT image
|
||||||
|
+ * to jump. To avoid this we check here if the provided xoffset and
|
||||||
|
+ * yoffset perfectly center the image and in that case we use them.
|
||||||
|
+ */
|
||||||
|
+ if (!have_panel_props && screen_scale == 1 &&
|
||||||
|
+ (screen_width - width ) / 2 == sysfs_x_offset &&
|
||||||
|
+ (screen_height - height) / 2 == sysfs_y_offset) {
|
||||||
|
+ x_offset = sysfs_x_offset;
|
||||||
|
+ y_offset = sysfs_y_offset;
|
||||||
|
+ }
|
||||||
|
+
|
||||||
|
ply_trace ("using %dx%d bgrt image centered at %dx%d for %dx%d screen",
|
||||||
|
width, height, x_offset, y_offset, screen_width, screen_height);
|
||||||
|
|
||||||
|
--
|
||||||
|
2.23.0
|
||||||
|
|
||||||
|
From 7bd45c782eb7a76005ab7f92de4b51fc2ecb8068 Mon Sep 17 00:00:00 2001
|
||||||
|
From: Hans de Goede <hdegoede@redhat.com>
|
||||||
|
Date: Fri, 6 Sep 2019 17:21:45 +0200
|
||||||
|
Subject: [PATCH 2/3] ply-device-manager: De-activate and close renderer on
|
||||||
|
device hot unplug
|
||||||
|
|
||||||
|
When a device gets hot unplugged, we should de-activate and close the
|
||||||
|
renderer before freeing it.
|
||||||
|
|
||||||
|
Speficially this fixes a problem where plymout's display "freezes" when
|
||||||
|
used with AMD GPUs which can be handled by both the radeon or the amdgpu
|
||||||
|
kms driver and the user has added kernel commandline options to prefer
|
||||||
|
the amdgpu driver. In this case the following happens:
|
||||||
|
|
||||||
|
1) radeon driver gets loaded, creates /dev/dri/card0 generating a udev
|
||||||
|
"add /dev/dri/card0" event.
|
||||||
|
2) radeon driver's "load" method checks if it should bind, comes to the
|
||||||
|
conclusion it should not, causing the probe of the device to be
|
||||||
|
aborted and /dev/dri/card0 to be unregistered generating a udev
|
||||||
|
"remove /dev/dri/card0" event
|
||||||
|
3) amdgpu driver loads, creates /dev/dri/card0 generating another udev
|
||||||
|
"add /dev/dri/card0" event.
|
||||||
|
4) plymouth loads while udev is still catching up with kernel events,
|
||||||
|
plymouth sees the first "add" event opens /dev/dri/card0 (which is
|
||||||
|
already managed by amdgpu at this time) finds outputs, shows splash
|
||||||
|
5) plymouth gets the "remove" events tears down the renderer, but
|
||||||
|
since deactivate and close were not called before this commit,
|
||||||
|
plymouth keeps the master rights attached to the old-fd which is
|
||||||
|
leaked at this point
|
||||||
|
6) plymouth gets the second "add" event brings everything back up, but
|
||||||
|
it cannot get master rights, since there can only be 1 master and the
|
||||||
|
leaked fd is the master. This leads to lots of:
|
||||||
|
"Couldn't set scan out buffer" errors and plymouth appears frozen
|
||||||
|
because of this, this is esp. bad with the disk encryption unlock screen
|
||||||
|
7) When gdm starts, it deactivates plymouth, but does not yet tell it to quit
|
||||||
|
to avoid flickering. The deactivate makes plymouth drop its master rights,
|
||||||
|
but this happens on the new broken fd which never got the master rights.
|
||||||
|
This also makes it impossible for gdm to get master rights, also breaking
|
||||||
|
gdm, with the following error in the logs:
|
||||||
|
"(EE) wl_drm@4: error 0: authenicate failed"
|
||||||
|
|
||||||
|
Note in theory calling ply_renderer_close() on teardown (from
|
||||||
|
ply_device_manager_free()) should be fine too, but for some reason
|
||||||
|
this causes the framebuffer to be all messed up (looks like wrong pitch /
|
||||||
|
tiling) when this is done after gdm has taken over the framebuffer.
|
||||||
|
|
||||||
|
Where as just exiting without calling drmModeRmFB, letting the kernel
|
||||||
|
clean up behind us, does not trigger this. This feels like it is a kernel
|
||||||
|
bug bug and I will file a bug for this. But for now we should not call
|
||||||
|
ply_renderer_close() on teardown to avoid this issue (which would be a
|
||||||
|
regression).
|
||||||
|
|
||||||
|
BugLink: https://bugzilla.redhat.com/show_bug.cgi?id=1490490
|
||||||
|
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
|
||||||
|
---
|
||||||
|
src/libply-splash-core/ply-device-manager.c | 14 +++++++++++---
|
||||||
|
1 file changed, 11 insertions(+), 3 deletions(-)
|
||||||
|
|
||||||
|
diff --git a/src/libply-splash-core/ply-device-manager.c b/src/libply-splash-core/ply-device-manager.c
|
||||||
|
index e2a9eae..160b8cb 100644
|
||||||
|
--- a/src/libply-splash-core/ply-device-manager.c
|
||||||
|
+++ b/src/libply-splash-core/ply-device-manager.c
|
||||||
|
@@ -169,7 +169,8 @@ free_keyboards_for_renderer (ply_device_manager_t *manager,
|
||||||
|
|
||||||
|
static void
|
||||||
|
free_devices_from_device_path (ply_device_manager_t *manager,
|
||||||
|
- const char *device_path)
|
||||||
|
+ const char *device_path,
|
||||||
|
+ bool close)
|
||||||
|
{
|
||||||
|
void *key = NULL;
|
||||||
|
void *renderer = NULL;
|
||||||
|
@@ -187,6 +188,13 @@ free_devices_from_device_path (ply_device_manager_t *manager,
|
||||||
|
|
||||||
|
ply_hashtable_remove (manager->renderers, (void *) device_path);
|
||||||
|
free (key);
|
||||||
|
+
|
||||||
|
+ if (manager->renderers_activated)
|
||||||
|
+ ply_renderer_deactivate (renderer);
|
||||||
|
+
|
||||||
|
+ if (close)
|
||||||
|
+ ply_renderer_close (renderer);
|
||||||
|
+
|
||||||
|
ply_renderer_free (renderer);
|
||||||
|
}
|
||||||
|
|
||||||
|
@@ -309,7 +317,7 @@ free_devices_for_udev_device (ply_device_manager_t *manager,
|
||||||
|
device_path = udev_device_get_devnode (device);
|
||||||
|
|
||||||
|
if (device_path != NULL)
|
||||||
|
- free_devices_from_device_path (manager, device_path);
|
||||||
|
+ free_devices_from_device_path (manager, device_path, true);
|
||||||
|
}
|
||||||
|
|
||||||
|
static bool
|
||||||
|
@@ -557,7 +565,7 @@ free_renderer (char *device_path,
|
||||||
|
ply_renderer_t *renderer,
|
||||||
|
ply_device_manager_t *manager)
|
||||||
|
{
|
||||||
|
- free_devices_from_device_path (manager, device_path);
|
||||||
|
+ free_devices_from_device_path (manager, device_path, false);
|
||||||
|
}
|
||||||
|
|
||||||
|
static void
|
||||||
|
--
|
||||||
|
2.23.0
|
||||||
|
|
||||||
|
From 443415375a58f2f58af9991a55a8f4aaff83a7c9 Mon Sep 17 00:00:00 2001
|
||||||
|
From: Hans de Goede <hdegoede@redhat.com>
|
||||||
|
Date: Fri, 6 Sep 2019 17:40:06 +0200
|
||||||
|
Subject: [PATCH 3/3] plymouthd.defaults: Bump default DeviceTimeout to 8
|
||||||
|
seconds
|
||||||
|
|
||||||
|
The amdgpu driver needs up to 5 seconds to initialize / boot some
|
||||||
|
AMD GPUs, which makes our 5 second timeout much too close to the actual
|
||||||
|
needed time, leading to systems where sometimes the boot is fine and the
|
||||||
|
next boot we fallback to a text based splash because we won the race from
|
||||||
|
the GPU probe code.
|
||||||
|
|
||||||
|
This commit bumps the default DeviceTimeout to 8 seconds giving us a
|
||||||
|
decent marging here without making it overly long.
|
||||||
|
|
||||||
|
BugLink: https://bugzilla.redhat.com/show_bug.cgi?id=1737221
|
||||||
|
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
|
||||||
|
---
|
||||||
|
src/plymouthd.defaults | 2 +-
|
||||||
|
1 file changed, 1 insertion(+), 1 deletion(-)
|
||||||
|
|
||||||
|
diff --git a/src/plymouthd.defaults b/src/plymouthd.defaults
|
||||||
|
index 9e3a342..7f247fd 100644
|
||||||
|
--- a/src/plymouthd.defaults
|
||||||
|
+++ b/src/plymouthd.defaults
|
||||||
|
@@ -3,4 +3,4 @@
|
||||||
|
[Daemon]
|
||||||
|
Theme=spinner
|
||||||
|
ShowDelay=0
|
||||||
|
-DeviceTimeout=5
|
||||||
|
+DeviceTimeout=8
|
||||||
|
--
|
||||||
|
2.23.0
|
||||||
|
|
||||||
|
|||||||
@ -1,7 +1,7 @@
|
|||||||
Summary: Graphical Boot Animation and Logger
|
Summary: Graphical Boot Animation and Logger
|
||||||
Name: plymouth
|
Name: plymouth
|
||||||
Version: 0.9.4
|
Version: 0.9.4
|
||||||
Release: 8%{?dist}
|
Release: 9%{?dist}
|
||||||
License: GPLv2+
|
License: GPLv2+
|
||||||
URL: http://www.freedesktop.org/wiki/Software/Plymouth
|
URL: http://www.freedesktop.org/wiki/Software/Plymouth
|
||||||
|
|
||||||
@ -415,6 +415,11 @@ fi
|
|||||||
|
|
||||||
|
|
||||||
%changelog
|
%changelog
|
||||||
|
* Sat Sep 7 2019 Hans de Goede <jwrdegoede@fedoraproject.org> - 0.9.4-9
|
||||||
|
- Add a patch fixing issues when using cards which default to the radeon
|
||||||
|
kms driver with the amdgpu kms driver (rhbz#1490490)
|
||||||
|
- Extend default DeviceTimeout to 8 seconds (rhbz#1737221)
|
||||||
|
|
||||||
* Fri Jul 26 2019 Fedora Release Engineering <releng@fedoraproject.org> - 0.9.4-8
|
* Fri Jul 26 2019 Fedora Release Engineering <releng@fedoraproject.org> - 0.9.4-8
|
||||||
- Rebuilt for https://fedoraproject.org/wiki/Fedora_31_Mass_Rebuild
|
- Rebuilt for https://fedoraproject.org/wiki/Fedora_31_Mass_Rebuild
|
||||||
|
|
||||||
|
|||||||
Loading…
Reference in New Issue
Block a user