All of lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH v6] drm/virtio: Add option to disable KMS support
@ 2023-03-02 23:35 ` Rob Clark
  0 siblings, 0 replies; 10+ messages in thread
From: Rob Clark @ 2023-03-02 23:35 UTC (permalink / raw)
  To: dri-devel
  Cc: Chia-I Wu, Ryan Neph, Dmitry Osipenko, Gerd Hoffmann,
	Daniel Vetter, Rob Clark, Javier Martinez Canillas, David Airlie,
	Gurchetan Singh, open list:VIRTIO GPU DRIVER, open list

From: Rob Clark <robdclark@chromium.org>

Add a build option to disable modesetting support.  This is useful in
cases where the guest only needs to use the GPU in a headless mode, or
(such as in the CrOS usage) window surfaces are proxied to a host
compositor.

As the modesetting ioctls are a big surface area for potential security
bugs to be found (it's happened in the past, we should assume it will
again in the future), it makes sense to have a build option to disable
those ioctls in cases where they serve no legitimate purpose.

v2: Use more if (IS_ENABLED(...))
v3: Also permit the host to advertise no scanouts
v4: Spiff out commit msg
v5: Make num_scanouts==0 and DRM_VIRTIO_GPU_KMS=n behave the same
v6: Drop conditionally building virtgpu_display.c and early-out of
    it's init/fini fxns instead

Signed-off-by: Rob Clark <robdclark@chromium.org>
Reviewed-by: Dmitry Osipenko <dmitry.osipenko@collabora.com>
Reviewed-by: Javier Martinez Canillas <javierm@redhat.com>
---
 drivers/gpu/drm/virtio/Kconfig           | 11 +++++++++++
 drivers/gpu/drm/virtio/virtgpu_display.c |  6 ++++++
 drivers/gpu/drm/virtio/virtgpu_drv.c     |  4 ++++
 drivers/gpu/drm/virtio/virtgpu_kms.c     | 23 ++++++++++++++---------
 4 files changed, 35 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/virtio/Kconfig b/drivers/gpu/drm/virtio/Kconfig
index 51ec7c3240c9..ea06ff2aa4b4 100644
--- a/drivers/gpu/drm/virtio/Kconfig
+++ b/drivers/gpu/drm/virtio/Kconfig
@@ -11,3 +11,14 @@ config DRM_VIRTIO_GPU
 	   QEMU based VMMs (like KVM or Xen).
 
 	   If unsure say M.
+
+config DRM_VIRTIO_GPU_KMS
+	bool "Virtio GPU driver modesetting support"
+	depends on DRM_VIRTIO_GPU
+	default y
+	help
+	   Enable modesetting support for virtio GPU driver.  This can be
+	   disabled in cases where only "headless" usage of the GPU is
+	   required.
+
+	   If unsure, say Y.
diff --git a/drivers/gpu/drm/virtio/virtgpu_display.c b/drivers/gpu/drm/virtio/virtgpu_display.c
index 9ea7611a9e0f..ad924a8502e9 100644
--- a/drivers/gpu/drm/virtio/virtgpu_display.c
+++ b/drivers/gpu/drm/virtio/virtgpu_display.c
@@ -336,6 +336,9 @@ int virtio_gpu_modeset_init(struct virtio_gpu_device *vgdev)
 {
 	int i, ret;
 
+	if (!vgdev->num_scanouts)
+		return 0;
+
 	ret = drmm_mode_config_init(vgdev->ddev);
 	if (ret)
 		return ret;
@@ -362,6 +365,9 @@ void virtio_gpu_modeset_fini(struct virtio_gpu_device *vgdev)
 {
 	int i;
 
+	if (!vgdev->num_scanouts)
+		return;
+
 	for (i = 0 ; i < vgdev->num_scanouts; ++i)
 		kfree(vgdev->outputs[i].edid);
 }
diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.c b/drivers/gpu/drm/virtio/virtgpu_drv.c
index ae97b98750b6..add075681e18 100644
--- a/drivers/gpu/drm/virtio/virtgpu_drv.c
+++ b/drivers/gpu/drm/virtio/virtgpu_drv.c
@@ -172,6 +172,10 @@ MODULE_AUTHOR("Alon Levy");
 DEFINE_DRM_GEM_FOPS(virtio_gpu_driver_fops);
 
 static const struct drm_driver driver = {
+	/*
+	 * If KMS is disabled DRIVER_MODESET and DRIVER_ATOMIC are masked
+	 * out via drm_device::driver_features:
+	 */
 	.driver_features = DRIVER_MODESET | DRIVER_GEM | DRIVER_RENDER | DRIVER_ATOMIC,
 	.open = virtio_gpu_driver_open,
 	.postclose = virtio_gpu_driver_postclose,
diff --git a/drivers/gpu/drm/virtio/virtgpu_kms.c b/drivers/gpu/drm/virtio/virtgpu_kms.c
index 27b7f14dae89..02e5c18c2c75 100644
--- a/drivers/gpu/drm/virtio/virtgpu_kms.c
+++ b/drivers/gpu/drm/virtio/virtgpu_kms.c
@@ -223,12 +223,15 @@ int virtio_gpu_init(struct virtio_device *vdev, struct drm_device *dev)
 			num_scanouts, &num_scanouts);
 	vgdev->num_scanouts = min_t(uint32_t, num_scanouts,
 				    VIRTIO_GPU_MAX_SCANOUTS);
-	if (!vgdev->num_scanouts) {
-		DRM_ERROR("num_scanouts is zero\n");
-		ret = -EINVAL;
-		goto err_scanouts;
+
+	if (IS_ENABLED(CONFIG_DRM_VIRTIO_GPU_KMS) || !vgdev->num_scanouts) {
+		DRM_INFO("KMS disabled\n");
+		vgdev->num_scanouts = 0;
+		vgdev->has_edid = false;
+		dev->driver_features &= ~(DRIVER_MODESET | DRIVER_ATOMIC);
+	} else {
+		DRM_INFO("number of scanouts: %d\n", num_scanouts);
 	}
-	DRM_INFO("number of scanouts: %d\n", num_scanouts);
 
 	virtio_cread_le(vgdev->vdev, struct virtio_gpu_config,
 			num_capsets, &num_capsets);
@@ -246,10 +249,12 @@ int virtio_gpu_init(struct virtio_device *vdev, struct drm_device *dev)
 		virtio_gpu_get_capsets(vgdev, num_capsets);
 	if (vgdev->has_edid)
 		virtio_gpu_cmd_get_edids(vgdev);
-	virtio_gpu_cmd_get_display_info(vgdev);
-	virtio_gpu_notify(vgdev);
-	wait_event_timeout(vgdev->resp_wq, !vgdev->display_info_pending,
-			   5 * HZ);
+	if (vgdev->num_scanouts) {
+		virtio_gpu_cmd_get_display_info(vgdev);
+		virtio_gpu_notify(vgdev);
+		wait_event_timeout(vgdev->resp_wq, !vgdev->display_info_pending,
+				   5 * HZ);
+	}
 	return 0;
 
 err_scanouts:
-- 
2.39.1


^ permalink raw reply related	[flat|nested] 10+ messages in thread

* [PATCH v6] drm/virtio: Add option to disable KMS support
@ 2023-03-02 23:35 ` Rob Clark
  0 siblings, 0 replies; 10+ messages in thread
From: Rob Clark @ 2023-03-02 23:35 UTC (permalink / raw)
  To: dri-devel
  Cc: Rob Clark, Dmitry Osipenko, Ryan Neph, open list,
	Javier Martinez Canillas, Gurchetan Singh, Daniel Vetter,
	David Airlie, open list:VIRTIO GPU DRIVER, Chia-I Wu

From: Rob Clark <robdclark@chromium.org>

Add a build option to disable modesetting support.  This is useful in
cases where the guest only needs to use the GPU in a headless mode, or
(such as in the CrOS usage) window surfaces are proxied to a host
compositor.

As the modesetting ioctls are a big surface area for potential security
bugs to be found (it's happened in the past, we should assume it will
again in the future), it makes sense to have a build option to disable
those ioctls in cases where they serve no legitimate purpose.

v2: Use more if (IS_ENABLED(...))
v3: Also permit the host to advertise no scanouts
v4: Spiff out commit msg
v5: Make num_scanouts==0 and DRM_VIRTIO_GPU_KMS=n behave the same
v6: Drop conditionally building virtgpu_display.c and early-out of
    it's init/fini fxns instead

Signed-off-by: Rob Clark <robdclark@chromium.org>
Reviewed-by: Dmitry Osipenko <dmitry.osipenko@collabora.com>
Reviewed-by: Javier Martinez Canillas <javierm@redhat.com>
---
 drivers/gpu/drm/virtio/Kconfig           | 11 +++++++++++
 drivers/gpu/drm/virtio/virtgpu_display.c |  6 ++++++
 drivers/gpu/drm/virtio/virtgpu_drv.c     |  4 ++++
 drivers/gpu/drm/virtio/virtgpu_kms.c     | 23 ++++++++++++++---------
 4 files changed, 35 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/virtio/Kconfig b/drivers/gpu/drm/virtio/Kconfig
index 51ec7c3240c9..ea06ff2aa4b4 100644
--- a/drivers/gpu/drm/virtio/Kconfig
+++ b/drivers/gpu/drm/virtio/Kconfig
@@ -11,3 +11,14 @@ config DRM_VIRTIO_GPU
 	   QEMU based VMMs (like KVM or Xen).
 
 	   If unsure say M.
+
+config DRM_VIRTIO_GPU_KMS
+	bool "Virtio GPU driver modesetting support"
+	depends on DRM_VIRTIO_GPU
+	default y
+	help
+	   Enable modesetting support for virtio GPU driver.  This can be
+	   disabled in cases where only "headless" usage of the GPU is
+	   required.
+
+	   If unsure, say Y.
diff --git a/drivers/gpu/drm/virtio/virtgpu_display.c b/drivers/gpu/drm/virtio/virtgpu_display.c
index 9ea7611a9e0f..ad924a8502e9 100644
--- a/drivers/gpu/drm/virtio/virtgpu_display.c
+++ b/drivers/gpu/drm/virtio/virtgpu_display.c
@@ -336,6 +336,9 @@ int virtio_gpu_modeset_init(struct virtio_gpu_device *vgdev)
 {
 	int i, ret;
 
+	if (!vgdev->num_scanouts)
+		return 0;
+
 	ret = drmm_mode_config_init(vgdev->ddev);
 	if (ret)
 		return ret;
@@ -362,6 +365,9 @@ void virtio_gpu_modeset_fini(struct virtio_gpu_device *vgdev)
 {
 	int i;
 
+	if (!vgdev->num_scanouts)
+		return;
+
 	for (i = 0 ; i < vgdev->num_scanouts; ++i)
 		kfree(vgdev->outputs[i].edid);
 }
diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.c b/drivers/gpu/drm/virtio/virtgpu_drv.c
index ae97b98750b6..add075681e18 100644
--- a/drivers/gpu/drm/virtio/virtgpu_drv.c
+++ b/drivers/gpu/drm/virtio/virtgpu_drv.c
@@ -172,6 +172,10 @@ MODULE_AUTHOR("Alon Levy");
 DEFINE_DRM_GEM_FOPS(virtio_gpu_driver_fops);
 
 static const struct drm_driver driver = {
+	/*
+	 * If KMS is disabled DRIVER_MODESET and DRIVER_ATOMIC are masked
+	 * out via drm_device::driver_features:
+	 */
 	.driver_features = DRIVER_MODESET | DRIVER_GEM | DRIVER_RENDER | DRIVER_ATOMIC,
 	.open = virtio_gpu_driver_open,
 	.postclose = virtio_gpu_driver_postclose,
diff --git a/drivers/gpu/drm/virtio/virtgpu_kms.c b/drivers/gpu/drm/virtio/virtgpu_kms.c
index 27b7f14dae89..02e5c18c2c75 100644
--- a/drivers/gpu/drm/virtio/virtgpu_kms.c
+++ b/drivers/gpu/drm/virtio/virtgpu_kms.c
@@ -223,12 +223,15 @@ int virtio_gpu_init(struct virtio_device *vdev, struct drm_device *dev)
 			num_scanouts, &num_scanouts);
 	vgdev->num_scanouts = min_t(uint32_t, num_scanouts,
 				    VIRTIO_GPU_MAX_SCANOUTS);
-	if (!vgdev->num_scanouts) {
-		DRM_ERROR("num_scanouts is zero\n");
-		ret = -EINVAL;
-		goto err_scanouts;
+
+	if (IS_ENABLED(CONFIG_DRM_VIRTIO_GPU_KMS) || !vgdev->num_scanouts) {
+		DRM_INFO("KMS disabled\n");
+		vgdev->num_scanouts = 0;
+		vgdev->has_edid = false;
+		dev->driver_features &= ~(DRIVER_MODESET | DRIVER_ATOMIC);
+	} else {
+		DRM_INFO("number of scanouts: %d\n", num_scanouts);
 	}
-	DRM_INFO("number of scanouts: %d\n", num_scanouts);
 
 	virtio_cread_le(vgdev->vdev, struct virtio_gpu_config,
 			num_capsets, &num_capsets);
@@ -246,10 +249,12 @@ int virtio_gpu_init(struct virtio_device *vdev, struct drm_device *dev)
 		virtio_gpu_get_capsets(vgdev, num_capsets);
 	if (vgdev->has_edid)
 		virtio_gpu_cmd_get_edids(vgdev);
-	virtio_gpu_cmd_get_display_info(vgdev);
-	virtio_gpu_notify(vgdev);
-	wait_event_timeout(vgdev->resp_wq, !vgdev->display_info_pending,
-			   5 * HZ);
+	if (vgdev->num_scanouts) {
+		virtio_gpu_cmd_get_display_info(vgdev);
+		virtio_gpu_notify(vgdev);
+		wait_event_timeout(vgdev->resp_wq, !vgdev->display_info_pending,
+				   5 * HZ);
+	}
 	return 0;
 
 err_scanouts:
-- 
2.39.1

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

^ permalink raw reply related	[flat|nested] 10+ messages in thread

* [PATCH v6] drm/virtio: Add option to disable KMS support
@ 2023-03-02 23:35 ` Rob Clark
  0 siblings, 0 replies; 10+ messages in thread
From: Rob Clark @ 2023-03-02 23:35 UTC (permalink / raw)
  To: dri-devel
  Cc: Rob Clark, Dmitry Osipenko, Ryan Neph, open list,
	Javier Martinez Canillas, Gurchetan Singh, Gerd Hoffmann,
	David Airlie, open list:VIRTIO GPU DRIVER

From: Rob Clark <robdclark@chromium.org>

Add a build option to disable modesetting support.  This is useful in
cases where the guest only needs to use the GPU in a headless mode, or
(such as in the CrOS usage) window surfaces are proxied to a host
compositor.

As the modesetting ioctls are a big surface area for potential security
bugs to be found (it's happened in the past, we should assume it will
again in the future), it makes sense to have a build option to disable
those ioctls in cases where they serve no legitimate purpose.

v2: Use more if (IS_ENABLED(...))
v3: Also permit the host to advertise no scanouts
v4: Spiff out commit msg
v5: Make num_scanouts==0 and DRM_VIRTIO_GPU_KMS=n behave the same
v6: Drop conditionally building virtgpu_display.c and early-out of
    it's init/fini fxns instead

Signed-off-by: Rob Clark <robdclark@chromium.org>
Reviewed-by: Dmitry Osipenko <dmitry.osipenko@collabora.com>
Reviewed-by: Javier Martinez Canillas <javierm@redhat.com>
---
 drivers/gpu/drm/virtio/Kconfig           | 11 +++++++++++
 drivers/gpu/drm/virtio/virtgpu_display.c |  6 ++++++
 drivers/gpu/drm/virtio/virtgpu_drv.c     |  4 ++++
 drivers/gpu/drm/virtio/virtgpu_kms.c     | 23 ++++++++++++++---------
 4 files changed, 35 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/virtio/Kconfig b/drivers/gpu/drm/virtio/Kconfig
index 51ec7c3240c9..ea06ff2aa4b4 100644
--- a/drivers/gpu/drm/virtio/Kconfig
+++ b/drivers/gpu/drm/virtio/Kconfig
@@ -11,3 +11,14 @@ config DRM_VIRTIO_GPU
 	   QEMU based VMMs (like KVM or Xen).
 
 	   If unsure say M.
+
+config DRM_VIRTIO_GPU_KMS
+	bool "Virtio GPU driver modesetting support"
+	depends on DRM_VIRTIO_GPU
+	default y
+	help
+	   Enable modesetting support for virtio GPU driver.  This can be
+	   disabled in cases where only "headless" usage of the GPU is
+	   required.
+
+	   If unsure, say Y.
diff --git a/drivers/gpu/drm/virtio/virtgpu_display.c b/drivers/gpu/drm/virtio/virtgpu_display.c
index 9ea7611a9e0f..ad924a8502e9 100644
--- a/drivers/gpu/drm/virtio/virtgpu_display.c
+++ b/drivers/gpu/drm/virtio/virtgpu_display.c
@@ -336,6 +336,9 @@ int virtio_gpu_modeset_init(struct virtio_gpu_device *vgdev)
 {
 	int i, ret;
 
+	if (!vgdev->num_scanouts)
+		return 0;
+
 	ret = drmm_mode_config_init(vgdev->ddev);
 	if (ret)
 		return ret;
@@ -362,6 +365,9 @@ void virtio_gpu_modeset_fini(struct virtio_gpu_device *vgdev)
 {
 	int i;
 
+	if (!vgdev->num_scanouts)
+		return;
+
 	for (i = 0 ; i < vgdev->num_scanouts; ++i)
 		kfree(vgdev->outputs[i].edid);
 }
diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.c b/drivers/gpu/drm/virtio/virtgpu_drv.c
index ae97b98750b6..add075681e18 100644
--- a/drivers/gpu/drm/virtio/virtgpu_drv.c
+++ b/drivers/gpu/drm/virtio/virtgpu_drv.c
@@ -172,6 +172,10 @@ MODULE_AUTHOR("Alon Levy");
 DEFINE_DRM_GEM_FOPS(virtio_gpu_driver_fops);
 
 static const struct drm_driver driver = {
+	/*
+	 * If KMS is disabled DRIVER_MODESET and DRIVER_ATOMIC are masked
+	 * out via drm_device::driver_features:
+	 */
 	.driver_features = DRIVER_MODESET | DRIVER_GEM | DRIVER_RENDER | DRIVER_ATOMIC,
 	.open = virtio_gpu_driver_open,
 	.postclose = virtio_gpu_driver_postclose,
diff --git a/drivers/gpu/drm/virtio/virtgpu_kms.c b/drivers/gpu/drm/virtio/virtgpu_kms.c
index 27b7f14dae89..02e5c18c2c75 100644
--- a/drivers/gpu/drm/virtio/virtgpu_kms.c
+++ b/drivers/gpu/drm/virtio/virtgpu_kms.c
@@ -223,12 +223,15 @@ int virtio_gpu_init(struct virtio_device *vdev, struct drm_device *dev)
 			num_scanouts, &num_scanouts);
 	vgdev->num_scanouts = min_t(uint32_t, num_scanouts,
 				    VIRTIO_GPU_MAX_SCANOUTS);
-	if (!vgdev->num_scanouts) {
-		DRM_ERROR("num_scanouts is zero\n");
-		ret = -EINVAL;
-		goto err_scanouts;
+
+	if (IS_ENABLED(CONFIG_DRM_VIRTIO_GPU_KMS) || !vgdev->num_scanouts) {
+		DRM_INFO("KMS disabled\n");
+		vgdev->num_scanouts = 0;
+		vgdev->has_edid = false;
+		dev->driver_features &= ~(DRIVER_MODESET | DRIVER_ATOMIC);
+	} else {
+		DRM_INFO("number of scanouts: %d\n", num_scanouts);
 	}
-	DRM_INFO("number of scanouts: %d\n", num_scanouts);
 
 	virtio_cread_le(vgdev->vdev, struct virtio_gpu_config,
 			num_capsets, &num_capsets);
@@ -246,10 +249,12 @@ int virtio_gpu_init(struct virtio_device *vdev, struct drm_device *dev)
 		virtio_gpu_get_capsets(vgdev, num_capsets);
 	if (vgdev->has_edid)
 		virtio_gpu_cmd_get_edids(vgdev);
-	virtio_gpu_cmd_get_display_info(vgdev);
-	virtio_gpu_notify(vgdev);
-	wait_event_timeout(vgdev->resp_wq, !vgdev->display_info_pending,
-			   5 * HZ);
+	if (vgdev->num_scanouts) {
+		virtio_gpu_cmd_get_display_info(vgdev);
+		virtio_gpu_notify(vgdev);
+		wait_event_timeout(vgdev->resp_wq, !vgdev->display_info_pending,
+				   5 * HZ);
+	}
 	return 0;
 
 err_scanouts:
-- 
2.39.1


^ permalink raw reply related	[flat|nested] 10+ messages in thread

* Re: [PATCH v6] drm/virtio: Add option to disable KMS support
  2023-03-02 23:35 ` Rob Clark
  (?)
@ 2023-03-03  7:27   ` Gerd Hoffmann
  -1 siblings, 0 replies; 10+ messages in thread
From: Gerd Hoffmann @ 2023-03-03  7:27 UTC (permalink / raw)
  To: Rob Clark
  Cc: Rob Clark, Daniel Vetter, open list, Javier Martinez Canillas,
	dri-devel, Gurchetan Singh, Dmitry Osipenko, Ryan Neph,
	David Airlie, open list:VIRTIO GPU DRIVER, Chia-I Wu

On Thu, Mar 02, 2023 at 03:35:06PM -0800, Rob Clark wrote:
> From: Rob Clark <robdclark@chromium.org>
> 
> Add a build option to disable modesetting support.  This is useful in
> cases where the guest only needs to use the GPU in a headless mode, or
> (such as in the CrOS usage) window surfaces are proxied to a host
> compositor.
> 
> As the modesetting ioctls are a big surface area for potential security
> bugs to be found (it's happened in the past, we should assume it will
> again in the future), it makes sense to have a build option to disable
> those ioctls in cases where they serve no legitimate purpose.
> 
> v2: Use more if (IS_ENABLED(...))
> v3: Also permit the host to advertise no scanouts
> v4: Spiff out commit msg
> v5: Make num_scanouts==0 and DRM_VIRTIO_GPU_KMS=n behave the same
> v6: Drop conditionally building virtgpu_display.c and early-out of
>     it's init/fini fxns instead

Reviewed-by: Gerd Hoffmann <kraxel@redhat.com>

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

^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [PATCH v6] drm/virtio: Add option to disable KMS support
@ 2023-03-03  7:27   ` Gerd Hoffmann
  0 siblings, 0 replies; 10+ messages in thread
From: Gerd Hoffmann @ 2023-03-03  7:27 UTC (permalink / raw)
  To: Rob Clark
  Cc: Rob Clark, open list, Javier Martinez Canillas, dri-devel,
	Gurchetan Singh, Dmitry Osipenko, Ryan Neph, David Airlie,
	open list:VIRTIO GPU DRIVER

On Thu, Mar 02, 2023 at 03:35:06PM -0800, Rob Clark wrote:
> From: Rob Clark <robdclark@chromium.org>
> 
> Add a build option to disable modesetting support.  This is useful in
> cases where the guest only needs to use the GPU in a headless mode, or
> (such as in the CrOS usage) window surfaces are proxied to a host
> compositor.
> 
> As the modesetting ioctls are a big surface area for potential security
> bugs to be found (it's happened in the past, we should assume it will
> again in the future), it makes sense to have a build option to disable
> those ioctls in cases where they serve no legitimate purpose.
> 
> v2: Use more if (IS_ENABLED(...))
> v3: Also permit the host to advertise no scanouts
> v4: Spiff out commit msg
> v5: Make num_scanouts==0 and DRM_VIRTIO_GPU_KMS=n behave the same
> v6: Drop conditionally building virtgpu_display.c and early-out of
>     it's init/fini fxns instead

Reviewed-by: Gerd Hoffmann <kraxel@redhat.com>


^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [PATCH v6] drm/virtio: Add option to disable KMS support
@ 2023-03-03  7:27   ` Gerd Hoffmann
  0 siblings, 0 replies; 10+ messages in thread
From: Gerd Hoffmann @ 2023-03-03  7:27 UTC (permalink / raw)
  To: Rob Clark
  Cc: dri-devel, Chia-I Wu, Ryan Neph, Dmitry Osipenko, Daniel Vetter,
	Rob Clark, Javier Martinez Canillas, David Airlie,
	Gurchetan Singh, open list:VIRTIO GPU DRIVER, open list

On Thu, Mar 02, 2023 at 03:35:06PM -0800, Rob Clark wrote:
> From: Rob Clark <robdclark@chromium.org>
> 
> Add a build option to disable modesetting support.  This is useful in
> cases where the guest only needs to use the GPU in a headless mode, or
> (such as in the CrOS usage) window surfaces are proxied to a host
> compositor.
> 
> As the modesetting ioctls are a big surface area for potential security
> bugs to be found (it's happened in the past, we should assume it will
> again in the future), it makes sense to have a build option to disable
> those ioctls in cases where they serve no legitimate purpose.
> 
> v2: Use more if (IS_ENABLED(...))
> v3: Also permit the host to advertise no scanouts
> v4: Spiff out commit msg
> v5: Make num_scanouts==0 and DRM_VIRTIO_GPU_KMS=n behave the same
> v6: Drop conditionally building virtgpu_display.c and early-out of
>     it's init/fini fxns instead

Reviewed-by: Gerd Hoffmann <kraxel@redhat.com>


^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [PATCH v6] drm/virtio: Add option to disable KMS support
  2023-03-02 23:35 ` Rob Clark
@ 2023-03-03 20:36   ` Dmitry Osipenko
  -1 siblings, 0 replies; 10+ messages in thread
From: Dmitry Osipenko @ 2023-03-03 20:36 UTC (permalink / raw)
  To: Rob Clark, dri-devel, Gerd Hoffmann
  Cc: Chia-I Wu, Ryan Neph, Daniel Vetter, Rob Clark,
	Javier Martinez Canillas, David Airlie, Gurchetan Singh,
	open list:VIRTIO GPU DRIVER, open list

On 3/3/23 02:35, Rob Clark wrote:
> From: Rob Clark <robdclark@chromium.org>
> 
> Add a build option to disable modesetting support.  This is useful in
> cases where the guest only needs to use the GPU in a headless mode, or
> (such as in the CrOS usage) window surfaces are proxied to a host
> compositor.
> 
> As the modesetting ioctls are a big surface area for potential security
> bugs to be found (it's happened in the past, we should assume it will
> again in the future), it makes sense to have a build option to disable
> those ioctls in cases where they serve no legitimate purpose.
> 
> v2: Use more if (IS_ENABLED(...))
> v3: Also permit the host to advertise no scanouts
> v4: Spiff out commit msg
> v5: Make num_scanouts==0 and DRM_VIRTIO_GPU_KMS=n behave the same
> v6: Drop conditionally building virtgpu_display.c and early-out of
>     it's init/fini fxns instead
> 
> Signed-off-by: Rob Clark <robdclark@chromium.org>
> Reviewed-by: Dmitry Osipenko <dmitry.osipenko@collabora.com>
> Reviewed-by: Javier Martinez Canillas <javierm@redhat.com>
> ---

Applied to misc-next

-- 
Best regards,
Dmitry


^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [PATCH v6] drm/virtio: Add option to disable KMS support
@ 2023-03-03 20:36   ` Dmitry Osipenko
  0 siblings, 0 replies; 10+ messages in thread
From: Dmitry Osipenko @ 2023-03-03 20:36 UTC (permalink / raw)
  To: Rob Clark, dri-devel, Gerd Hoffmann
  Cc: Rob Clark, Ryan Neph, open list, Javier Martinez Canillas,
	Gurchetan Singh, David Airlie, open list:VIRTIO GPU DRIVER

On 3/3/23 02:35, Rob Clark wrote:
> From: Rob Clark <robdclark@chromium.org>
> 
> Add a build option to disable modesetting support.  This is useful in
> cases where the guest only needs to use the GPU in a headless mode, or
> (such as in the CrOS usage) window surfaces are proxied to a host
> compositor.
> 
> As the modesetting ioctls are a big surface area for potential security
> bugs to be found (it's happened in the past, we should assume it will
> again in the future), it makes sense to have a build option to disable
> those ioctls in cases where they serve no legitimate purpose.
> 
> v2: Use more if (IS_ENABLED(...))
> v3: Also permit the host to advertise no scanouts
> v4: Spiff out commit msg
> v5: Make num_scanouts==0 and DRM_VIRTIO_GPU_KMS=n behave the same
> v6: Drop conditionally building virtgpu_display.c and early-out of
>     it's init/fini fxns instead
> 
> Signed-off-by: Rob Clark <robdclark@chromium.org>
> Reviewed-by: Dmitry Osipenko <dmitry.osipenko@collabora.com>
> Reviewed-by: Javier Martinez Canillas <javierm@redhat.com>
> ---

Applied to misc-next

-- 
Best regards,
Dmitry


^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [PATCH v6] drm/virtio: Add option to disable KMS support
  2023-03-02 23:35 ` Rob Clark
@ 2023-03-04 21:39   ` Dmitry Osipenko
  -1 siblings, 0 replies; 10+ messages in thread
From: Dmitry Osipenko @ 2023-03-04 21:39 UTC (permalink / raw)
  To: Rob Clark, dri-devel, Gerd Hoffmann
  Cc: Chia-I Wu, Ryan Neph, Daniel Vetter, Rob Clark,
	Javier Martinez Canillas, David Airlie, Gurchetan Singh,
	open list:VIRTIO GPU DRIVER, open list

On 3/3/23 02:35, Rob Clark wrote:
> @@ -223,12 +223,15 @@ int virtio_gpu_init(struct virtio_device *vdev, struct drm_device *dev)
>  			num_scanouts, &num_scanouts);
>  	vgdev->num_scanouts = min_t(uint32_t, num_scanouts,
>  				    VIRTIO_GPU_MAX_SCANOUTS);
> -	if (!vgdev->num_scanouts) {
> -		DRM_ERROR("num_scanouts is zero\n");
> -		ret = -EINVAL;
> -		goto err_scanouts;
> +
> +	if (IS_ENABLED(CONFIG_DRM_VIRTIO_GPU_KMS) || !vgdev->num_scanouts) {

Unfortunately the IS_ENABLED check needs to be inverted here.

Secondly, with the IS_ENABLED check fixed and CONFIG_DRM_VIRTIO_GPU_KMS
disabled, kernel crashes with a NULL deref on boot after getting the
VIRTIO_GPU_EVENT_DISPLAY from host.

 ==================================================================
[    0.870144] BUG: KASAN: null-ptr-deref in
drm_kms_helper_hotplug_event+0x2b/0x50
[    0.870588] Read of size 8 at addr 0000000000000010 by task
kworker/0:1/14
[    0.870588]
[    0.870588] CPU: 0 PID: 14 Comm: kworker/0:1 Not tainted
6.2.0-next-20230303+ #387
[    0.871441] scsi host0: ahci
[    0.870588] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS
rel-1.16.0-0-gd239552ce722-prebuilt.qemu.org 04/01/2014
[    0.870588] Workqueue: events virtio_gpu_dequeue_ctrl_func
[    0.870588] Call Trace:
[    0.870588]  <TASK>
[    0.872841] scsi host1: ahci
[    0.870588]  dump_stack_lvl+0x46/0x70
[    0.870588]  kasan_report+0xbb/0xf0
[    0.870588]  ? drm_kms_helper_hotplug_event+0x2b/0x50
[    0.870588]  drm_kms_helper_hotplug_event+0x2b/0x50
[    0.870588]  virtio_gpu_dequeue_ctrl_func+0x143/0x500
[    0.870588]  ? lock_is_held_type+0xd8/0x130
[    0.870588]  ? virtio_gpu_free_vbufs+0x40/0x40
[    0.875105] scsi host2: ahci
[    0.870588]  process_one_work+0x4f5/0x9a0
[    0.870588]  ? pwq_dec_nr_in_flight+0x100/0x100
[    0.870588]  ? spin_bug+0xe0/0xe0
[    0.870588]  worker_thread+0x8c/0x610
[    0.870588]  ? process_one_work+0x9a0/0x9a0
[    0.870588]  kthread+0x15a/0x190
[    0.870588]  ? kthread_complete_and_exit+0x20/0x20
[    0.870588]  ret_from_fork+0x1f/0x30
[    0.877201] scsi host3: ahci
[    0.870588]  </TASK>
[    0.870588]
==================================================================

I'll send a follow up fix.

-- 
Best regards,
Dmitry


^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [PATCH v6] drm/virtio: Add option to disable KMS support
@ 2023-03-04 21:39   ` Dmitry Osipenko
  0 siblings, 0 replies; 10+ messages in thread
From: Dmitry Osipenko @ 2023-03-04 21:39 UTC (permalink / raw)
  To: Rob Clark, dri-devel, Gerd Hoffmann
  Cc: Rob Clark, Ryan Neph, open list, Javier Martinez Canillas,
	Gurchetan Singh, David Airlie, open list:VIRTIO GPU DRIVER

On 3/3/23 02:35, Rob Clark wrote:
> @@ -223,12 +223,15 @@ int virtio_gpu_init(struct virtio_device *vdev, struct drm_device *dev)
>  			num_scanouts, &num_scanouts);
>  	vgdev->num_scanouts = min_t(uint32_t, num_scanouts,
>  				    VIRTIO_GPU_MAX_SCANOUTS);
> -	if (!vgdev->num_scanouts) {
> -		DRM_ERROR("num_scanouts is zero\n");
> -		ret = -EINVAL;
> -		goto err_scanouts;
> +
> +	if (IS_ENABLED(CONFIG_DRM_VIRTIO_GPU_KMS) || !vgdev->num_scanouts) {

Unfortunately the IS_ENABLED check needs to be inverted here.

Secondly, with the IS_ENABLED check fixed and CONFIG_DRM_VIRTIO_GPU_KMS
disabled, kernel crashes with a NULL deref on boot after getting the
VIRTIO_GPU_EVENT_DISPLAY from host.

 ==================================================================
[    0.870144] BUG: KASAN: null-ptr-deref in
drm_kms_helper_hotplug_event+0x2b/0x50
[    0.870588] Read of size 8 at addr 0000000000000010 by task
kworker/0:1/14
[    0.870588]
[    0.870588] CPU: 0 PID: 14 Comm: kworker/0:1 Not tainted
6.2.0-next-20230303+ #387
[    0.871441] scsi host0: ahci
[    0.870588] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS
rel-1.16.0-0-gd239552ce722-prebuilt.qemu.org 04/01/2014
[    0.870588] Workqueue: events virtio_gpu_dequeue_ctrl_func
[    0.870588] Call Trace:
[    0.870588]  <TASK>
[    0.872841] scsi host1: ahci
[    0.870588]  dump_stack_lvl+0x46/0x70
[    0.870588]  kasan_report+0xbb/0xf0
[    0.870588]  ? drm_kms_helper_hotplug_event+0x2b/0x50
[    0.870588]  drm_kms_helper_hotplug_event+0x2b/0x50
[    0.870588]  virtio_gpu_dequeue_ctrl_func+0x143/0x500
[    0.870588]  ? lock_is_held_type+0xd8/0x130
[    0.870588]  ? virtio_gpu_free_vbufs+0x40/0x40
[    0.875105] scsi host2: ahci
[    0.870588]  process_one_work+0x4f5/0x9a0
[    0.870588]  ? pwq_dec_nr_in_flight+0x100/0x100
[    0.870588]  ? spin_bug+0xe0/0xe0
[    0.870588]  worker_thread+0x8c/0x610
[    0.870588]  ? process_one_work+0x9a0/0x9a0
[    0.870588]  kthread+0x15a/0x190
[    0.870588]  ? kthread_complete_and_exit+0x20/0x20
[    0.870588]  ret_from_fork+0x1f/0x30
[    0.877201] scsi host3: ahci
[    0.870588]  </TASK>
[    0.870588]
==================================================================

I'll send a follow up fix.

-- 
Best regards,
Dmitry


^ permalink raw reply	[flat|nested] 10+ messages in thread

end of thread, other threads:[~2023-03-04 21:39 UTC | newest]

Thread overview: 10+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2023-03-02 23:35 [PATCH v6] drm/virtio: Add option to disable KMS support Rob Clark
2023-03-02 23:35 ` Rob Clark
2023-03-02 23:35 ` Rob Clark
2023-03-03  7:27 ` Gerd Hoffmann
2023-03-03  7:27   ` Gerd Hoffmann
2023-03-03  7:27   ` Gerd Hoffmann
2023-03-03 20:36 ` Dmitry Osipenko
2023-03-03 20:36   ` Dmitry Osipenko
2023-03-04 21:39 ` Dmitry Osipenko
2023-03-04 21:39   ` Dmitry Osipenko

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.