Hi,
This series aims to check panel device dependency upon DRM device before
managing device link between them. It fixes eariler patches in v6.7-rc1
which tried to manage the link. Without this series, the link fails to
be added for dependent panel devices and hence relevant panel bridges
fail to be attached. A real broken panel is "novatek,nt35510" defined
in arch/arm/boot/dts/st/ste-ux500-samsung-skomer.dts as reported by
Linus Walleij.
Patch 1 exports device_is_dependent() to modules as needed by patch 2.
Patch 2 checks device dependency before managing the device link.
Note that patch 2 is already in drm-misc/drm-misc-fixes and
drm-misc/for-linux-next-fixes. Patch 1 needs to be reviewed and picked up.
v2:
* Introduce patch 1 to export device_is_dependent() to modules as needed by
patch 2.
Liu Ying (2):
driver core: Export device_is_dependent() to modules
drm/bridge: panel: Check device dependency before managing device link
drivers/base/core.c | 1 +
drivers/gpu/drm/bridge/panel.c | 27 ++++++++++++++++++---------
2 files changed, 19 insertions(+), 9 deletions(-)
--
2.37.1
Export device_is_dependent() since the drm_kms_helper module is starting
to use it.
Signed-off-by: Liu Ying <[email protected]>
---
v2:
* Newly introduced as needed by patch 2.
drivers/base/core.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/drivers/base/core.c b/drivers/base/core.c
index 67ba592afc77..bfd2bf0364b7 100644
--- a/drivers/base/core.c
+++ b/drivers/base/core.c
@@ -328,6 +328,7 @@ int device_is_dependent(struct device *dev, void *target)
}
return ret;
}
+EXPORT_SYMBOL_GPL(device_is_dependent);
static void device_link_init_status(struct device_link *link,
struct device *consumer,
--
2.37.1
Some panel devices already depend on DRM device, like the panel in
arch/arm/boot/dts/st/ste-ux500-samsung-skomer.dts, because DRM device is
the ancestor of those panel devices. device_link_add() would fail by
returning a NULL pointer for those panel devices because of the existing
dependency. So, check the dependency by calling device_is_dependent()
before adding or deleting device link between panel device and DRM device
so that the link is managed only for independent panel devices.
Fixes: 887878014534 ("drm/bridge: panel: Fix device link for DRM_BRIDGE_ATTACH_NO_CONNECTOR")
Fixes: 199cf07ebd2b ("drm/bridge: panel: Add a device link between drm device and panel device")
Reported-by: Linus Walleij <[email protected]>
Closes: https://lore.kernel.org/lkml/CACRpkdaGzXD6HbiX7mVUNJAJtMEPG00Pp6+nJ1P0JrfJ-ArMvQ@mail.gmail.com/T/
Tested-by: Linus Walleij <[email protected]>
Signed-off-by: Liu Ying <[email protected]>
---
v2:
* No change.
drivers/gpu/drm/bridge/panel.c | 27 ++++++++++++++++++---------
1 file changed, 18 insertions(+), 9 deletions(-)
diff --git a/drivers/gpu/drm/bridge/panel.c b/drivers/gpu/drm/bridge/panel.c
index e48823a4f1ed..5e8980023407 100644
--- a/drivers/gpu/drm/bridge/panel.c
+++ b/drivers/gpu/drm/bridge/panel.c
@@ -23,6 +23,7 @@ struct panel_bridge {
struct drm_panel *panel;
struct device_link *link;
u32 connector_type;
+ bool is_independent;
};
static inline struct panel_bridge *
@@ -67,12 +68,17 @@ static int panel_bridge_attach(struct drm_bridge *bridge,
struct drm_device *drm_dev = bridge->dev;
int ret;
- panel_bridge->link = device_link_add(drm_dev->dev, panel->dev,
- DL_FLAG_STATELESS);
- if (!panel_bridge->link) {
- DRM_ERROR("Failed to add device link between %s and %s\n",
- dev_name(drm_dev->dev), dev_name(panel->dev));
- return -EINVAL;
+ panel_bridge->is_independent = !device_is_dependent(drm_dev->dev,
+ panel->dev);
+
+ if (panel_bridge->is_independent) {
+ panel_bridge->link = device_link_add(drm_dev->dev, panel->dev,
+ DL_FLAG_STATELESS);
+ if (!panel_bridge->link) {
+ DRM_ERROR("Failed to add device link between %s and %s\n",
+ dev_name(drm_dev->dev), dev_name(panel->dev));
+ return -EINVAL;
+ }
}
if (flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR)
@@ -80,7 +86,8 @@ static int panel_bridge_attach(struct drm_bridge *bridge,
if (!bridge->encoder) {
DRM_ERROR("Missing encoder\n");
- device_link_del(panel_bridge->link);
+ if (panel_bridge->is_independent)
+ device_link_del(panel_bridge->link);
return -ENODEV;
}
@@ -92,7 +99,8 @@ static int panel_bridge_attach(struct drm_bridge *bridge,
panel_bridge->connector_type);
if (ret) {
DRM_ERROR("Failed to initialize connector\n");
- device_link_del(panel_bridge->link);
+ if (panel_bridge->is_independent)
+ device_link_del(panel_bridge->link);
return ret;
}
@@ -115,7 +123,8 @@ static void panel_bridge_detach(struct drm_bridge *bridge)
struct panel_bridge *panel_bridge = drm_bridge_to_panel_bridge(bridge);
struct drm_connector *connector = &panel_bridge->connector;
- device_link_del(panel_bridge->link);
+ if (panel_bridge->is_independent)
+ device_link_del(panel_bridge->link);
/*
* Cleanup the connector if we know it was initialized.
--
2.37.1
On Mon, Nov 27, 2023 at 6:10 AM Liu Ying <[email protected]> wrote:
> This series aims to check panel device dependency upon DRM device before
> managing device link between them. It fixes eariler patches in v6.7-rc1
> which tried to manage the link. Without this series, the link fails to
> be added for dependent panel devices and hence relevant panel bridges
> fail to be attached. A real broken panel is "novatek,nt35510" defined
> in arch/arm/boot/dts/st/ste-ux500-samsung-skomer.dts as reported by
> Linus Walleij.
>
> Patch 1 exports device_is_dependent() to modules as needed by patch 2.
> Patch 2 checks device dependency before managing the device link.
>
> Note that patch 2 is already in drm-misc/drm-misc-fixes and
> drm-misc/for-linux-next-fixes. Patch 1 needs to be reviewed and picked up.
>
> v2:
> * Introduce patch 1 to export device_is_dependent() to modules as needed by
> patch 2.
>
> Liu Ying (2):
> driver core: Export device_is_dependent() to modules
> drm/bridge: panel: Check device dependency before managing device link
I just applied patch 1 directly to the drm-misc-fixes so we don't have to
revert and then re-apply patches, because that is a bigger evil. (We can't
rebase these branches...)
Yours,
Linus Walleij
On Mon, Nov 27, 2023 at 05:03:53PM +0100, Linus Walleij wrote:
> On Mon, Nov 27, 2023 at 6:10 AM Liu Ying <[email protected]> wrote:
>
> > This series aims to check panel device dependency upon DRM device before
> > managing device link between them. It fixes eariler patches in v6.7-rc1
> > which tried to manage the link. Without this series, the link fails to
> > be added for dependent panel devices and hence relevant panel bridges
> > fail to be attached. A real broken panel is "novatek,nt35510" defined
> > in arch/arm/boot/dts/st/ste-ux500-samsung-skomer.dts as reported by
> > Linus Walleij.
> >
> > Patch 1 exports device_is_dependent() to modules as needed by patch 2.
> > Patch 2 checks device dependency before managing the device link.
> >
> > Note that patch 2 is already in drm-misc/drm-misc-fixes and
> > drm-misc/for-linux-next-fixes. Patch 1 needs to be reviewed and picked up.
> >
> > v2:
> > * Introduce patch 1 to export device_is_dependent() to modules as needed by
> > patch 2.
> >
> > Liu Ying (2):
> > driver core: Export device_is_dependent() to modules
> > drm/bridge: panel: Check device dependency before managing device link
>
> I just applied patch 1 directly to the drm-misc-fixes so we don't have to
> revert and then re-apply patches, because that is a bigger evil. (We can't
> rebase these branches...)
Erm, you did wait for GKH or Rafael's ACK to do that, right?
Maxime
Greg, Rafael,
On Mon, Nov 27, 2023 at 01:14:13PM +0800, Liu Ying wrote:
> Export device_is_dependent() since the drm_kms_helper module is starting
> to use it.
>
> Signed-off-by: Liu Ying <[email protected]>
> ---
> v2:
> * Newly introduced as needed by patch 2.
>
> drivers/base/core.c | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/drivers/base/core.c b/drivers/base/core.c
> index 67ba592afc77..bfd2bf0364b7 100644
> --- a/drivers/base/core.c
> +++ b/drivers/base/core.c
> @@ -328,6 +328,7 @@ int device_is_dependent(struct device *dev, void *target)
> }
> return ret;
> }
> +EXPORT_SYMBOL_GPL(device_is_dependent);
So, a committer just applied this to drm-misc-fixes without your
approval. Could you ack it? If you don't want to, we'll fix it.
Maxime
On Mon, Nov 27, 2023 at 05:38:13PM +0100, Maxime Ripard wrote:
> Greg, Rafael,
>
> On Mon, Nov 27, 2023 at 01:14:13PM +0800, Liu Ying wrote:
> > Export device_is_dependent() since the drm_kms_helper module is starting
> > to use it.
> >
> > Signed-off-by: Liu Ying <[email protected]>
> > ---
> > v2:
> > * Newly introduced as needed by patch 2.
> >
> > drivers/base/core.c | 1 +
> > 1 file changed, 1 insertion(+)
> >
> > diff --git a/drivers/base/core.c b/drivers/base/core.c
> > index 67ba592afc77..bfd2bf0364b7 100644
> > --- a/drivers/base/core.c
> > +++ b/drivers/base/core.c
> > @@ -328,6 +328,7 @@ int device_is_dependent(struct device *dev, void *target)
> > }
> > return ret;
> > }
> > +EXPORT_SYMBOL_GPL(device_is_dependent);
>
> So, a committer just applied this to drm-misc-fixes without your
> approval. Could you ack it? If you don't want to, we'll fix it.
Wait, why exactly is this needed? Nothing outside of the driver core
should be needing this function, it shouldn't be public at all (I missed
that before.)
So please, revert it for now, let's figure out why DRM thinks this is
needed for it's devices, and yet no other bus/subsystem does.
thanks,
greg k-h
On Mon, Nov 27, 2023 at 5:29 PM Maxime Ripard <[email protected]> wrote:
> On Mon, Nov 27, 2023 at 05:03:53PM +0100, Linus Walleij wrote:
> > > Liu Ying (2):
> > > driver core: Export device_is_dependent() to modules
> > > drm/bridge: panel: Check device dependency before managing device link
> >
> > I just applied patch 1 directly to the drm-misc-fixes so we don't have to
> > revert and then re-apply patches, because that is a bigger evil. (We can't
> > rebase these branches...)
>
> Erm, you did wait for GKH or Rafael's ACK to do that, right?
No.
It is a bigger evil to leave the tree broken than to enforce formal process,
and it is pretty self-evident. If any of them get annoyed about it we can
revert the patch, or both.
Yours,
Linus Walleij
On Mon, Nov 27, 2023 at 7:20 PM Greg KH <[email protected]> wrote:
[Maxime]
> > So, a committer just applied this to drm-misc-fixes without your
> > approval.
That was me, mea culpa.
I learned a lesson. Or two.
> Wait, why exactly is this needed? Nothing outside of the driver core
> should be needing this function, it shouldn't be public at all (I missed
> that before.)
>
> So please, revert it for now, let's figure out why DRM thinks this is
> needed for it's devices, and yet no other bus/subsystem does.
I'm preparing a revert series back to the original patch causing
the issue so we can clear the original bug out of the way, take
a step back and fix this properly.
Yours,
Linus Walleij
Hi Linus,
On Mon, Nov 27, 2023 at 11:13:31PM +0100, Linus Walleij wrote:
> On Mon, Nov 27, 2023 at 5:29 PM Maxime Ripard <[email protected]> wrote:
> > On Mon, Nov 27, 2023 at 05:03:53PM +0100, Linus Walleij wrote:
>
> > > > Liu Ying (2):
> > > > driver core: Export device_is_dependent() to modules
> > > > drm/bridge: panel: Check device dependency before managing device link
> > >
> > > I just applied patch 1 directly to the drm-misc-fixes so we don't have to
> > > revert and then re-apply patches, because that is a bigger evil. (We can't
> > > rebase these branches...)
> >
> > Erm, you did wait for GKH or Rafael's ACK to do that, right?
>
> No.
>
> It is a bigger evil to leave the tree broken than to enforce formal process,
> and it is pretty self-evident. If any of them get annoyed about it we can
> revert the patch, or both.
Yeah, I definitely understand why you did it, but I don't think it's
something we would encourage in drm-misc.
We've discussed it with Sima yesterday, and I think we would even need
the extra check in dim to make sure that a committer shouldn't do that
without dim complaining.
Sima played a bit with it, and it should be doable to get something
fairly reliable if you use get_maintainers.pl to retrieve the git tree
(through scripts/get_maintainer.pl --no-email --no-l --scm) and figuring
out if only drm.git, drm-intel.git or drm-misc.git is involved.
If it isn't, then we should check that there's the ack of one of the
maintainers.
Could you write a patch to do so?
Thanks!
Maxime
On Wed, Nov 29, 2023 at 1:32 PM Maxime Ripard <[email protected]> wrote:
[Me]
> > It is a bigger evil to leave the tree broken than to enforce formal process,
> > and it is pretty self-evident. If any of them get annoyed about it we can
> > revert the patch, or both.
>
> Yeah, I definitely understand why you did it, but I don't think it's
> something we would encourage in drm-misc.
Hm OK I guess, it can be debated but no point in debating it either.
> We've discussed it with Sima yesterday, and I think we would even need
> the extra check in dim to make sure that a committer shouldn't do that
> without dim complaining.
(...)
> Sima played a bit with it, and it should be doable to get something
> fairly reliable if you use get_maintainers.pl to retrieve the git tree
> (through scripts/get_maintainer.pl --no-email --no-l --scm) and figuring
> out if only drm.git, drm-intel.git or drm-misc.git is involved.
>
> If it isn't, then we should check that there's the ack of one of the
> maintainers.
So check for any code that is hitting namespaces outside drivers/gpu/*
Documentation/gpu/* or include/[uapi/]drm/* that it is ACKed by the respective
maintainer for that area of the kernel?
> Could you write a patch to do so?
Patch dim? Well my bash skills are a bit so-so. But I guess I could
learn it. But did you say there is already a prototype?
Yours,
Linus Walleij
Hi Linus,
On Wed, Nov 29, 2023 at 03:38:35PM +0100, Linus Walleij wrote:
> On Wed, Nov 29, 2023 at 1:32 PM Maxime Ripard <[email protected]> wrote:
> [Me]
> > > It is a bigger evil to leave the tree broken than to enforce formal process,
> > > and it is pretty self-evident. If any of them get annoyed about it we can
> > > revert the patch, or both.
> >
> > Yeah, I definitely understand why you did it, but I don't think it's
> > something we would encourage in drm-misc.
>
> Hm OK I guess, it can be debated but no point in debating it either.
>
> > We've discussed it with Sima yesterday, and I think we would even need
> > the extra check in dim to make sure that a committer shouldn't do that
> > without dim complaining.
> (...)
> > Sima played a bit with it, and it should be doable to get something
> > fairly reliable if you use get_maintainers.pl to retrieve the git tree
> > (through scripts/get_maintainer.pl --no-email --no-l --scm) and figuring
> > out if only drm.git, drm-intel.git or drm-misc.git is involved.
> >
> > If it isn't, then we should check that there's the ack of one of the
> > maintainers.
>
> So check for any code that is hitting namespaces outside drivers/gpu/*
> Documentation/gpu/* or include/[uapi/]drm/* that it is ACKed by the respective
> maintainer for that area of the kernel?
We can have something more reliable if we just check the git tree listed
in MAINTAINERS (and returned by get_maintainers --scm). That way we
don't have to whitelist anything, and it will always by in sync with
MAINTAINERS.
And if it's not one of drm trees, then it requires an ack from someone
else get_maintainers will also tell you about.
> > Could you write a patch to do so?
>
> Patch dim? Well my bash skills are a bit so-so. But I guess I could
> learn it. But did you say there is already a prototype?
My shell skills are also fairly limited, so we just discussed the
solution but didn't do a prototype yet :)
Maxime