| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <CAKMK7uGscg0YDgNZh61PAhnkF8xnASepo2HK2Y51wROPSkqJLA@mail.gmail.com>
Date: Fri, 25 Oct 2019 21:03:12 +0200
From: Daniel Vetter <daniel@...ll.ch>
To: Thierry Reding <thierry.reding@...il.com>
Cc: Rajat Jain <rajatxjain@...il.com>, Rajat Jain <rajatja@...gle.com>,
Maarten Lankhorst <maarten.lankhorst@...ux.intel.com>,
Maxime Ripard <mripard@...nel.org>,
Sean Paul <sean@...rly.run>, David Airlie <airlied@...ux.ie>,
Jani Nikula <jani.nikula@...ux.intel.com>,
Joonas Lahtinen <joonas.lahtinen@...ux.intel.com>,
Rodrigo Vivi <rodrigo.vivi@...el.com>,
Ville Syrjälä <ville.syrjala@...ux.intel.com>,
Chris Wilson <chris@...is-wilson.co.uk>,
Imre Deak <imre.deak@...el.com>,
José Roberto de Souza <jose.souza@...el.com>,
Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
dri-devel <dri-devel@...ts.freedesktop.org>,
intel-gfx <intel-gfx@...ts.freedesktop.org>,
Greg KH <gregkh@...uxfoundation.org>,
Mat King <mathewk@...gle.com>,
Daniel Thompson <daniel.thompson@...aro.org>,
Jonathan Corbet <corbet@....net>, Pavel Machek <pavel@...x.de>,
Sean Paul <seanpaul@...gle.com>,
Duncan Laurie <dlaurie@...gle.com>,
Jesse Barnes <jsbarnes@...gle.com>
Subject: Re: [PATCH] drm: Add support for integrated privacy screens
On Fri, Oct 25, 2019 at 1:36 PM Thierry Reding <thierry.reding@...il.com> wrote:
>
> On Thu, Oct 24, 2019 at 01:45:16PM -0700, Rajat Jain wrote:
> > Hi,
> >
> > Thanks for your review and comments. Please see inline below.
> >
> > On Thu, Oct 24, 2019 at 4:20 AM Thierry Reding <thierry.reding@...il.com> wrote:
> > >
> > > On Tue, Oct 22, 2019 at 05:12:06PM -0700, Rajat Jain wrote:
> > > > Certain laptops now come with panels that have integrated privacy
> > > > screens on them. This patch adds support for such panels by adding
> > > > a privacy-screen property to the drm_connector for the panel, that
> > > > the userspace can then use to control and check the status. The idea
> > > > was discussed here:
> > > >
> > > > https://lkml.org/lkml/2019/10/1/786
> > > >
> > > > ACPI methods are used to identify, query and control privacy screen:
> > > >
> > > > * Identifying an ACPI object corresponding to the panel: The patch
> > > > follows ACPI Spec 6.3 (available at
> > > > https://uefi.org/sites/default/files/resources/ACPI_6_3_final_Jan30.pdf).
> > > > Pages 1119 - 1123 describe what I believe, is a standard way of
> > > > identifying / addressing "display panels" in the ACPI tables, thus
> > > > allowing kernel to attach ACPI nodes to the panel. IMHO, this ability
> > > > to identify and attach ACPI nodes to drm connectors may be useful for
> > > > reasons other privacy-screens, in future.
> > > >
> > > > * Identifying the presence of privacy screen, and controlling it, is done
> > > > via ACPI _DSM methods.
> > > >
> > > > Currently, this is done only for the Intel display ports. But in future,
> > > > this can be done for any other ports if the hardware becomes available
> > > > (e.g. external monitors supporting integrated privacy screens?).
> > > >
> > > > Also, this code can be extended in future to support non-ACPI methods
> > > > (e.g. using a kernel GPIO driver to toggle a gpio that controls the
> > > > privacy-screen).
> > > >
> > > > Signed-off-by: Rajat Jain <rajatja@...gle.com>
> > > > ---
> > > > drivers/gpu/drm/Makefile | 1 +
> > > > drivers/gpu/drm/drm_atomic_uapi.c | 5 +
> > > > drivers/gpu/drm/drm_connector.c | 38 +++++
> > > > drivers/gpu/drm/drm_privacy_screen.c | 176 ++++++++++++++++++++++++
> > > > drivers/gpu/drm/i915/display/intel_dp.c | 3 +
> > > > include/drm/drm_connector.h | 18 +++
> > > > include/drm/drm_mode_config.h | 7 +
> > > > include/drm/drm_privacy_screen.h | 33 +++++
> > > > 8 files changed, 281 insertions(+)
> > > > create mode 100644 drivers/gpu/drm/drm_privacy_screen.c
> > > > create mode 100644 include/drm/drm_privacy_screen.h
> > >
> > > I like this much better than the prior proposal to use sysfs. However
> > > the support currently looks a bit tangled. I realize that we only have a
> > > single implementation for this in hardware right now, so there's no use
> > > in over-engineering things, but I think we can do a better job from the
> > > start without getting into too many abstractions. See below.
> > >
> > > > diff --git a/drivers/gpu/drm/Makefile b/drivers/gpu/drm/Makefile
> > > > index 82ff826b33cc..e1fc33d69bb7 100644
> > > > --- a/drivers/gpu/drm/Makefile
> > > > +++ b/drivers/gpu/drm/Makefile
> > > > @@ -19,6 +19,7 @@ drm-y := drm_auth.o drm_cache.o \
> > > > drm_syncobj.o drm_lease.o drm_writeback.o drm_client.o \
> > > > drm_client_modeset.o drm_atomic_uapi.o drm_hdcp.o
> > > >
> > > > +drm-$(CONFIG_ACPI) += drm_privacy_screen.o
> > > > drm-$(CONFIG_DRM_LEGACY) += drm_legacy_misc.o drm_bufs.o drm_context.o drm_dma.o drm_scatter.o drm_lock.o
> > > > drm-$(CONFIG_DRM_LIB_RANDOM) += lib/drm_random.o
> > > > drm-$(CONFIG_DRM_VM) += drm_vm.o
> > > > diff --git a/drivers/gpu/drm/drm_atomic_uapi.c b/drivers/gpu/drm/drm_atomic_uapi.c
> > > > index 7a26bfb5329c..44131165e4ea 100644
> > > > --- a/drivers/gpu/drm/drm_atomic_uapi.c
> > > > +++ b/drivers/gpu/drm/drm_atomic_uapi.c
> > > > @@ -30,6 +30,7 @@
> > > > #include <drm/drm_atomic.h>
> > > > #include <drm/drm_print.h>
> > > > #include <drm/drm_drv.h>
> > > > +#include <drm/drm_privacy_screen.h>
> > > > #include <drm/drm_writeback.h>
> > > > #include <drm/drm_vblank.h>
> > > >
> > > > @@ -766,6 +767,8 @@ static int drm_atomic_connector_set_property(struct drm_connector *connector,
> > > > fence_ptr);
> > > > } else if (property == connector->max_bpc_property) {
> > > > state->max_requested_bpc = val;
> > > > + } else if (property == config->privacy_screen_property) {
> > > > + drm_privacy_screen_set_val(connector, val);
> > >
> > > This doesn't look right. Shouldn't you store the value in the connector
> > > state and then leave it up to the connector driver to set it
> > > appropriately? I think that also has the advantage of untangling this
> > > support a little.
> >
> > Hopefully this gets answered in my explanations below.
> >
> > >
> > > > } else if (connector->funcs->atomic_set_property) {
> > > > return connector->funcs->atomic_set_property(connector,
> > > > state, property, val);
> > > > @@ -842,6 +845,8 @@ drm_atomic_connector_get_property(struct drm_connector *connector,
> > > > *val = 0;
> > > > } else if (property == connector->max_bpc_property) {
> > > > *val = state->max_requested_bpc;
> > > > + } else if (property == config->privacy_screen_property) {
> > > > + *val = drm_privacy_screen_get_val(connector);
> > >
> > > Similarly, I think this can just return the atomic state's value for
> > > this.
> >
> > I did think about having a state variable in software to get and set
> > this. However, I think it is not very far fetched that some platforms
> > may have "hardware kill" switches that allow hardware to switch
> > privacy-screen on and off directly, in addition to the software
> > control that we are implementing. Privacy is a touchy subject in
> > enterprise, and anything that reduces the possibility of having any
> > inconsistency between software state and hardware state is desirable.
> > So in this case, I chose to not have a state in software about this -
> > we just report the hardware state everytime we are asked for it.
>
> So this doesn't really work with atomic KMS, then. The main idea behind
> atomic KMS is that you apply a configuration either completely or not at
> all. So at least for setting this property you'd have to go through the
> state object.
>
> Now, for reading out the property you might be able to get away with the
> above. I'm not sure if that's enough to keep the state up-to-date,
> though. Is there some way for a kill switch to trigger an interrupt or
> other event of some sort so that the state could be kept up-to-date?
>
> Daniel (or anyone else), do you know of any precedent for state that
> might get modified behind the atomic helpers' back? Seems to me like we
> need to find some point where we can actually read back the current
> "hardware value" of this privacy screen property and store that back
> into the state.
We have atomic properties that the driver also updates, not just userspace:
- link status
- hdcp machinery
But both still use the state structures like anything else, and then
update that state structure through some worker.
And yes I totally didn't catch this here, this breaks atomic, badly :-)
>
> >
> > >
> > > > } else if (connector->funcs->atomic_get_property) {
> > > > return connector->funcs->atomic_get_property(connector,
> > > > state, property, val);
> > > > diff --git a/drivers/gpu/drm/drm_connector.c b/drivers/gpu/drm/drm_connector.c
> > > > index 4c766624b20d..a31e0382132b 100644
> > > > --- a/drivers/gpu/drm/drm_connector.c
> > > > +++ b/drivers/gpu/drm/drm_connector.c
> > > > @@ -821,6 +821,11 @@ static const struct drm_prop_enum_list drm_panel_orientation_enum_list[] = {
> > > > { DRM_MODE_PANEL_ORIENTATION_RIGHT_UP, "Right Side Up" },
> > > > };
> > > >
> > > > +static const struct drm_prop_enum_list drm_privacy_screen_enum_list[] = {
> > > > + { DRM_PRIVACY_SCREEN_DISABLED, "Disabled" },
> > > > + { DRM_PRIVACY_SCREEN_ENABLED, "Enabled" },
> > > > +};
> > > > +
> > > > static const struct drm_prop_enum_list drm_dvi_i_select_enum_list[] = {
> > > > { DRM_MODE_SUBCONNECTOR_Automatic, "Automatic" }, /* DVI-I and TV-out */
> > > > { DRM_MODE_SUBCONNECTOR_DVID, "DVI-D" }, /* DVI-I */
> > > > @@ -2253,6 +2258,39 @@ static void drm_tile_group_free(struct kref *kref)
> > > > kfree(tg);
> > > > }
> > > >
> > > > +/**
> > > > + * drm_connector_init_privacy_screen_property -
> > > > + * create and attach the connecter's privacy-screen property.
> > > > + * @connector: connector for which to init the privacy-screen property.
> > > > + *
> > > > + * This function creates and attaches the "privacy-screen" property to the
> > > > + * connector. Initial state of privacy-screen is set to disabled.
> > > > + *
> > > > + * Returns:
> > > > + * Zero on success, negative errno on failure.
> > > > + */
> > > > +int drm_connector_init_privacy_screen_property(struct drm_connector *connector)
> > > > +{
> > > > + struct drm_device *dev = connector->dev;
> > > > + struct drm_property *prop;
> > > > +
> > > > + prop = dev->mode_config.privacy_screen_property;
> > > > + if (!prop) {
> > > > + prop = drm_property_create_enum(dev, DRM_MODE_PROP_ENUM,
> > > > + "privacy-screen", drm_privacy_screen_enum_list,
> > >
> > > Seems to me like the -screen suffix here is somewhat redundant. Yes, the
> > > thing that we enable/disable may be called a "privacy screen", but the
> > > property that we enable/disable on the connector is the "privacy" of the
> > > user. I'd reflect that in all the related variable names and so on as
> > > well.
> >
> > IMHO a property called "privacy" may be a little generic for the users
> > to understand what it is. For e.g. when I started looking at code, I
> > found the "Content Protection" property and I got confused thinking
> > may be it provides something similar to what I'm trying to do. I think
> > "privacy-screen" conveys the complete context without being long, so
> > there is no confusion or ambiguity. But I don't mind changing it if a
> > property "privacy" is what people think is better to convey what it
> > is, as long as it is clear to user.
>
> This being a property of a display connector it doesn't seem very
> ambiguous to me what this is. How this ends up being presented to the
> users is mostly orthogonal anyway. We've got a bunch of properties whose
> purpose may not be clear to the average user. The properties, while they
> are UABI, don't typically face the user directly. They're still part of
> an API, so as long as they are properly documented there shouldn't be
> any ambiguities.
Sometimes we go the cheap way with connector properties and say that
those should be set by the user directly, through something like
xrandr (mostly for stuff to hack around broken monitors). But ofc
nicer if the property comes with some proper userspace integration.
> > > > + ARRAY_SIZE(drm_privacy_screen_enum_list));
> > > > + if (!prop)
> > > > + return -ENOMEM;
> > > > +
> > > > + dev->mode_config.privacy_screen_property = prop;
> > > > + }
> > > > +
> > > > + drm_object_attach_property(&connector->base, prop,
> > > > + DRM_PRIVACY_SCREEN_DISABLED);
> > > > + return 0;
> > > > +}
> > > > +EXPORT_SYMBOL(drm_connector_init_privacy_screen_property);
> > > > +
> > > > /**
> > > > * drm_mode_put_tile_group - drop a reference to a tile group.
> > > > * @dev: DRM device
> > > > diff --git a/drivers/gpu/drm/drm_privacy_screen.c b/drivers/gpu/drm/drm_privacy_screen.c
> > > > new file mode 100644
> > > > index 000000000000..1d68e8aa6c5f
> > > > --- /dev/null
> > > > +++ b/drivers/gpu/drm/drm_privacy_screen.c
> > > > @@ -0,0 +1,176 @@
> > > > +// SPDX-License-Identifier: GPL-2.0-or-later
> > > > +/*
> > > > + * DRM privacy Screen code
> > > > + *
> > > > + * Copyright © 2019 Google Inc.
> > > > + */
> > > > +
> > > > +#include <linux/acpi.h>
> > > > +#include <linux/pci.h>
> > > > +
> > > > +#include <drm/drm_connector.h>
> > > > +#include <drm/drm_device.h>
> > > > +#include <drm/drm_print.h>
> > > > +
> > > > +#define DRM_CONN_DSM_REVID 1
> > > > +
> > > > +#define DRM_CONN_DSM_FN_PRIVACY_GET_STATUS 1
> > > > +#define DRM_CONN_DSM_FN_PRIVACY_ENABLE 2
> > > > +#define DRM_CONN_DSM_FN_PRIVACY_DISABLE 3
> > > > +
> > > > +static const guid_t drm_conn_dsm_guid =
> > > > + GUID_INIT(0xC7033113, 0x8720, 0x4CEB,
> > > > + 0x90, 0x90, 0x9D, 0x52, 0xB3, 0xE5, 0x2D, 0x73);
> > > > +
> > > > +/*
> > > > + * Makes _DSM call to set privacy screen status or get privacy screen. Return
> > > > + * value matters only for PRIVACY_GET_STATUS case. Returns 0 if disabled, 1 if
> > > > + * enabled.
> > > > + */
> > > > +static int acpi_privacy_screen_call_dsm(acpi_handle conn_handle, u64 func)
> > > > +{
> > > > + union acpi_object *obj;
> > > > + int ret = 0;
> > > > +
> > > > + obj = acpi_evaluate_dsm(conn_handle, &drm_conn_dsm_guid,
> > > > + DRM_CONN_DSM_REVID, func, NULL);
> > > > + if (!obj) {
> > > > + DRM_DEBUG_DRIVER("failed to evaluate _DSM for fn %llx\n", func);
> > > > + /* Can't do much. For get_val, assume privacy_screen disabled */
> > > > + goto done;
> > > > + }
> > > > +
> > > > + if (func == DRM_CONN_DSM_FN_PRIVACY_GET_STATUS &&
> > > > + obj->type == ACPI_TYPE_INTEGER)
> > > > + ret = !!obj->integer.value;
> > > > +done:
> > > > + ACPI_FREE(obj);
> > > > + return ret;
> > > > +}
> > > > +
> > > > +void drm_privacy_screen_set_val(struct drm_connector *connector,
> > > > + enum drm_privacy_screen val)
> > > > +{
> > > > + acpi_handle acpi_handle = connector->privacy_screen_handle;
> > > > +
> > > > + if (!acpi_handle)
> > > > + return;
> > > > +
> > > > + if (val == DRM_PRIVACY_SCREEN_DISABLED)
> > > > + acpi_privacy_screen_call_dsm(acpi_handle,
> > > > + DRM_CONN_DSM_FN_PRIVACY_DISABLE);
> > > > + else if (val == DRM_PRIVACY_SCREEN_ENABLED)
> > > > + acpi_privacy_screen_call_dsm(acpi_handle,
> > > > + DRM_CONN_DSM_FN_PRIVACY_ENABLE);
> > > > +}
> > > > +
> > > > +enum drm_privacy_screen drm_privacy_screen_get_val(struct drm_connector
> > > > + *connector)
> > > > +{
> > > > + acpi_handle acpi_handle = connector->privacy_screen_handle;
> > > > +
> > > > + if (acpi_handle &&
> > > > + acpi_privacy_screen_call_dsm(acpi_handle,
> > > > + DRM_CONN_DSM_FN_PRIVACY_GET_STATUS))
> > > > + return DRM_PRIVACY_SCREEN_ENABLED;
> > > > +
> > > > + return DRM_PRIVACY_SCREEN_DISABLED;
> > > > +}
> > > > +
> > > > +/*
> > > > + * See ACPI Spec v6.3, Table B-2, "Display Type" for details.
> > > > + * In short, these macros define the base _ADR values for ACPI nodes
> > > > + */
> > > > +#define ACPI_BASE_ADR_FOR_OTHERS (0ULL << 8)
> > > > +#define ACPI_BASE_ADR_FOR_VGA (1ULL << 8)
> > > > +#define ACPI_BASE_ADR_FOR_TV (2ULL << 8)
> > > > +#define ACPI_BASE_ADR_FOR_EXT_MON (3ULL << 8)
> > > > +#define ACPI_BASE_ADR_FOR_INTEGRATED (4ULL << 8)
> > > > +
> > > > +#define ACPI_DEVICE_ID_SCHEME (1ULL << 31)
> > > > +#define ACPI_FIRMWARE_CAN_DETECT (1ULL << 16)
> > > > +
> > > > +/*
> > > > + * Ref: ACPI Spec 6.3
> > > > + * https://uefi.org/sites/default/files/resources/ACPI_6_3_final_Jan30.pdf
> > > > + * Pages 1119 - 1123 describe, what I believe, a standard way of
> > > > + * identifying / addressing "display panels" in the ACPI. Thus it provides
> > > > + * a way for the ACPI to define devices for the display panels attached
> > > > + * to the system. It thus provides a way for the BIOS to export any panel
> > > > + * specific properties to the system via ACPI (like device trees).
> > > > + *
> > > > + * The following function looks up the ACPI node for a connector and links
> > > > + * to it. Technically it is independent from the privacy_screen code, and
> > > > + * ideally may be called for all connectors. It is generally a good idea to
> > > > + * be able to attach an ACPI node to describe anything if needed. (This can
> > > > + * help in future for other panel specific features maybe). However, it
> > > > + * needs a "port index" which I believe is the index within a particular
> > > > + * type of port (Ref to the pages of spec mentioned above). This port index
> > > > + * unfortunately is not available in DRM code, so currently its call is
> > > > + * originated from i915 driver.
> > > > + */
> > > > +static int drm_connector_attach_acpi_node(struct drm_connector *connector,
> > > > + u8 port_index)
> > > > +{
> > > > + struct device *dev = &connector->dev->pdev->dev;
> > > > + struct acpi_device *conn_dev;
> > > > + u64 conn_addr;
> > > > +
> > > > + /*
> > > > + * Determine what _ADR to look for, depending on device type and
> > > > + * port number. Potentially we only care about the
> > > > + * eDP / integrated displays?
> > > > + */
> > > > + switch (connector->connector_type) {
> > > > + case DRM_MODE_CONNECTOR_eDP:
> > > > + conn_addr = ACPI_BASE_ADR_FOR_INTEGRATED + port_index;
> > > > + break;
> > > > + case DRM_MODE_CONNECTOR_VGA:
> > > > + conn_addr = ACPI_BASE_ADR_FOR_VGA + port_index;
> > > > + break;
> > > > + case DRM_MODE_CONNECTOR_TV:
> > > > + conn_addr = ACPI_BASE_ADR_FOR_TV + port_index;
> > > > + break;
> > > > + case DRM_MODE_CONNECTOR_DisplayPort:
> > > > + conn_addr = ACPI_BASE_ADR_FOR_EXT_MON + port_index;
> > > > + break;
> > > > + default:
> > > > + return -ENOTSUPP;
> > > > + }
> > > > +
> > > > + conn_addr |= ACPI_DEVICE_ID_SCHEME;
> > > > + conn_addr |= ACPI_FIRMWARE_CAN_DETECT;
> > > > +
> > > > + DRM_DEV_DEBUG(dev, "%s: Finding drm_connector ACPI node at _ADR=%llX\n",
> > > > + __func__, conn_addr);
> > > > +
> > > > + /* Look up the connector device, under the PCI device */
> > > > + conn_dev = acpi_find_child_device(ACPI_COMPANION(dev),
> > > > + conn_addr, false);
> > > > + if (!conn_dev)
> > > > + return -ENODEV;
> > > > +
> > > > + connector->privacy_screen_handle = conn_dev->handle;
> > > > + return 0;
> > > > +}
> > > > +
> > > > +bool drm_privacy_screen_present(struct drm_connector *connector, u8 port_index)
> > >
> > > This is the main part that I think is a little tangled. This is a very
> > > specific implementation that hides in a generic API.
> >
> > I agree that this is an ACPI specific implementation, but other than
> > that, I think it does not have any driver specific details. More
> > detailed thoughts on this below.
>
> Well, the port_index kind of ties this to i915 because that uses this
> concept. Other drivers may not.
>
> Also, I'm wondering if you couldn't somehow derive the port_index from
> the connector. If all this does is to find an ACPI node corresponding to
> a connector, shouldn't the connector really be all that you need?
Yeah port should be stored already plenty of times in i915 connector structs.
> > > I we store the privacy setting in the atomic state, there isn't really a
> > > reason to store the privacy handle in the connector. Instead it could be
> > > simply stored in the driver that supports this.
> > >
> > > Ideally I think we'd have a very small drm_privacy_screen object type
> > > that would just wrap this, but perhaps we don't need that right away,
> > > given that we only have a single implementation so far.
> >
> > Yes, agreed.
> >
> > >
> > > However, I think if we just pushed this specific implementation into the
> > > drivers that'd help pave the way for something more generic later on
> > > without a lot of extra work up front.
> > >
> > > For example you could turn the drm_connector_attach_acpi_node() into a
> > > helper that simply returns the ACPI handle, something like this perhaps:
> > >
> > > struct acpi_handle *drm_acpi_find_privacy_screen(struct drm_connector *connector,
> > > unsigned int port)
> > > {
> > > ...
> > > }
> >
> > Yes, I like that idea of making it a helper function. In fact, finding
> > an ACPI node for the connector doesn't have to do anything with
> > privacy screen (so it can be used for other purposes also, in future).
>
> Looks like I misunderstood the purpose of that function. You store the
> ACPI handle as connector->privacy_screen_handle, so I was assuming that
> it was getting an ACPI handle for some privacy screen subdevice.
>
> > > That the i915 driver would then call and store the returned value
> > > internally. When it commits the atomic state for the connector it can
> > > then call the drm_acpi_set_privacy() (I think that'd be a better name
> > > for your drm_privacy_screen_set_val()) by passing that handle and the
> > > value from the atomic state.
> > >
> > > The above has the advantage that we don't clutter the generic core with
> > > something that's not at all generic. If eventually we see that these
> > > types of privacy screens are implemented in more device we can always
> > > refactor this into something really generic and maybe even decide to put
> > > it into the drm_connector directly.
> >
> > This is where I think I'm in slight disagreement. I think the
> > functionality we're adding is still "generic", just that the only
> > hardware *I have today* to test is using Intel eDP ports. But I don't
> > see why AMD CPU laptops can't have it (For E.g. HP's Elitebook 745 G5
> > seems to use AMD and has integrated privacy screen feature
> > http://www8.hp.com/h20195/V2/GetPDF.aspx/4aa7-2802eee) .
> > My worry is that if we don't make it generic today, we might see
> > duplicate / similar-but-different / different ways of this in other
> > places (e.g. https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=110ea1d833ad)
> > because unless it is generic to start with, it is difficult for some
> > one to change later for the fear of breaking hardware that is already
> > in market given that
> > * hardware may not be available to new developer to test for
> > regressions (also there is very little motivation to check any
> > hardware other than your own).
> > * specially for a code that relies on firmware ACPI (firmware
> > upgrades in field are always costly).
> >
> > My understanding is that we're adding 2 functionalities here:
> >
> > 1) Identify and Attach ACPI node to DRM connector. Since this is
> > following ACPI spec, I think this is generic enough.
>
> It's probably worth making this a separate patch in that case. If the
> ACPI handle really represents the connector itself, then it seems fine
> to store it in the connector. But it shouldn't be called privacy_screen
> in that case.
Probably still should be in i915. We have a few firmware handle things
in there already. If other drivers need this, we could make an acpi
helper library that drivers can call to find the right acpi handle for
their connector.
> > 2) Use ACPI _DSM mthods to identify screen, set and get values. This
> > is where I think we're setting (generic) expectations for the ACPI
> > methods in how they should behave if ACPI is to be used to control
> > privacy screen. If we put this in generic code today, future
> > developers can look at this to understand how their ACPI for new
> > platforms is to behave if they want to use this generic code. If we
> > put it in i915 specific code, this will be seen as driver specific
> > behavior and developers may choose some other behavior in their
> > driver.
>
> I think it's fine to put this functionality into generic code. What I
> don't think is good to do is to have this code called from generic code.
> It's opt-in functionality that drivers should call if they know that the
> connector has an associated ACPI handle that can be used for privacy
> screen control.
>
> After reading the patch again and realizing that you're not actually
> dealing with an ACPI handle to the privacy screen directly but one for
> the connector, I think this is okay. You do in fact call into this from
> the i915 only. I still don't think the naming is great, and it'd be nice
> to see ACPI as part of the function name to make that explicit. We could
> always address that at a later point, but may as well do it right from
> the start.
>
> > I agree that the functionality we're adding is ACPI specific (today -
> > but can be extended to gpio in future for non x86 platforms), but not
> > necessarily driver specific. Actually the only reason, I had to call
> > the drm_privacy_screen_present() (and the
> > drm_init_privacy_screen_property()) function from i915 code is because
> > we need a port_index to lookup ACPI node. If we had that available in
> > drm code, we wouldn't need to call anything from i915 at all.
>
> You're kind of proving my point about this API being driver-specific, or
> at least ACPI specific. Non-ACPI devices (maybe even non-i915 devices?)
> may not have a concept of a port index. Encoding this in the API makes
> the API non-generic.
>
> As I mentioned above, if we could derive the port index from the
> connector, that'd be much better. Could you perhaps use drm_connector's
> index field?
>
> Unless there's a way to reliably detect this type of functionality from
> generic code, I think it should always be called from the driver.
Maybe time to pitch the old classic again:
https://blog.ffwll.ch/2016/12/midlayers-once-more-with-feeling.html
Especially all the links in the first part.
> > So, for the reasons stated above, IMHO it is better to retain this
> > functionality in drm code instead of i915 driver. But I'm new to the
> > drm / i915 code, and would be happy to change my code if people have
> > strong opinions about this. Let me know.
>
> Maybe I was being unclear. I wasn't arguing that all the code should be
> moved into the i915 driver. All I was saying that instead of storing the
> ACPI handle inside struct drm_connector, we should maybe store it inside
> the i915 driver's connector structure. struct drm_connector is a very
> generic concept and each and every connector object on every platform
> will get that ACPI handle pointer if you add it there. I don't think an
> ACPI handle belongs there. For example, on ARM SoCs it's common to have
> connectors be backed by a struct device (or struct platform_device more
> specifically). But we don't put that information into drm_connector
> because PCI graphics adapters don't have a struct device that represents
> the connector.
+1 on all the stuff Thierry says here essentially.
Cheers, Daniel
>
> Thierry
>
> >
> > Thanks & Best Regards,
> >
> > Rajat
> >
> > >
> > > > +{
> > > > + acpi_handle handle;
> > > > +
> > > > + if (drm_connector_attach_acpi_node(connector, port_index))
> > > > + return false;
> > > > +
> > > > + handle = connector->privacy_screen_handle;
> > > > + if (!acpi_check_dsm(handle, &drm_conn_dsm_guid,
> > > > + DRM_CONN_DSM_REVID,
> > > > + 1 << DRM_CONN_DSM_FN_PRIVACY_GET_STATUS |
> > > > + 1 << DRM_CONN_DSM_FN_PRIVACY_ENABLE |
> > > > + 1 << DRM_CONN_DSM_FN_PRIVACY_DISABLE)) {
> > > > + DRM_WARN("%s: Odd, connector ACPI node but no privacy scrn?\n",
> > > > + connector->dev->dev);
> > > > + return false;
> > > > + }
> > > > + DRM_DEV_INFO(connector->dev->dev, "supports privacy screen\n");
> > > > + return true;
> > > > +}
> > > > diff --git a/drivers/gpu/drm/i915/display/intel_dp.c b/drivers/gpu/drm/i915/display/intel_dp.c
> > > > index 57e9f0ba331b..3ff3962d27db 100644
> > > > --- a/drivers/gpu/drm/i915/display/intel_dp.c
> > > > +++ b/drivers/gpu/drm/i915/display/intel_dp.c
> > > > @@ -39,6 +39,7 @@
> > > > #include <drm/drm_dp_helper.h>
> > > > #include <drm/drm_edid.h>
> > > > #include <drm/drm_hdcp.h>
> > > > +#include <drm/drm_privacy_screen.h>
> > > > #include <drm/drm_probe_helper.h>
> > > > #include <drm/i915_drm.h>
> > > >
> > > > @@ -6354,6 +6355,8 @@ intel_dp_add_properties(struct intel_dp *intel_dp, struct drm_connector *connect
> > > >
> > > > connector->state->scaling_mode = DRM_MODE_SCALE_ASPECT;
> > > >
> > > > + if (drm_privacy_screen_present(connector, port - PORT_A))
> > > > + drm_connector_init_privacy_screen_property(connector);
> > > > }
> > > > }
> > > >
> > > > diff --git a/include/drm/drm_connector.h b/include/drm/drm_connector.h
> > > > index 681cb590f952..63b8318bd68c 100644
> > > > --- a/include/drm/drm_connector.h
> > > > +++ b/include/drm/drm_connector.h
> > > > @@ -225,6 +225,20 @@ enum drm_link_status {
> > > > DRM_LINK_STATUS_BAD = DRM_MODE_LINK_STATUS_BAD,
> > > > };
> > > >
> > > > +/**
> > > > + * enum drm_privacy_screen - privacy_screen status
> > > > + *
> > > > + * This enum is used to track and control the state of the privacy screen.
> > > > + * There are no separate #defines for the uapi!
> > > > + *
> > > > + * @DRM_PRIVACY_SCREEN_DISABLED: The privacy-screen on the panel is disabled
> > > > + * @DRM_PRIVACY_SCREEN_ENABLED: The privacy-screen on the panel is enabled
> > > > + */
> > > > +enum drm_privacy_screen {
> > > > + DRM_PRIVACY_SCREEN_DISABLED = 0,
> > > > + DRM_PRIVACY_SCREEN_ENABLED = 1,
> > > > +};
> > > > +
> > >
> > > Shouldn't this go into include/uapi/drm/drm_mode.h? That would have the
> > > advantage of giving userspace symbolic names to use when setting the
> > > property.
> > >
> > > Maybe also rename these to something like:
> > >
> > > #define DRM_MODE_PRIVACY_DISABLED 0
> > > #define DRM_MODE_PRIVACY_ENABLED 1
> > >
> > > for consistency with other properties.
> > >
> > > Thierry
> > >
> > > > /**
> > > > * enum drm_panel_orientation - panel_orientation info for &drm_display_info
> > > > *
> > > > @@ -1410,6 +1424,9 @@ struct drm_connector {
> > > >
> > > > /** @hdr_sink_metadata: HDR Metadata Information read from sink */
> > > > struct hdr_sink_metadata hdr_sink_metadata;
> > > > +
> > > > + /* Handle used by privacy screen code */
> > > > + void *privacy_screen_handle;
> > > > };
> > > >
> > > > #define obj_to_connector(x) container_of(x, struct drm_connector, base)
> > > > @@ -1543,6 +1560,7 @@ int drm_connector_init_panel_orientation_property(
> > > > struct drm_connector *connector, int width, int height);
> > > > int drm_connector_attach_max_bpc_property(struct drm_connector *connector,
> > > > int min, int max);
> > > > +int drm_connector_init_privacy_screen_property(struct drm_connector *connector);
> > > >
> > > > /**
> > > > * struct drm_tile_group - Tile group metadata
> > > > diff --git a/include/drm/drm_mode_config.h b/include/drm/drm_mode_config.h
> > > > index 3bcbe30339f0..6d5d23da90d4 100644
> > > > --- a/include/drm/drm_mode_config.h
> > > > +++ b/include/drm/drm_mode_config.h
> > > > @@ -813,6 +813,13 @@ struct drm_mode_config {
> > > > */
> > > > struct drm_property *panel_orientation_property;
> > > >
> > > > + /**
> > > > + * @privacy_screen_property: Optional connector property to indicate
> > > > + * and control the state (enabled / disabled) of privacy-screen on the
> > > > + * panel, if present.
> > > > + */
> > > > + struct drm_property *privacy_screen_property;
> > > > +
> > > > /**
> > > > * @writeback_fb_id_property: Property for writeback connectors, storing
> > > > * the ID of the output framebuffer.
> > > > diff --git a/include/drm/drm_privacy_screen.h b/include/drm/drm_privacy_screen.h
> > > > new file mode 100644
> > > > index 000000000000..c589bbc47656
> > > > --- /dev/null
> > > > +++ b/include/drm/drm_privacy_screen.h
> > > > @@ -0,0 +1,33 @@
> > > > +/* SPDX-License-Identifier: GPL-2.0-or-later */
> > > > +/*
> > > > + * Copyright © 2019 Google Inc.
> > > > + */
> > > > +
> > > > +#ifndef __DRM_PRIVACY_SCREEN_H__
> > > > +#define __DRM_PRIVACY_SCREEN_H__
> > > > +
> > > > +#ifdef CONFIG_ACPI
> > > > +bool drm_privacy_screen_present(struct drm_connector *connector, u8 port);
> > > > +void drm_privacy_screen_set_val(struct drm_connector *connector,
> > > > + enum drm_privacy_screen val);
> > > > +enum drm_privacy_screen drm_privacy_screen_get_val(struct drm_connector
> > > > + *connector);
> > > > +#else
> > > > +static inline bool drm_privacy_screen_present(struct drm_connector *connector,
> > > > + u8 port)
> > > > +{
> > > > + return false;
> > > > +}
> > > > +
> > > > +void drm_privacy_screen_set_val(struct drm_connector *connector,
> > > > + enum drm_privacy_screen val)
> > > > +{ }
> > > > +
> > > > +enum drm_privacy_screen drm_privacy_screen_get_val(
> > > > + struct drm_connector *connector)
> > > > +{
> > > > + return DRM_PRIVACY_SCREEN_DISABLED;
> > > > +}
> > > > +#endif /* CONFIG_ACPI */
> > > > +
> > > > +#endif /* __DRM_PRIVACY_SCREEN_H__ */
> > > > --
> > > > 2.23.0.866.gb869b98d4c-goog
> > > >
--
Daniel Vetter
Software Engineer, Intel Corporation
+41 (0) 79 365 57 48 - http://blog.ffwll.ch
Powered by blists - more mailing lists