Skip to contentSkip to conversation

drm: pl111: enable render node

PATCH In review Peter Collingbourne 6 participants 24 messages Apr 23, 2020 – May 4, 2020

Subsystems (20)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 drm: pl111: enable render node

The render node is required by Android which does not support the legacy drmAuth authentication process.

---
 drivers/gpu/drm/pl111/pl111_drv.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/gpu/drm/pl111/pl111_drv.c +1 -1
index aa8aa8d9e405..246662129715 100644
@@ -225,7 +225,7 @@ DEFINE_DRM_GEM_CMA_FOPS(drm_fops);
 
 static struct drm_driver pl111_drm_driver = {
 	.driver_features =
-		DRIVER_MODESET | DRIVER_GEM | DRIVER_ATOMIC,
+		DRIVER_MODESET | DRIVER_GEM | DRIVER_ATOMIC | DRIVER_RENDER,
 	.ioctls = NULL,
 	.fops = &drm_fops,
 	.name = "pl111",
Signature · 6 lines
2.26.2.303.gf8c07b1a785-goog

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel
#1 lore raw
Quoted 3 lines · Peter Collingbourne

The render node is required by Android which does not support the legacy drmAuth authentication process.

There's no rendering engine on pl111, so only the control node makes sense for this HW since all this driver does is display. Can you clarify what you're trying to do with pl111?
_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel

#2 lore raw
In reply to Peter Collingbourne
Quoted 9 lines · Peter Collingbourne

The render node is required by Android which does not support the legacy drmAuth authentication process.

---
 drivers/gpu/drm/pl111/pl111_drv.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

The summary talks about drmAuth, yet exposes a render node. Even through there's no rendering engine in the HW, as mentioned by Eric.

AFAICT the only way drmAuth is relevant with pl111 is when you want to export/import dma bufs.
Although that is handled in core and the artificial DRM_AUTH restriction has been lifted with commit [1].

-Emil

[1] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?h=v5.7-rc2&id=30a958526d2cc6df38347336a602479d048d92e7
_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel

#3 lore raw
Quoted 14 lines · Emil Velikov
Quoted 9 lines · Peter Collingbourne

The render node is required by Android which does not support the legacy drmAuth authentication process.

---
 drivers/gpu/drm/pl111/pl111_drv.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

The summary talks about drmAuth, yet exposes a render node. Even through there's no rendering engine in the HW, as mentioned by Eric.

Quoted 8 lines

AFAICT the only way drmAuth is relevant with pl111 is when you want to export/import dma bufs.
Although that is handled in core and the artificial DRM_AUTH restriction has been lifted with commit [1].

-Emil

[1] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?h=v5.7-rc2&id=30a958526d2cc6df38347336a602479d048d92e7

Okay, most likely drmAuth is irrelevant here (I don't know much about it to be honest; I know that Android uses render nodes, so I figured that drmAuth must therefore be the thing that it doesn't use). Sorry for the confusion. Here is a better explanation of why I needed this change.

Android has a composer process that opens the primary node and uses DRM_IOCTL_MODE_ATOMIC to switch between frame buffers, and a renderer process (surfaceflinger) that opens the render node, prepares frame buffers and sends them to the composer. One idea for adapting this architecture to devices without render nodes is to have the renderer process open the primary node instead. But this runs into a problem:
suppose that the renderer process starts before the composer process. In this case, the kernel makes the renderer the DRM master, so the composer can't change the frame buffer. Render nodes don't have this problem because opening them doesn't affect the master.

I considered fixing this by having the composer issue DRM_IOCTL_SET_MASTER, but this requires root privileges. If we require drivers to provide render nodes and control access to the primary node while opening up the render node, we can ensure that only authorized processes can become the master without requiring them to be root.

Peter
_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel

#4 lore raw
Quoted 42 lines · Peter Collingbourne
Quoted 14 lines · Emil Velikov
Quoted 9 lines · Peter Collingbourne

The render node is required by Android which does not support the legacy drmAuth authentication process.

---
 drivers/gpu/drm/pl111/pl111_drv.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

The summary talks about drmAuth, yet exposes a render node. Even through there's no rendering engine in the HW, as mentioned by Eric.

Quoted 8 lines

AFAICT the only way drmAuth is relevant with pl111 is when you want to export/import dma bufs.
Although that is handled in core and the artificial DRM_AUTH restriction has been lifted with commit [1].

-Emil

[1] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?h=v5.7-rc2&id=30a958526d2cc6df38347336a602479d048d92e7

Okay, most likely drmAuth is irrelevant here (I don't know much about it to be honest; I know that Android uses render nodes, so I figured that drmAuth must therefore be the thing that it doesn't use). Sorry for the confusion. Here is a better explanation of why I needed this change.

Android has a composer process that opens the primary node and uses DRM_IOCTL_MODE_ATOMIC to switch between frame buffers, and a renderer process (surfaceflinger) that opens the render node, prepares frame buffers and sends them to the composer. One idea for adapting this architecture to devices without render nodes is to have the renderer process open the primary node instead. But this runs into a problem:
suppose that the renderer process starts before the composer process. In this case, the kernel makes the renderer the DRM master, so the composer can't change the frame buffer. Render nodes don't have this problem because opening them doesn't affect the master.

Quoted 6 lines

I considered fixing this by having the composer issue DRM_IOCTL_SET_MASTER, but this requires root privileges. If we require drivers to provide render nodes and control access to the primary node while opening up the render node, we can ensure that only authorized processes can become the master without requiring them to be root.

I think that the crux, is trying to open a _primary_ node for _rendering_ purposes.
While that may work - as you outline above - it's usually a bad idea.

To step back a bit, in practise we have three SoC combinations:
- complete lack of render device - then you default to software rendering [1]
- display and render device are one and the same - no change needed,
things should just work
- display and render devices are separate - surfaceflinger should
open the correct _rendering_ device node.

[1] Mesa's libEGL (don't recall exact name on Android) can open VGEM render node, to work with the kms-swrast_dri.so software rasteriser module.

While I did not try [1] with Android, it was working fine with CrOS. I suggest giving it a try and reporting bugs.

Thanks
Emil
_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel

#5 lore raw
Quoted 61 lines · Emil Velikov
Quoted 42 lines · Peter Collingbourne
Quoted 14 lines · Emil Velikov
Quoted 9 lines · Peter Collingbourne

The render node is required by Android which does not support the legacy drmAuth authentication process.

---
 drivers/gpu/drm/pl111/pl111_drv.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

The summary talks about drmAuth, yet exposes a render node. Even through there's no rendering engine in the HW, as mentioned by Eric.

Quoted 8 lines

AFAICT the only way drmAuth is relevant with pl111 is when you want to export/import dma bufs.
Although that is handled in core and the artificial DRM_AUTH restriction has been lifted with commit [1].

-Emil

[1] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?h=v5.7-rc2&id=30a958526d2cc6df38347336a602479d048d92e7

Okay, most likely drmAuth is irrelevant here (I don't know much about it to be honest; I know that Android uses render nodes, so I figured that drmAuth must therefore be the thing that it doesn't use). Sorry for the confusion. Here is a better explanation of why I needed this change.

Android has a composer process that opens the primary node and uses DRM_IOCTL_MODE_ATOMIC to switch between frame buffers, and a renderer process (surfaceflinger) that opens the render node, prepares frame buffers and sends them to the composer. One idea for adapting this architecture to devices without render nodes is to have the renderer process open the primary node instead. But this runs into a problem:
suppose that the renderer process starts before the composer process. In this case, the kernel makes the renderer the DRM master, so the composer can't change the frame buffer. Render nodes don't have this problem because opening them doesn't affect the master.

Quoted 6 lines

I considered fixing this by having the composer issue DRM_IOCTL_SET_MASTER, but this requires root privileges. If we require drivers to provide render nodes and control access to the primary node while opening up the render node, we can ensure that only authorized processes can become the master without requiring them to be root.

I think that the crux, is trying to open a _primary_ node for _rendering_ purposes.
While that may work - as you outline above - it's usually a bad idea.

To step back a bit, in practise we have three SoC combinations:
- complete lack of render device - then you default to software rendering [1]
- display and render device are one and the same - no change needed,
things should just work
- display and render devices are separate - surfaceflinger should
open the correct _rendering_ device node.

Quoted 6 lines

[1] Mesa's libEGL (don't recall exact name on Android) can open VGEM render node, to work with the kms-swrast_dri.so software rasteriser module.

While I did not try [1] with Android, it was working fine with CrOS. I suggest giving it a try and reporting bugs.

VGEM is not a solution, because it doesn't coordinate caching behavior with your display node and so you get corruption if you try to to share dma-bufs between them. In cros it's used only for some tests as far as I've heard, and it's been a source of pain.

If we want to go the route of "kms-only nodes also provide a render node interface for doing swrast on", I think that would be a good idea, but we should do this at a core DRM level (probably "does this driver expose dma-buf? then also expose render nodes") rather than per kms driver.
_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel

#8 lore raw

Re: [PATCH] drm: enable render nodes wherever buffer sharing is supported

In reply to Peter Collingbourne
Quoted 10 lines · Peter Collingbourne

Render nodes are not just useful for devices supporting GPU hardware acceleration. Even on devices that only support dumb frame buffers, they are useful in situations where composition (using software rasterization) and KMS are done in different processes with buffer sharing being used to send frame buffers between them. This is the situation on Android, where surfaceflinger is the compositor and the composer HAL uses KMS to display the buffers. Thus it is beneficial to expose render nodes on all devices that support buffer sharing.

Quoted 5 lines

Because all drivers that currently support render nodes also support buffer sharing, the DRIVER_RENDER flag is no longer necessary to mark devices as supporting render nodes, so remove it and just rely on the presence of a prime_handle_to_fd function pointer to determine whether buffer sharing is supported.

I'm definitely interested in seeing a patch like this land, as I think the current state is an ugly historical artifact. We just have to be careful.

Were there any instances of drivers with render engines exposing PRIME but not RENDER? We should be careful to make sure that we're not exposing new privileges for those through adding render nodes.

There's a UAPI risk I see here. Right now, on a system with a single renderer GPU, we can just open /dev/dri/renderD128 and get the GPU for rendering, and various things are relying on that (such as libwaffle,
used in piglit among other things) Adding render nodes for kms-only
drivers could displace the actual GPU to 129, and the question is whether this will be disruptive. For Mesa, I think this works out, because kmsro should load on the kms device's node and then share buffers over to the real GPU that it digs around to find at init time. Just saying, I'm not sure I know all of the userspace well enough to say "this should be safe despite that"

(And, maybe, if we decide that it's not safe enough, we could punt kms-only drivers to a higher starting number?)

drivers/gpu/drm/vc4/vc4_drv.c static int vc4_drm_bind(struct device *dev) · line 263

Quoted 13 lines
drivers/gpu/drm/vc4/vc4_drv.c +0 -6
@@ -260,12 +258,6 @@ static int vc4_drm_bind(struct device *dev)
        if (!vc4)
                return -ENOMEM;
 
-       /* If VC4 V3D is missing, don't advertise render nodes. */
-       node = of_find_matching_node_and_match(NULL, vc4_v3d_dt_match, NULL);
-       if (!node || !of_device_is_available(node))
-               vc4_drm_driver.driver_features &= ~DRIVER_RENDER;
-       of_node_put(node);
-
        drm = drm_dev_alloc(&vc4_drm_driver, dev);
        if (IS_ERR(drm))
                return PTR_ERR(drm);

Looks like dropping this code from vc4 should be fine -- kmsro looks for a render node from each driver name it supports in turn, so even if v3d moves from renderD128 to renderD129, things should still work.
_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel

#10 lore raw

Re: [PATCH] drm: enable render nodes wherever buffer sharing is supported

In reply to Peter Collingbourne
Quoted 92 lines · Peter Collingbourne
Quoted 18 lines · Eric Anholt
Quoted 10 lines · Peter Collingbourne

Render nodes are not just useful for devices supporting GPU hardware acceleration. Even on devices that only support dumb frame buffers, they are useful in situations where composition (using software rasterization) and KMS are done in different processes with buffer sharing being used to send frame buffers between them. This is the situation on Android, where surfaceflinger is the compositor and the composer HAL uses KMS to display the buffers. Thus it is beneficial to expose render nodes on all devices that support buffer sharing.

Quoted 5 lines

Because all drivers that currently support render nodes also support buffer sharing, the DRIVER_RENDER flag is no longer necessary to mark devices as supporting render nodes, so remove it and just rely on the presence of a prime_handle_to_fd function pointer to determine whether buffer sharing is supported.

Quoted 7 lines

I'm definitely interested in seeing a patch like this land, as I think the current state is an ugly historical artifact. We just have to be careful.

Were there any instances of drivers with render engines exposing PRIME but not RENDER? We should be careful to make sure that we're not exposing new privileges for those through adding render nodes.

These are the drivers that we'd be adding render nodes for with this change:

$ git grep -l prime_handle_to_fd (git grep -L DRIVER_RENDER (git grep
-l '\.driver_features'))
drivers/gpu/drm/arc/arcpgu_drv.c
drivers/gpu/drm/arm/display/komeda/komeda_kms.c
drivers/gpu/drm/arm/hdlcd_drv.c
drivers/gpu/drm/arm/malidp_drv.c
drivers/gpu/drm/armada/armada_drv.c
drivers/gpu/drm/aspeed/aspeed_gfx_drv.c
drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_dc.c
drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_drv.c
drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c
drivers/gpu/drm/imx/imx-drm-core.c
drivers/gpu/drm/ingenic/ingenic-drm.c
drivers/gpu/drm/mcde/mcde_drv.c
drivers/gpu/drm/mediatek/mtk_drm_drv.c
drivers/gpu/drm/meson/meson_drv.c
drivers/gpu/drm/mxsfb/mxsfb_drv.c
drivers/gpu/drm/pl111/pl111_drv.c
drivers/gpu/drm/qxl/qxl_drv.c
drivers/gpu/drm/rcar-du/rcar_du_drv.c
drivers/gpu/drm/rockchip/rockchip_drm_drv.c
drivers/gpu/drm/shmobile/shmob_drm_drv.c
drivers/gpu/drm/sti/sti_drv.c
drivers/gpu/drm/stm/drv.c
drivers/gpu/drm/tilcdc/tilcdc_drv.c
drivers/gpu/drm/tve200/tve200_drv.c
drivers/gpu/drm/xen/xen_drm_front.c
drivers/gpu/drm/zte/zx_drm_drv.c

Some of the drivers provide custom ioctls but they are already protected from render nodes by not setting DRM_RENDER_ALLOW. Another thing to check for is drivers providing custom fops that might expose something undesirable in the render node:

$ git grep -L 'DEFINE_DRM_GEM_CMA_FOPS\|DEFINE_DRM_GEM_FOPS' (git grep
-l prime_handle_to_fd (git grep -L DRIVER_RENDER (git grep -l
'\.driver_features')))
drivers/gpu/drm/mediatek/mtk_drm_drv.c
drivers/gpu/drm/rockchip/rockchip_drm_drv.c
drivers/gpu/drm/xen/xen_drm_front.c

But looking at those drivers in detail, I see that each of them is using the standard DRM fops handlers (which presumably already handle render nodes correctly) with the exception of mmap, which they provide wrappers for that mostly just wrap drm_gem_mmap.

So unless I'm missing something significant (which seems likely -- I'm not a DRM expert), I don't see a problem so far.

Quoted 11 lines

There's a UAPI risk I see here. Right now, on a system with a single renderer GPU, we can just open /dev/dri/renderD128 and get the GPU for rendering, and various things are relying on that (such as libwaffle,
used in piglit among other things) Adding render nodes for kms-only
drivers could displace the actual GPU to 129, and the question is whether this will be disruptive. For Mesa, I think this works out, because kmsro should load on the kms device's node and then share buffers over to the real GPU that it digs around to find at init time. Just saying, I'm not sure I know all of the userspace well enough to say "this should be safe despite that"

Quoted 2 lines

(And, maybe, if we decide that it's not safe enough, we could punt kms-only drivers to a higher starting number?)

Quoted 7 lines

Android (minigbm) similarly tries to open /dev/dri/renderD$N in a loop with 128 <= N < 192 and assumes that the first non-blacklisted (i.e. not vgem) one that it can open corresponds to the real GPU [1]. I think that the risk of breaking something on Android is low since Android's architecture basically already depends on there being a render node, and it seems unlikely for a device to have more than one GPU, one of which would be non-functional.

Would Juno suffer from this issue? It has 2 HDLCD drivers (both can be active at the same time if you want) and you could run panfrost for the actual GPU. Depending on the probing order, HDLCD render nodes would be registered before the GPU's.

Best regards,
Liviu

Quoted 19 lines

It's also worth bearing in mind that render nodes were added to vgem in commit 3a6eb795 from 2018. To the extent that exposing additional render nodes would lead to widespread breakage, this would seem to me to be a likely way in which it could have happened (since I would expect that it could cause many machines to go from having one render node to having more than one), so perhaps the argument can be made that if we hadn't seen widespread breakage as a result of that change, we'd be unlikely to see it as a result of this one.

This would be conditional on the userspace code not blacklisting the vgem render node like minigbm does -- at a glance I couldn't find such code in Mesa (there does appear to be some code that looks for the vgem driver name, but it seems to only be used on primary nodes, not render nodes) or libwaffle.

Peter

[1] https://cs.android.com/android/platform/superproject/+/master:external/minigbm/cros_gralloc/cros_gralloc_driver.cc;l=48

Signature · 12 lines
====================
| I would like to |
| fix the world,  |
| but they're not |
| giving me the   |
 \ source code!  /
  ---------------
    ¯\_(ツ)_/¯
_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel
#7 lore raw

PATCH 1/1 drm: enable render nodes wherever buffer sharing is supported

In reply to Peter Collingbourne

Render nodes are not just useful for devices supporting GPU hardware acceleration. Even on devices that only support dumb frame buffers, they are useful in situations where composition (using software rasterization) and KMS are done in different processes with buffer sharing being used to send frame buffers between them. This is the situation on Android, where surfaceflinger is the compositor and the composer HAL uses KMS to display the buffers. Thus it is beneficial to expose render nodes on all devices that support buffer sharing.

Because all drivers that currently support render nodes also support buffer sharing, the DRIVER_RENDER flag is no longer necessary to mark devices as supporting render nodes, so remove it and just rely on the presence of a prime_handle_to_fd function pointer to determine whether buffer sharing is supported.

---
 Documentation/gpu/drm-uapi.rst          | 9 +++++----
 drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 2 +-
 drivers/gpu/drm/drm_drv.c               | 2 +-
 drivers/gpu/drm/etnaviv/etnaviv_drv.c   | 2 +-
 drivers/gpu/drm/exynos/exynos_drm_drv.c | 2 +-
 drivers/gpu/drm/i915/i915_drv.c         | 2 +-
 drivers/gpu/drm/lima/lima_drv.c         | 2 +-
 drivers/gpu/drm/msm/msm_drv.c           | 1 -
 drivers/gpu/drm/nouveau/nouveau_drm.c   | 2 +-
 drivers/gpu/drm/omapdrm/omap_drv.c      | 2 +-
 drivers/gpu/drm/panfrost/panfrost_drv.c | 2 +-
 drivers/gpu/drm/radeon/radeon_drv.c     | 3 +--
 drivers/gpu/drm/tegra/drm.c             | 2 +-
 drivers/gpu/drm/v3d/v3d_drv.c           | 1 -
 drivers/gpu/drm/vc4/vc4_drv.c           | 8 --------
 drivers/gpu/drm/vgem/vgem_drv.c         | 2 +-
 drivers/gpu/drm/virtio/virtgpu_drv.c    | 2 +-
 drivers/gpu/drm/vmwgfx/vmwgfx_drv.c     | 2 +-
 include/drm/drm_drv.h                   | 7 -------
 19 files changed, 19 insertions(+), 36 deletions(-)
Documentation/gpu/drm-uapi.rst +5 -4
index 56fec6ed1ad8..2769714ae75a 100644
@@ -129,10 +129,11 @@ authenticate to a DRM-Master prior to getting GPU access. To avoid this
 step and to grant clients GPU access without authenticating, render
 nodes were introduced. Render nodes solely serve render clients, that
 is, no modesetting or privileged ioctls can be issued on render nodes.
-Only non-global rendering commands are allowed. If a driver supports
-render nodes, it must advertise it via the DRIVER_RENDER DRM driver
-capability. If not supported, the primary node must be used for render
-clients together with the legacy drmAuth authentication procedure.
+Only non-global rendering commands are allowed. Drivers that support
+:ref:`PRIME buffer sharing <prime_buffer_sharing>` automatically
+support render nodes. If a driver does not support render nodes,
+the primary node must be used for render clients together with the
+legacy drmAuth authentication procedure.
 
 If a driver advertises render node support, DRM core will create a
 separate render node called renderD<num>. There will be one render node
drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +1 -1
index 8ea86ffdea0d..46460620bc37 100644
@@ -1426,7 +1426,7 @@ static struct drm_driver kms_driver = {
 	.driver_features =
 	    DRIVER_ATOMIC |
 	    DRIVER_GEM |
-	    DRIVER_RENDER | DRIVER_MODESET | DRIVER_SYNCOBJ |
+	    DRIVER_MODESET | DRIVER_SYNCOBJ |
 	    DRIVER_SYNCOBJ_TIMELINE,
 	.open = amdgpu_driver_open_kms,
 	.postclose = amdgpu_driver_postclose_kms,
drivers/gpu/drm/drm_drv.c +1 -1
index 7b1a628d1f6e..8861a30920e5 100644
@@ -651,7 +651,7 @@ int drm_dev_init(struct drm_device *dev,
 		goto err_free;
 	}
 
-	if (drm_core_check_feature(dev, DRIVER_RENDER)) {
+	if (driver->prime_handle_to_fd) {
 		ret = drm_minor_alloc(dev, DRM_MINOR_RENDER);
 		if (ret)
 			goto err_minors;
drivers/gpu/drm/etnaviv/etnaviv_drv.c +1 -1
index a8685b2e1803..abfb143334ac 100644
@@ -497,7 +497,7 @@ static const struct file_operations fops = {
 };
 
 static struct drm_driver etnaviv_drm_driver = {
-	.driver_features    = DRIVER_GEM | DRIVER_RENDER,
+	.driver_features    = DRIVER_GEM,
 	.open               = etnaviv_open,
 	.postclose           = etnaviv_postclose,
 	.gem_free_object_unlocked = etnaviv_gem_free_object,
drivers/gpu/drm/exynos/exynos_drm_drv.c +1 -1
index 57defeb44522..834eb5713fe4 100644
@@ -121,7 +121,7 @@ static const struct file_operations exynos_drm_driver_fops = {
 
 static struct drm_driver exynos_drm_driver = {
 	.driver_features	= DRIVER_MODESET | DRIVER_GEM
-				  | DRIVER_ATOMIC | DRIVER_RENDER,
+				  | DRIVER_ATOMIC,
 	.open			= exynos_drm_open,
 	.lastclose		= drm_fb_helper_lastclose,
 	.postclose		= exynos_drm_postclose,
drivers/gpu/drm/i915/i915_drv.c +1 -1
index 81a4621853db..b028a32fcac5 100644
@@ -1834,7 +1834,7 @@ static struct drm_driver driver = {
 	 */
 	.driver_features =
 	    DRIVER_GEM |
-	    DRIVER_RENDER | DRIVER_MODESET | DRIVER_ATOMIC | DRIVER_SYNCOBJ,
+	    DRIVER_MODESET | DRIVER_ATOMIC | DRIVER_SYNCOBJ,
 	.release = i915_driver_release,
 	.open = i915_driver_open,
 	.lastclose = i915_driver_lastclose,
drivers/gpu/drm/lima/lima_drv.c +1 -1
index 2daac64d8955..bd8b6ad25ac0 100644
@@ -252,7 +252,7 @@ DEFINE_DRM_GEM_FOPS(lima_drm_driver_fops);
  */
 
 static struct drm_driver lima_drm_driver = {
-	.driver_features    = DRIVER_RENDER | DRIVER_GEM | DRIVER_SYNCOBJ,
+	.driver_features    = DRIVER_GEM | DRIVER_SYNCOBJ,
 	.open               = lima_drm_driver_open,
 	.postclose          = lima_drm_driver_postclose,
 	.ioctls             = lima_drm_driver_ioctls,
drivers/gpu/drm/msm/msm_drv.c +0 -1
index 29295dee2a2e..061e62d4b691 100644
@@ -1000,7 +1000,6 @@ static const struct file_operations fops = {
 
 static struct drm_driver msm_driver = {
 	.driver_features    = DRIVER_GEM |
-				DRIVER_RENDER |
 				DRIVER_ATOMIC |
 				DRIVER_MODESET,
 	.open               = msm_open,
drivers/gpu/drm/nouveau/nouveau_drm.c +1 -1
index ca4087f5a15b..3ba8c9789292 100644
@@ -1169,7 +1169,7 @@ nouveau_driver_fops = {
 static struct drm_driver
 driver_stub = {
 	.driver_features =
-		DRIVER_GEM | DRIVER_MODESET | DRIVER_RENDER
+		DRIVER_GEM | DRIVER_MODESET
 #if defined(CONFIG_NOUVEAU_LEGACY_CTX_SUPPORT)
 		| DRIVER_KMS_LEGACY_CONTEXT
 #endif
drivers/gpu/drm/omapdrm/omap_drv.c +1 -1
index cdafd7ef1c32..fe25b352a755 100644
@@ -558,7 +558,7 @@ static const struct file_operations omapdriver_fops = {
 
 static struct drm_driver omap_drm_driver = {
 	.driver_features = DRIVER_MODESET | DRIVER_GEM  |
-		DRIVER_ATOMIC | DRIVER_RENDER,
+		DRIVER_ATOMIC,
 	.open = dev_open,
 	.lastclose = drm_fb_helper_lastclose,
 #ifdef CONFIG_DEBUG_FS
drivers/gpu/drm/panfrost/panfrost_drv.c +1 -1
index 882fecc33fdb..058765d5d5d5 100644
@@ -550,7 +550,7 @@ DEFINE_DRM_GEM_FOPS(panfrost_drm_driver_fops);
  * - 1.1 - adds HEAP and NOEXEC flags for CREATE_BO
  */
 static struct drm_driver panfrost_drm_driver = {
-	.driver_features	= DRIVER_RENDER | DRIVER_GEM | DRIVER_SYNCOBJ,
+	.driver_features	= DRIVER_GEM | DRIVER_SYNCOBJ,
 	.open			= panfrost_open,
 	.postclose		= panfrost_postclose,
 	.ioctls			= panfrost_drm_driver_ioctls,
drivers/gpu/drm/radeon/radeon_drv.c +1 -2
index 59f8186a2415..9457b2f8f81e 100644
@@ -600,8 +600,7 @@ static const struct file_operations radeon_driver_kms_fops = {
 };
 
 static struct drm_driver kms_driver = {
-	.driver_features =
-	    DRIVER_GEM | DRIVER_RENDER,
+	.driver_features = DRIVER_GEM,
 	.load = radeon_driver_load_kms,
 	.open = radeon_driver_open_kms,
 	.postclose = radeon_driver_postclose_kms,
drivers/gpu/drm/tegra/drm.c +1 -1
index bd268028fb3d..55190c582946 100644
@@ -849,7 +849,7 @@ static int tegra_debugfs_init(struct drm_minor *minor)
 
 static struct drm_driver tegra_drm_driver = {
 	.driver_features = DRIVER_MODESET | DRIVER_GEM |
-			   DRIVER_ATOMIC | DRIVER_RENDER,
+			   DRIVER_ATOMIC,
 	.open = tegra_drm_open,
 	.postclose = tegra_drm_postclose,
 	.lastclose = drm_fb_helper_lastclose,
drivers/gpu/drm/v3d/v3d_drv.c +0 -1
index eaa8e9682373..96db702fd648 100644
@@ -195,7 +195,6 @@ static const struct drm_ioctl_desc v3d_drm_ioctls[] = {
 
 static struct drm_driver v3d_drm_driver = {
 	.driver_features = (DRIVER_GEM |
-			    DRIVER_RENDER |
 			    DRIVER_SYNCOBJ),
 
 	.open = v3d_open,
drivers/gpu/drm/vc4/vc4_drv.c +0 -8
index 76f93b662766..bf143d1e0d2e 100644
@@ -181,7 +181,6 @@ static struct drm_driver vc4_drm_driver = {
 	.driver_features = (DRIVER_MODESET |
 			    DRIVER_ATOMIC |
 			    DRIVER_GEM |
-			    DRIVER_RENDER |
 			    DRIVER_SYNCOBJ),
 	.open = vc4_open,
 	.postclose = vc4_close,
@@ -251,7 +250,6 @@ static int vc4_drm_bind(struct device *dev)
 	struct platform_device *pdev = to_platform_device(dev);
 	struct drm_device *drm;
 	struct vc4_dev *vc4;
-	struct device_node *node;
 	int ret = 0;
 
 	dev->coherent_dma_mask = DMA_BIT_MASK(32);
@@ -260,12 +258,6 @@ static int vc4_drm_bind(struct device *dev)
 	if (!vc4)
 		return -ENOMEM;
 
-	/* If VC4 V3D is missing, don't advertise render nodes. */
-	node = of_find_matching_node_and_match(NULL, vc4_v3d_dt_match, NULL);
-	if (!node || !of_device_is_available(node))
-		vc4_drm_driver.driver_features &= ~DRIVER_RENDER;
-	of_node_put(node);
-
 	drm = drm_dev_alloc(&vc4_drm_driver, dev);
 	if (IS_ERR(drm))
 		return PTR_ERR(drm);
drivers/gpu/drm/vgem/vgem_drv.c +1 -1
index 909eba43664a..abc7345a2e96 100644
@@ -437,7 +437,7 @@ static void vgem_release(struct drm_device *dev)
 }
 
 static struct drm_driver vgem_driver = {
-	.driver_features		= DRIVER_GEM | DRIVER_RENDER,
+	.driver_features		= DRIVER_GEM,
 	.release			= vgem_release,
 	.open				= vgem_open,
 	.postclose			= vgem_postclose,
drivers/gpu/drm/virtio/virtgpu_drv.c +1 -1
index ab4bed78e656..22195e008f0d 100644
@@ -189,7 +189,7 @@ MODULE_AUTHOR("Alon Levy");
 DEFINE_DRM_GEM_FOPS(virtio_gpu_driver_fops);
 
 static struct drm_driver driver = {
-	.driver_features = DRIVER_MODESET | DRIVER_GEM | DRIVER_RENDER | DRIVER_ATOMIC,
+	.driver_features = DRIVER_MODESET | DRIVER_GEM | DRIVER_ATOMIC,
 	.open = virtio_gpu_driver_open,
 	.postclose = virtio_gpu_driver_postclose,
 
drivers/gpu/drm/vmwgfx/vmwgfx_drv.c +1 -1
index c2247a893ed4..415c3f8ea907 100644
@@ -1435,7 +1435,7 @@ static const struct file_operations vmwgfx_driver_fops = {
 
 static struct drm_driver driver = {
 	.driver_features =
-	DRIVER_MODESET | DRIVER_RENDER | DRIVER_ATOMIC,
+	DRIVER_MODESET | DRIVER_ATOMIC,
 	.ioctls = vmw_ioctls,
 	.num_ioctls = ARRAY_SIZE(vmw_ioctls),
 	.master_set = vmw_master_set,
include/drm/drm_drv.h +0 -7
index 97109df5beac..f0277d3f89fe 100644
@@ -61,13 +61,6 @@ enum drm_driver_feature {
 	 * Driver supports mode setting interfaces (KMS).
 	 */
 	DRIVER_MODESET			= BIT(1),
-	/**
-	 * @DRIVER_RENDER:
-	 *
-	 * Driver supports dedicated render nodes. See also the :ref:`section on
-	 * render nodes <drm_render_node>` for details.
-	 */
-	DRIVER_RENDER			= BIT(3),
 	/**
 	 * @DRIVER_ATOMIC:
 	 *
Signature · 6 lines
2.26.2.303.gf8c07b1a785-goog

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel
#9 lore raw

Re: [PATCH] drm: enable render nodes wherever buffer sharing is supported

In reply to Eric Anholt
Quoted 18 lines · Eric Anholt
Quoted 10 lines · Peter Collingbourne

Render nodes are not just useful for devices supporting GPU hardware acceleration. Even on devices that only support dumb frame buffers, they are useful in situations where composition (using software rasterization) and KMS are done in different processes with buffer sharing being used to send frame buffers between them. This is the situation on Android, where surfaceflinger is the compositor and the composer HAL uses KMS to display the buffers. Thus it is beneficial to expose render nodes on all devices that support buffer sharing.

Quoted 5 lines

Because all drivers that currently support render nodes also support buffer sharing, the DRIVER_RENDER flag is no longer necessary to mark devices as supporting render nodes, so remove it and just rely on the presence of a prime_handle_to_fd function pointer to determine whether buffer sharing is supported.

Quoted 7 lines

I'm definitely interested in seeing a patch like this land, as I think the current state is an ugly historical artifact. We just have to be careful.

Were there any instances of drivers with render engines exposing PRIME but not RENDER? We should be careful to make sure that we're not exposing new privileges for those through adding render nodes.

These are the drivers that we'd be adding render nodes for with this change:

$ git grep -l prime_handle_to_fd (git grep -L DRIVER_RENDER (git grep
-l '\.driver_features'))
drivers/gpu/drm/arc/arcpgu_drv.c
drivers/gpu/drm/arm/display/komeda/komeda_kms.c
drivers/gpu/drm/arm/hdlcd_drv.c
drivers/gpu/drm/arm/malidp_drv.c
drivers/gpu/drm/armada/armada_drv.c
drivers/gpu/drm/aspeed/aspeed_gfx_drv.c
drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_dc.c
drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_drv.c
drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c
drivers/gpu/drm/imx/imx-drm-core.c
drivers/gpu/drm/ingenic/ingenic-drm.c
drivers/gpu/drm/mcde/mcde_drv.c
drivers/gpu/drm/mediatek/mtk_drm_drv.c
drivers/gpu/drm/meson/meson_drv.c
drivers/gpu/drm/mxsfb/mxsfb_drv.c
drivers/gpu/drm/pl111/pl111_drv.c
drivers/gpu/drm/qxl/qxl_drv.c
drivers/gpu/drm/rcar-du/rcar_du_drv.c
drivers/gpu/drm/rockchip/rockchip_drm_drv.c
drivers/gpu/drm/shmobile/shmob_drm_drv.c
drivers/gpu/drm/sti/sti_drv.c
drivers/gpu/drm/stm/drv.c
drivers/gpu/drm/tilcdc/tilcdc_drv.c
drivers/gpu/drm/tve200/tve200_drv.c
drivers/gpu/drm/xen/xen_drm_front.c
drivers/gpu/drm/zte/zx_drm_drv.c

Some of the drivers provide custom ioctls but they are already protected from render nodes by not setting DRM_RENDER_ALLOW. Another thing to check for is drivers providing custom fops that might expose something undesirable in the render node:

$ git grep -L 'DEFINE_DRM_GEM_CMA_FOPS\|DEFINE_DRM_GEM_FOPS' (git grep
-l prime_handle_to_fd (git grep -L DRIVER_RENDER (git grep -l
'\.driver_features')))
drivers/gpu/drm/mediatek/mtk_drm_drv.c
drivers/gpu/drm/rockchip/rockchip_drm_drv.c
drivers/gpu/drm/xen/xen_drm_front.c

But looking at those drivers in detail, I see that each of them is using the standard DRM fops handlers (which presumably already handle render nodes correctly) with the exception of mmap, which they provide wrappers for that mostly just wrap drm_gem_mmap.

So unless I'm missing something significant (which seems likely -- I'm not a DRM expert), I don't see a problem so far.

Quoted 11 lines

There's a UAPI risk I see here. Right now, on a system with a single renderer GPU, we can just open /dev/dri/renderD128 and get the GPU for rendering, and various things are relying on that (such as libwaffle,
used in piglit among other things) Adding render nodes for kms-only
drivers could displace the actual GPU to 129, and the question is whether this will be disruptive. For Mesa, I think this works out, because kmsro should load on the kms device's node and then share buffers over to the real GPU that it digs around to find at init time. Just saying, I'm not sure I know all of the userspace well enough to say "this should be safe despite that"

Quoted 2 lines

(And, maybe, if we decide that it's not safe enough, we could punt kms-only drivers to a higher starting number?)

Android (minigbm) similarly tries to open /dev/dri/renderD$N in a loop with 128 <= N < 192 and assumes that the first non-blacklisted (i.e. not vgem) one that it can open corresponds to the real GPU [1]. I think that the risk of breaking something on Android is low since Android's architecture basically already depends on there being a render node, and it seems unlikely for a device to have more than one GPU, one of which would be non-functional.

It's also worth bearing in mind that render nodes were added to vgem in commit 3a6eb795 from 2018. To the extent that exposing additional render nodes would lead to widespread breakage, this would seem to me to be a likely way in which it could have happened (since I would expect that it could cause many machines to go from having one render node to having more than one), so perhaps the argument can be made that if we hadn't seen widespread breakage as a result of that change, we'd be unlikely to see it as a result of this one.

This would be conditional on the userspace code not blacklisting the vgem render node like minigbm does -- at a glance I couldn't find such code in Mesa (there does appear to be some code that looks for the vgem driver name, but it seems to only be used on primary nodes, not render nodes) or libwaffle.

Peter

[1] https://cs.android.com/android/platform/superproject/+/master:external/minigbm/cros_gralloc/cros_gralloc_driver.cc;l=48
_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel

#22 lore raw
In reply to Emil Velikov
Quoted 54 lines · Emil Velikov
Quoted 42 lines · Peter Collingbourne
Quoted 14 lines · Emil Velikov
Quoted 9 lines · Peter Collingbourne

The render node is required by Android which does not support the legacy drmAuth authentication process.

---
 drivers/gpu/drm/pl111/pl111_drv.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

The summary talks about drmAuth, yet exposes a render node. Even through there's no rendering engine in the HW, as mentioned by Eric.

Quoted 8 lines

AFAICT the only way drmAuth is relevant with pl111 is when you want to export/import dma bufs.
Although that is handled in core and the artificial DRM_AUTH restriction has been lifted with commit [1].

-Emil

[1] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?h=v5.7-rc2&id=30a958526d2cc6df38347336a602479d048d92e7

Okay, most likely drmAuth is irrelevant here (I don't know much about it to be honest; I know that Android uses render nodes, so I figured that drmAuth must therefore be the thing that it doesn't use). Sorry for the confusion. Here is a better explanation of why I needed this change.

Android has a composer process that opens the primary node and uses DRM_IOCTL_MODE_ATOMIC to switch between frame buffers, and a renderer process (surfaceflinger) that opens the render node, prepares frame buffers and sends them to the composer. One idea for adapting this architecture to devices without render nodes is to have the renderer process open the primary node instead. But this runs into a problem:
suppose that the renderer process starts before the composer process. In this case, the kernel makes the renderer the DRM master, so the composer can't change the frame buffer. Render nodes don't have this problem because opening them doesn't affect the master.

Quoted 6 lines

I considered fixing this by having the composer issue DRM_IOCTL_SET_MASTER, but this requires root privileges. If we require drivers to provide render nodes and control access to the primary node while opening up the render node, we can ensure that only authorized processes can become the master without requiring them to be root.

I think that the crux, is trying to open a _primary_ node for _rendering_ purposes.
While that may work - as you outline above - it's usually a bad idea.

Quoted 8 lines

To step back a bit, in practise we have three SoC combinations:
- complete lack of render device - then you default to software rendering [1]
- display and render device are one and the same - no change needed,
things should just work
- display and render devices are separate - surfaceflinger should
open the correct _rendering_ device node.

[1] Mesa's libEGL (don't recall exact name on Android) can open VGEM

(Android uses SwiftShader for software rendering, not Mesa, FWIW.)

Quoted 5 lines

render node, to work with the kms-swrast_dri.so software rasteriser module.

While I did not try [1] with Android, it was working fine with CrOS. I suggest giving it a try and reporting bugs.

I'm not sure I understand your suggestion, or at least how it would work on Android. The composer process wouldn't be able to use DRM_IOCTL_MODE_ATOMIC to display a frame buffer produced by the surfaceflinger process using a vgem render node, would it? It's a different driver so the memory region used for the frame buffer wouldn't necessarily be compatible with the pl111 device. As far as I know, the frame buffer would need to be produced using a pl111 render node.

Peter
_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel

#6 lore raw
In reply to Eric Anholt
Quoted 5 lines · Eric Anholt

If we want to go the route of "kms-only nodes also provide a render node interface for doing swrast on", I think that would be a good idea, but we should do this at a core DRM level (probably "does this driver expose dma-buf? then also expose render nodes") rather than per kms driver.

This sounds like a good idea to me. I will send a patch shortly.

Peter
_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel

#20 lore raw
In reply to Eric Anholt
Quoted 74 lines · Eric Anholt
Quoted 61 lines · Emil Velikov
Quoted 42 lines · Peter Collingbourne
Quoted 14 lines · Emil Velikov
Quoted 9 lines · Peter Collingbourne

The render node is required by Android which does not support the legacy drmAuth authentication process.

---
 drivers/gpu/drm/pl111/pl111_drv.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

The summary talks about drmAuth, yet exposes a render node. Even through there's no rendering engine in the HW, as mentioned by Eric.

Quoted 8 lines

AFAICT the only way drmAuth is relevant with pl111 is when you want to export/import dma bufs.
Although that is handled in core and the artificial DRM_AUTH restriction has been lifted with commit [1].

-Emil

[1] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?h=v5.7-rc2&id=30a958526d2cc6df38347336a602479d048d92e7

Okay, most likely drmAuth is irrelevant here (I don't know much about it to be honest; I know that Android uses render nodes, so I figured that drmAuth must therefore be the thing that it doesn't use). Sorry for the confusion. Here is a better explanation of why I needed this change.

Android has a composer process that opens the primary node and uses DRM_IOCTL_MODE_ATOMIC to switch between frame buffers, and a renderer process (surfaceflinger) that opens the render node, prepares frame buffers and sends them to the composer. One idea for adapting this architecture to devices without render nodes is to have the renderer process open the primary node instead. But this runs into a problem:
suppose that the renderer process starts before the composer process. In this case, the kernel makes the renderer the DRM master, so the composer can't change the frame buffer. Render nodes don't have this problem because opening them doesn't affect the master.

Quoted 6 lines

I considered fixing this by having the composer issue DRM_IOCTL_SET_MASTER, but this requires root privileges. If we require drivers to provide render nodes and control access to the primary node while opening up the render node, we can ensure that only authorized processes can become the master without requiring them to be root.

I think that the crux, is trying to open a _primary_ node for _rendering_ purposes.
While that may work - as you outline above - it's usually a bad idea.

To step back a bit, in practise we have three SoC combinations:
- complete lack of render device - then you default to software rendering [1]
- display and render device are one and the same - no change needed,
things should just work
- display and render devices are separate - surfaceflinger should
open the correct _rendering_ device node.

Quoted 6 lines

[1] Mesa's libEGL (don't recall exact name on Android) can open VGEM render node, to work with the kms-swrast_dri.so software rasteriser module.

While I did not try [1] with Android, it was working fine with CrOS. I suggest giving it a try and reporting bugs.

VGEM is not a solution, because it doesn't coordinate caching behavior with your display node and so you get corruption if you try to to share dma-bufs between them. In cros it's used only for some tests as far as I've heard, and it's been a source of pain.

Quoted 5 lines

If we want to go the route of "kms-only nodes also provide a render node interface for doing swrast on", I think that would be a good idea, but we should do this at a core DRM level (probably "does this driver expose dma-buf? then also expose render nodes") rather than per kms driver.

Hm I thought to fix that vgem was switched over to wc mmap?

Probably still broken in some hilarious ways somewhere. -Daniel

Signature · 7 lines
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch
_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel
#11 lore raw

Re: [PATCH] drm: enable render nodes wherever buffer sharing is supported

In reply to Liviu Dudau
Quoted 102 lines · Liviu Dudau
Quoted 92 lines · Peter Collingbourne
Quoted 18 lines · Eric Anholt
Quoted 10 lines · Peter Collingbourne

Render nodes are not just useful for devices supporting GPU hardware acceleration. Even on devices that only support dumb frame buffers, they are useful in situations where composition (using software rasterization) and KMS are done in different processes with buffer sharing being used to send frame buffers between them. This is the situation on Android, where surfaceflinger is the compositor and the composer HAL uses KMS to display the buffers. Thus it is beneficial to expose render nodes on all devices that support buffer sharing.

Quoted 5 lines

Because all drivers that currently support render nodes also support buffer sharing, the DRIVER_RENDER flag is no longer necessary to mark devices as supporting render nodes, so remove it and just rely on the presence of a prime_handle_to_fd function pointer to determine whether buffer sharing is supported.

Quoted 7 lines

I'm definitely interested in seeing a patch like this land, as I think the current state is an ugly historical artifact. We just have to be careful.

Were there any instances of drivers with render engines exposing PRIME but not RENDER? We should be careful to make sure that we're not exposing new privileges for those through adding render nodes.

These are the drivers that we'd be adding render nodes for with this change:

$ git grep -l prime_handle_to_fd (git grep -L DRIVER_RENDER (git grep
-l '\.driver_features'))
drivers/gpu/drm/arc/arcpgu_drv.c
drivers/gpu/drm/arm/display/komeda/komeda_kms.c
drivers/gpu/drm/arm/hdlcd_drv.c
drivers/gpu/drm/arm/malidp_drv.c
drivers/gpu/drm/armada/armada_drv.c
drivers/gpu/drm/aspeed/aspeed_gfx_drv.c
drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_dc.c
drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_drv.c
drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c
drivers/gpu/drm/imx/imx-drm-core.c
drivers/gpu/drm/ingenic/ingenic-drm.c
drivers/gpu/drm/mcde/mcde_drv.c
drivers/gpu/drm/mediatek/mtk_drm_drv.c
drivers/gpu/drm/meson/meson_drv.c
drivers/gpu/drm/mxsfb/mxsfb_drv.c
drivers/gpu/drm/pl111/pl111_drv.c
drivers/gpu/drm/qxl/qxl_drv.c
drivers/gpu/drm/rcar-du/rcar_du_drv.c
drivers/gpu/drm/rockchip/rockchip_drm_drv.c
drivers/gpu/drm/shmobile/shmob_drm_drv.c
drivers/gpu/drm/sti/sti_drv.c
drivers/gpu/drm/stm/drv.c
drivers/gpu/drm/tilcdc/tilcdc_drv.c
drivers/gpu/drm/tve200/tve200_drv.c
drivers/gpu/drm/xen/xen_drm_front.c
drivers/gpu/drm/zte/zx_drm_drv.c

Some of the drivers provide custom ioctls but they are already protected from render nodes by not setting DRM_RENDER_ALLOW. Another thing to check for is drivers providing custom fops that might expose something undesirable in the render node:

$ git grep -L 'DEFINE_DRM_GEM_CMA_FOPS\|DEFINE_DRM_GEM_FOPS' (git grep
-l prime_handle_to_fd (git grep -L DRIVER_RENDER (git grep -l
'\.driver_features')))
drivers/gpu/drm/mediatek/mtk_drm_drv.c
drivers/gpu/drm/rockchip/rockchip_drm_drv.c
drivers/gpu/drm/xen/xen_drm_front.c

But looking at those drivers in detail, I see that each of them is using the standard DRM fops handlers (which presumably already handle render nodes correctly) with the exception of mmap, which they provide wrappers for that mostly just wrap drm_gem_mmap.

So unless I'm missing something significant (which seems likely -- I'm not a DRM expert), I don't see a problem so far.

Quoted 11 lines

There's a UAPI risk I see here. Right now, on a system with a single renderer GPU, we can just open /dev/dri/renderD128 and get the GPU for rendering, and various things are relying on that (such as libwaffle,
used in piglit among other things) Adding render nodes for kms-only
drivers could displace the actual GPU to 129, and the question is whether this will be disruptive. For Mesa, I think this works out, because kmsro should load on the kms device's node and then share buffers over to the real GPU that it digs around to find at init time. Just saying, I'm not sure I know all of the userspace well enough to say "this should be safe despite that"

Quoted 2 lines

(And, maybe, if we decide that it's not safe enough, we could punt kms-only drivers to a higher starting number?)

Quoted 7 lines

Android (minigbm) similarly tries to open /dev/dri/renderD$N in a loop with 128 <= N < 192 and assumes that the first non-blacklisted (i.e. not vgem) one that it can open corresponds to the real GPU [1]. I think that the risk of breaking something on Android is low since Android's architecture basically already depends on there being a render node, and it seems unlikely for a device to have more than one GPU, one of which would be non-functional.

Quoted 4 lines

Would Juno suffer from this issue? It has 2 HDLCD drivers (both can be active at the same time if you want) and you could run panfrost for the actual GPU. Depending on the probing order, HDLCD render nodes would be registered before the GPU's.

If I'm reading the Mesa code [1] correctly, it will loop through the render nodes and attempt to match the driver name against its list of known devices. So if there were two hdlcd render nodes and one panfrost render node it should find the latter regardless of probing order, because it doesn't know about hdlcd.

Peter

[1] https://cs.android.com/android/platform/superproject/+/master:external/mesa3d/src/gallium/auxiliary/pipe-loader/pipe_loader_drm.c;l=214

Quoted 26 lines

Best regards,
Liviu

Quoted 16 lines

It's also worth bearing in mind that render nodes were added to vgem in commit 3a6eb795 from 2018. To the extent that exposing additional render nodes would lead to widespread breakage, this would seem to me to be a likely way in which it could have happened (since I would expect that it could cause many machines to go from having one render node to having more than one), so perhaps the argument can be made that if we hadn't seen widespread breakage as a result of that change, we'd be unlikely to see it as a result of this one.

This would be conditional on the userspace code not blacklisting the vgem render node like minigbm does -- at a glance I couldn't find such code in Mesa (there does appear to be some code that looks for the vgem driver name, but it seems to only be used on primary nodes, not render nodes) or libwaffle.

Quoted 3 lines

Peter

[1] https://cs.android.com/android/platform/superproject/+/master:external/minigbm/cros_gralloc/cros_gralloc_driver.cc;l=48

--
====================

Quoted 7 lines
| I would like to |
| fix the world,  |
| but they're not |
| giving me the   |
 \ source code!  /
  ---------------
    ¯\_(ツ)_/¯

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel

#12 lore raw

Re: [PATCH] drm: enable render nodes wherever buffer sharing is supported

In reply to Peter Collingbourne

Hi Peter,

Quoted 8 lines · Peter Collingbourne

Render nodes are not just useful for devices supporting GPU hardware acceleration. Even on devices that only support dumb frame buffers, they are useful in situations where composition (using software rasterization) and KMS are done in different processes with buffer sharing being used to send frame buffers between them. This is the situation on Android, where surfaceflinger is the compositor and the composer HAL uses KMS to display the buffers. Thus it is beneficial to expose render nodes on all devices that support buffer sharing.

If I understand your problem right, the issue is that you're getting your buffers in minigbm via pl111, which means you want a render node just for buffer allocation? Then HWComposer will import those on the primary node for displaying.

I'm not at all familiar with how minigbm sits in Android, but on our platforms where the Display and GPU devices are different, we allocate via ion in Gralloc, and import those into both the GPU and Display. I think that should work on pl111 too - if you can allocate contiguous memory from ion (or dma-buf heaps) in minigbm, then you don't need the render node.

Cheers,
-Brian
_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel

#14 lore raw

Re: [PATCH] drm: enable render nodes wherever buffer sharing is supported

In reply to Peter Collingbourne
Quoted 28 lines · Peter Collingbourne
Quoted 13 lines · Brian Starkey

Hi Peter,

Quoted 8 lines · Peter Collingbourne

Render nodes are not just useful for devices supporting GPU hardware acceleration. Even on devices that only support dumb frame buffers, they are useful in situations where composition (using software rasterization) and KMS are done in different processes with buffer sharing being used to send frame buffers between them. This is the situation on Android, where surfaceflinger is the compositor and the composer HAL uses KMS to display the buffers. Thus it is beneficial to expose render nodes on all devices that support buffer sharing.

Quoted 4 lines

If I understand your problem right, the issue is that you're getting your buffers in minigbm via pl111, which means you want a render node just for buffer allocation? Then HWComposer will import those on the primary node for displaying.

Correct.

Quoted 6 lines

I'm not at all familiar with how minigbm sits in Android, but on our platforms where the Display and GPU devices are different, we allocate via ion in Gralloc, and import those into both the GPU and Display. I think that should work on pl111 too - if you can allocate contiguous memory from ion (or dma-buf heaps) in minigbm, then you don't need the render node.

Quoted 8 lines

Those contiguous memory regions would not necessarily be compatible with the pl111 device as far as I know -- according to [1], on some devices, a designated memory region must be used for the framebuffer and therefore the memory region allocated in CMA would not be compatible. That being said, FVP doesn't appear to be one of those devices, so in principle that might work for FVP (provided that the user provides a sufficiently large cma=X kernel command line argument), but not for those other devices.

We have been doing that with mali-dp drivers for years. Because most of our dev environments are on FPGAs, we needed to use the local RAM on those boards so we've had to assign a memory region to the driver that in turn was using CMA. We've made heavy use of 'reserved-memory' and 'memory-region' nodes in the DT to link the two together, but things worked out quite well. Because the 'reserved-memory' sub-node was marked as 'compatible="shared-dma-pool"', gralloc and ION could be used to allocate memory from it.

Best regards,
Liviu

Quoted 4 lines

Peter

[1] https://www.kernel.org/doc/Documentation/devicetree/bindings/display/arm%2Cpl11x.txt

Signature · 12 lines
====================
| I would like to |
| fix the world,  |
| but they're not |
| giving me the   |
 \ source code!  /
  ---------------
    ¯\_(ツ)_/¯
_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel
#13 lore raw

Re: [PATCH] drm: enable render nodes wherever buffer sharing is supported

In reply to Brian Starkey
Quoted 13 lines · Brian Starkey

Hi Peter,

Quoted 8 lines · Peter Collingbourne

Render nodes are not just useful for devices supporting GPU hardware acceleration. Even on devices that only support dumb frame buffers, they are useful in situations where composition (using software rasterization) and KMS are done in different processes with buffer sharing being used to send frame buffers between them. This is the situation on Android, where surfaceflinger is the compositor and the composer HAL uses KMS to display the buffers. Thus it is beneficial to expose render nodes on all devices that support buffer sharing.

Quoted 4 lines

If I understand your problem right, the issue is that you're getting your buffers in minigbm via pl111, which means you want a render node just for buffer allocation? Then HWComposer will import those on the primary node for displaying.

Correct.

Quoted 6 lines

I'm not at all familiar with how minigbm sits in Android, but on our platforms where the Display and GPU devices are different, we allocate via ion in Gralloc, and import those into both the GPU and Display. I think that should work on pl111 too - if you can allocate contiguous memory from ion (or dma-buf heaps) in minigbm, then you don't need the render node.

Those contiguous memory regions would not necessarily be compatible with the pl111 device as far as I know -- according to [1], on some devices, a designated memory region must be used for the framebuffer and therefore the memory region allocated in CMA would not be compatible. That being said, FVP doesn't appear to be one of those devices, so in principle that might work for FVP (provided that the user provides a sufficiently large cma=X kernel command line argument), but not for those other devices.

Peter

[1] https://www.kernel.org/doc/Documentation/devicetree/bindings/display/arm%2Cpl11x.txt
_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel

#15 lore raw

Re: [PATCH] drm: enable render nodes wherever buffer sharing is supported

In reply to Liviu Dudau
Quoted 29 lines · Liviu Dudau
Quoted 28 lines · Peter Collingbourne
Quoted 13 lines · Brian Starkey

Hi Peter,

Quoted 8 lines · Peter Collingbourne

Render nodes are not just useful for devices supporting GPU hardware acceleration. Even on devices that only support dumb frame buffers, they are useful in situations where composition (using software rasterization) and KMS are done in different processes with buffer sharing being used to send frame buffers between them. This is the situation on Android, where surfaceflinger is the compositor and the composer HAL uses KMS to display the buffers. Thus it is beneficial to expose render nodes on all devices that support buffer sharing.

Quoted 4 lines

If I understand your problem right, the issue is that you're getting your buffers in minigbm via pl111, which means you want a render node just for buffer allocation? Then HWComposer will import those on the primary node for displaying.

Correct.

Quoted 6 lines

I'm not at all familiar with how minigbm sits in Android, but on our platforms where the Display and GPU devices are different, we allocate via ion in Gralloc, and import those into both the GPU and Display. I think that should work on pl111 too - if you can allocate contiguous memory from ion (or dma-buf heaps) in minigbm, then you don't need the render node.

Quoted 8 lines
Quoted 8 lines

Those contiguous memory regions would not necessarily be compatible with the pl111 device as far as I know -- according to [1], on some devices, a designated memory region must be used for the framebuffer and therefore the memory region allocated in CMA would not be compatible. That being said, FVP doesn't appear to be one of those devices, so in principle that might work for FVP (provided that the user provides a sufficiently large cma=X kernel command line argument), but not for those other devices.

Yeah I'd be surprised if FVP cares about anything other than it being contiguous.

My understanding of how "most" Android devices implement this is to have ion heaps representing whatever dedicated memory regions there are. Gralloc can then pick the appropriate one based on the gralloc usage flags. So allocations wouldn't go via the DRM driver.

It looks like the checks in pl111 can't support that usage model if use_device_memory is set (though that wouldn't matter on FVP).

Quoted 9 lines

We have been doing that with mali-dp drivers for years. Because most of our dev environments are on FPGAs, we needed to use the local RAM on those boards so we've had to assign a memory region to the driver that in turn was using CMA. We've made heavy use of 'reserved-memory' and 'memory-region' nodes in the DT to link the two together, but things worked out quite well. Because the 'reserved-memory' sub-node was marked as 'compatible="shared-dma-pool"', gralloc and ION could be used to allocate memory from it.

This is a little imperfect because ion doesn't have a way to access the dev pointer needed to allocate from device-attached CMA regions, so some hackery is required.

I think dma-heaps would let you expose device-specific CMA regions.

Even if you did that and allocated from the right place, the check in pl111 would reject any attempt to import it if it's set to use_device_memory.

I don't know if there's a better way to tell if the memory is compatible, but that check seems like a bit of a sledgehammer. It looks like it not only forces the memory to be from the right place, it also forces it to have been allocated via pl111.

On FVP though, I reckon any old CMA memory should be fine.

Cheers,
-Brian

Quoted 18 lines

Best regards,
Liviu

Quoted 4 lines

Peter

[1] https://www.kernel.org/doc/Documentation/devicetree/bindings/display/arm%2Cpl11x.txt

Signature · 8 lines
====================
| I would like to |
| fix the world,  |
| but they're not |
| giving me the   |
 \ source code!  /
  ---------------
    ¯\_(ツ)_/¯

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel

#16 lore raw

Re: [PATCH] drm: enable render nodes wherever buffer sharing is supported

Quoted 61 lines · Brian Starkey
Quoted 29 lines · Liviu Dudau
Quoted 28 lines · Peter Collingbourne
Quoted 13 lines · Brian Starkey

Hi Peter,

Quoted 8 lines · Peter Collingbourne

Render nodes are not just useful for devices supporting GPU hardware acceleration. Even on devices that only support dumb frame buffers, they are useful in situations where composition (using software rasterization) and KMS are done in different processes with buffer sharing being used to send frame buffers between them. This is the situation on Android, where surfaceflinger is the compositor and the composer HAL uses KMS to display the buffers. Thus it is beneficial to expose render nodes on all devices that support buffer sharing.

Quoted 4 lines

If I understand your problem right, the issue is that you're getting your buffers in minigbm via pl111, which means you want a render node just for buffer allocation? Then HWComposer will import those on the primary node for displaying.

Correct.

Quoted 6 lines

I'm not at all familiar with how minigbm sits in Android, but on our platforms where the Display and GPU devices are different, we allocate via ion in Gralloc, and import those into both the GPU and Display. I think that should work on pl111 too - if you can allocate contiguous memory from ion (or dma-buf heaps) in minigbm, then you don't need the render node.

Quoted 8 lines
Quoted 8 lines

Those contiguous memory regions would not necessarily be compatible with the pl111 device as far as I know -- according to [1], on some devices, a designated memory region must be used for the framebuffer and therefore the memory region allocated in CMA would not be compatible. That being said, FVP doesn't appear to be one of those devices, so in principle that might work for FVP (provided that the user provides a sufficiently large cma=X kernel command line argument), but not for those other devices.

Yeah I'd be surprised if FVP cares about anything other than it being contiguous.

My understanding of how "most" Android devices implement this is to have ion heaps representing whatever dedicated memory regions there are. Gralloc can then pick the appropriate one based on the gralloc usage flags. So allocations wouldn't go via the DRM driver.

It looks like the checks in pl111 can't support that usage model if use_device_memory is set (though that wouldn't matter on FVP).

Quoted 9 lines

We have been doing that with mali-dp drivers for years. Because most of our dev environments are on FPGAs, we needed to use the local RAM on those boards so we've had to assign a memory region to the driver that in turn was using CMA. We've made heavy use of 'reserved-memory' and 'memory-region' nodes in the DT to link the two together, but things worked out quite well. Because the 'reserved-memory' sub-node was marked as 'compatible="shared-dma-pool"', gralloc and ION could be used to allocate memory from it.

Quoted 5 lines

This is a little imperfect because ion doesn't have a way to access the dev pointer needed to allocate from device-attached CMA regions, so some hackery is required.

I think dma-heaps would let you expose device-specific CMA regions.

The plan was that each device in sysfs which needs special dma memory would have a pointer to the corresponding dma-heap. That way userspace knows where to allocate. Just need some userspace to use these, and kernel patch to expose those links. I think it's defacto already there through the dma pointers of struct device.
-Daniel

Quoted 37 lines

Even if you did that and allocated from the right place, the check in pl111 would reject any attempt to import it if it's set to use_device_memory.

I don't know if there's a better way to tell if the memory is compatible, but that check seems like a bit of a sledgehammer. It looks like it not only forces the memory to be from the right place, it also forces it to have been allocated via pl111.

On FVP though, I reckon any old CMA memory should be fine.

Cheers,
-Brian

Quoted 11 lines

Best regards,
Liviu

Quoted 4 lines

Peter

[1] https://www.kernel.org/doc/Documentation/devicetree/bindings/display/arm%2Cpl11x.txt

--
====================

Quoted 7 lines
| I would like to |
| fix the world,  |
| but they're not |
| giving me the   |
 \ source code!  /
  ---------------
    ¯\_(ツ)_/¯

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel

Signature · 7 lines
Daniel Vetter
Software Engineer, Intel Corporation
+41 (0) 79 365 57 48 - http://blog.ffwll.ch
_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel
#17 lore raw

Re: [PATCH] drm: enable render nodes wherever buffer sharing is supported

In reply to Peter Collingbourne
Quoted 10 lines · Peter Collingbourne

Render nodes are not just useful for devices supporting GPU hardware acceleration. Even on devices that only support dumb frame buffers, they are useful in situations where composition (using software rasterization) and KMS are done in different processes with buffer sharing being used to send frame buffers between them. This is the situation on Android, where surfaceflinger is the compositor and the composer HAL uses KMS to display the buffers. Thus it is beneficial to expose render nodes on all devices that support buffer sharing.

Quoted 5 lines

Because all drivers that currently support render nodes also support buffer sharing, the DRIVER_RENDER flag is no longer necessary to mark devices as supporting render nodes, so remove it and just rely on the presence of a prime_handle_to_fd function pointer to determine whether buffer sharing is supported.

The idea behind render nodes is that you can freely pass these to unpriviledged users, and nothing bad happens. That's why we have gpu reset code in drivers, proper pagetables, and also (in at least the solid drivers) ban code so that repeat offenders from userspace who constantly submit endless batch buffers and funny stuff like that can't DOS the gpu.

Ofc in practice there's hw issues and fun stuff like that sometimes, and driver bugs, and all that. But that's the aspiration.

Now many of these display-only drivers need contiguous buffers, and there's not endless amounts of that around. So if you allow random clients to allocate buffers, they can easily exhaust that, and not just upset the render side of the gpu, but essentially make it impossible for a compositor to allocate more framebuffers. I don't think that's a good idea.

I know there's hw like vc4 which needs contiguous buffers for everything, but that's kinda the places where aspiration falls a bit short.

So from that pov I'm a rather worried with handing out render rights to everyone for these display-only buffers. It's not entirely harmless.
-Daniel

Quoted 305 lines
---
 Documentation/gpu/drm-uapi.rst          | 9 +++++----
 drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 2 +-
 drivers/gpu/drm/drm_drv.c               | 2 +-
 drivers/gpu/drm/etnaviv/etnaviv_drv.c   | 2 +-
 drivers/gpu/drm/exynos/exynos_drm_drv.c | 2 +-
 drivers/gpu/drm/i915/i915_drv.c         | 2 +-
 drivers/gpu/drm/lima/lima_drv.c         | 2 +-
 drivers/gpu/drm/msm/msm_drv.c           | 1 -
 drivers/gpu/drm/nouveau/nouveau_drm.c   | 2 +-
 drivers/gpu/drm/omapdrm/omap_drv.c      | 2 +-
 drivers/gpu/drm/panfrost/panfrost_drv.c | 2 +-
 drivers/gpu/drm/radeon/radeon_drv.c     | 3 +--
 drivers/gpu/drm/tegra/drm.c             | 2 +-
 drivers/gpu/drm/v3d/v3d_drv.c           | 1 -
 drivers/gpu/drm/vc4/vc4_drv.c           | 8 --------
 drivers/gpu/drm/vgem/vgem_drv.c         | 2 +-
 drivers/gpu/drm/virtio/virtgpu_drv.c    | 2 +-
 drivers/gpu/drm/vmwgfx/vmwgfx_drv.c     | 2 +-
 include/drm/drm_drv.h                   | 7 -------
 19 files changed, 19 insertions(+), 36 deletions(-)
Documentation/gpu/drm-uapi.rst +5 -4
index 56fec6ed1ad8..2769714ae75a 100644
@@ -129,10 +129,11 @@ authenticate to a DRM-Master prior to getting GPU access. To avoid this
 step and to grant clients GPU access without authenticating, render
 nodes were introduced. Render nodes solely serve render clients, that
 is, no modesetting or privileged ioctls can be issued on render nodes.
-Only non-global rendering commands are allowed. If a driver supports
-render nodes, it must advertise it via the DRIVER_RENDER DRM driver
-capability. If not supported, the primary node must be used for render
-clients together with the legacy drmAuth authentication procedure.
+Only non-global rendering commands are allowed. Drivers that support
+:ref:`PRIME buffer sharing <prime_buffer_sharing>` automatically
+support render nodes. If a driver does not support render nodes,
+the primary node must be used for render clients together with the
+legacy drmAuth authentication procedure.
 
 If a driver advertises render node support, DRM core will create a
 separate render node called renderD<num>. There will be one render node
drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +1 -1
index 8ea86ffdea0d..46460620bc37 100644
@@ -1426,7 +1426,7 @@ static struct drm_driver kms_driver = {
        .driver_features =
            DRIVER_ATOMIC |
            DRIVER_GEM |
-           DRIVER_RENDER | DRIVER_MODESET | DRIVER_SYNCOBJ |
+           DRIVER_MODESET | DRIVER_SYNCOBJ |
            DRIVER_SYNCOBJ_TIMELINE,
        .open = amdgpu_driver_open_kms,
        .postclose = amdgpu_driver_postclose_kms,
drivers/gpu/drm/drm_drv.c +1 -1
index 7b1a628d1f6e..8861a30920e5 100644
@@ -651,7 +651,7 @@ int drm_dev_init(struct drm_device *dev,
                goto err_free;
        }
 
-       if (drm_core_check_feature(dev, DRIVER_RENDER)) {
+       if (driver->prime_handle_to_fd) {
                ret = drm_minor_alloc(dev, DRM_MINOR_RENDER);
                if (ret)
                        goto err_minors;
drivers/gpu/drm/etnaviv/etnaviv_drv.c +1 -1
index a8685b2e1803..abfb143334ac 100644
@@ -497,7 +497,7 @@ static const struct file_operations fops = {
 };
 
 static struct drm_driver etnaviv_drm_driver = {
-       .driver_features    = DRIVER_GEM | DRIVER_RENDER,
+       .driver_features    = DRIVER_GEM,
        .open               = etnaviv_open,
        .postclose           = etnaviv_postclose,
        .gem_free_object_unlocked = etnaviv_gem_free_object,
drivers/gpu/drm/exynos/exynos_drm_drv.c +1 -1
index 57defeb44522..834eb5713fe4 100644
@@ -121,7 +121,7 @@ static const struct file_operations exynos_drm_driver_fops = {
 
 static struct drm_driver exynos_drm_driver = {
        .driver_features        = DRIVER_MODESET | DRIVER_GEM
-                                 | DRIVER_ATOMIC | DRIVER_RENDER,
+                                 | DRIVER_ATOMIC,
        .open                   = exynos_drm_open,
        .lastclose              = drm_fb_helper_lastclose,
        .postclose              = exynos_drm_postclose,
drivers/gpu/drm/i915/i915_drv.c +1 -1
index 81a4621853db..b028a32fcac5 100644
@@ -1834,7 +1834,7 @@ static struct drm_driver driver = {
         */
        .driver_features =
            DRIVER_GEM |
-           DRIVER_RENDER | DRIVER_MODESET | DRIVER_ATOMIC | DRIVER_SYNCOBJ,
+           DRIVER_MODESET | DRIVER_ATOMIC | DRIVER_SYNCOBJ,
        .release = i915_driver_release,
        .open = i915_driver_open,
        .lastclose = i915_driver_lastclose,
drivers/gpu/drm/lima/lima_drv.c +1 -1
index 2daac64d8955..bd8b6ad25ac0 100644
@@ -252,7 +252,7 @@ DEFINE_DRM_GEM_FOPS(lima_drm_driver_fops);
  */
 
 static struct drm_driver lima_drm_driver = {
-       .driver_features    = DRIVER_RENDER | DRIVER_GEM | DRIVER_SYNCOBJ,
+       .driver_features    = DRIVER_GEM | DRIVER_SYNCOBJ,
        .open               = lima_drm_driver_open,
        .postclose          = lima_drm_driver_postclose,
        .ioctls             = lima_drm_driver_ioctls,
drivers/gpu/drm/msm/msm_drv.c +0 -1
index 29295dee2a2e..061e62d4b691 100644
@@ -1000,7 +1000,6 @@ static const struct file_operations fops = {
 
 static struct drm_driver msm_driver = {
        .driver_features    = DRIVER_GEM |
-                               DRIVER_RENDER |
                                DRIVER_ATOMIC |
                                DRIVER_MODESET,
        .open               = msm_open,
drivers/gpu/drm/nouveau/nouveau_drm.c +1 -1
index ca4087f5a15b..3ba8c9789292 100644
@@ -1169,7 +1169,7 @@ nouveau_driver_fops = {
 static struct drm_driver
 driver_stub = {
        .driver_features =
-               DRIVER_GEM | DRIVER_MODESET | DRIVER_RENDER
+               DRIVER_GEM | DRIVER_MODESET
 #if defined(CONFIG_NOUVEAU_LEGACY_CTX_SUPPORT)
                | DRIVER_KMS_LEGACY_CONTEXT
 #endif
drivers/gpu/drm/omapdrm/omap_drv.c +1 -1
index cdafd7ef1c32..fe25b352a755 100644
@@ -558,7 +558,7 @@ static const struct file_operations omapdriver_fops = {
 
 static struct drm_driver omap_drm_driver = {
        .driver_features = DRIVER_MODESET | DRIVER_GEM  |
-               DRIVER_ATOMIC | DRIVER_RENDER,
+               DRIVER_ATOMIC,
        .open = dev_open,
        .lastclose = drm_fb_helper_lastclose,
 #ifdef CONFIG_DEBUG_FS
drivers/gpu/drm/panfrost/panfrost_drv.c +1 -1
index 882fecc33fdb..058765d5d5d5 100644
@@ -550,7 +550,7 @@ DEFINE_DRM_GEM_FOPS(panfrost_drm_driver_fops);
  * - 1.1 - adds HEAP and NOEXEC flags for CREATE_BO
  */
 static struct drm_driver panfrost_drm_driver = {
-       .driver_features        = DRIVER_RENDER | DRIVER_GEM | DRIVER_SYNCOBJ,
+       .driver_features        = DRIVER_GEM | DRIVER_SYNCOBJ,
        .open                   = panfrost_open,
        .postclose              = panfrost_postclose,
        .ioctls                 = panfrost_drm_driver_ioctls,
drivers/gpu/drm/radeon/radeon_drv.c +1 -2
index 59f8186a2415..9457b2f8f81e 100644
@@ -600,8 +600,7 @@ static const struct file_operations radeon_driver_kms_fops = {
 };
 
 static struct drm_driver kms_driver = {
-       .driver_features =
-           DRIVER_GEM | DRIVER_RENDER,
+       .driver_features = DRIVER_GEM,
        .load = radeon_driver_load_kms,
        .open = radeon_driver_open_kms,
        .postclose = radeon_driver_postclose_kms,
drivers/gpu/drm/tegra/drm.c +1 -1
index bd268028fb3d..55190c582946 100644
@@ -849,7 +849,7 @@ static int tegra_debugfs_init(struct drm_minor *minor)
 
 static struct drm_driver tegra_drm_driver = {
        .driver_features = DRIVER_MODESET | DRIVER_GEM |
-                          DRIVER_ATOMIC | DRIVER_RENDER,
+                          DRIVER_ATOMIC,
        .open = tegra_drm_open,
        .postclose = tegra_drm_postclose,
        .lastclose = drm_fb_helper_lastclose,
drivers/gpu/drm/v3d/v3d_drv.c +0 -1
index eaa8e9682373..96db702fd648 100644
@@ -195,7 +195,6 @@ static const struct drm_ioctl_desc v3d_drm_ioctls[] = {
 
 static struct drm_driver v3d_drm_driver = {
        .driver_features = (DRIVER_GEM |
-                           DRIVER_RENDER |
                            DRIVER_SYNCOBJ),
 
        .open = v3d_open,
drivers/gpu/drm/vc4/vc4_drv.c +0 -8
index 76f93b662766..bf143d1e0d2e 100644
@@ -181,7 +181,6 @@ static struct drm_driver vc4_drm_driver = {
        .driver_features = (DRIVER_MODESET |
                            DRIVER_ATOMIC |
                            DRIVER_GEM |
-                           DRIVER_RENDER |
                            DRIVER_SYNCOBJ),
        .open = vc4_open,
        .postclose = vc4_close,
@@ -251,7 +250,6 @@ static int vc4_drm_bind(struct device *dev)
        struct platform_device *pdev = to_platform_device(dev);
        struct drm_device *drm;
        struct vc4_dev *vc4;
-       struct device_node *node;
        int ret = 0;
 
        dev->coherent_dma_mask = DMA_BIT_MASK(32);
@@ -260,12 +258,6 @@ static int vc4_drm_bind(struct device *dev)
        if (!vc4)
                return -ENOMEM;
 
-       /* If VC4 V3D is missing, don't advertise render nodes. */
-       node = of_find_matching_node_and_match(NULL, vc4_v3d_dt_match, NULL);
-       if (!node || !of_device_is_available(node))
-               vc4_drm_driver.driver_features &= ~DRIVER_RENDER;
-       of_node_put(node);
-
        drm = drm_dev_alloc(&vc4_drm_driver, dev);
        if (IS_ERR(drm))
                return PTR_ERR(drm);
drivers/gpu/drm/vgem/vgem_drv.c +1 -1
index 909eba43664a..abc7345a2e96 100644
@@ -437,7 +437,7 @@ static void vgem_release(struct drm_device *dev)
 }
 
 static struct drm_driver vgem_driver = {
-       .driver_features                = DRIVER_GEM | DRIVER_RENDER,
+       .driver_features                = DRIVER_GEM,
        .release                        = vgem_release,
        .open                           = vgem_open,
        .postclose                      = vgem_postclose,
drivers/gpu/drm/virtio/virtgpu_drv.c +1 -1
index ab4bed78e656..22195e008f0d 100644
@@ -189,7 +189,7 @@ MODULE_AUTHOR("Alon Levy");
 DEFINE_DRM_GEM_FOPS(virtio_gpu_driver_fops);
 
 static struct drm_driver driver = {
-       .driver_features = DRIVER_MODESET | DRIVER_GEM | DRIVER_RENDER | DRIVER_ATOMIC,
+       .driver_features = DRIVER_MODESET | DRIVER_GEM | DRIVER_ATOMIC,
        .open = virtio_gpu_driver_open,
        .postclose = virtio_gpu_driver_postclose,
 
drivers/gpu/drm/vmwgfx/vmwgfx_drv.c +1 -1
index c2247a893ed4..415c3f8ea907 100644
@@ -1435,7 +1435,7 @@ static const struct file_operations vmwgfx_driver_fops = {
 
 static struct drm_driver driver = {
        .driver_features =
-       DRIVER_MODESET | DRIVER_RENDER | DRIVER_ATOMIC,
+       DRIVER_MODESET | DRIVER_ATOMIC,
        .ioctls = vmw_ioctls,
        .num_ioctls = ARRAY_SIZE(vmw_ioctls),
        .master_set = vmw_master_set,
include/drm/drm_drv.h +0 -8
index 97109df5beac..f0277d3f89fe 100644
@@ -61,13 +61,6 @@ enum drm_driver_feature {
         * Driver supports mode setting interfaces (KMS).
         */
        DRIVER_MODESET                  = BIT(1),
-       /**
-        * @DRIVER_RENDER:
-        *
-        * Driver supports dedicated render nodes. See also the :ref:`section on
-        * render nodes <drm_render_node>` for details.
-        */
-       DRIVER_RENDER                   = BIT(3),
        /**
         * @DRIVER_ATOMIC:
         *
--

2.26.2.303.gf8c07b1a785-goog

Quoted 4 lines

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel

--
Daniel Vetter
Software Engineer, Intel Corporation
+41 (0) 79 365 57 48 - http://blog.ffwll.ch
_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel

#21 lore raw
In reply to Eric Anholt
Quoted 67 lines · Eric Anholt
Quoted 65 lines · Emil Velikov
Quoted 42 lines · Peter Collingbourne
Quoted 14 lines · Emil Velikov
Quoted 9 lines · Peter Collingbourne

The render node is required by Android which does not support the legacy drmAuth authentication process.

---
 drivers/gpu/drm/pl111/pl111_drv.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

The summary talks about drmAuth, yet exposes a render node. Even through there's no rendering engine in the HW, as mentioned by Eric.

Quoted 8 lines

AFAICT the only way drmAuth is relevant with pl111 is when you want to export/import dma bufs.
Although that is handled in core and the artificial DRM_AUTH restriction has been lifted with commit [1].

-Emil

[1] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?h=v5.7-rc2&id=30a958526d2cc6df38347336a602479d048d92e7

Okay, most likely drmAuth is irrelevant here (I don't know much about it to be honest; I know that Android uses render nodes, so I figured that drmAuth must therefore be the thing that it doesn't use). Sorry for the confusion. Here is a better explanation of why I needed this change.

Android has a composer process that opens the primary node and uses DRM_IOCTL_MODE_ATOMIC to switch between frame buffers, and a renderer process (surfaceflinger) that opens the render node, prepares frame buffers and sends them to the composer. One idea for adapting this architecture to devices without render nodes is to have the renderer process open the primary node instead. But this runs into a problem:
suppose that the renderer process starts before the composer process. In this case, the kernel makes the renderer the DRM master, so the composer can't change the frame buffer. Render nodes don't have this problem because opening them doesn't affect the master.

Quoted 6 lines

I considered fixing this by having the composer issue DRM_IOCTL_SET_MASTER, but this requires root privileges. If we require drivers to provide render nodes and control access to the primary node while opening up the render node, we can ensure that only authorized processes can become the master without requiring them to be root.

I think that the crux, is trying to open a _primary_ node for _rendering_ purposes.
While that may work - as you outline above - it's usually a bad idea.

To step back a bit, in practise we have three SoC combinations:
- complete lack of render device - then you default to software rendering [1]
- display and render device are one and the same - no change needed,
things should just work
- display and render devices are separate - surfaceflinger should
open the correct _rendering_ device node.

[1] Mesa's libEGL (don't recall exact name on Android) can open VGEM render node, to work with the kms-swrast_dri.so software rasteriser module.

Quoted 8 lines
Quoted 2 lines

While I did not try [1] with Android, it was working fine with CrOS. I suggest giving it a try and reporting bugs.

VGEM is not a solution, because it doesn't coordinate caching behavior with your display node and so you get corruption if you try to to share dma-bufs between them. In cros it's used only for some tests as far as I've heard, and it's been a source of pain.

My understanding is that dma_buf_{begin,end}_cpu_access should be used to handle the caching concerns.
Perhaps we're missing some calls, apart from the wc mmap Daniel mentioned?

Fwiw I've played around with CrOS for 30 minutes w/o any corruption. Mind you it was a boot + casual web browsing.

Quoted 5 lines

If we want to go the route of "kms-only nodes also provide a render node interface for doing swrast on", I think that would be a good idea, but we should do this at a core DRM level (probably "does this driver expose dma-buf? then also expose render nodes") rather than per kms driver.

This sounds doable, although I'm concerned about existing applications, which do not expect this behaviour. Be that mesa, compositors or otherwise.

-Emil
_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel

#23 lore raw
In reply to Peter Collingbourne
Quoted 63 lines · Peter Collingbourne
Quoted 61 lines · Emil Velikov
Quoted 42 lines · Peter Collingbourne
Quoted 14 lines · Emil Velikov
Quoted 9 lines · Peter Collingbourne

The render node is required by Android which does not support the legacy drmAuth authentication process.

---
 drivers/gpu/drm/pl111/pl111_drv.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

The summary talks about drmAuth, yet exposes a render node. Even through there's no rendering engine in the HW, as mentioned by Eric.

Quoted 8 lines

AFAICT the only way drmAuth is relevant with pl111 is when you want to export/import dma bufs.
Although that is handled in core and the artificial DRM_AUTH restriction has been lifted with commit [1].

-Emil

[1] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?h=v5.7-rc2&id=30a958526d2cc6df38347336a602479d048d92e7

Okay, most likely drmAuth is irrelevant here (I don't know much about it to be honest; I know that Android uses render nodes, so I figured that drmAuth must therefore be the thing that it doesn't use). Sorry for the confusion. Here is a better explanation of why I needed this change.

Android has a composer process that opens the primary node and uses DRM_IOCTL_MODE_ATOMIC to switch between frame buffers, and a renderer process (surfaceflinger) that opens the render node, prepares frame buffers and sends them to the composer. One idea for adapting this architecture to devices without render nodes is to have the renderer process open the primary node instead. But this runs into a problem:
suppose that the renderer process starts before the composer process. In this case, the kernel makes the renderer the DRM master, so the composer can't change the frame buffer. Render nodes don't have this problem because opening them doesn't affect the master.

Quoted 6 lines

I considered fixing this by having the composer issue DRM_IOCTL_SET_MASTER, but this requires root privileges. If we require drivers to provide render nodes and control access to the primary node while opening up the render node, we can ensure that only authorized processes can become the master without requiring them to be root.

I think that the crux, is trying to open a _primary_ node for _rendering_ purposes.
While that may work - as you outline above - it's usually a bad idea.

To step back a bit, in practise we have three SoC combinations:
- complete lack of render device - then you default to software rendering [1]
- display and render device are one and the same - no change needed,
things should just work
- display and render devices are separate - surfaceflinger should
open the correct _rendering_ device node.

Quoted 4 lines
Quoted 1 line

[1] Mesa's libEGL (don't recall exact name on Android) can open VGEM

(Android uses SwiftShader for software rendering, not Mesa, FWIW.)

I don't know much about SwiftShader.

Quoted 6 lines
Quoted 5 lines

render node, to work with the kms-swrast_dri.so software rasteriser module.

While I did not try [1] with Android, it was working fine with CrOS. I suggest giving it a try and reporting bugs.

Quoted 9 lines

I'm not sure I understand your suggestion, or at least how it would work on Android. The composer process wouldn't be able to use DRM_IOCTL_MODE_ATOMIC to display a frame buffer produced by the surfaceflinger process using a vgem render node, would it? It's a different driver so the memory region used for the frame buffer wouldn't necessarily be compatible with the pl111 device. As far as I know, the frame buffer would need to be produced using a pl111 render node.

The pl111 will create a buffer, exports it. Renderer will import and draw onto it.
Upon completion, the composer will DRM_IOCTL_MODE_ATOMIC the pl111 buffer.

I believe this approach has been used for a few years now.

-Emil
_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel

#18 lore raw

Re: [PATCH] drm: enable render nodes wherever buffer sharing is supported

In reply to Daniel Vetter
Quoted 35 lines · Daniel Vetter
Quoted 10 lines · Peter Collingbourne

Render nodes are not just useful for devices supporting GPU hardware acceleration. Even on devices that only support dumb frame buffers, they are useful in situations where composition (using software rasterization) and KMS are done in different processes with buffer sharing being used to send frame buffers between them. This is the situation on Android, where surfaceflinger is the compositor and the composer HAL uses KMS to display the buffers. Thus it is beneficial to expose render nodes on all devices that support buffer sharing.

Quoted 5 lines

Because all drivers that currently support render nodes also support buffer sharing, the DRIVER_RENDER flag is no longer necessary to mark devices as supporting render nodes, so remove it and just rely on the presence of a prime_handle_to_fd function pointer to determine whether buffer sharing is supported.

The idea behind render nodes is that you can freely pass these to unpriviledged users, and nothing bad happens. That's why we have gpu reset code in drivers, proper pagetables, and also (in at least the solid drivers) ban code so that repeat offenders from userspace who constantly submit endless batch buffers and funny stuff like that can't DOS the gpu.

Ofc in practice there's hw issues and fun stuff like that sometimes, and driver bugs, and all that. But that's the aspiration.

Now many of these display-only drivers need contiguous buffers, and there's not endless amounts of that around. So if you allow random clients to allocate buffers, they can easily exhaust that, and not just upset the render side of the gpu, but essentially make it impossible for a compositor to allocate more framebuffers. I don't think that's a good idea.

Quoted 7 lines

I know there's hw like vc4 which needs contiguous buffers for everything, but that's kinda the places where aspiration falls a bit short.

So from that pov I'm a rather worried with handing out render rights to everyone for these display-only buffers. It's not entirely harmless.

This doesn't feel like a contiguous-mem-specific concern to me. We don't have resource limits on renderer GPU nodes today, you can allocate memory there to fill up and DOS the system, and unless something changed since last time I was looking, we don't even tell the OOM killer about our allocations so they can kill the right app! (my compositor always got picked, in my experience)

And, keep in mind what we're fixing at a system level here: the current workaround is you either get your compositor to hand you a GPU fd, at which point you can DOS the system, or you open the master node yourself and try to drop master before the compositor comes up, then DOS the system. "But there are access controls for the compositor or the card node!" you say: yes, but there are for render nodes too. I know my distro doesn't default to open access to /dev/dri/render*
_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel

#19 lore raw

Re: [PATCH] drm: enable render nodes wherever buffer sharing is supported

Quoted 51 lines · Eric Anholt
Quoted 35 lines · Daniel Vetter
Quoted 10 lines · Peter Collingbourne

Render nodes are not just useful for devices supporting GPU hardware acceleration. Even on devices that only support dumb frame buffers, they are useful in situations where composition (using software rasterization) and KMS are done in different processes with buffer sharing being used to send frame buffers between them. This is the situation on Android, where surfaceflinger is the compositor and the composer HAL uses KMS to display the buffers. Thus it is beneficial to expose render nodes on all devices that support buffer sharing.

Quoted 5 lines

Because all drivers that currently support render nodes also support buffer sharing, the DRIVER_RENDER flag is no longer necessary to mark devices as supporting render nodes, so remove it and just rely on the presence of a prime_handle_to_fd function pointer to determine whether buffer sharing is supported.

The idea behind render nodes is that you can freely pass these to unpriviledged users, and nothing bad happens. That's why we have gpu reset code in drivers, proper pagetables, and also (in at least the solid drivers) ban code so that repeat offenders from userspace who constantly submit endless batch buffers and funny stuff like that can't DOS the gpu.

Ofc in practice there's hw issues and fun stuff like that sometimes, and driver bugs, and all that. But that's the aspiration.

Now many of these display-only drivers need contiguous buffers, and there's not endless amounts of that around. So if you allow random clients to allocate buffers, they can easily exhaust that, and not just upset the render side of the gpu, but essentially make it impossible for a compositor to allocate more framebuffers. I don't think that's a good idea.

Quoted 7 lines

I know there's hw like vc4 which needs contiguous buffers for everything, but that's kinda the places where aspiration falls a bit short.

So from that pov I'm a rather worried with handing out render rights to everyone for these display-only buffers. It's not entirely harmless.

This doesn't feel like a contiguous-mem-specific concern to me. We don't have resource limits on renderer GPU nodes today, you can allocate memory there to fill up and DOS the system, and unless something changed since last time I was looking, we don't even tell the OOM killer about our allocations so they can kill the right app! (my compositor always got picked, in my experience)

Quoted 7 lines

And, keep in mind what we're fixing at a system level here: the current workaround is you either get your compositor to hand you a GPU fd, at which point you can DOS the system, or you open the master node yourself and try to drop master before the compositor comes up, then DOS the system. "But there are access controls for the compositor or the card node!" you say: yes, but there are for render nodes too. I know my distro doesn't default to open access to /dev/dri/render*

Hm indeed, I thought we've limited dumb buffer creation to master only on the primary device.

But that kinda leaves me wondering, how do you allocate buffers on these then? Exposing dumb on render nodes gets the Dave "it's not for rendering" Airlie nack, at least for drivers which do have real userspace sitting on the same drm driver. And exposing dumb only for these display-only devices feels somewhat silly and inconsistent too.

So not sure.
-Daniel

Signature · 7 lines
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch
_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel