[Bug 1882851] [NEW] QEMU video freezes with "Guest disabled display" (virtio driver)

[Bug 1882851] [NEW] QEMU video freezes with "Guest disabled display" (virtio driver)

[Diego Viola]

Public bug reported:

I am using Arch Linux as my Guest and Host OS, after starting qemu with
the following command:

  $ qemu-system-x86_64 -enable-kvm -hda arch-zoom.qcow2 -m 4G -vga
virtio

and waiting for a screen blank, I get this message:

  Guest disabled display

And nothing happens after that, I can move the mouse or hit any key, and
the message is still there.

I can still reboot the VM but that's not optimal.

I can reproduce this with the latest QEMU release (5.0.0) or git master, 
I also tried this with older releases (4.0.0, 3.0.0) and the issue is still there.

I can't reproduce this with other video drivers (std, qxl).

With std/qxl the screen will blank a bit and then continue as normal.

** Affects: qemu
     Importance: Undecided
         Status: New

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1882851

Title:
  QEMU video freezes with "Guest disabled display" (virtio driver)

Status in QEMU:
  New

Bug description:
  I am using Arch Linux as my Guest and Host OS, after starting qemu
  with the following command:

    $ qemu-system-x86_64 -enable-kvm -hda arch-zoom.qcow2 -m 4G -vga
  virtio

  and waiting for a screen blank, I get this message:

    Guest disabled display

  And nothing happens after that, I can move the mouse or hit any key,
  and the message is still there.

  I can still reboot the VM but that's not optimal.

  I can reproduce this with the latest QEMU release (5.0.0) or git master, 
  I also tried this with older releases (4.0.0, 3.0.0) and the issue is still there.

  I can't reproduce this with other video drivers (std, qxl).

  With std/qxl the screen will blank a bit and then continue as normal.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1882851/+subscriptions

[Bug 1882851] Re: QEMU video freezes with "Guest disabled display" (virtio driver)

[Diego Viola]

OK, I found a workaround: sendkey ctrl-alt-f1 from the QEMU console
(ctrl alt 2) then I can switch back to X and continue from where I left
off.

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1882851

Title:
  QEMU video freezes with "Guest disabled display" (virtio driver)

Status in QEMU:
  New

Bug description:
  I am using Arch Linux as my Guest and Host OS, after starting qemu
  with the following command:

    $ qemu-system-x86_64 -enable-kvm -hda arch-zoom.qcow2 -m 4G -vga
  virtio

  and waiting for a screen blank, I get this message:

    Guest disabled display

  And nothing happens after that, I can move the mouse or hit any key,
  and the message is still there.

  I can still reboot the VM but that's not optimal.

  I can reproduce this with the latest QEMU release (5.0.0) or git master, 
  I also tried this with older releases (4.0.0, 3.0.0) and the issue is still there.

  I can't reproduce this with other video drivers (std, qxl).

  With std/qxl the screen will blank a bit and then continue as normal.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1882851/+subscriptions

[Bug 1882851] Re: QEMU video freezes with "Guest disabled display" (virtio driver)

[Diego Viola]

Strange, that workaround doesn't work anymore.

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1882851

Title:
  QEMU video freezes with "Guest disabled display" (virtio driver)

Status in QEMU:
  New

Bug description:
  I am using Arch Linux as my Guest and Host OS, after starting qemu
  with the following command:

    $ qemu-system-x86_64 -enable-kvm -hda arch-zoom.qcow2 -m 4G -vga
  virtio

  and waiting for a screen blank, I get this message:

    Guest disabled display

  And nothing happens after that, I can move the mouse or hit any key,
  and the message is still there.

  I can still reboot the VM but that's not optimal.

  I can reproduce this with the latest QEMU release (5.0.0) or git master, 
  I also tried this with older releases (4.0.0, 3.0.0) and the issue is still there.

  I can't reproduce this with other video drivers (std, qxl).

  With std/qxl the screen will blank a bit and then continue as normal.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1882851/+subscriptions

[Bug 1882851] Re: QEMU video freezes with "Guest disabled display" (virtio driver)

[Diego Viola]

My bad, the workaround works, it's just a bit tricky.

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1882851

Title:
  QEMU video freezes with "Guest disabled display" (virtio driver)

Status in QEMU:
  New

Bug description:
  I am using Arch Linux as my Guest and Host OS, after starting qemu
  with the following command:

    $ qemu-system-x86_64 -enable-kvm -hda arch-zoom.qcow2 -m 4G -vga
  virtio

  and waiting for a screen blank, I get this message:

    Guest disabled display

  And nothing happens after that, I can move the mouse or hit any key,
  and the message is still there.

  I can still reboot the VM but that's not optimal.

  I can reproduce this with the latest QEMU release (5.0.0) or git master, 
  I also tried this with older releases (4.0.0, 3.0.0) and the issue is still there.

  I can't reproduce this with other video drivers (std, qxl).

  With std/qxl the screen will blank a bit and then continue as normal.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1882851/+subscriptions

[Bug 1882851] Re: QEMU video freezes with "Guest disabled display" (virtio driver)

[Diego Viola]

`xset dpms force off' on the guest is a good way to reproduce it.

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1882851

Title:
  QEMU video freezes with "Guest disabled display" (virtio driver)

Status in QEMU:
  New

Bug description:
  I am using Arch Linux as my Guest and Host OS, after starting qemu
  with the following command:

    $ qemu-system-x86_64 -enable-kvm -hda arch-zoom.qcow2 -m 4G -vga
  virtio

  and waiting for a screen blank, I get this message:

    Guest disabled display

  And nothing happens after that, I can move the mouse or hit any key,
  and the message is still there.

  I can still reboot the VM but that's not optimal.

  I can reproduce this with the latest QEMU release (5.0.0) or git master, 
  I also tried this with older releases (4.0.0, 3.0.0) and the issue is still there.

  I can't reproduce this with other video drivers (std, qxl).

  With std/qxl the screen will blank a bit and then continue as normal.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1882851/+subscriptions

[Bug 1882851] Re: QEMU video freezes with "Guest disabled display" (virtio driver)

[Gerd Hoffmann]

Hmm, happens with xorg only.
Wayland behaves as expected (any mouse/kbd event wakes up the screen).
Which pretty much implies this is a guest bug.

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1882851

Title:
  QEMU video freezes with "Guest disabled display" (virtio driver)

Status in QEMU:
  New

Bug description:
  I am using Arch Linux as my Guest and Host OS, after starting qemu
  with the following command:

    $ qemu-system-x86_64 -enable-kvm -hda arch-zoom.qcow2 -m 4G -vga
  virtio

  and waiting for a screen blank, I get this message:

    Guest disabled display

  And nothing happens after that, I can move the mouse or hit any key,
  and the message is still there.

  I can still reboot the VM but that's not optimal.

  I can reproduce this with the latest QEMU release (5.0.0) or git master, 
  I also tried this with older releases (4.0.0, 3.0.0) and the issue is still there.

  I can't reproduce this with other video drivers (std, qxl).

  With std/qxl the screen will blank a bit and then continue as normal.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1882851/+subscriptions

[Bug 1882851] Re: QEMU video freezes with "Guest disabled display" (virtio driver)

[Diego Viola]

> Hmm, happens with xorg only.

Nope, I can reproduce it with sway as well (which is another Wayland
compositor).

To reproduce it with sway, just do: swaymsg "output * dpms off" and then
should you see "Guest disabled display", at that point I'm unable to get
back image.

I tried the sendkey ctrl-alt-f2 and then switch back to TTY1 but the
"Guest disabled display" remains.

Can you please tell me which compositor you used?

Thanks.

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1882851

Title:
  QEMU video freezes with "Guest disabled display" (virtio driver)

Status in QEMU:
  New

Bug description:
  I am using Arch Linux as my Guest and Host OS, after starting qemu
  with the following command:

    $ qemu-system-x86_64 -enable-kvm -hda arch-zoom.qcow2 -m 4G -vga
  virtio

  and waiting for a screen blank, I get this message:

    Guest disabled display

  And nothing happens after that, I can move the mouse or hit any key,
  and the message is still there.

  I can still reboot the VM but that's not optimal.

  I can reproduce this with the latest QEMU release (5.0.0) or git master, 
  I also tried this with older releases (4.0.0, 3.0.0) and the issue is still there.

  I can't reproduce this with other video drivers (std, qxl).

  With std/qxl the screen will blank a bit and then continue as normal.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1882851/+subscriptions

[Bug 1882851] Re: QEMU video freezes with "Guest disabled display" (virtio driver)

[Diego Viola]

I can't wake up the screen after hitting keys or moving the mouse after
turning off the screen with sway.

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1882851

Title:
  QEMU video freezes with "Guest disabled display" (virtio driver)

Status in QEMU:
  New

Bug description:
  I am using Arch Linux as my Guest and Host OS, after starting qemu
  with the following command:

    $ qemu-system-x86_64 -enable-kvm -hda arch-zoom.qcow2 -m 4G -vga
  virtio

  and waiting for a screen blank, I get this message:

    Guest disabled display

  And nothing happens after that, I can move the mouse or hit any key,
  and the message is still there.

  I can still reboot the VM but that's not optimal.

  I can reproduce this with the latest QEMU release (5.0.0) or git master, 
  I also tried this with older releases (4.0.0, 3.0.0) and the issue is still there.

  I can't reproduce this with other video drivers (std, qxl).

  With std/qxl the screen will blank a bit and then continue as normal.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1882851/+subscriptions

[Bug 1882851] Re: QEMU video freezes with "Guest disabled display" (virtio driver)

[Diego Viola]

Gerd: I tried the LTS kernel on Arch Linux (5.4.46-1-lts) and I can't
reproduce the bug with this kernel.

It works as expected: `xset dpms force off' triggers the "Guest disabled
display" and it disappears after moving the mouse.

Could it be a regression in virtio_gpu?

I guess I'll try the latest linux git and if it's an issue in master,
I'll bisect it.

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1882851

Title:
  QEMU video freezes with "Guest disabled display" (virtio driver)

Status in QEMU:
  New

Bug description:
  I am using Arch Linux as my Guest and Host OS, after starting qemu
  with the following command:

    $ qemu-system-x86_64 -enable-kvm -hda arch-zoom.qcow2 -m 4G -vga
  virtio

  and waiting for a screen blank, I get this message:

    Guest disabled display

  And nothing happens after that, I can move the mouse or hit any key,
  and the message is still there.

  I can still reboot the VM but that's not optimal.

  I can reproduce this with the latest QEMU release (5.0.0) or git master, 
  I also tried this with older releases (4.0.0, 3.0.0) and the issue is still there.

  I can't reproduce this with other video drivers (std, qxl).

  With std/qxl the screen will blank a bit and then continue as normal.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1882851/+subscriptions

[Bug 1882851] Re: QEMU video freezes with "Guest disabled display" (virtio driver)

[Diego Viola]

I can reproduce it with current linux git master[1].

1. git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1882851

Title:
  QEMU video freezes with "Guest disabled display" (virtio driver)

Status in QEMU:
  New

Bug description:
  I am using Arch Linux as my Guest and Host OS, after starting qemu
  with the following command:

    $ qemu-system-x86_64 -enable-kvm -hda arch-zoom.qcow2 -m 4G -vga
  virtio

  and waiting for a screen blank, I get this message:

    Guest disabled display

  And nothing happens after that, I can move the mouse or hit any key,
  and the message is still there.

  I can still reboot the VM but that's not optimal.

  I can reproduce this with the latest QEMU release (5.0.0) or git master, 
  I also tried this with older releases (4.0.0, 3.0.0) and the issue is still there.

  I can't reproduce this with other video drivers (std, qxl).

  With std/qxl the screen will blank a bit and then continue as normal.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1882851/+subscriptions

[Bug 1882851] Re: QEMU video freezes with "Guest disabled display" (virtio driver)

[Diego Viola]

OK, I was able to bisect, here is the result:

[diego@arch-zoom linux]$ git bisect bad
3954ff10e06e4fbc548fc02ff1fcaaac3228fed5 is the first bad commit
commit 3954ff10e06e4fbc548fc02ff1fcaaac3228fed5
Author: Gerd Hoffmann <kraxel@redhat.com>
Date:   Thu Dec 12 13:53:44 2019 +0100

    drm/virtio: skip set_scanout if framebuffer didn't change

    v2: also check src rect (Chia-I Wu).

    Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
    Reviewed-by: Chia-I Wu <olvaffe@gmail.com>
    Link: http://patchwork.freedesktop.org/patch/msgid/20191212125346.8334-2-kraxel@redhat.com

 drivers/gpu/drm/virtio/virtgpu_plane.c | 35 ++++++++++++++++++++--------------
 1 file changed, 21 insertions(+), 14 deletions(-)
[diego@arch-zoom linux]$

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1882851

Title:
  QEMU video freezes with "Guest disabled display" (virtio driver)

Status in QEMU:
  New

Bug description:
  I am using Arch Linux as my Guest and Host OS, after starting qemu
  with the following command:

    $ qemu-system-x86_64 -enable-kvm -hda arch-zoom.qcow2 -m 4G -vga
  virtio

  and waiting for a screen blank, I get this message:

    Guest disabled display

  And nothing happens after that, I can move the mouse or hit any key,
  and the message is still there.

  I can still reboot the VM but that's not optimal.

  I can reproduce this with the latest QEMU release (5.0.0) or git master, 
  I also tried this with older releases (4.0.0, 3.0.0) and the issue is still there.

  I can't reproduce this with other video drivers (std, qxl).

  With std/qxl the screen will blank a bit and then continue as normal.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1882851/+subscriptions

[Bug 1882851] Re: QEMU video freezes with "Guest disabled display" (virtio driver)

[Diego Viola]

** Attachment added: "bisectlog.txt"
   https://bugs.launchpad.net/qemu/+bug/1882851/+attachment/5383160/+files/bisectlog.txt

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1882851

Title:
  QEMU video freezes with "Guest disabled display" (virtio driver)

Status in QEMU:
  New

Bug description:
  I am using Arch Linux as my Guest and Host OS, after starting qemu
  with the following command:

    $ qemu-system-x86_64 -enable-kvm -hda arch-zoom.qcow2 -m 4G -vga
  virtio

  and waiting for a screen blank, I get this message:

    Guest disabled display

  And nothing happens after that, I can move the mouse or hit any key,
  and the message is still there.

  I can still reboot the VM but that's not optimal.

  I can reproduce this with the latest QEMU release (5.0.0) or git master, 
  I also tried this with older releases (4.0.0, 3.0.0) and the issue is still there.

  I can't reproduce this with other video drivers (std, qxl).

  With std/qxl the screen will blank a bit and then continue as normal.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1882851/+subscriptions

[Bug 1882851] Re: QEMU video freezes with "Guest disabled display" (virtio driver)

[Diego Viola]

I just sanity checked, and can confirm the bad commit causes it, and
going back one commit makes it work.

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1882851

Title:
  QEMU video freezes with "Guest disabled display" (virtio driver)

Status in QEMU:
  New

Bug description:
  I am using Arch Linux as my Guest and Host OS, after starting qemu
  with the following command:

    $ qemu-system-x86_64 -enable-kvm -hda arch-zoom.qcow2 -m 4G -vga
  virtio

  and waiting for a screen blank, I get this message:

    Guest disabled display

  And nothing happens after that, I can move the mouse or hit any key,
  and the message is still there.

  I can still reboot the VM but that's not optimal.

  I can reproduce this with the latest QEMU release (5.0.0) or git master, 
  I also tried this with older releases (4.0.0, 3.0.0) and the issue is still there.

  I can't reproduce this with other video drivers (std, qxl).

  With std/qxl the screen will blank a bit and then continue as normal.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1882851/+subscriptions

[PATCH] drm/virtio: fix unblank

[Gerd Hoffmann]

When going through a disable/enable cycle without changing the framebuffer
the optimization added by commit 3954ff10e06e causes the screen stay
blank.  Add a bool to force an update to fix that.

Cc: 1882851@bugs.launchpad.net
Fixes: 3954ff10e06e ("drm/virtio: skip set_scanout if framebuffer didn't change")
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
---
 drivers/gpu/drm/virtio/virtgpu_drv.h     | 1 +
 drivers/gpu/drm/virtio/virtgpu_display.c | 1 +
 drivers/gpu/drm/virtio/virtgpu_plane.c   | 4 +++-
 3 files changed, 5 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h
index 7879ff58236f..6d5410d5dd84 100644
--- a/drivers/gpu/drm/virtio/virtgpu_drv.h
+++ b/drivers/gpu/drm/virtio/virtgpu_drv.h
@@ -138,6 +138,7 @@ struct virtio_gpu_output {
 	int cur_x;
 	int cur_y;
 	bool enabled;
+	bool need_update;
 };
 #define drm_crtc_to_virtio_gpu_output(x) \
 	container_of(x, struct virtio_gpu_output, crtc)
diff --git a/drivers/gpu/drm/virtio/virtgpu_display.c b/drivers/gpu/drm/virtio/virtgpu_display.c
index 2b7e6ae65546..44e9c7b874f5 100644
--- a/drivers/gpu/drm/virtio/virtgpu_display.c
+++ b/drivers/gpu/drm/virtio/virtgpu_display.c
@@ -99,6 +99,7 @@ static void virtio_gpu_crtc_atomic_enable(struct drm_crtc *crtc,
 	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
 
 	output->enabled = true;
+	output->need_update = true;
 }
 
 static void virtio_gpu_crtc_atomic_disable(struct drm_crtc *crtc,
diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c
index 52d24179bcec..5948031a9ce8 100644
--- a/drivers/gpu/drm/virtio/virtgpu_plane.c
+++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
@@ -163,7 +163,8 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
 	    plane->state->src_w != old_state->src_w ||
 	    plane->state->src_h != old_state->src_h ||
 	    plane->state->src_x != old_state->src_x ||
-	    plane->state->src_y != old_state->src_y) {
+	    plane->state->src_y != old_state->src_y ||
+	    output->need_update) {
 		DRM_DEBUG("handle 0x%x, crtc %dx%d+%d+%d, src %dx%d+%d+%d\n",
 			  bo->hw_res_handle,
 			  plane->state->crtc_w, plane->state->crtc_h,
@@ -178,6 +179,7 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
 					   plane->state->src_h >> 16,
 					   plane->state->src_x >> 16,
 					   plane->state->src_y >> 16);
+		output->need_update = false;
 	}
 
 	virtio_gpu_cmd_resource_flush(vgdev, bo->hw_res_handle,
-- 
2.18.4

[Bug 1882851] [PATCH] drm/virtio: fix unblank

[Gerd Hoffmann]

When going through a disable/enable cycle without changing the framebuffer
the optimization added by commit 3954ff10e06e causes the screen stay
blank.  Add a bool to force an update to fix that.

Cc: 1882851@bugs.launchpad.net
Fixes: 3954ff10e06e ("drm/virtio: skip set_scanout if framebuffer didn't change")
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
---
 drivers/gpu/drm/virtio/virtgpu_drv.h     | 1 +
 drivers/gpu/drm/virtio/virtgpu_display.c | 1 +
 drivers/gpu/drm/virtio/virtgpu_plane.c   | 4 +++-
 3 files changed, 5 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h
index 7879ff58236f..6d5410d5dd84 100644
--- a/drivers/gpu/drm/virtio/virtgpu_drv.h
+++ b/drivers/gpu/drm/virtio/virtgpu_drv.h
@@ -138,6 +138,7 @@ struct virtio_gpu_output {
 	int cur_x;
 	int cur_y;
 	bool enabled;
+	bool need_update;
 };
 #define drm_crtc_to_virtio_gpu_output(x) \
 	container_of(x, struct virtio_gpu_output, crtc)
diff --git a/drivers/gpu/drm/virtio/virtgpu_display.c b/drivers/gpu/drm/virtio/virtgpu_display.c
index 2b7e6ae65546..44e9c7b874f5 100644
--- a/drivers/gpu/drm/virtio/virtgpu_display.c
+++ b/drivers/gpu/drm/virtio/virtgpu_display.c
@@ -99,6 +99,7 @@ static void virtio_gpu_crtc_atomic_enable(struct drm_crtc *crtc,
 	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
 
 	output->enabled = true;
+	output->need_update = true;
 }
 
 static void virtio_gpu_crtc_atomic_disable(struct drm_crtc *crtc,
diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c
index 52d24179bcec..5948031a9ce8 100644
--- a/drivers/gpu/drm/virtio/virtgpu_plane.c
+++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
@@ -163,7 +163,8 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
 	    plane->state->src_w != old_state->src_w ||
 	    plane->state->src_h != old_state->src_h ||
 	    plane->state->src_x != old_state->src_x ||
-	    plane->state->src_y != old_state->src_y) {
+	    plane->state->src_y != old_state->src_y ||
+	    output->need_update) {
 		DRM_DEBUG("handle 0x%x, crtc %dx%d+%d+%d, src %dx%d+%d+%d\n",
 			  bo->hw_res_handle,
 			  plane->state->crtc_w, plane->state->crtc_h,
@@ -178,6 +179,7 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
 					   plane->state->src_h >> 16,
 					   plane->state->src_x >> 16,
 					   plane->state->src_y >> 16);
+		output->need_update = false;
 	}
 
 	virtio_gpu_cmd_resource_flush(vgdev, bo->hw_res_handle,
-- 
2.18.4

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1882851

Title:
  QEMU video freezes with "Guest disabled display" (virtio driver)

Status in QEMU:
  New

Bug description:
  I am using Arch Linux as my Guest and Host OS, after starting qemu
  with the following command:

    $ qemu-system-x86_64 -enable-kvm -hda arch-zoom.qcow2 -m 4G -vga
  virtio

  and waiting for a screen blank, I get this message:

    Guest disabled display

  And nothing happens after that, I can move the mouse or hit any key,
  and the message is still there.

  I can still reboot the VM but that's not optimal.

  I can reproduce this with the latest QEMU release (5.0.0) or git master, 
  I also tried this with older releases (4.0.0, 3.0.0) and the issue is still there.

  I can't reproduce this with other video drivers (std, qxl).

  With std/qxl the screen will blank a bit and then continue as normal.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1882851/+subscriptions

[PATCH] drm/virtio: fix unblank

[Gerd Hoffmann]

When going through a disable/enable cycle without changing the framebuffer
the optimization added by commit 3954ff10e06e causes the screen stay
blank.  Add a bool to force an update to fix that.

Cc: 1882851@bugs.launchpad.net
Fixes: 3954ff10e06e ("drm/virtio: skip set_scanout if framebuffer didn't change")
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
---
 drivers/gpu/drm/virtio/virtgpu_drv.h     | 1 +
 drivers/gpu/drm/virtio/virtgpu_display.c | 1 +
 drivers/gpu/drm/virtio/virtgpu_plane.c   | 4 +++-
 3 files changed, 5 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h
index 7879ff58236f..6d5410d5dd84 100644
--- a/drivers/gpu/drm/virtio/virtgpu_drv.h
+++ b/drivers/gpu/drm/virtio/virtgpu_drv.h
@@ -138,6 +138,7 @@ struct virtio_gpu_output {
 	int cur_x;
 	int cur_y;
 	bool enabled;
+	bool need_update;
 };
 #define drm_crtc_to_virtio_gpu_output(x) \
 	container_of(x, struct virtio_gpu_output, crtc)
diff --git a/drivers/gpu/drm/virtio/virtgpu_display.c b/drivers/gpu/drm/virtio/virtgpu_display.c
index 2b7e6ae65546..44e9c7b874f5 100644
--- a/drivers/gpu/drm/virtio/virtgpu_display.c
+++ b/drivers/gpu/drm/virtio/virtgpu_display.c
@@ -99,6 +99,7 @@ static void virtio_gpu_crtc_atomic_enable(struct drm_crtc *crtc,
 	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
 
 	output->enabled = true;
+	output->need_update = true;
 }
 
 static void virtio_gpu_crtc_atomic_disable(struct drm_crtc *crtc,
diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c
index 52d24179bcec..5948031a9ce8 100644
--- a/drivers/gpu/drm/virtio/virtgpu_plane.c
+++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
@@ -163,7 +163,8 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
 	    plane->state->src_w != old_state->src_w ||
 	    plane->state->src_h != old_state->src_h ||
 	    plane->state->src_x != old_state->src_x ||
-	    plane->state->src_y != old_state->src_y) {
+	    plane->state->src_y != old_state->src_y ||
+	    output->need_update) {
 		DRM_DEBUG("handle 0x%x, crtc %dx%d+%d+%d, src %dx%d+%d+%d\n",
 			  bo->hw_res_handle,
 			  plane->state->crtc_w, plane->state->crtc_h,
@@ -178,6 +179,7 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
 					   plane->state->src_h >> 16,
 					   plane->state->src_x >> 16,
 					   plane->state->src_y >> 16);
+		output->need_update = false;
 	}
 
 	virtio_gpu_cmd_resource_flush(vgdev, bo->hw_res_handle,
-- 
2.18.4

[PATCH] drm/virtio: fix unblank

[Gerd Hoffmann]

When going through a disable/enable cycle without changing the framebuffer
the optimization added by commit 3954ff10e06e causes the screen stay
blank.  Add a bool to force an update to fix that.

Cc: 1882851@bugs.launchpad.net
Fixes: 3954ff10e06e ("drm/virtio: skip set_scanout if framebuffer didn't change")
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
---
 drivers/gpu/drm/virtio/virtgpu_drv.h     | 1 +
 drivers/gpu/drm/virtio/virtgpu_display.c | 1 +
 drivers/gpu/drm/virtio/virtgpu_plane.c   | 4 +++-
 3 files changed, 5 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h
index 7879ff58236f..6d5410d5dd84 100644
--- a/drivers/gpu/drm/virtio/virtgpu_drv.h
+++ b/drivers/gpu/drm/virtio/virtgpu_drv.h
@@ -138,6 +138,7 @@ struct virtio_gpu_output {
 	int cur_x;
 	int cur_y;
 	bool enabled;
+	bool need_update;
 };
 #define drm_crtc_to_virtio_gpu_output(x) \
 	container_of(x, struct virtio_gpu_output, crtc)
diff --git a/drivers/gpu/drm/virtio/virtgpu_display.c b/drivers/gpu/drm/virtio/virtgpu_display.c
index 2b7e6ae65546..44e9c7b874f5 100644
--- a/drivers/gpu/drm/virtio/virtgpu_display.c
+++ b/drivers/gpu/drm/virtio/virtgpu_display.c
@@ -99,6 +99,7 @@ static void virtio_gpu_crtc_atomic_enable(struct drm_crtc *crtc,
 	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
 
 	output->enabled = true;
+	output->need_update = true;
 }
 
 static void virtio_gpu_crtc_atomic_disable(struct drm_crtc *crtc,
diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c
index 52d24179bcec..5948031a9ce8 100644
--- a/drivers/gpu/drm/virtio/virtgpu_plane.c
+++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
@@ -163,7 +163,8 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
 	    plane->state->src_w != old_state->src_w ||
 	    plane->state->src_h != old_state->src_h ||
 	    plane->state->src_x != old_state->src_x ||
-	    plane->state->src_y != old_state->src_y) {
+	    plane->state->src_y != old_state->src_y ||
+	    output->need_update) {
 		DRM_DEBUG("handle 0x%x, crtc %dx%d+%d+%d, src %dx%d+%d+%d\n",
 			  bo->hw_res_handle,
 			  plane->state->crtc_w, plane->state->crtc_h,
@@ -178,6 +179,7 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
 					   plane->state->src_h >> 16,
 					   plane->state->src_x >> 16,
 					   plane->state->src_y >> 16);
+		output->need_update = false;
 	}
 
 	virtio_gpu_cmd_resource_flush(vgdev, bo->hw_res_handle,
-- 
2.18.4

_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

[Bug 1882851] Re: QEMU video freezes with "Guest disabled display" (virtio driver)

[Diego Viola]

Gerd, thanks. I can confirm your patch fixes the problem with X, but
Wayland (sway) is still affected.

I tested X and wayland, and while the "Guest disabled display" no longer
hangs on X, it still does hangs under wayland.

Should I bisect again?

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1882851

Title:
  QEMU video freezes with "Guest disabled display" (virtio driver)

Status in QEMU:
  New

Bug description:
  I am using Arch Linux as my Guest and Host OS, after starting qemu
  with the following command:

    $ qemu-system-x86_64 -enable-kvm -hda arch-zoom.qcow2 -m 4G -vga
  virtio

  and waiting for a screen blank, I get this message:

    Guest disabled display

  And nothing happens after that, I can move the mouse or hit any key,
  and the message is still there.

  I can still reboot the VM but that's not optimal.

  I can reproduce this with the latest QEMU release (5.0.0) or git master, 
  I also tried this with older releases (4.0.0, 3.0.0) and the issue is still there.

  I can't reproduce this with other video drivers (std, qxl).

  With std/qxl the screen will blank a bit and then continue as normal.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1882851/+subscriptions

[Bug 1882851] Re: QEMU video freezes with "Guest disabled display" (virtio driver)

[Diego Viola]

Sway log after the crash.

** Attachment added: "swaylog.txt"
   https://bugs.launchpad.net/qemu/+bug/1882851/+attachment/5383276/+files/swaylog.txt

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1882851

Title:
  QEMU video freezes with "Guest disabled display" (virtio driver)

Status in QEMU:
  New

Bug description:
  I am using Arch Linux as my Guest and Host OS, after starting qemu
  with the following command:

    $ qemu-system-x86_64 -enable-kvm -hda arch-zoom.qcow2 -m 4G -vga
  virtio

  and waiting for a screen blank, I get this message:

    Guest disabled display

  And nothing happens after that, I can move the mouse or hit any key,
  and the message is still there.

  I can still reboot the VM but that's not optimal.

  I can reproduce this with the latest QEMU release (5.0.0) or git master, 
  I also tried this with older releases (4.0.0, 3.0.0) and the issue is still there.

  I can't reproduce this with other video drivers (std, qxl).

  With std/qxl the screen will blank a bit and then continue as normal.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1882851/+subscriptions

[Bug 1882851] Re: QEMU video freezes with "Guest disabled display" (virtio driver)

[Diego Viola]

It looks like sway requires swayidle to wake up from sleep[1].

This works:

swayidle timeout 2 'swaymsg "output * dpms off"' resume 'swaymsg "output
* dpms on"'

1. https://github.com/swaywm/sway/issues/2914

** Bug watch added: github.com/swaywm/sway/issues #2914
   https://github.com/swaywm/sway/issues/2914

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1882851

Title:
  QEMU video freezes with "Guest disabled display" (virtio driver)

Status in QEMU:
  New

Bug description:
  I am using Arch Linux as my Guest and Host OS, after starting qemu
  with the following command:

    $ qemu-system-x86_64 -enable-kvm -hda arch-zoom.qcow2 -m 4G -vga
  virtio

  and waiting for a screen blank, I get this message:

    Guest disabled display

  And nothing happens after that, I can move the mouse or hit any key,
  and the message is still there.

  I can still reboot the VM but that's not optimal.

  I can reproduce this with the latest QEMU release (5.0.0) or git master, 
  I also tried this with older releases (4.0.0, 3.0.0) and the issue is still there.

  I can't reproduce this with other video drivers (std, qxl).

  With std/qxl the screen will blank a bit and then continue as normal.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1882851/+subscriptions

[Bug 1882851] Re: QEMU video freezes with "Guest disabled display" (virtio driver)

[Diego Viola]

Yeah, I can reproduce the same exact behavior outside of QEMU with sway
and it's consistent to what I observed in QEMU.

> Hmm, happens with xorg only.

I think you were right all along about this, sorry.

Thanks for fixing this bug, feel free to close this bug as fixed.

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1882851

Title:
  QEMU video freezes with "Guest disabled display" (virtio driver)

Status in QEMU:
  New

Bug description:
  I am using Arch Linux as my Guest and Host OS, after starting qemu
  with the following command:

    $ qemu-system-x86_64 -enable-kvm -hda arch-zoom.qcow2 -m 4G -vga
  virtio

  and waiting for a screen blank, I get this message:

    Guest disabled display

  And nothing happens after that, I can move the mouse or hit any key,
  and the message is still there.

  I can still reboot the VM but that's not optimal.

  I can reproduce this with the latest QEMU release (5.0.0) or git master, 
  I also tried this with older releases (4.0.0, 3.0.0) and the issue is still there.

  I can't reproduce this with other video drivers (std, qxl).

  With std/qxl the screen will blank a bit and then continue as normal.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1882851/+subscriptions

[Bug 1882851] Re: QEMU video freezes with "Guest disabled display" (virtio driver)

[Diego Viola]

Will the patch make it for 5.8?

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1882851

Title:
  QEMU video freezes with "Guest disabled display" (virtio driver)

Status in QEMU:
  New

Bug description:
  I am using Arch Linux as my Guest and Host OS, after starting qemu
  with the following command:

    $ qemu-system-x86_64 -enable-kvm -hda arch-zoom.qcow2 -m 4G -vga
  virtio

  and waiting for a screen blank, I get this message:

    Guest disabled display

  And nothing happens after that, I can move the mouse or hit any key,
  and the message is still there.

  I can still reboot the VM but that's not optimal.

  I can reproduce this with the latest QEMU release (5.0.0) or git master, 
  I also tried this with older releases (4.0.0, 3.0.0) and the issue is still there.

  I can't reproduce this with other video drivers (std, qxl).

  With std/qxl the screen will blank a bit and then continue as normal.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1882851/+subscriptions

[PATCH] drm/virtio: fix unblank

[Gerd Hoffmann]

When going through a disable/enable cycle without changing the
framebuffer the optimization added by commit 3954ff10e06e ("drm/virtio:
skip set_scanout if framebuffer didn't change") causes the screen stay
blank.  Add a bool to force an update to fix that.

Cc: 1882851@bugs.launchpad.net
Fixes: 3954ff10e06e ("drm/virtio: skip set_scanout if framebuffer didn't change")
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
---
 drivers/gpu/drm/virtio/virtgpu_drv.h     | 1 +
 drivers/gpu/drm/virtio/virtgpu_display.c | 1 +
 drivers/gpu/drm/virtio/virtgpu_plane.c   | 4 +++-
 3 files changed, 5 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h
index 9ff9f4ac0522..7b0c319f23c9 100644
--- a/drivers/gpu/drm/virtio/virtgpu_drv.h
+++ b/drivers/gpu/drm/virtio/virtgpu_drv.h
@@ -138,6 +138,7 @@ struct virtio_gpu_output {
 	int cur_x;
 	int cur_y;
 	bool enabled;
+	bool need_update;
 };
 #define drm_crtc_to_virtio_gpu_output(x) \
 	container_of(x, struct virtio_gpu_output, crtc)
diff --git a/drivers/gpu/drm/virtio/virtgpu_display.c b/drivers/gpu/drm/virtio/virtgpu_display.c
index cc7fd957a307..378be5956b30 100644
--- a/drivers/gpu/drm/virtio/virtgpu_display.c
+++ b/drivers/gpu/drm/virtio/virtgpu_display.c
@@ -100,6 +100,7 @@ static void virtio_gpu_crtc_atomic_enable(struct drm_crtc *crtc,
 	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
 
 	output->enabled = true;
+	output->need_update = true;
 }
 
 static void virtio_gpu_crtc_atomic_disable(struct drm_crtc *crtc,
diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c
index 52d24179bcec..5948031a9ce8 100644
--- a/drivers/gpu/drm/virtio/virtgpu_plane.c
+++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
@@ -163,7 +163,8 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
 	    plane->state->src_w != old_state->src_w ||
 	    plane->state->src_h != old_state->src_h ||
 	    plane->state->src_x != old_state->src_x ||
-	    plane->state->src_y != old_state->src_y) {
+	    plane->state->src_y != old_state->src_y ||
+	    output->need_update) {
 		DRM_DEBUG("handle 0x%x, crtc %dx%d+%d+%d, src %dx%d+%d+%d\n",
 			  bo->hw_res_handle,
 			  plane->state->crtc_w, plane->state->crtc_h,
@@ -178,6 +179,7 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
 					   plane->state->src_h >> 16,
 					   plane->state->src_x >> 16,
 					   plane->state->src_y >> 16);
+		output->need_update = false;
 	}
 
 	virtio_gpu_cmd_resource_flush(vgdev, bo->hw_res_handle,
-- 
2.18.4

[Bug 1882851] [PATCH] drm/virtio: fix unblank

[Gerd Hoffmann]

When going through a disable/enable cycle without changing the
framebuffer the optimization added by commit 3954ff10e06e ("drm/virtio:
skip set_scanout if framebuffer didn't change") causes the screen stay
blank.  Add a bool to force an update to fix that.

Cc: 1882851@bugs.launchpad.net
Fixes: 3954ff10e06e ("drm/virtio: skip set_scanout if framebuffer didn't change")
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
---
 drivers/gpu/drm/virtio/virtgpu_drv.h     | 1 +
 drivers/gpu/drm/virtio/virtgpu_display.c | 1 +
 drivers/gpu/drm/virtio/virtgpu_plane.c   | 4 +++-
 3 files changed, 5 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h
index 9ff9f4ac0522..7b0c319f23c9 100644
--- a/drivers/gpu/drm/virtio/virtgpu_drv.h
+++ b/drivers/gpu/drm/virtio/virtgpu_drv.h
@@ -138,6 +138,7 @@ struct virtio_gpu_output {
 	int cur_x;
 	int cur_y;
 	bool enabled;
+	bool need_update;
 };
 #define drm_crtc_to_virtio_gpu_output(x) \
 	container_of(x, struct virtio_gpu_output, crtc)
diff --git a/drivers/gpu/drm/virtio/virtgpu_display.c b/drivers/gpu/drm/virtio/virtgpu_display.c
index cc7fd957a307..378be5956b30 100644
--- a/drivers/gpu/drm/virtio/virtgpu_display.c
+++ b/drivers/gpu/drm/virtio/virtgpu_display.c
@@ -100,6 +100,7 @@ static void virtio_gpu_crtc_atomic_enable(struct drm_crtc *crtc,
 	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
 
 	output->enabled = true;
+	output->need_update = true;
 }
 
 static void virtio_gpu_crtc_atomic_disable(struct drm_crtc *crtc,
diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c
index 52d24179bcec..5948031a9ce8 100644
--- a/drivers/gpu/drm/virtio/virtgpu_plane.c
+++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
@@ -163,7 +163,8 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
 	    plane->state->src_w != old_state->src_w ||
 	    plane->state->src_h != old_state->src_h ||
 	    plane->state->src_x != old_state->src_x ||
-	    plane->state->src_y != old_state->src_y) {
+	    plane->state->src_y != old_state->src_y ||
+	    output->need_update) {
 		DRM_DEBUG("handle 0x%x, crtc %dx%d+%d+%d, src %dx%d+%d+%d\n",
 			  bo->hw_res_handle,
 			  plane->state->crtc_w, plane->state->crtc_h,
@@ -178,6 +179,7 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
 					   plane->state->src_h >> 16,
 					   plane->state->src_x >> 16,
 					   plane->state->src_y >> 16);
+		output->need_update = false;
 	}
 
 	virtio_gpu_cmd_resource_flush(vgdev, bo->hw_res_handle,
-- 
2.18.4

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1882851

Title:
  QEMU video freezes with "Guest disabled display" (virtio driver)

Status in QEMU:
  New

Bug description:
  I am using Arch Linux as my Guest and Host OS, after starting qemu
  with the following command:

    $ qemu-system-x86_64 -enable-kvm -hda arch-zoom.qcow2 -m 4G -vga
  virtio

  and waiting for a screen blank, I get this message:

    Guest disabled display

  And nothing happens after that, I can move the mouse or hit any key,
  and the message is still there.

  I can still reboot the VM but that's not optimal.

  I can reproduce this with the latest QEMU release (5.0.0) or git master, 
  I also tried this with older releases (4.0.0, 3.0.0) and the issue is still there.

  I can't reproduce this with other video drivers (std, qxl).

  With std/qxl the screen will blank a bit and then continue as normal.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1882851/+subscriptions

[PATCH] drm/virtio: fix unblank

[Gerd Hoffmann]

When going through a disable/enable cycle without changing the
framebuffer the optimization added by commit 3954ff10e06e ("drm/virtio:
skip set_scanout if framebuffer didn't change") causes the screen stay
blank.  Add a bool to force an update to fix that.

Cc: 1882851@bugs.launchpad.net
Fixes: 3954ff10e06e ("drm/virtio: skip set_scanout if framebuffer didn't change")
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
---
 drivers/gpu/drm/virtio/virtgpu_drv.h     | 1 +
 drivers/gpu/drm/virtio/virtgpu_display.c | 1 +
 drivers/gpu/drm/virtio/virtgpu_plane.c   | 4 +++-
 3 files changed, 5 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h
index 9ff9f4ac0522..7b0c319f23c9 100644
--- a/drivers/gpu/drm/virtio/virtgpu_drv.h
+++ b/drivers/gpu/drm/virtio/virtgpu_drv.h
@@ -138,6 +138,7 @@ struct virtio_gpu_output {
 	int cur_x;
 	int cur_y;
 	bool enabled;
+	bool need_update;
 };
 #define drm_crtc_to_virtio_gpu_output(x) \
 	container_of(x, struct virtio_gpu_output, crtc)
diff --git a/drivers/gpu/drm/virtio/virtgpu_display.c b/drivers/gpu/drm/virtio/virtgpu_display.c
index cc7fd957a307..378be5956b30 100644
--- a/drivers/gpu/drm/virtio/virtgpu_display.c
+++ b/drivers/gpu/drm/virtio/virtgpu_display.c
@@ -100,6 +100,7 @@ static void virtio_gpu_crtc_atomic_enable(struct drm_crtc *crtc,
 	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
 
 	output->enabled = true;
+	output->need_update = true;
 }
 
 static void virtio_gpu_crtc_atomic_disable(struct drm_crtc *crtc,
diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c
index 52d24179bcec..5948031a9ce8 100644
--- a/drivers/gpu/drm/virtio/virtgpu_plane.c
+++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
@@ -163,7 +163,8 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
 	    plane->state->src_w != old_state->src_w ||
 	    plane->state->src_h != old_state->src_h ||
 	    plane->state->src_x != old_state->src_x ||
-	    plane->state->src_y != old_state->src_y) {
+	    plane->state->src_y != old_state->src_y ||
+	    output->need_update) {
 		DRM_DEBUG("handle 0x%x, crtc %dx%d+%d+%d, src %dx%d+%d+%d\n",
 			  bo->hw_res_handle,
 			  plane->state->crtc_w, plane->state->crtc_h,
@@ -178,6 +179,7 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
 					   plane->state->src_h >> 16,
 					   plane->state->src_x >> 16,
 					   plane->state->src_y >> 16);
+		output->need_update = false;
 	}
 
 	virtio_gpu_cmd_resource_flush(vgdev, bo->hw_res_handle,
-- 
2.18.4

_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

[PATCH] drm/virtio: fix unblank

[Gerd Hoffmann]

When going through a disable/enable cycle without changing the
framebuffer the optimization added by commit 3954ff10e06e ("drm/virtio:
skip set_scanout if framebuffer didn't change") causes the screen stay
blank.  Add a bool to force an update to fix that.

Cc: 1882851@bugs.launchpad.net
Fixes: 3954ff10e06e ("drm/virtio: skip set_scanout if framebuffer didn't change")
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
---
 drivers/gpu/drm/virtio/virtgpu_drv.h     | 1 +
 drivers/gpu/drm/virtio/virtgpu_display.c | 1 +
 drivers/gpu/drm/virtio/virtgpu_plane.c   | 4 +++-
 3 files changed, 5 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h
index 9ff9f4ac0522..7b0c319f23c9 100644
--- a/drivers/gpu/drm/virtio/virtgpu_drv.h
+++ b/drivers/gpu/drm/virtio/virtgpu_drv.h
@@ -138,6 +138,7 @@ struct virtio_gpu_output {
 	int cur_x;
 	int cur_y;
 	bool enabled;
+	bool need_update;
 };
 #define drm_crtc_to_virtio_gpu_output(x) \
 	container_of(x, struct virtio_gpu_output, crtc)
diff --git a/drivers/gpu/drm/virtio/virtgpu_display.c b/drivers/gpu/drm/virtio/virtgpu_display.c
index cc7fd957a307..378be5956b30 100644
--- a/drivers/gpu/drm/virtio/virtgpu_display.c
+++ b/drivers/gpu/drm/virtio/virtgpu_display.c
@@ -100,6 +100,7 @@ static void virtio_gpu_crtc_atomic_enable(struct drm_crtc *crtc,
 	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
 
 	output->enabled = true;
+	output->need_update = true;
 }
 
 static void virtio_gpu_crtc_atomic_disable(struct drm_crtc *crtc,
diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c
index 52d24179bcec..5948031a9ce8 100644
--- a/drivers/gpu/drm/virtio/virtgpu_plane.c
+++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
@@ -163,7 +163,8 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
 	    plane->state->src_w != old_state->src_w ||
 	    plane->state->src_h != old_state->src_h ||
 	    plane->state->src_x != old_state->src_x ||
-	    plane->state->src_y != old_state->src_y) {
+	    plane->state->src_y != old_state->src_y ||
+	    output->need_update) {
 		DRM_DEBUG("handle 0x%x, crtc %dx%d+%d+%d, src %dx%d+%d+%d\n",
 			  bo->hw_res_handle,
 			  plane->state->crtc_w, plane->state->crtc_h,
@@ -178,6 +179,7 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
 					   plane->state->src_h >> 16,
 					   plane->state->src_x >> 16,
 					   plane->state->src_y >> 16);
+		output->need_update = false;
 	}
 
 	virtio_gpu_cmd_resource_flush(vgdev, bo->hw_res_handle,
-- 
2.18.4

_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

Re: [PATCH] drm/virtio: fix unblank

[Daniel Vetter]

On Fri, Aug 07, 2020 at 12:54:29PM +0200, Gerd Hoffmann wrote:
> When going through a disable/enable cycle without changing the
> framebuffer the optimization added by commit 3954ff10e06e ("drm/virtio:
> skip set_scanout if framebuffer didn't change") causes the screen stay
> blank.  Add a bool to force an update to fix that.
> 
> Cc: 1882851@bugs.launchpad.net
> Fixes: 3954ff10e06e ("drm/virtio: skip set_scanout if framebuffer didn't change")
> Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
> ---
>  drivers/gpu/drm/virtio/virtgpu_drv.h     | 1 +
>  drivers/gpu/drm/virtio/virtgpu_display.c | 1 +
>  drivers/gpu/drm/virtio/virtgpu_plane.c   | 4 +++-
>  3 files changed, 5 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h
> index 9ff9f4ac0522..7b0c319f23c9 100644
> --- a/drivers/gpu/drm/virtio/virtgpu_drv.h
> +++ b/drivers/gpu/drm/virtio/virtgpu_drv.h
> @@ -138,6 +138,7 @@ struct virtio_gpu_output {
>  	int cur_x;
>  	int cur_y;
>  	bool enabled;
> +	bool need_update;
>  };
>  #define drm_crtc_to_virtio_gpu_output(x) \
>  	container_of(x, struct virtio_gpu_output, crtc)
> diff --git a/drivers/gpu/drm/virtio/virtgpu_display.c b/drivers/gpu/drm/virtio/virtgpu_display.c
> index cc7fd957a307..378be5956b30 100644
> --- a/drivers/gpu/drm/virtio/virtgpu_display.c
> +++ b/drivers/gpu/drm/virtio/virtgpu_display.c
> @@ -100,6 +100,7 @@ static void virtio_gpu_crtc_atomic_enable(struct drm_crtc *crtc,
>  	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
>  
>  	output->enabled = true;
> +	output->need_update = true;
>  }
>  
>  static void virtio_gpu_crtc_atomic_disable(struct drm_crtc *crtc,
> diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c
> index 52d24179bcec..5948031a9ce8 100644
> --- a/drivers/gpu/drm/virtio/virtgpu_plane.c
> +++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
> @@ -163,7 +163,8 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
>  	    plane->state->src_w != old_state->src_w ||
>  	    plane->state->src_h != old_state->src_h ||
>  	    plane->state->src_x != old_state->src_x ||
> -	    plane->state->src_y != old_state->src_y) {
> +	    plane->state->src_y != old_state->src_y ||
> +	    output->need_update) {

Uh instead of hand-rolling what's essentially a drm_crtc_needs_modeset
check, why not use that one? atomic helpers try to keep the usual suspects
for state transitions already handy, to avoid every driver rolling their
own. Or do I miss something here?
-Daniel


>  		DRM_DEBUG("handle 0x%x, crtc %dx%d+%d+%d, src %dx%d+%d+%d\n",
>  			  bo->hw_res_handle,
>  			  plane->state->crtc_w, plane->state->crtc_h,
> @@ -178,6 +179,7 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
>  					   plane->state->src_h >> 16,
>  					   plane->state->src_x >> 16,
>  					   plane->state->src_y >> 16);
> +		output->need_update = false;
>  	}
>  
>  	virtio_gpu_cmd_resource_flush(vgdev, bo->hw_res_handle,
> -- 
> 2.18.4
> 

-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch

Re: [PATCH] drm/virtio: fix unblank

[Daniel Vetter]

On Fri, Aug 07, 2020 at 12:54:29PM +0200, Gerd Hoffmann wrote:
> When going through a disable/enable cycle without changing the
> framebuffer the optimization added by commit 3954ff10e06e ("drm/virtio:
> skip set_scanout if framebuffer didn't change") causes the screen stay
> blank.  Add a bool to force an update to fix that.
> 
> Cc: 1882851@bugs.launchpad.net
> Fixes: 3954ff10e06e ("drm/virtio: skip set_scanout if framebuffer didn't change")
> Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
> ---
>  drivers/gpu/drm/virtio/virtgpu_drv.h     | 1 +
>  drivers/gpu/drm/virtio/virtgpu_display.c | 1 +
>  drivers/gpu/drm/virtio/virtgpu_plane.c   | 4 +++-
>  3 files changed, 5 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h
> index 9ff9f4ac0522..7b0c319f23c9 100644
> --- a/drivers/gpu/drm/virtio/virtgpu_drv.h
> +++ b/drivers/gpu/drm/virtio/virtgpu_drv.h
> @@ -138,6 +138,7 @@ struct virtio_gpu_output {
>  	int cur_x;
>  	int cur_y;
>  	bool enabled;
> +	bool need_update;
>  };
>  #define drm_crtc_to_virtio_gpu_output(x) \
>  	container_of(x, struct virtio_gpu_output, crtc)
> diff --git a/drivers/gpu/drm/virtio/virtgpu_display.c b/drivers/gpu/drm/virtio/virtgpu_display.c
> index cc7fd957a307..378be5956b30 100644
> --- a/drivers/gpu/drm/virtio/virtgpu_display.c
> +++ b/drivers/gpu/drm/virtio/virtgpu_display.c
> @@ -100,6 +100,7 @@ static void virtio_gpu_crtc_atomic_enable(struct drm_crtc *crtc,
>  	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
>  
>  	output->enabled = true;
> +	output->need_update = true;
>  }
>  
>  static void virtio_gpu_crtc_atomic_disable(struct drm_crtc *crtc,
> diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c
> index 52d24179bcec..5948031a9ce8 100644
> --- a/drivers/gpu/drm/virtio/virtgpu_plane.c
> +++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
> @@ -163,7 +163,8 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
>  	    plane->state->src_w != old_state->src_w ||
>  	    plane->state->src_h != old_state->src_h ||
>  	    plane->state->src_x != old_state->src_x ||
> -	    plane->state->src_y != old_state->src_y) {
> +	    plane->state->src_y != old_state->src_y ||
> +	    output->need_update) {

Uh instead of hand-rolling what's essentially a drm_crtc_needs_modeset
check, why not use that one? atomic helpers try to keep the usual suspects
for state transitions already handy, to avoid every driver rolling their
own. Or do I miss something here?
-Daniel


>  		DRM_DEBUG("handle 0x%x, crtc %dx%d+%d+%d, src %dx%d+%d+%d\n",
>  			  bo->hw_res_handle,
>  			  plane->state->crtc_w, plane->state->crtc_h,
> @@ -178,6 +179,7 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
>  					   plane->state->src_h >> 16,
>  					   plane->state->src_x >> 16,
>  					   plane->state->src_y >> 16);
> +		output->need_update = false;
>  	}
>  
>  	virtio_gpu_cmd_resource_flush(vgdev, bo->hw_res_handle,
> -- 
> 2.18.4
> 

-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch
_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

Re: [PATCH] drm/virtio: fix unblank

[Daniel Vetter]

On Fri, Aug 07, 2020 at 12:54:29PM +0200, Gerd Hoffmann wrote:
> When going through a disable/enable cycle without changing the
> framebuffer the optimization added by commit 3954ff10e06e ("drm/virtio:
> skip set_scanout if framebuffer didn't change") causes the screen stay
> blank.  Add a bool to force an update to fix that.
> 
> Cc: 1882851@bugs.launchpad.net
> Fixes: 3954ff10e06e ("drm/virtio: skip set_scanout if framebuffer didn't change")
> Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
> ---
>  drivers/gpu/drm/virtio/virtgpu_drv.h     | 1 +
>  drivers/gpu/drm/virtio/virtgpu_display.c | 1 +
>  drivers/gpu/drm/virtio/virtgpu_plane.c   | 4 +++-
>  3 files changed, 5 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h
> index 9ff9f4ac0522..7b0c319f23c9 100644
> --- a/drivers/gpu/drm/virtio/virtgpu_drv.h
> +++ b/drivers/gpu/drm/virtio/virtgpu_drv.h
> @@ -138,6 +138,7 @@ struct virtio_gpu_output {
>  	int cur_x;
>  	int cur_y;
>  	bool enabled;
> +	bool need_update;
>  };
>  #define drm_crtc_to_virtio_gpu_output(x) \
>  	container_of(x, struct virtio_gpu_output, crtc)
> diff --git a/drivers/gpu/drm/virtio/virtgpu_display.c b/drivers/gpu/drm/virtio/virtgpu_display.c
> index cc7fd957a307..378be5956b30 100644
> --- a/drivers/gpu/drm/virtio/virtgpu_display.c
> +++ b/drivers/gpu/drm/virtio/virtgpu_display.c
> @@ -100,6 +100,7 @@ static void virtio_gpu_crtc_atomic_enable(struct drm_crtc *crtc,
>  	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
>  
>  	output->enabled = true;
> +	output->need_update = true;
>  }
>  
>  static void virtio_gpu_crtc_atomic_disable(struct drm_crtc *crtc,
> diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c
> index 52d24179bcec..5948031a9ce8 100644
> --- a/drivers/gpu/drm/virtio/virtgpu_plane.c
> +++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
> @@ -163,7 +163,8 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
>  	    plane->state->src_w != old_state->src_w ||
>  	    plane->state->src_h != old_state->src_h ||
>  	    plane->state->src_x != old_state->src_x ||
> -	    plane->state->src_y != old_state->src_y) {
> +	    plane->state->src_y != old_state->src_y ||
> +	    output->need_update) {

Uh instead of hand-rolling what's essentially a drm_crtc_needs_modeset
check, why not use that one? atomic helpers try to keep the usual suspects
for state transitions already handy, to avoid every driver rolling their
own. Or do I miss something here?
-Daniel


>  		DRM_DEBUG("handle 0x%x, crtc %dx%d+%d+%d, src %dx%d+%d+%d\n",
>  			  bo->hw_res_handle,
>  			  plane->state->crtc_w, plane->state->crtc_h,
> @@ -178,6 +179,7 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
>  					   plane->state->src_h >> 16,
>  					   plane->state->src_x >> 16,
>  					   plane->state->src_y >> 16);
> +		output->need_update = false;
>  	}
>  
>  	virtio_gpu_cmd_resource_flush(vgdev, bo->hw_res_handle,
> -- 
> 2.18.4
> 

-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

Re: [PATCH] drm/virtio: fix unblank

[Gerd Hoffmann]

  Hi,

> > --- a/drivers/gpu/drm/virtio/virtgpu_display.c
> > +++ b/drivers/gpu/drm/virtio/virtgpu_display.c
> > @@ -100,6 +100,7 @@ static void virtio_gpu_crtc_atomic_enable(struct drm_crtc *crtc,
> >  	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
> >  
> >  	output->enabled = true;
> > +	output->need_update = true;

> > --- a/drivers/gpu/drm/virtio/virtgpu_plane.c
> > +++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
> > @@ -163,7 +163,8 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
> >  	    plane->state->src_w != old_state->src_w ||
> >  	    plane->state->src_h != old_state->src_h ||
> >  	    plane->state->src_x != old_state->src_x ||
> > -	    plane->state->src_y != old_state->src_y) {
> > +	    plane->state->src_y != old_state->src_y ||
> > +	    output->need_update) {
> 
> Uh instead of hand-rolling what's essentially a drm_crtc_needs_modeset
> check, why not use that one? atomic helpers try to keep the usual suspects
> for state transitions already handy, to avoid every driver rolling their
> own. Or do I miss something here?

Well, the virtio-gpu virtual hardware can't do plane updates and crtc
updates independant from each other.  So the crtc callbacks handle
disable only (we don't need a fb for that) and leave the enable to the
plane update.

I suspect calling drm_atomic_crtc_needs_modeset() in plane update isn't
going to fly ...

take care,
  Gerd

[Bug 1882851] Re: [PATCH] drm/virtio: fix unblank

[Gerd Hoffmann]

Hi,

> > --- a/drivers/gpu/drm/virtio/virtgpu_display.c
> > +++ b/drivers/gpu/drm/virtio/virtgpu_display.c
> > @@ -100,6 +100,7 @@ static void virtio_gpu_crtc_atomic_enable(struct drm_crtc *crtc,
> >  	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
> >  
> >  	output->enabled = true;
> > +	output->need_update = true;

> > --- a/drivers/gpu/drm/virtio/virtgpu_plane.c
> > +++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
> > @@ -163,7 +163,8 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
> >  	    plane->state->src_w != old_state->src_w ||
> >  	    plane->state->src_h != old_state->src_h ||
> >  	    plane->state->src_x != old_state->src_x ||
> > -	    plane->state->src_y != old_state->src_y) {
> > +	    plane->state->src_y != old_state->src_y ||
> > +	    output->need_update) {
> 
> Uh instead of hand-rolling what's essentially a drm_crtc_needs_modeset
> check, why not use that one? atomic helpers try to keep the usual suspects
> for state transitions already handy, to avoid every driver rolling their
> own. Or do I miss something here?

Well, the virtio-gpu virtual hardware can't do plane updates and crtc
updates independant from each other.  So the crtc callbacks handle
disable only (we don't need a fb for that) and leave the enable to the
plane update.

I suspect calling drm_atomic_crtc_needs_modeset() in plane update isn't
going to fly ...

take care,
  Gerd

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1882851

Title:
  QEMU video freezes with "Guest disabled display" (virtio driver)

Status in QEMU:
  New

Bug description:
  I am using Arch Linux as my Guest and Host OS, after starting qemu
  with the following command:

    $ qemu-system-x86_64 -enable-kvm -hda arch-zoom.qcow2 -m 4G -vga
  virtio

  and waiting for a screen blank, I get this message:

    Guest disabled display

  And nothing happens after that, I can move the mouse or hit any key,
  and the message is still there.

  I can still reboot the VM but that's not optimal.

  I can reproduce this with the latest QEMU release (5.0.0) or git master, 
  I also tried this with older releases (4.0.0, 3.0.0) and the issue is still there.

  I can't reproduce this with other video drivers (std, qxl).

  With std/qxl the screen will blank a bit and then continue as normal.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1882851/+subscriptions

Re: [PATCH] drm/virtio: fix unblank

[Gerd Hoffmann]

  Hi,

> > --- a/drivers/gpu/drm/virtio/virtgpu_display.c
> > +++ b/drivers/gpu/drm/virtio/virtgpu_display.c
> > @@ -100,6 +100,7 @@ static void virtio_gpu_crtc_atomic_enable(struct drm_crtc *crtc,
> >  	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
> >  
> >  	output->enabled = true;
> > +	output->need_update = true;

> > --- a/drivers/gpu/drm/virtio/virtgpu_plane.c
> > +++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
> > @@ -163,7 +163,8 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
> >  	    plane->state->src_w != old_state->src_w ||
> >  	    plane->state->src_h != old_state->src_h ||
> >  	    plane->state->src_x != old_state->src_x ||
> > -	    plane->state->src_y != old_state->src_y) {
> > +	    plane->state->src_y != old_state->src_y ||
> > +	    output->need_update) {
> 
> Uh instead of hand-rolling what's essentially a drm_crtc_needs_modeset
> check, why not use that one? atomic helpers try to keep the usual suspects
> for state transitions already handy, to avoid every driver rolling their
> own. Or do I miss something here?

Well, the virtio-gpu virtual hardware can't do plane updates and crtc
updates independant from each other.  So the crtc callbacks handle
disable only (we don't need a fb for that) and leave the enable to the
plane update.

I suspect calling drm_atomic_crtc_needs_modeset() in plane update isn't
going to fly ...

take care,
  Gerd

_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

Re: [PATCH] drm/virtio: fix unblank

[Gerd Hoffmann]

  Hi,

> > --- a/drivers/gpu/drm/virtio/virtgpu_display.c
> > +++ b/drivers/gpu/drm/virtio/virtgpu_display.c
> > @@ -100,6 +100,7 @@ static void virtio_gpu_crtc_atomic_enable(struct drm_crtc *crtc,
> >  	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
> >  
> >  	output->enabled = true;
> > +	output->need_update = true;

> > --- a/drivers/gpu/drm/virtio/virtgpu_plane.c
> > +++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
> > @@ -163,7 +163,8 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
> >  	    plane->state->src_w != old_state->src_w ||
> >  	    plane->state->src_h != old_state->src_h ||
> >  	    plane->state->src_x != old_state->src_x ||
> > -	    plane->state->src_y != old_state->src_y) {
> > +	    plane->state->src_y != old_state->src_y ||
> > +	    output->need_update) {
> 
> Uh instead of hand-rolling what's essentially a drm_crtc_needs_modeset
> check, why not use that one? atomic helpers try to keep the usual suspects
> for state transitions already handy, to avoid every driver rolling their
> own. Or do I miss something here?

Well, the virtio-gpu virtual hardware can't do plane updates and crtc
updates independant from each other.  So the crtc callbacks handle
disable only (we don't need a fb for that) and leave the enable to the
plane update.

I suspect calling drm_atomic_crtc_needs_modeset() in plane update isn't
going to fly ...

take care,
  Gerd

_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

Re: [PATCH] drm/virtio: fix unblank

[Gerd Hoffmann]

On Mon, Aug 17, 2020 at 11:03:42AM +0200, Gerd Hoffmann wrote:
>   Hi,
> 
> > > --- a/drivers/gpu/drm/virtio/virtgpu_display.c
> > > +++ b/drivers/gpu/drm/virtio/virtgpu_display.c
> > > @@ -100,6 +100,7 @@ static void virtio_gpu_crtc_atomic_enable(struct drm_crtc *crtc,
> > >  	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
> > >  
> > >  	output->enabled = true;
> > > +	output->need_update = true;
> 
> > > --- a/drivers/gpu/drm/virtio/virtgpu_plane.c
> > > +++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
> > > @@ -163,7 +163,8 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
> > >  	    plane->state->src_w != old_state->src_w ||
> > >  	    plane->state->src_h != old_state->src_h ||
> > >  	    plane->state->src_x != old_state->src_x ||
> > > -	    plane->state->src_y != old_state->src_y) {
> > > +	    plane->state->src_y != old_state->src_y ||
> > > +	    output->need_update) {
> > 
> > Uh instead of hand-rolling what's essentially a drm_crtc_needs_modeset
> > check, why not use that one? atomic helpers try to keep the usual suspects
> > for state transitions already handy, to avoid every driver rolling their
> > own. Or do I miss something here?
> 
> Well, the virtio-gpu virtual hardware can't do plane updates and crtc
> updates independant from each other.  So the crtc callbacks handle
> disable only (we don't need a fb for that) and leave the enable to the
> plane update.
> 
> I suspect calling drm_atomic_crtc_needs_modeset() in plane update isn't
> going to fly ...

Digged a bit more, seems crtc_state->*_changed is cleared after modeset
so the following plane update wouldn't see it.  Which I think means
there is no way around tracking that in need_update.

output->enabled is probably not needed though, seems I can replace that
by either output->crtc.state->enable or ->active.  Not fully sure which
one, probably ->active.

take care,
  Gerd

[Bug 1882851] Re: [PATCH] drm/virtio: fix unblank

[Gerd Hoffmann]

On Mon, Aug 17, 2020 at 11:03:42AM +0200, Gerd Hoffmann wrote:
>   Hi,
> 
> > > --- a/drivers/gpu/drm/virtio/virtgpu_display.c
> > > +++ b/drivers/gpu/drm/virtio/virtgpu_display.c
> > > @@ -100,6 +100,7 @@ static void virtio_gpu_crtc_atomic_enable(struct drm_crtc *crtc,
> > >  	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
> > >  
> > >  	output->enabled = true;
> > > +	output->need_update = true;
> 
> > > --- a/drivers/gpu/drm/virtio/virtgpu_plane.c
> > > +++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
> > > @@ -163,7 +163,8 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
> > >  	    plane->state->src_w != old_state->src_w ||
> > >  	    plane->state->src_h != old_state->src_h ||
> > >  	    plane->state->src_x != old_state->src_x ||
> > > -	    plane->state->src_y != old_state->src_y) {
> > > +	    plane->state->src_y != old_state->src_y ||
> > > +	    output->need_update) {
> > 
> > Uh instead of hand-rolling what's essentially a drm_crtc_needs_modeset
> > check, why not use that one? atomic helpers try to keep the usual suspects
> > for state transitions already handy, to avoid every driver rolling their
> > own. Or do I miss something here?
> 
> Well, the virtio-gpu virtual hardware can't do plane updates and crtc
> updates independant from each other.  So the crtc callbacks handle
> disable only (we don't need a fb for that) and leave the enable to the
> plane update.
> 
> I suspect calling drm_atomic_crtc_needs_modeset() in plane update isn't
> going to fly ...

Digged a bit more, seems crtc_state->*_changed is cleared after modeset
so the following plane update wouldn't see it.  Which I think means
there is no way around tracking that in need_update.

output->enabled is probably not needed though, seems I can replace that
by either output->crtc.state->enable or ->active.  Not fully sure which
one, probably ->active.

take care,
  Gerd

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1882851

Title:
  QEMU video freezes with "Guest disabled display" (virtio driver)

Status in QEMU:
  New

Bug description:
  I am using Arch Linux as my Guest and Host OS, after starting qemu
  with the following command:

    $ qemu-system-x86_64 -enable-kvm -hda arch-zoom.qcow2 -m 4G -vga
  virtio

  and waiting for a screen blank, I get this message:

    Guest disabled display

  And nothing happens after that, I can move the mouse or hit any key,
  and the message is still there.

  I can still reboot the VM but that's not optimal.

  I can reproduce this with the latest QEMU release (5.0.0) or git master, 
  I also tried this with older releases (4.0.0, 3.0.0) and the issue is still there.

  I can't reproduce this with other video drivers (std, qxl).

  With std/qxl the screen will blank a bit and then continue as normal.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1882851/+subscriptions

Re: [PATCH] drm/virtio: fix unblank

[Gerd Hoffmann]

On Mon, Aug 17, 2020 at 11:03:42AM +0200, Gerd Hoffmann wrote:
>   Hi,
> 
> > > --- a/drivers/gpu/drm/virtio/virtgpu_display.c
> > > +++ b/drivers/gpu/drm/virtio/virtgpu_display.c
> > > @@ -100,6 +100,7 @@ static void virtio_gpu_crtc_atomic_enable(struct drm_crtc *crtc,
> > >  	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
> > >  
> > >  	output->enabled = true;
> > > +	output->need_update = true;
> 
> > > --- a/drivers/gpu/drm/virtio/virtgpu_plane.c
> > > +++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
> > > @@ -163,7 +163,8 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
> > >  	    plane->state->src_w != old_state->src_w ||
> > >  	    plane->state->src_h != old_state->src_h ||
> > >  	    plane->state->src_x != old_state->src_x ||
> > > -	    plane->state->src_y != old_state->src_y) {
> > > +	    plane->state->src_y != old_state->src_y ||
> > > +	    output->need_update) {
> > 
> > Uh instead of hand-rolling what's essentially a drm_crtc_needs_modeset
> > check, why not use that one? atomic helpers try to keep the usual suspects
> > for state transitions already handy, to avoid every driver rolling their
> > own. Or do I miss something here?
> 
> Well, the virtio-gpu virtual hardware can't do plane updates and crtc
> updates independant from each other.  So the crtc callbacks handle
> disable only (we don't need a fb for that) and leave the enable to the
> plane update.
> 
> I suspect calling drm_atomic_crtc_needs_modeset() in plane update isn't
> going to fly ...

Digged a bit more, seems crtc_state->*_changed is cleared after modeset
so the following plane update wouldn't see it.  Which I think means
there is no way around tracking that in need_update.

output->enabled is probably not needed though, seems I can replace that
by either output->crtc.state->enable or ->active.  Not fully sure which
one, probably ->active.

take care,
  Gerd

_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

Re: [PATCH] drm/virtio: fix unblank

[Gerd Hoffmann]

On Mon, Aug 17, 2020 at 11:03:42AM +0200, Gerd Hoffmann wrote:
>   Hi,
> 
> > > --- a/drivers/gpu/drm/virtio/virtgpu_display.c
> > > +++ b/drivers/gpu/drm/virtio/virtgpu_display.c
> > > @@ -100,6 +100,7 @@ static void virtio_gpu_crtc_atomic_enable(struct drm_crtc *crtc,
> > >  	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
> > >  
> > >  	output->enabled = true;
> > > +	output->need_update = true;
> 
> > > --- a/drivers/gpu/drm/virtio/virtgpu_plane.c
> > > +++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
> > > @@ -163,7 +163,8 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
> > >  	    plane->state->src_w != old_state->src_w ||
> > >  	    plane->state->src_h != old_state->src_h ||
> > >  	    plane->state->src_x != old_state->src_x ||
> > > -	    plane->state->src_y != old_state->src_y) {
> > > +	    plane->state->src_y != old_state->src_y ||
> > > +	    output->need_update) {
> > 
> > Uh instead of hand-rolling what's essentially a drm_crtc_needs_modeset
> > check, why not use that one? atomic helpers try to keep the usual suspects
> > for state transitions already handy, to avoid every driver rolling their
> > own. Or do I miss something here?
> 
> Well, the virtio-gpu virtual hardware can't do plane updates and crtc
> updates independant from each other.  So the crtc callbacks handle
> disable only (we don't need a fb for that) and leave the enable to the
> plane update.
> 
> I suspect calling drm_atomic_crtc_needs_modeset() in plane update isn't
> going to fly ...

Digged a bit more, seems crtc_state->*_changed is cleared after modeset
so the following plane update wouldn't see it.  Which I think means
there is no way around tracking that in need_update.

output->enabled is probably not needed though, seems I can replace that
by either output->crtc.state->enable or ->active.  Not fully sure which
one, probably ->active.

take care,
  Gerd

_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

[PATCH 0/2] drm/virtio: fix unblank

[Gerd Hoffmann]

Gerd Hoffmann (2):
  drm/virtio: fix unblank
  drm/virtio: drop virtio_gpu_output->enabled

 drivers/gpu/drm/virtio/virtgpu_drv.h     |  2 +-
 drivers/gpu/drm/virtio/virtgpu_display.c | 15 +++++++++++----
 drivers/gpu/drm/virtio/virtgpu_plane.c   |  6 ++++--
 3 files changed, 16 insertions(+), 7 deletions(-)

-- 
2.18.4

_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

[PATCH 1/2] drm/virtio: fix unblank

[Gerd Hoffmann]

When going through a disable/enable cycle without changing the
framebuffer the optimization added by commit 3954ff10e06e ("drm/virtio:
skip set_scanout if framebuffer didn't change") causes the screen stay
blank.  Add a bool to force an update to fix that.

v2: use drm_atomic_crtc_needs_modeset() (Daniel).

Cc: 1882851@bugs.launchpad.net
Fixes: 3954ff10e06e ("drm/virtio: skip set_scanout if framebuffer didn't change")
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
---
 drivers/gpu/drm/virtio/virtgpu_drv.h     |  1 +
 drivers/gpu/drm/virtio/virtgpu_display.c | 11 +++++++++++
 drivers/gpu/drm/virtio/virtgpu_plane.c   |  4 +++-
 3 files changed, 15 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h
index 9ff9f4ac0522..4ab1b0ba2925 100644
--- a/drivers/gpu/drm/virtio/virtgpu_drv.h
+++ b/drivers/gpu/drm/virtio/virtgpu_drv.h
@@ -138,6 +138,7 @@ struct virtio_gpu_output {
 	int cur_x;
 	int cur_y;
 	bool enabled;
+	bool needs_modeset;
 };
 #define drm_crtc_to_virtio_gpu_output(x) \
 	container_of(x, struct virtio_gpu_output, crtc)
diff --git a/drivers/gpu/drm/virtio/virtgpu_display.c b/drivers/gpu/drm/virtio/virtgpu_display.c
index 2c2742b8d657..6c26b41f4e0d 100644
--- a/drivers/gpu/drm/virtio/virtgpu_display.c
+++ b/drivers/gpu/drm/virtio/virtgpu_display.c
@@ -123,6 +123,17 @@ static int virtio_gpu_crtc_atomic_check(struct drm_crtc *crtc,
 static void virtio_gpu_crtc_atomic_flush(struct drm_crtc *crtc,
 					 struct drm_crtc_state *old_state)
 {
+	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
+
+	/*
+	 * virtio-gpu can't do modeset and plane update operations
+	 * independant from each other.  So the actual modeset happens
+	 * in the plane update callback, and here we just check
+	 * whenever we must force the modeset.
+	 */
+	if (drm_atomic_crtc_needs_modeset(crtc->state)) {
+		output->needs_modeset = true;
+	}
 }
 
 static const struct drm_crtc_helper_funcs virtio_gpu_crtc_helper_funcs = {
diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c
index 52d24179bcec..65757409d9ed 100644
--- a/drivers/gpu/drm/virtio/virtgpu_plane.c
+++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
@@ -163,7 +163,9 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
 	    plane->state->src_w != old_state->src_w ||
 	    plane->state->src_h != old_state->src_h ||
 	    plane->state->src_x != old_state->src_x ||
-	    plane->state->src_y != old_state->src_y) {
+	    plane->state->src_y != old_state->src_y ||
+	    output->needs_modeset) {
+		output->needs_modeset = false;
 		DRM_DEBUG("handle 0x%x, crtc %dx%d+%d+%d, src %dx%d+%d+%d\n",
 			  bo->hw_res_handle,
 			  plane->state->crtc_w, plane->state->crtc_h,
-- 
2.18.4

[Bug 1882851] [PATCH 1/2] drm/virtio: fix unblank

[Gerd Hoffmann]

When going through a disable/enable cycle without changing the
framebuffer the optimization added by commit 3954ff10e06e ("drm/virtio:
skip set_scanout if framebuffer didn't change") causes the screen stay
blank.  Add a bool to force an update to fix that.

v2: use drm_atomic_crtc_needs_modeset() (Daniel).

Cc: 1882851@bugs.launchpad.net
Fixes: 3954ff10e06e ("drm/virtio: skip set_scanout if framebuffer didn't change")
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
---
 drivers/gpu/drm/virtio/virtgpu_drv.h     |  1 +
 drivers/gpu/drm/virtio/virtgpu_display.c | 11 +++++++++++
 drivers/gpu/drm/virtio/virtgpu_plane.c   |  4 +++-
 3 files changed, 15 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h
index 9ff9f4ac0522..4ab1b0ba2925 100644
--- a/drivers/gpu/drm/virtio/virtgpu_drv.h
+++ b/drivers/gpu/drm/virtio/virtgpu_drv.h
@@ -138,6 +138,7 @@ struct virtio_gpu_output {
 	int cur_x;
 	int cur_y;
 	bool enabled;
+	bool needs_modeset;
 };
 #define drm_crtc_to_virtio_gpu_output(x) \
 	container_of(x, struct virtio_gpu_output, crtc)
diff --git a/drivers/gpu/drm/virtio/virtgpu_display.c b/drivers/gpu/drm/virtio/virtgpu_display.c
index 2c2742b8d657..6c26b41f4e0d 100644
--- a/drivers/gpu/drm/virtio/virtgpu_display.c
+++ b/drivers/gpu/drm/virtio/virtgpu_display.c
@@ -123,6 +123,17 @@ static int virtio_gpu_crtc_atomic_check(struct drm_crtc *crtc,
 static void virtio_gpu_crtc_atomic_flush(struct drm_crtc *crtc,
 					 struct drm_crtc_state *old_state)
 {
+	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
+
+	/*
+	 * virtio-gpu can't do modeset and plane update operations
+	 * independant from each other.  So the actual modeset happens
+	 * in the plane update callback, and here we just check
+	 * whenever we must force the modeset.
+	 */
+	if (drm_atomic_crtc_needs_modeset(crtc->state)) {
+		output->needs_modeset = true;
+	}
 }
 
 static const struct drm_crtc_helper_funcs virtio_gpu_crtc_helper_funcs = {
diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c
index 52d24179bcec..65757409d9ed 100644
--- a/drivers/gpu/drm/virtio/virtgpu_plane.c
+++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
@@ -163,7 +163,9 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
 	    plane->state->src_w != old_state->src_w ||
 	    plane->state->src_h != old_state->src_h ||
 	    plane->state->src_x != old_state->src_x ||
-	    plane->state->src_y != old_state->src_y) {
+	    plane->state->src_y != old_state->src_y ||
+	    output->needs_modeset) {
+		output->needs_modeset = false;
 		DRM_DEBUG("handle 0x%x, crtc %dx%d+%d+%d, src %dx%d+%d+%d\n",
 			  bo->hw_res_handle,
 			  plane->state->crtc_w, plane->state->crtc_h,
-- 
2.18.4

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1882851

Title:
  QEMU video freezes with "Guest disabled display" (virtio driver)

Status in QEMU:
  New

Bug description:
  I am using Arch Linux as my Guest and Host OS, after starting qemu
  with the following command:

    $ qemu-system-x86_64 -enable-kvm -hda arch-zoom.qcow2 -m 4G -vga
  virtio

  and waiting for a screen blank, I get this message:

    Guest disabled display

  And nothing happens after that, I can move the mouse or hit any key,
  and the message is still there.

  I can still reboot the VM but that's not optimal.

  I can reproduce this with the latest QEMU release (5.0.0) or git master, 
  I also tried this with older releases (4.0.0, 3.0.0) and the issue is still there.

  I can't reproduce this with other video drivers (std, qxl).

  With std/qxl the screen will blank a bit and then continue as normal.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1882851/+subscriptions

[PATCH 1/2] drm/virtio: fix unblank

[Gerd Hoffmann]

When going through a disable/enable cycle without changing the
framebuffer the optimization added by commit 3954ff10e06e ("drm/virtio:
skip set_scanout if framebuffer didn't change") causes the screen stay
blank.  Add a bool to force an update to fix that.

v2: use drm_atomic_crtc_needs_modeset() (Daniel).

Cc: 1882851@bugs.launchpad.net
Fixes: 3954ff10e06e ("drm/virtio: skip set_scanout if framebuffer didn't change")
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
---
 drivers/gpu/drm/virtio/virtgpu_drv.h     |  1 +
 drivers/gpu/drm/virtio/virtgpu_display.c | 11 +++++++++++
 drivers/gpu/drm/virtio/virtgpu_plane.c   |  4 +++-
 3 files changed, 15 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h
index 9ff9f4ac0522..4ab1b0ba2925 100644
--- a/drivers/gpu/drm/virtio/virtgpu_drv.h
+++ b/drivers/gpu/drm/virtio/virtgpu_drv.h
@@ -138,6 +138,7 @@ struct virtio_gpu_output {
 	int cur_x;
 	int cur_y;
 	bool enabled;
+	bool needs_modeset;
 };
 #define drm_crtc_to_virtio_gpu_output(x) \
 	container_of(x, struct virtio_gpu_output, crtc)
diff --git a/drivers/gpu/drm/virtio/virtgpu_display.c b/drivers/gpu/drm/virtio/virtgpu_display.c
index 2c2742b8d657..6c26b41f4e0d 100644
--- a/drivers/gpu/drm/virtio/virtgpu_display.c
+++ b/drivers/gpu/drm/virtio/virtgpu_display.c
@@ -123,6 +123,17 @@ static int virtio_gpu_crtc_atomic_check(struct drm_crtc *crtc,
 static void virtio_gpu_crtc_atomic_flush(struct drm_crtc *crtc,
 					 struct drm_crtc_state *old_state)
 {
+	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
+
+	/*
+	 * virtio-gpu can't do modeset and plane update operations
+	 * independant from each other.  So the actual modeset happens
+	 * in the plane update callback, and here we just check
+	 * whenever we must force the modeset.
+	 */
+	if (drm_atomic_crtc_needs_modeset(crtc->state)) {
+		output->needs_modeset = true;
+	}
 }
 
 static const struct drm_crtc_helper_funcs virtio_gpu_crtc_helper_funcs = {
diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c
index 52d24179bcec..65757409d9ed 100644
--- a/drivers/gpu/drm/virtio/virtgpu_plane.c
+++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
@@ -163,7 +163,9 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
 	    plane->state->src_w != old_state->src_w ||
 	    plane->state->src_h != old_state->src_h ||
 	    plane->state->src_x != old_state->src_x ||
-	    plane->state->src_y != old_state->src_y) {
+	    plane->state->src_y != old_state->src_y ||
+	    output->needs_modeset) {
+		output->needs_modeset = false;
 		DRM_DEBUG("handle 0x%x, crtc %dx%d+%d+%d, src %dx%d+%d+%d\n",
 			  bo->hw_res_handle,
 			  plane->state->crtc_w, plane->state->crtc_h,
-- 
2.18.4

_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

[PATCH 1/2] drm/virtio: fix unblank

[Gerd Hoffmann]

When going through a disable/enable cycle without changing the
framebuffer the optimization added by commit 3954ff10e06e ("drm/virtio:
skip set_scanout if framebuffer didn't change") causes the screen stay
blank.  Add a bool to force an update to fix that.

v2: use drm_atomic_crtc_needs_modeset() (Daniel).

Cc: 1882851@bugs.launchpad.net
Fixes: 3954ff10e06e ("drm/virtio: skip set_scanout if framebuffer didn't change")
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
---
 drivers/gpu/drm/virtio/virtgpu_drv.h     |  1 +
 drivers/gpu/drm/virtio/virtgpu_display.c | 11 +++++++++++
 drivers/gpu/drm/virtio/virtgpu_plane.c   |  4 +++-
 3 files changed, 15 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h
index 9ff9f4ac0522..4ab1b0ba2925 100644
--- a/drivers/gpu/drm/virtio/virtgpu_drv.h
+++ b/drivers/gpu/drm/virtio/virtgpu_drv.h
@@ -138,6 +138,7 @@ struct virtio_gpu_output {
 	int cur_x;
 	int cur_y;
 	bool enabled;
+	bool needs_modeset;
 };
 #define drm_crtc_to_virtio_gpu_output(x) \
 	container_of(x, struct virtio_gpu_output, crtc)
diff --git a/drivers/gpu/drm/virtio/virtgpu_display.c b/drivers/gpu/drm/virtio/virtgpu_display.c
index 2c2742b8d657..6c26b41f4e0d 100644
--- a/drivers/gpu/drm/virtio/virtgpu_display.c
+++ b/drivers/gpu/drm/virtio/virtgpu_display.c
@@ -123,6 +123,17 @@ static int virtio_gpu_crtc_atomic_check(struct drm_crtc *crtc,
 static void virtio_gpu_crtc_atomic_flush(struct drm_crtc *crtc,
 					 struct drm_crtc_state *old_state)
 {
+	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
+
+	/*
+	 * virtio-gpu can't do modeset and plane update operations
+	 * independant from each other.  So the actual modeset happens
+	 * in the plane update callback, and here we just check
+	 * whenever we must force the modeset.
+	 */
+	if (drm_atomic_crtc_needs_modeset(crtc->state)) {
+		output->needs_modeset = true;
+	}
 }
 
 static const struct drm_crtc_helper_funcs virtio_gpu_crtc_helper_funcs = {
diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c
index 52d24179bcec..65757409d9ed 100644
--- a/drivers/gpu/drm/virtio/virtgpu_plane.c
+++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
@@ -163,7 +163,9 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
 	    plane->state->src_w != old_state->src_w ||
 	    plane->state->src_h != old_state->src_h ||
 	    plane->state->src_x != old_state->src_x ||
-	    plane->state->src_y != old_state->src_y) {
+	    plane->state->src_y != old_state->src_y ||
+	    output->needs_modeset) {
+		output->needs_modeset = false;
 		DRM_DEBUG("handle 0x%x, crtc %dx%d+%d+%d, src %dx%d+%d+%d\n",
 			  bo->hw_res_handle,
 			  plane->state->crtc_w, plane->state->crtc_h,
-- 
2.18.4

_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

[PATCH 2/2] drm/virtio: drop virtio_gpu_output->enabled

[Gerd Hoffmann]

Not needed, already tracked by drm_crtc_state->active.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
---
 drivers/gpu/drm/virtio/virtgpu_drv.h     | 1 -
 drivers/gpu/drm/virtio/virtgpu_display.c | 4 ----
 drivers/gpu/drm/virtio/virtgpu_plane.c   | 2 +-
 3 files changed, 1 insertion(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h
index 4ab1b0ba2925..fbc04272db4f 100644
--- a/drivers/gpu/drm/virtio/virtgpu_drv.h
+++ b/drivers/gpu/drm/virtio/virtgpu_drv.h
@@ -137,7 +137,6 @@ struct virtio_gpu_output {
 	struct edid *edid;
 	int cur_x;
 	int cur_y;
-	bool enabled;
 	bool needs_modeset;
 };
 #define drm_crtc_to_virtio_gpu_output(x) \
diff --git a/drivers/gpu/drm/virtio/virtgpu_display.c b/drivers/gpu/drm/virtio/virtgpu_display.c
index 6c26b41f4e0d..86a3a800d12e 100644
--- a/drivers/gpu/drm/virtio/virtgpu_display.c
+++ b/drivers/gpu/drm/virtio/virtgpu_display.c
@@ -97,9 +97,6 @@ static void virtio_gpu_crtc_mode_set_nofb(struct drm_crtc *crtc)
 static void virtio_gpu_crtc_atomic_enable(struct drm_crtc *crtc,
 					  struct drm_crtc_state *old_state)
 {
-	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
-
-	output->enabled = true;
 }
 
 static void virtio_gpu_crtc_atomic_disable(struct drm_crtc *crtc,
@@ -111,7 +108,6 @@ static void virtio_gpu_crtc_atomic_disable(struct drm_crtc *crtc,
 
 	virtio_gpu_cmd_set_scanout(vgdev, output->index, 0, 0, 0, 0, 0);
 	virtio_gpu_notify(vgdev);
-	output->enabled = false;
 }
 
 static int virtio_gpu_crtc_atomic_check(struct drm_crtc *crtc,
diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c
index 65757409d9ed..6a311cd93440 100644
--- a/drivers/gpu/drm/virtio/virtgpu_plane.c
+++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
@@ -142,7 +142,7 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
 	if (WARN_ON(!output))
 		return;
 
-	if (!plane->state->fb || !output->enabled) {
+	if (!plane->state->fb || !output->crtc.state->active) {
 		DRM_DEBUG("nofb\n");
 		virtio_gpu_cmd_set_scanout(vgdev, output->index, 0,
 					   plane->state->src_w >> 16,
-- 
2.18.4

[PATCH 2/2] drm/virtio: drop virtio_gpu_output->enabled

[Gerd Hoffmann]

Not needed, already tracked by drm_crtc_state->active.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
---
 drivers/gpu/drm/virtio/virtgpu_drv.h     | 1 -
 drivers/gpu/drm/virtio/virtgpu_display.c | 4 ----
 drivers/gpu/drm/virtio/virtgpu_plane.c   | 2 +-
 3 files changed, 1 insertion(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h
index 4ab1b0ba2925..fbc04272db4f 100644
--- a/drivers/gpu/drm/virtio/virtgpu_drv.h
+++ b/drivers/gpu/drm/virtio/virtgpu_drv.h
@@ -137,7 +137,6 @@ struct virtio_gpu_output {
 	struct edid *edid;
 	int cur_x;
 	int cur_y;
-	bool enabled;
 	bool needs_modeset;
 };
 #define drm_crtc_to_virtio_gpu_output(x) \
diff --git a/drivers/gpu/drm/virtio/virtgpu_display.c b/drivers/gpu/drm/virtio/virtgpu_display.c
index 6c26b41f4e0d..86a3a800d12e 100644
--- a/drivers/gpu/drm/virtio/virtgpu_display.c
+++ b/drivers/gpu/drm/virtio/virtgpu_display.c
@@ -97,9 +97,6 @@ static void virtio_gpu_crtc_mode_set_nofb(struct drm_crtc *crtc)
 static void virtio_gpu_crtc_atomic_enable(struct drm_crtc *crtc,
 					  struct drm_crtc_state *old_state)
 {
-	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
-
-	output->enabled = true;
 }
 
 static void virtio_gpu_crtc_atomic_disable(struct drm_crtc *crtc,
@@ -111,7 +108,6 @@ static void virtio_gpu_crtc_atomic_disable(struct drm_crtc *crtc,
 
 	virtio_gpu_cmd_set_scanout(vgdev, output->index, 0, 0, 0, 0, 0);
 	virtio_gpu_notify(vgdev);
-	output->enabled = false;
 }
 
 static int virtio_gpu_crtc_atomic_check(struct drm_crtc *crtc,
diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c
index 65757409d9ed..6a311cd93440 100644
--- a/drivers/gpu/drm/virtio/virtgpu_plane.c
+++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
@@ -142,7 +142,7 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
 	if (WARN_ON(!output))
 		return;
 
-	if (!plane->state->fb || !output->enabled) {
+	if (!plane->state->fb || !output->crtc.state->active) {
 		DRM_DEBUG("nofb\n");
 		virtio_gpu_cmd_set_scanout(vgdev, output->index, 0,
 					   plane->state->src_w >> 16,
-- 
2.18.4

_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

[PATCH 2/2] drm/virtio: drop virtio_gpu_output->enabled

[Gerd Hoffmann]

Not needed, already tracked by drm_crtc_state->active.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
---
 drivers/gpu/drm/virtio/virtgpu_drv.h     | 1 -
 drivers/gpu/drm/virtio/virtgpu_display.c | 4 ----
 drivers/gpu/drm/virtio/virtgpu_plane.c   | 2 +-
 3 files changed, 1 insertion(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h
index 4ab1b0ba2925..fbc04272db4f 100644
--- a/drivers/gpu/drm/virtio/virtgpu_drv.h
+++ b/drivers/gpu/drm/virtio/virtgpu_drv.h
@@ -137,7 +137,6 @@ struct virtio_gpu_output {
 	struct edid *edid;
 	int cur_x;
 	int cur_y;
-	bool enabled;
 	bool needs_modeset;
 };
 #define drm_crtc_to_virtio_gpu_output(x) \
diff --git a/drivers/gpu/drm/virtio/virtgpu_display.c b/drivers/gpu/drm/virtio/virtgpu_display.c
index 6c26b41f4e0d..86a3a800d12e 100644
--- a/drivers/gpu/drm/virtio/virtgpu_display.c
+++ b/drivers/gpu/drm/virtio/virtgpu_display.c
@@ -97,9 +97,6 @@ static void virtio_gpu_crtc_mode_set_nofb(struct drm_crtc *crtc)
 static void virtio_gpu_crtc_atomic_enable(struct drm_crtc *crtc,
 					  struct drm_crtc_state *old_state)
 {
-	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
-
-	output->enabled = true;
 }
 
 static void virtio_gpu_crtc_atomic_disable(struct drm_crtc *crtc,
@@ -111,7 +108,6 @@ static void virtio_gpu_crtc_atomic_disable(struct drm_crtc *crtc,
 
 	virtio_gpu_cmd_set_scanout(vgdev, output->index, 0, 0, 0, 0, 0);
 	virtio_gpu_notify(vgdev);
-	output->enabled = false;
 }
 
 static int virtio_gpu_crtc_atomic_check(struct drm_crtc *crtc,
diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c
index 65757409d9ed..6a311cd93440 100644
--- a/drivers/gpu/drm/virtio/virtgpu_plane.c
+++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
@@ -142,7 +142,7 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
 	if (WARN_ON(!output))
 		return;
 
-	if (!plane->state->fb || !output->enabled) {
+	if (!plane->state->fb || !output->crtc.state->active) {
 		DRM_DEBUG("nofb\n");
 		virtio_gpu_cmd_set_scanout(vgdev, output->index, 0,
 					   plane->state->src_w >> 16,
-- 
2.18.4

_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

Re: [PATCH 1/2] drm/virtio: fix unblank

[Jiri Slaby]

On 18. 08. 20, 9:25, Gerd Hoffmann wrote:
> When going through a disable/enable cycle without changing the
> framebuffer the optimization added by commit 3954ff10e06e ("drm/virtio:
> skip set_scanout if framebuffer didn't change") causes the screen stay
> blank.  Add a bool to force an update to fix that.
> 
> v2: use drm_atomic_crtc_needs_modeset() (Daniel).
> 
> Cc: 1882851@bugs.launchpad.net
> Fixes: 3954ff10e06e ("drm/virtio: skip set_scanout if framebuffer didn't change")
> Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>

Tested-by: Jiri Slaby <jirislaby@kernel.org>

> ---
>  drivers/gpu/drm/virtio/virtgpu_drv.h     |  1 +
>  drivers/gpu/drm/virtio/virtgpu_display.c | 11 +++++++++++
>  drivers/gpu/drm/virtio/virtgpu_plane.c   |  4 +++-
>  3 files changed, 15 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h
> index 9ff9f4ac0522..4ab1b0ba2925 100644
> --- a/drivers/gpu/drm/virtio/virtgpu_drv.h
> +++ b/drivers/gpu/drm/virtio/virtgpu_drv.h
> @@ -138,6 +138,7 @@ struct virtio_gpu_output {
>  	int cur_x;
>  	int cur_y;
>  	bool enabled;
> +	bool needs_modeset;
>  };
>  #define drm_crtc_to_virtio_gpu_output(x) \
>  	container_of(x, struct virtio_gpu_output, crtc)
> diff --git a/drivers/gpu/drm/virtio/virtgpu_display.c b/drivers/gpu/drm/virtio/virtgpu_display.c
> index 2c2742b8d657..6c26b41f4e0d 100644
> --- a/drivers/gpu/drm/virtio/virtgpu_display.c
> +++ b/drivers/gpu/drm/virtio/virtgpu_display.c
> @@ -123,6 +123,17 @@ static int virtio_gpu_crtc_atomic_check(struct drm_crtc *crtc,
>  static void virtio_gpu_crtc_atomic_flush(struct drm_crtc *crtc,
>  					 struct drm_crtc_state *old_state)
>  {
> +	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
> +
> +	/*
> +	 * virtio-gpu can't do modeset and plane update operations
> +	 * independant from each other.  So the actual modeset happens
> +	 * in the plane update callback, and here we just check
> +	 * whenever we must force the modeset.
> +	 */
> +	if (drm_atomic_crtc_needs_modeset(crtc->state)) {
> +		output->needs_modeset = true;
> +	}
>  }
>  
>  static const struct drm_crtc_helper_funcs virtio_gpu_crtc_helper_funcs = {
> diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c
> index 52d24179bcec..65757409d9ed 100644
> --- a/drivers/gpu/drm/virtio/virtgpu_plane.c
> +++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
> @@ -163,7 +163,9 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
>  	    plane->state->src_w != old_state->src_w ||
>  	    plane->state->src_h != old_state->src_h ||
>  	    plane->state->src_x != old_state->src_x ||
> -	    plane->state->src_y != old_state->src_y) {
> +	    plane->state->src_y != old_state->src_y ||
> +	    output->needs_modeset) {
> +		output->needs_modeset = false;
>  		DRM_DEBUG("handle 0x%x, crtc %dx%d+%d+%d, src %dx%d+%d+%d\n",
>  			  bo->hw_res_handle,
>  			  plane->state->crtc_w, plane->state->crtc_h,
> 


-- 
js
suse labs

Re: [PATCH 1/2] drm/virtio: fix unblank

[Jiri Slaby]

On 18. 08. 20, 9:25, Gerd Hoffmann wrote:
> When going through a disable/enable cycle without changing the
> framebuffer the optimization added by commit 3954ff10e06e ("drm/virtio:
> skip set_scanout if framebuffer didn't change") causes the screen stay
> blank.  Add a bool to force an update to fix that.
> 
> v2: use drm_atomic_crtc_needs_modeset() (Daniel).
> 
> Cc: 1882851@bugs.launchpad.net
> Fixes: 3954ff10e06e ("drm/virtio: skip set_scanout if framebuffer didn't change")
> Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>

Tested-by: Jiri Slaby <jirislaby@kernel.org>

> ---
>  drivers/gpu/drm/virtio/virtgpu_drv.h     |  1 +
>  drivers/gpu/drm/virtio/virtgpu_display.c | 11 +++++++++++
>  drivers/gpu/drm/virtio/virtgpu_plane.c   |  4 +++-
>  3 files changed, 15 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h
> index 9ff9f4ac0522..4ab1b0ba2925 100644
> --- a/drivers/gpu/drm/virtio/virtgpu_drv.h
> +++ b/drivers/gpu/drm/virtio/virtgpu_drv.h
> @@ -138,6 +138,7 @@ struct virtio_gpu_output {
>  	int cur_x;
>  	int cur_y;
>  	bool enabled;
> +	bool needs_modeset;
>  };
>  #define drm_crtc_to_virtio_gpu_output(x) \
>  	container_of(x, struct virtio_gpu_output, crtc)
> diff --git a/drivers/gpu/drm/virtio/virtgpu_display.c b/drivers/gpu/drm/virtio/virtgpu_display.c
> index 2c2742b8d657..6c26b41f4e0d 100644
> --- a/drivers/gpu/drm/virtio/virtgpu_display.c
> +++ b/drivers/gpu/drm/virtio/virtgpu_display.c
> @@ -123,6 +123,17 @@ static int virtio_gpu_crtc_atomic_check(struct drm_crtc *crtc,
>  static void virtio_gpu_crtc_atomic_flush(struct drm_crtc *crtc,
>  					 struct drm_crtc_state *old_state)
>  {
> +	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
> +
> +	/*
> +	 * virtio-gpu can't do modeset and plane update operations
> +	 * independant from each other.  So the actual modeset happens
> +	 * in the plane update callback, and here we just check
> +	 * whenever we must force the modeset.
> +	 */
> +	if (drm_atomic_crtc_needs_modeset(crtc->state)) {
> +		output->needs_modeset = true;
> +	}
>  }
>  
>  static const struct drm_crtc_helper_funcs virtio_gpu_crtc_helper_funcs = {
> diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c
> index 52d24179bcec..65757409d9ed 100644
> --- a/drivers/gpu/drm/virtio/virtgpu_plane.c
> +++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
> @@ -163,7 +163,9 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
>  	    plane->state->src_w != old_state->src_w ||
>  	    plane->state->src_h != old_state->src_h ||
>  	    plane->state->src_x != old_state->src_x ||
> -	    plane->state->src_y != old_state->src_y) {
> +	    plane->state->src_y != old_state->src_y ||
> +	    output->needs_modeset) {
> +		output->needs_modeset = false;
>  		DRM_DEBUG("handle 0x%x, crtc %dx%d+%d+%d, src %dx%d+%d+%d\n",
>  			  bo->hw_res_handle,
>  			  plane->state->crtc_w, plane->state->crtc_h,
> 


-- 
js
suse labs
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

[PATCH] drm/virtio: fix unblank

[Diego Viola]

From: Gerd Hoffmann <kraxel@redhat.com>

When going through a disable/enable cycle without changing the
framebuffer the optimization added by commit 3954ff10e06e ("drm/virtio:
skip set_scanout if framebuffer didn't change") causes the screen stay
blank.  Add a bool to force an update to fix that.

v2: use drm_atomic_crtc_needs_modeset() (Daniel).

Cc: 1882851@bugs.launchpad.net
Fixes: 3954ff10e06e ("drm/virtio: skip set_scanout if framebuffer didn't change")
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Tested-by: Jiri Slaby <jirislaby@kernel.org>
Tested-by: Diego Viola <diego.viola@gmail.com>
---
 drivers/gpu/drm/virtio/virtgpu_display.c | 11 +++++++++++
 drivers/gpu/drm/virtio/virtgpu_drv.h     |  1 +
 drivers/gpu/drm/virtio/virtgpu_plane.c   |  4 +++-
 3 files changed, 15 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/virtio/virtgpu_display.c b/drivers/gpu/drm/virtio/virtgpu_display.c
index af55b334be2f..35b5c80f5d85 100644
--- a/drivers/gpu/drm/virtio/virtgpu_display.c
+++ b/drivers/gpu/drm/virtio/virtgpu_display.c
@@ -123,6 +123,17 @@ static int virtio_gpu_crtc_atomic_check(struct drm_crtc *crtc,
 static void virtio_gpu_crtc_atomic_flush(struct drm_crtc *crtc,
 					 struct drm_crtc_state *old_state)
 {
+	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
+
+	/*
+	 * virtio-gpu can't do modeset and plane update operations
+	 * independant from each other.  So the actual modeset happens
+	 * in the plane update callback, and here we just check
+	 * whenever we must force the modeset.
+	 */
+	if (drm_atomic_crtc_needs_modeset(crtc->state)) {
+		output->needs_modeset = true;
+	}
 }
 
 static const struct drm_crtc_helper_funcs virtio_gpu_crtc_helper_funcs = {
diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h
index 9ff9f4ac0522..4ab1b0ba2925 100644
--- a/drivers/gpu/drm/virtio/virtgpu_drv.h
+++ b/drivers/gpu/drm/virtio/virtgpu_drv.h
@@ -138,6 +138,7 @@ struct virtio_gpu_output {
 	int cur_x;
 	int cur_y;
 	bool enabled;
+	bool needs_modeset;
 };
 #define drm_crtc_to_virtio_gpu_output(x) \
 	container_of(x, struct virtio_gpu_output, crtc)
diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c
index 52d24179bcec..65757409d9ed 100644
--- a/drivers/gpu/drm/virtio/virtgpu_plane.c
+++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
@@ -163,7 +163,9 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
 	    plane->state->src_w != old_state->src_w ||
 	    plane->state->src_h != old_state->src_h ||
 	    plane->state->src_x != old_state->src_x ||
-	    plane->state->src_y != old_state->src_y) {
+	    plane->state->src_y != old_state->src_y ||
+	    output->needs_modeset) {
+		output->needs_modeset = false;
 		DRM_DEBUG("handle 0x%x, crtc %dx%d+%d+%d, src %dx%d+%d+%d\n",
 			  bo->hw_res_handle,
 			  plane->state->crtc_w, plane->state->crtc_h,
-- 
2.28.0

[Bug 1882851] [PATCH] drm/virtio: fix unblank

[Diego Viola]

From: Gerd Hoffmann <kraxel@redhat.com>

When going through a disable/enable cycle without changing the
framebuffer the optimization added by commit 3954ff10e06e ("drm/virtio:
skip set_scanout if framebuffer didn't change") causes the screen stay
blank.  Add a bool to force an update to fix that.

v2: use drm_atomic_crtc_needs_modeset() (Daniel).

Cc: 1882851@bugs.launchpad.net
Fixes: 3954ff10e06e ("drm/virtio: skip set_scanout if framebuffer didn't change")
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Tested-by: Jiri Slaby <jirislaby@kernel.org>
Tested-by: Diego Viola <diego.viola@gmail.com>
---
 drivers/gpu/drm/virtio/virtgpu_display.c | 11 +++++++++++
 drivers/gpu/drm/virtio/virtgpu_drv.h     |  1 +
 drivers/gpu/drm/virtio/virtgpu_plane.c   |  4 +++-
 3 files changed, 15 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/virtio/virtgpu_display.c b/drivers/gpu/drm/virtio/virtgpu_display.c
index af55b334be2f..35b5c80f5d85 100644
--- a/drivers/gpu/drm/virtio/virtgpu_display.c
+++ b/drivers/gpu/drm/virtio/virtgpu_display.c
@@ -123,6 +123,17 @@ static int virtio_gpu_crtc_atomic_check(struct drm_crtc *crtc,
 static void virtio_gpu_crtc_atomic_flush(struct drm_crtc *crtc,
 					 struct drm_crtc_state *old_state)
 {
+	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
+
+	/*
+	 * virtio-gpu can't do modeset and plane update operations
+	 * independant from each other.  So the actual modeset happens
+	 * in the plane update callback, and here we just check
+	 * whenever we must force the modeset.
+	 */
+	if (drm_atomic_crtc_needs_modeset(crtc->state)) {
+		output->needs_modeset = true;
+	}
 }
 
 static const struct drm_crtc_helper_funcs virtio_gpu_crtc_helper_funcs = {
diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h
index 9ff9f4ac0522..4ab1b0ba2925 100644
--- a/drivers/gpu/drm/virtio/virtgpu_drv.h
+++ b/drivers/gpu/drm/virtio/virtgpu_drv.h
@@ -138,6 +138,7 @@ struct virtio_gpu_output {
 	int cur_x;
 	int cur_y;
 	bool enabled;
+	bool needs_modeset;
 };
 #define drm_crtc_to_virtio_gpu_output(x) \
 	container_of(x, struct virtio_gpu_output, crtc)
diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c
index 52d24179bcec..65757409d9ed 100644
--- a/drivers/gpu/drm/virtio/virtgpu_plane.c
+++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
@@ -163,7 +163,9 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
 	    plane->state->src_w != old_state->src_w ||
 	    plane->state->src_h != old_state->src_h ||
 	    plane->state->src_x != old_state->src_x ||
-	    plane->state->src_y != old_state->src_y) {
+	    plane->state->src_y != old_state->src_y ||
+	    output->needs_modeset) {
+		output->needs_modeset = false;
 		DRM_DEBUG("handle 0x%x, crtc %dx%d+%d+%d, src %dx%d+%d+%d\n",
 			  bo->hw_res_handle,
 			  plane->state->crtc_w, plane->state->crtc_h,
-- 
2.28.0

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1882851

Title:
  QEMU video freezes with "Guest disabled display" (virtio driver)

Status in QEMU:
  New

Bug description:
  I am using Arch Linux as my Guest and Host OS, after starting qemu
  with the following command:

    $ qemu-system-x86_64 -enable-kvm -hda arch-zoom.qcow2 -m 4G -vga
  virtio

  and waiting for a screen blank, I get this message:

    Guest disabled display

  And nothing happens after that, I can move the mouse or hit any key,
  and the message is still there.

  I can still reboot the VM but that's not optimal.

  I can reproduce this with the latest QEMU release (5.0.0) or git master, 
  I also tried this with older releases (4.0.0, 3.0.0) and the issue is still there.

  I can't reproduce this with other video drivers (std, qxl).

  With std/qxl the screen will blank a bit and then continue as normal.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1882851/+subscriptions

[PATCH] drm/virtio: fix unblank

[Diego Viola]

From: Gerd Hoffmann <kraxel@redhat.com>

When going through a disable/enable cycle without changing the
framebuffer the optimization added by commit 3954ff10e06e ("drm/virtio:
skip set_scanout if framebuffer didn't change") causes the screen stay
blank.  Add a bool to force an update to fix that.

v2: use drm_atomic_crtc_needs_modeset() (Daniel).

Cc: 1882851@bugs.launchpad.net
Fixes: 3954ff10e06e ("drm/virtio: skip set_scanout if framebuffer didn't change")
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Tested-by: Jiri Slaby <jirislaby@kernel.org>
Tested-by: Diego Viola <diego.viola@gmail.com>
---
 drivers/gpu/drm/virtio/virtgpu_display.c | 11 +++++++++++
 drivers/gpu/drm/virtio/virtgpu_drv.h     |  1 +
 drivers/gpu/drm/virtio/virtgpu_plane.c   |  4 +++-
 3 files changed, 15 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/virtio/virtgpu_display.c b/drivers/gpu/drm/virtio/virtgpu_display.c
index af55b334be2f..35b5c80f5d85 100644
--- a/drivers/gpu/drm/virtio/virtgpu_display.c
+++ b/drivers/gpu/drm/virtio/virtgpu_display.c
@@ -123,6 +123,17 @@ static int virtio_gpu_crtc_atomic_check(struct drm_crtc *crtc,
 static void virtio_gpu_crtc_atomic_flush(struct drm_crtc *crtc,
 					 struct drm_crtc_state *old_state)
 {
+	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
+
+	/*
+	 * virtio-gpu can't do modeset and plane update operations
+	 * independant from each other.  So the actual modeset happens
+	 * in the plane update callback, and here we just check
+	 * whenever we must force the modeset.
+	 */
+	if (drm_atomic_crtc_needs_modeset(crtc->state)) {
+		output->needs_modeset = true;
+	}
 }
 
 static const struct drm_crtc_helper_funcs virtio_gpu_crtc_helper_funcs = {
diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h
index 9ff9f4ac0522..4ab1b0ba2925 100644
--- a/drivers/gpu/drm/virtio/virtgpu_drv.h
+++ b/drivers/gpu/drm/virtio/virtgpu_drv.h
@@ -138,6 +138,7 @@ struct virtio_gpu_output {
 	int cur_x;
 	int cur_y;
 	bool enabled;
+	bool needs_modeset;
 };
 #define drm_crtc_to_virtio_gpu_output(x) \
 	container_of(x, struct virtio_gpu_output, crtc)
diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c
index 52d24179bcec..65757409d9ed 100644
--- a/drivers/gpu/drm/virtio/virtgpu_plane.c
+++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
@@ -163,7 +163,9 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
 	    plane->state->src_w != old_state->src_w ||
 	    plane->state->src_h != old_state->src_h ||
 	    plane->state->src_x != old_state->src_x ||
-	    plane->state->src_y != old_state->src_y) {
+	    plane->state->src_y != old_state->src_y ||
+	    output->needs_modeset) {
+		output->needs_modeset = false;
 		DRM_DEBUG("handle 0x%x, crtc %dx%d+%d+%d, src %dx%d+%d+%d\n",
 			  bo->hw_res_handle,
 			  plane->state->crtc_w, plane->state->crtc_h,
-- 
2.28.0

_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

Re: [PATCH 1/2] drm/virtio: fix unblank

[Gerd Hoffmann]

On Mon, Aug 24, 2020 at 09:24:40AM +0200, Jiri Slaby wrote:
> On 18. 08. 20, 9:25, Gerd Hoffmann wrote:
> > When going through a disable/enable cycle without changing the
> > framebuffer the optimization added by commit 3954ff10e06e ("drm/virtio:
> > skip set_scanout if framebuffer didn't change") causes the screen stay
> > blank.  Add a bool to force an update to fix that.
> > 
> > v2: use drm_atomic_crtc_needs_modeset() (Daniel).
> > 
> > Cc: 1882851@bugs.launchpad.net
> > Fixes: 3954ff10e06e ("drm/virtio: skip set_scanout if framebuffer didn't change")
> > Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
> 
> Tested-by: Jiri Slaby <jirislaby@kernel.org>

Ping.  Need an ack or an review to merge this.

thanks,
  Gerd

[Bug 1882851] Re: [PATCH 1/2] drm/virtio: fix unblank

[Gerd Hoffmann]

On Mon, Aug 24, 2020 at 09:24:40AM +0200, Jiri Slaby wrote:
> On 18. 08. 20, 9:25, Gerd Hoffmann wrote:
> > When going through a disable/enable cycle without changing the
> > framebuffer the optimization added by commit 3954ff10e06e ("drm/virtio:
> > skip set_scanout if framebuffer didn't change") causes the screen stay
> > blank.  Add a bool to force an update to fix that.
> > 
> > v2: use drm_atomic_crtc_needs_modeset() (Daniel).
> > 
> > Cc: 1882851@bugs.launchpad.net
> > Fixes: 3954ff10e06e ("drm/virtio: skip set_scanout if framebuffer didn't change")
> > Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
> 
> Tested-by: Jiri Slaby <jirislaby@kernel.org>

Ping.  Need an ack or an review to merge this.

thanks,
  Gerd

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1882851

Title:
  QEMU video freezes with "Guest disabled display" (virtio driver)

Status in QEMU:
  New

Bug description:
  I am using Arch Linux as my Guest and Host OS, after starting qemu
  with the following command:

    $ qemu-system-x86_64 -enable-kvm -hda arch-zoom.qcow2 -m 4G -vga
  virtio

  and waiting for a screen blank, I get this message:

    Guest disabled display

  And nothing happens after that, I can move the mouse or hit any key,
  and the message is still there.

  I can still reboot the VM but that's not optimal.

  I can reproduce this with the latest QEMU release (5.0.0) or git master, 
  I also tried this with older releases (4.0.0, 3.0.0) and the issue is still there.

  I can't reproduce this with other video drivers (std, qxl).

  With std/qxl the screen will blank a bit and then continue as normal.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1882851/+subscriptions

Re: [PATCH 1/2] drm/virtio: fix unblank

[Gerd Hoffmann]

On Mon, Aug 24, 2020 at 09:24:40AM +0200, Jiri Slaby wrote:
> On 18. 08. 20, 9:25, Gerd Hoffmann wrote:
> > When going through a disable/enable cycle without changing the
> > framebuffer the optimization added by commit 3954ff10e06e ("drm/virtio:
> > skip set_scanout if framebuffer didn't change") causes the screen stay
> > blank.  Add a bool to force an update to fix that.
> > 
> > v2: use drm_atomic_crtc_needs_modeset() (Daniel).
> > 
> > Cc: 1882851@bugs.launchpad.net
> > Fixes: 3954ff10e06e ("drm/virtio: skip set_scanout if framebuffer didn't change")
> > Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
> 
> Tested-by: Jiri Slaby <jirislaby@kernel.org>

Ping.  Need an ack or an review to merge this.

thanks,
  Gerd

_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

Re: [PATCH 1/2] drm/virtio: fix unblank

[Gerd Hoffmann]

On Mon, Aug 24, 2020 at 09:24:40AM +0200, Jiri Slaby wrote:
> On 18. 08. 20, 9:25, Gerd Hoffmann wrote:
> > When going through a disable/enable cycle without changing the
> > framebuffer the optimization added by commit 3954ff10e06e ("drm/virtio:
> > skip set_scanout if framebuffer didn't change") causes the screen stay
> > blank.  Add a bool to force an update to fix that.
> > 
> > v2: use drm_atomic_crtc_needs_modeset() (Daniel).
> > 
> > Cc: 1882851@bugs.launchpad.net
> > Fixes: 3954ff10e06e ("drm/virtio: skip set_scanout if framebuffer didn't change")
> > Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
> 
> Tested-by: Jiri Slaby <jirislaby@kernel.org>

Ping.  Need an ack or an review to merge this.

thanks,
  Gerd

_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

Re: [PATCH 1/2] drm/virtio: fix unblank

[Daniel Vetter]

On Fri, Aug 28, 2020 at 01:27:59PM +0200, Gerd Hoffmann wrote:
> On Mon, Aug 24, 2020 at 09:24:40AM +0200, Jiri Slaby wrote:
> > On 18. 08. 20, 9:25, Gerd Hoffmann wrote:
> > > When going through a disable/enable cycle without changing the
> > > framebuffer the optimization added by commit 3954ff10e06e ("drm/virtio:
> > > skip set_scanout if framebuffer didn't change") causes the screen stay
> > > blank.  Add a bool to force an update to fix that.
> > > 
> > > v2: use drm_atomic_crtc_needs_modeset() (Daniel).
> > > 
> > > Cc: 1882851@bugs.launchpad.net
> > > Fixes: 3954ff10e06e ("drm/virtio: skip set_scanout if framebuffer didn't change")
> > > Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
> > 
> > Tested-by: Jiri Slaby <jirislaby@kernel.org>
> 
> Ping.  Need an ack or an review to merge this.

In case you still need that, on both patches:

Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
> 
> thanks,
>   Gerd
> 
> _______________________________________________
> dri-devel mailing list
> dri-devel@lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/dri-devel

-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch

Re: [PATCH 1/2] drm/virtio: fix unblank

[Daniel Vetter]

On Fri, Aug 28, 2020 at 01:27:59PM +0200, Gerd Hoffmann wrote:
> On Mon, Aug 24, 2020 at 09:24:40AM +0200, Jiri Slaby wrote:
> > On 18. 08. 20, 9:25, Gerd Hoffmann wrote:
> > > When going through a disable/enable cycle without changing the
> > > framebuffer the optimization added by commit 3954ff10e06e ("drm/virtio:
> > > skip set_scanout if framebuffer didn't change") causes the screen stay
> > > blank.  Add a bool to force an update to fix that.
> > > 
> > > v2: use drm_atomic_crtc_needs_modeset() (Daniel).
> > > 
> > > Cc: 1882851@bugs.launchpad.net
> > > Fixes: 3954ff10e06e ("drm/virtio: skip set_scanout if framebuffer didn't change")
> > > Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
> > 
> > Tested-by: Jiri Slaby <jirislaby@kernel.org>
> 
> Ping.  Need an ack or an review to merge this.

In case you still need that, on both patches:

Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
> 
> thanks,
>   Gerd
> 
> _______________________________________________
> dri-devel mailing list
> dri-devel@lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/dri-devel

-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch
_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

Re: [PATCH 1/2] drm/virtio: fix unblank

[Daniel Vetter]

On Fri, Aug 28, 2020 at 01:27:59PM +0200, Gerd Hoffmann wrote:
> On Mon, Aug 24, 2020 at 09:24:40AM +0200, Jiri Slaby wrote:
> > On 18. 08. 20, 9:25, Gerd Hoffmann wrote:
> > > When going through a disable/enable cycle without changing the
> > > framebuffer the optimization added by commit 3954ff10e06e ("drm/virtio:
> > > skip set_scanout if framebuffer didn't change") causes the screen stay
> > > blank.  Add a bool to force an update to fix that.
> > > 
> > > v2: use drm_atomic_crtc_needs_modeset() (Daniel).
> > > 
> > > Cc: 1882851@bugs.launchpad.net
> > > Fixes: 3954ff10e06e ("drm/virtio: skip set_scanout if framebuffer didn't change")
> > > Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
> > 
> > Tested-by: Jiri Slaby <jirislaby@kernel.org>
> 
> Ping.  Need an ack or an review to merge this.

In case you still need that, on both patches:

Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
> 
> thanks,
>   Gerd
> 
> _______________________________________________
> dri-devel mailing list
> dri-devel@lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/dri-devel

-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

Re: [PATCH 1/2] drm/virtio: fix unblank

[Daniel Vetter]

On Tue, Aug 18, 2020 at 09:25:10AM +0200, Gerd Hoffmann wrote:
> When going through a disable/enable cycle without changing the
> framebuffer the optimization added by commit 3954ff10e06e ("drm/virtio:
> skip set_scanout if framebuffer didn't change") causes the screen stay
> blank.  Add a bool to force an update to fix that.
> 
> v2: use drm_atomic_crtc_needs_modeset() (Daniel).
> 
> Cc: 1882851@bugs.launchpad.net
> Fixes: 3954ff10e06e ("drm/virtio: skip set_scanout if framebuffer didn't change")
> Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
> ---
>  drivers/gpu/drm/virtio/virtgpu_drv.h     |  1 +
>  drivers/gpu/drm/virtio/virtgpu_display.c | 11 +++++++++++
>  drivers/gpu/drm/virtio/virtgpu_plane.c   |  4 +++-
>  3 files changed, 15 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h
> index 9ff9f4ac0522..4ab1b0ba2925 100644
> --- a/drivers/gpu/drm/virtio/virtgpu_drv.h
> +++ b/drivers/gpu/drm/virtio/virtgpu_drv.h
> @@ -138,6 +138,7 @@ struct virtio_gpu_output {
>  	int cur_x;
>  	int cur_y;
>  	bool enabled;
> +	bool needs_modeset;

Maybe for a follow-up in -next: The clean atomic way of doing this is to
put this into a virtio_crtc_state, compute it in atomic_check, and then
fish it out (through old_state->state lookup, somewhat contrived I know)
in the commit side. Putting random atomic commit state tracking stuff into
non-state structures without appropriate amounts of locks is kinda iffy
and means more work for reviewers pondering whether it all works
correctly.

Cheers, Daniel


>  };
>  #define drm_crtc_to_virtio_gpu_output(x) \
>  	container_of(x, struct virtio_gpu_output, crtc)
> diff --git a/drivers/gpu/drm/virtio/virtgpu_display.c b/drivers/gpu/drm/virtio/virtgpu_display.c
> index 2c2742b8d657..6c26b41f4e0d 100644
> --- a/drivers/gpu/drm/virtio/virtgpu_display.c
> +++ b/drivers/gpu/drm/virtio/virtgpu_display.c
> @@ -123,6 +123,17 @@ static int virtio_gpu_crtc_atomic_check(struct drm_crtc *crtc,
>  static void virtio_gpu_crtc_atomic_flush(struct drm_crtc *crtc,
>  					 struct drm_crtc_state *old_state)
>  {
> +	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
> +
> +	/*
> +	 * virtio-gpu can't do modeset and plane update operations
> +	 * independant from each other.  So the actual modeset happens
> +	 * in the plane update callback, and here we just check
> +	 * whenever we must force the modeset.
> +	 */
> +	if (drm_atomic_crtc_needs_modeset(crtc->state)) {
> +		output->needs_modeset = true;
> +	}
>  }
>  
>  static const struct drm_crtc_helper_funcs virtio_gpu_crtc_helper_funcs = {
> diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c
> index 52d24179bcec..65757409d9ed 100644
> --- a/drivers/gpu/drm/virtio/virtgpu_plane.c
> +++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
> @@ -163,7 +163,9 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
>  	    plane->state->src_w != old_state->src_w ||
>  	    plane->state->src_h != old_state->src_h ||
>  	    plane->state->src_x != old_state->src_x ||
> -	    plane->state->src_y != old_state->src_y) {
> +	    plane->state->src_y != old_state->src_y ||
> +	    output->needs_modeset) {
> +		output->needs_modeset = false;
>  		DRM_DEBUG("handle 0x%x, crtc %dx%d+%d+%d, src %dx%d+%d+%d\n",
>  			  bo->hw_res_handle,
>  			  plane->state->crtc_w, plane->state->crtc_h,
> -- 
> 2.18.4
> 

-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch

Re: [PATCH 1/2] drm/virtio: fix unblank

[Daniel Vetter]

On Tue, Aug 18, 2020 at 09:25:10AM +0200, Gerd Hoffmann wrote:
> When going through a disable/enable cycle without changing the
> framebuffer the optimization added by commit 3954ff10e06e ("drm/virtio:
> skip set_scanout if framebuffer didn't change") causes the screen stay
> blank.  Add a bool to force an update to fix that.
> 
> v2: use drm_atomic_crtc_needs_modeset() (Daniel).
> 
> Cc: 1882851@bugs.launchpad.net
> Fixes: 3954ff10e06e ("drm/virtio: skip set_scanout if framebuffer didn't change")
> Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
> ---
>  drivers/gpu/drm/virtio/virtgpu_drv.h     |  1 +
>  drivers/gpu/drm/virtio/virtgpu_display.c | 11 +++++++++++
>  drivers/gpu/drm/virtio/virtgpu_plane.c   |  4 +++-
>  3 files changed, 15 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h
> index 9ff9f4ac0522..4ab1b0ba2925 100644
> --- a/drivers/gpu/drm/virtio/virtgpu_drv.h
> +++ b/drivers/gpu/drm/virtio/virtgpu_drv.h
> @@ -138,6 +138,7 @@ struct virtio_gpu_output {
>  	int cur_x;
>  	int cur_y;
>  	bool enabled;
> +	bool needs_modeset;

Maybe for a follow-up in -next: The clean atomic way of doing this is to
put this into a virtio_crtc_state, compute it in atomic_check, and then
fish it out (through old_state->state lookup, somewhat contrived I know)
in the commit side. Putting random atomic commit state tracking stuff into
non-state structures without appropriate amounts of locks is kinda iffy
and means more work for reviewers pondering whether it all works
correctly.

Cheers, Daniel


>  };
>  #define drm_crtc_to_virtio_gpu_output(x) \
>  	container_of(x, struct virtio_gpu_output, crtc)
> diff --git a/drivers/gpu/drm/virtio/virtgpu_display.c b/drivers/gpu/drm/virtio/virtgpu_display.c
> index 2c2742b8d657..6c26b41f4e0d 100644
> --- a/drivers/gpu/drm/virtio/virtgpu_display.c
> +++ b/drivers/gpu/drm/virtio/virtgpu_display.c
> @@ -123,6 +123,17 @@ static int virtio_gpu_crtc_atomic_check(struct drm_crtc *crtc,
>  static void virtio_gpu_crtc_atomic_flush(struct drm_crtc *crtc,
>  					 struct drm_crtc_state *old_state)
>  {
> +	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
> +
> +	/*
> +	 * virtio-gpu can't do modeset and plane update operations
> +	 * independant from each other.  So the actual modeset happens
> +	 * in the plane update callback, and here we just check
> +	 * whenever we must force the modeset.
> +	 */
> +	if (drm_atomic_crtc_needs_modeset(crtc->state)) {
> +		output->needs_modeset = true;
> +	}
>  }
>  
>  static const struct drm_crtc_helper_funcs virtio_gpu_crtc_helper_funcs = {
> diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c
> index 52d24179bcec..65757409d9ed 100644
> --- a/drivers/gpu/drm/virtio/virtgpu_plane.c
> +++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
> @@ -163,7 +163,9 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
>  	    plane->state->src_w != old_state->src_w ||
>  	    plane->state->src_h != old_state->src_h ||
>  	    plane->state->src_x != old_state->src_x ||
> -	    plane->state->src_y != old_state->src_y) {
> +	    plane->state->src_y != old_state->src_y ||
> +	    output->needs_modeset) {
> +		output->needs_modeset = false;
>  		DRM_DEBUG("handle 0x%x, crtc %dx%d+%d+%d, src %dx%d+%d+%d\n",
>  			  bo->hw_res_handle,
>  			  plane->state->crtc_w, plane->state->crtc_h,
> -- 
> 2.18.4
> 

-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch
_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

Re: [PATCH 1/2] drm/virtio: fix unblank

[Daniel Vetter]

On Tue, Aug 18, 2020 at 09:25:10AM +0200, Gerd Hoffmann wrote:
> When going through a disable/enable cycle without changing the
> framebuffer the optimization added by commit 3954ff10e06e ("drm/virtio:
> skip set_scanout if framebuffer didn't change") causes the screen stay
> blank.  Add a bool to force an update to fix that.
> 
> v2: use drm_atomic_crtc_needs_modeset() (Daniel).
> 
> Cc: 1882851@bugs.launchpad.net
> Fixes: 3954ff10e06e ("drm/virtio: skip set_scanout if framebuffer didn't change")
> Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
> ---
>  drivers/gpu/drm/virtio/virtgpu_drv.h     |  1 +
>  drivers/gpu/drm/virtio/virtgpu_display.c | 11 +++++++++++
>  drivers/gpu/drm/virtio/virtgpu_plane.c   |  4 +++-
>  3 files changed, 15 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h
> index 9ff9f4ac0522..4ab1b0ba2925 100644
> --- a/drivers/gpu/drm/virtio/virtgpu_drv.h
> +++ b/drivers/gpu/drm/virtio/virtgpu_drv.h
> @@ -138,6 +138,7 @@ struct virtio_gpu_output {
>  	int cur_x;
>  	int cur_y;
>  	bool enabled;
> +	bool needs_modeset;

Maybe for a follow-up in -next: The clean atomic way of doing this is to
put this into a virtio_crtc_state, compute it in atomic_check, and then
fish it out (through old_state->state lookup, somewhat contrived I know)
in the commit side. Putting random atomic commit state tracking stuff into
non-state structures without appropriate amounts of locks is kinda iffy
and means more work for reviewers pondering whether it all works
correctly.

Cheers, Daniel


>  };
>  #define drm_crtc_to_virtio_gpu_output(x) \
>  	container_of(x, struct virtio_gpu_output, crtc)
> diff --git a/drivers/gpu/drm/virtio/virtgpu_display.c b/drivers/gpu/drm/virtio/virtgpu_display.c
> index 2c2742b8d657..6c26b41f4e0d 100644
> --- a/drivers/gpu/drm/virtio/virtgpu_display.c
> +++ b/drivers/gpu/drm/virtio/virtgpu_display.c
> @@ -123,6 +123,17 @@ static int virtio_gpu_crtc_atomic_check(struct drm_crtc *crtc,
>  static void virtio_gpu_crtc_atomic_flush(struct drm_crtc *crtc,
>  					 struct drm_crtc_state *old_state)
>  {
> +	struct virtio_gpu_output *output = drm_crtc_to_virtio_gpu_output(crtc);
> +
> +	/*
> +	 * virtio-gpu can't do modeset and plane update operations
> +	 * independant from each other.  So the actual modeset happens
> +	 * in the plane update callback, and here we just check
> +	 * whenever we must force the modeset.
> +	 */
> +	if (drm_atomic_crtc_needs_modeset(crtc->state)) {
> +		output->needs_modeset = true;
> +	}
>  }
>  
>  static const struct drm_crtc_helper_funcs virtio_gpu_crtc_helper_funcs = {
> diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c
> index 52d24179bcec..65757409d9ed 100644
> --- a/drivers/gpu/drm/virtio/virtgpu_plane.c
> +++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
> @@ -163,7 +163,9 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
>  	    plane->state->src_w != old_state->src_w ||
>  	    plane->state->src_h != old_state->src_h ||
>  	    plane->state->src_x != old_state->src_x ||
> -	    plane->state->src_y != old_state->src_y) {
> +	    plane->state->src_y != old_state->src_y ||
> +	    output->needs_modeset) {
> +		output->needs_modeset = false;
>  		DRM_DEBUG("handle 0x%x, crtc %dx%d+%d+%d, src %dx%d+%d+%d\n",
>  			  bo->hw_res_handle,
>  			  plane->state->crtc_w, plane->state->crtc_h,
> -- 
> 2.18.4
> 

-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

[Bug 1882851] Re: QEMU video freezes with "Guest disabled display" (virtio driver)

[Diego Viola]

This bug is now fixed with Linux 5.9-rc5, thank you.

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1882851

Title:
  QEMU video freezes with "Guest disabled display" (virtio driver)

Status in QEMU:
  New

Bug description:
  I am using Arch Linux as my Guest and Host OS, after starting qemu
  with the following command:

    $ qemu-system-x86_64 -enable-kvm -hda arch-zoom.qcow2 -m 4G -vga
  virtio

  and waiting for a screen blank, I get this message:

    Guest disabled display

  And nothing happens after that, I can move the mouse or hit any key,
  and the message is still there.

  I can still reboot the VM but that's not optimal.

  I can reproduce this with the latest QEMU release (5.0.0) or git master, 
  I also tried this with older releases (4.0.0, 3.0.0) and the issue is still there.

  I can't reproduce this with other video drivers (std, qxl).

  With std/qxl the screen will blank a bit and then continue as normal.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1882851/+subscriptions

[Bug 1882851] Re: QEMU video freezes with "Guest disabled display" (virtio driver)

[Thomas Huth]

** Changed in: qemu
       Status: New => Fix Released

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1882851

Title:
  QEMU video freezes with "Guest disabled display" (virtio driver)

Status in QEMU:
  Fix Released

Bug description:
  I am using Arch Linux as my Guest and Host OS, after starting qemu
  with the following command:

    $ qemu-system-x86_64 -enable-kvm -hda arch-zoom.qcow2 -m 4G -vga
  virtio

  and waiting for a screen blank, I get this message:

    Guest disabled display

  And nothing happens after that, I can move the mouse or hit any key,
  and the message is still there.

  I can still reboot the VM but that's not optimal.

  I can reproduce this with the latest QEMU release (5.0.0) or git master, 
  I also tried this with older releases (4.0.0, 3.0.0) and the issue is still there.

  I can't reproduce this with other video drivers (std, qxl).

  With std/qxl the screen will blank a bit and then continue as normal.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1882851/+subscriptions