[PATCH] drm/appletbdrm: Add suspend and resume support

[PATCH] drm/appletbdrm: Add suspend and resume support

[Vasilij Markin via B4 Relay]

From: Vasilij Markin <vasilij.kernel@tototo.si>

The driver has no USB power management callbacks, so the USB core
unbinds it before system suspend and binds it again on resume. The DRM
device is unregistered and registered again on every suspend and resume
cycle, userspace driving the Touch Bar through it (such as tiny-dfr)
loses its device, and the Touch Bar stays dark after resume.

Implement suspend and resume with drm_mode_config_helper_suspend() and
drm_mode_config_helper_resume(), so the device stays registered and the
last committed state is restored on resume. If the device was reset
while suspended, repeat the probe-time handshake first.

Fixes: 0670c2f56e45 ("drm/tiny: add driver for Apple Touch Bars in x86 Macs")
Assisted-by: Claude:claude-opus-5-5
Signed-off-by: Vasilij Markin <vasilij.kernel@tototo.si>
---
Tested on a MacBookPro16,2 with this change backported to 7.2.7, deep (S3)
suspend by closing the lid, tiny-dfr 0.3.7. Without it the driver is probed
again on every resume and systemd stops tiny-dfr; with it tiny-dfr keeps
running, the driver is not probed again and the Touch Bar shows its content
right after resume. The reset_resume path is untested: the device never came
back reset.

Kernels before commit 3d95f93b30b8 ("drm/appletbdrm: Convert to atomic_create_state")
also need the plane reset hook to free the existing state, otherwise
drm_mode_config_reset() on resume hits WARN_ON(plane->state) in
appletbdrm_primary_plane_reset() and leaks the old state. 7.2.y is such a
kernel; the tested backport carries that change as a separate patch.

Built on drm-misc-next (c4c54ffd6) with W=1 without warnings.
---
 drivers/gpu/drm/tiny/appletbdrm.c | 43 +++++++++++++++++++++++++++++++++++++++
 1 file changed, 43 insertions(+)

diff --git a/drivers/gpu/drm/tiny/appletbdrm.c b/drivers/gpu/drm/tiny/appletbdrm.c
index 4297d9c11..024ad8bf6 100644
--- a/drivers/gpu/drm/tiny/appletbdrm.c
+++ b/drivers/gpu/drm/tiny/appletbdrm.c
@@ -30,6 +30,7 @@
 #include <drm/drm_gem_atomic_helper.h>
 #include <drm/drm_gem_framebuffer_helper.h>
 #include <drm/drm_gem_shmem_helper.h>
+#include <drm/drm_modeset_helper.h>
 #include <drm/drm_plane.h>
 #include <drm/drm_print.h>
 #include <drm/drm_probe_helper.h>
@@ -821,6 +822,45 @@ static void appletbdrm_shutdown(struct usb_interface *intf)
 	drm_atomic_helper_shutdown(&adev->drm);
 }
 
+static int appletbdrm_suspend(struct usb_interface *intf, pm_message_t message)
+{
+	struct appletbdrm_device *adev = usb_get_intfdata(intf);
+
+	return drm_mode_config_helper_suspend(&adev->drm);
+}
+
+static int appletbdrm_resume(struct usb_interface *intf)
+{
+	struct appletbdrm_device *adev = usb_get_intfdata(intf);
+
+	return drm_mode_config_helper_resume(&adev->drm);
+}
+
+static int appletbdrm_reset_resume(struct usb_interface *intf)
+{
+	struct appletbdrm_device *adev = usb_get_intfdata(intf);
+	struct drm_device *drm = &adev->drm;
+	int ret;
+
+	/*
+	 * A reset puts the device back into its initial state, where it
+	 * expects the same handshake as during probe
+	 */
+	ret = appletbdrm_get_information(adev);
+	if (ret) {
+		drm_err(drm, "Failed to get display information\n");
+		return ret;
+	}
+
+	ret = appletbdrm_signal_readiness(adev);
+	if (ret) {
+		drm_err(drm, "Failed to signal readiness\n");
+		return ret;
+	}
+
+	return drm_mode_config_helper_resume(drm);
+}
+
 static const struct usb_device_id appletbdrm_usb_id_table[] = {
 	{ USB_DEVICE_INTERFACE_CLASS(0x05ac, 0x8302, USB_CLASS_AUDIO_VIDEO) },
 	{}
@@ -832,6 +872,9 @@ static struct usb_driver appletbdrm_usb_driver = {
 	.probe		= appletbdrm_probe,
 	.disconnect	= appletbdrm_disconnect,
 	.shutdown	= appletbdrm_shutdown,
+	.suspend	= appletbdrm_suspend,
+	.resume		= appletbdrm_resume,
+	.reset_resume	= appletbdrm_reset_resume,
 	.id_table	= appletbdrm_usb_id_table,
 };
 module_usb_driver(appletbdrm_usb_driver);

---
base-commit: c4c54ffd65555ac5316f8faf4d7dc8c815877937
change-id: 20260924-appletbdrm-pm-6bd220385beb

Best regards,
-- 
Vasilij Markin <vasilij.kernel@tototo.si>

Re: [PATCH] drm/appletbdrm: Add suspend and resume support

[Aditya Garg]

Hi Vasilij

Thanks for the patch

> From: Vasilij Markin <vasilij.kernel@tototo.si>
> 
> The driver has no USB power management callbacks, so the USB core
> unbinds it before system suspend and binds it again on resume. The DRM
> device is unregistered and registered again on every suspend and resume
> cycle, userspace driving the Touch Bar through it (such as tiny-dfr)
> loses its device, and the Touch Bar stays dark after resume.
> 
> Implement suspend and resume with drm_mode_config_helper_suspend() and
> drm_mode_config_helper_resume(), so the device stays registered and the
> last committed state is restored on resume. If the device was reset
> while suspended, repeat the probe-time handshake first.
> 
> Fixes: 0670c2f56e45 ("drm/tiny: add driver for Apple Touch Bars in x86 Macs")
> Assisted-by: Claude:claude-opus-5-5
> Signed-off-by: Vasilij Markin <vasilij.kernel@tototo.si>
> ---
> Tested on a MacBookPro16,2 with this change backported to 7.2.7, deep (S3)
> suspend by closing the lid, tiny-dfr 0.3.7. Without it the driver is probed
> again on every resume and systemd stops tiny-dfr; with it tiny-dfr keeps
> running, the driver is not probed again and the Touch Bar shows its content
> right after resume. The reset_resume path is untested: the device never came
> back reset.
> 
> Kernels before commit 3d95f93b30b8 ("drm/appletbdrm: Convert to atomic_create_state")
> also need the plane reset hook to free the existing state, otherwise
> drm_mode_config_reset() on resume hits WARN_ON(plane->state) in
> appletbdrm_primary_plane_reset() and leaks the old state. 7.2.y is such a
> kernel; the tested backport carries that change as a separate patch.
> 
> Built on drm-misc-next (c4c54ffd6) with W=1 without warnings.
> ---
>   drivers/gpu/drm/tiny/appletbdrm.c | 43 +++++++++++++++++++++++++++++++++++++++
>   1 file changed, 43 insertions(+)
> 
> diff --git a/drivers/gpu/drm/tiny/appletbdrm.c b/drivers/gpu/drm/tiny/appletbdrm.c
> index 4297d9c11..024ad8bf6 100644
> --- a/drivers/gpu/drm/tiny/appletbdrm.c
> +++ b/drivers/gpu/drm/tiny/appletbdrm.c
> @@ -30,6 +30,7 @@
>   #include <drm/drm_gem_atomic_helper.h>
>   #include <drm/drm_gem_framebuffer_helper.h>
>   #include <drm/drm_gem_shmem_helper.h>
> +#include <drm/drm_modeset_helper.h>
>   #include <drm/drm_plane.h>
>   #include <drm/drm_print.h>
>   #include <drm/drm_probe_helper.h>
> @@ -821,6 +822,45 @@ static void appletbdrm_shutdown(struct usb_interface *intf)
>   	drm_atomic_helper_shutdown(&adev->drm);
>   }
>   
> +static int appletbdrm_suspend(struct usb_interface *intf, pm_message_t message)
> +{
> +	struct appletbdrm_device *adev = usb_get_intfdata(intf);
> +
> +	return drm_mode_config_helper_suspend(&adev->drm);
> +}
> +
> +static int appletbdrm_resume(struct usb_interface *intf)
> +{
> +	struct appletbdrm_device *adev = usb_get_intfdata(intf);
> +
> +	return drm_mode_config_helper_resume(&adev->drm);
> +}
> +
> +static int appletbdrm_reset_resume(struct usb_interface *intf)

We don't need a reset_resume function as the USB device is virtually 
exposed by the t2bce driver and current implementation doesn't reset the 
device in any case. If the need arises, it can be added in a later patch.

> +{
> +	struct appletbdrm_device *adev = usb_get_intfdata(intf);
> +	struct drm_device *drm = &adev->drm;
> +	int ret;
> +
> +	/*
> +	 * A reset puts the device back into its initial state, where it
> +	 * expects the same handshake as during probe
> +	 */
> +	ret = appletbdrm_get_information(adev);
> +	if (ret) {
> +		drm_err(drm, "Failed to get display information\n");
> +		return ret;
> +	}
> +
> +	ret = appletbdrm_signal_readiness(adev);
> +	if (ret) {
> +		drm_err(drm, "Failed to signal readiness\n");
> +		return ret;
> +	}
> +
> +	return drm_mode_config_helper_resume(drm);
> +}
> +
>   static const struct usb_device_id appletbdrm_usb_id_table[] = {
>   	{ USB_DEVICE_INTERFACE_CLASS(0x05ac, 0x8302, USB_CLASS_AUDIO_VIDEO) },
>   	{}
> @@ -832,6 +872,9 @@ static struct usb_driver appletbdrm_usb_driver = {
>   	.probe		= appletbdrm_probe,
>   	.disconnect	= appletbdrm_disconnect,
>   	.shutdown	= appletbdrm_shutdown,
> +	.suspend	= appletbdrm_suspend,
> +	.resume		= appletbdrm_resume,
> +	.reset_resume	= appletbdrm_reset_resume,
>   	.id_table	= appletbdrm_usb_id_table,
>   };
>   module_usb_driver(appletbdrm_usb_driver);
> 
> ---
> base-commit: c4c54ffd65555ac5316f8faf4d7dc8c815877937
> change-id: 20260924-appletbdrm-pm-6bd220385beb
> 
> Best regards,

[PATCH v2] drm/appletbdrm: Add suspend and resume support

[Aditya Garg]

From: Vasilij Markin <vasilij.kernel@tototo.si>

The driver has no USB power management callbacks, so the USB core
unbinds it before system suspend and binds it again on resume. The DRM
device is unregistered and registered again on every suspend and resume
cycle, userspace driving the Touch Bar through it (such as tiny-dfr)
loses its device, and the Touch Bar stays dark after resume.

Implement suspend and resume with drm_mode_config_helper_suspend() and
drm_mode_config_helper_resume(), so the device stays registered and the
last committed state is restored on resume. If the device was reset
while suspended, repeat the probe-time handshake first.

Fixes: 0670c2f56e45 ("drm/tiny: add driver for Apple Touch Bars in x86 Macs")
Assisted-by: Claude:claude-opus-5-5
Signed-off-by: Vasilij Markin <vasilij.kernel@tototo.si>
Signed-off-by: Aditya Garg <aditya.garg@linux.dev>
---
v2: Remove reset_resume

 drivers/gpu/drm/tiny/appletbdrm.c | 17 +++++++++++++++++
 1 file changed, 17 insertions(+)

diff --git a/drivers/gpu/drm/tiny/appletbdrm.c b/drivers/gpu/drm/tiny/appletbdrm.c
index ef71b9957..c94ed1068 100644
--- a/drivers/gpu/drm/tiny/appletbdrm.c
+++ b/drivers/gpu/drm/tiny/appletbdrm.c
@@ -30,6 +30,7 @@
 #include <drm/drm_gem_atomic_helper.h>
 #include <drm/drm_gem_framebuffer_helper.h>
 #include <drm/drm_gem_shmem_helper.h>
+#include <drm/drm_modeset_helper.h>
 #include <drm/drm_plane.h>
 #include <drm/drm_print.h>
 #include <drm/drm_probe_helper.h>
@@ -821,6 +822,20 @@ static void appletbdrm_shutdown(struct usb_interface *intf)
 	drm_atomic_helper_shutdown(&adev->drm);
 }
 
+static int appletbdrm_suspend(struct usb_interface *intf, pm_message_t message)
+{
+	struct appletbdrm_device *adev = usb_get_intfdata(intf);
+
+	return drm_mode_config_helper_suspend(&adev->drm);
+}
+
+static int appletbdrm_resume(struct usb_interface *intf)
+{
+	struct appletbdrm_device *adev = usb_get_intfdata(intf);
+
+	return drm_mode_config_helper_resume(&adev->drm);
+}
+
 static const struct usb_device_id appletbdrm_usb_id_table[] = {
 	{ USB_DEVICE_INTERFACE_CLASS(0x05ac, 0x8302, USB_CLASS_AUDIO_VIDEO) },
 	{}
@@ -832,6 +847,8 @@ static struct usb_driver appletbdrm_usb_driver = {
 	.probe		= appletbdrm_probe,
 	.disconnect	= appletbdrm_disconnect,
 	.shutdown	= appletbdrm_shutdown,
+	.suspend	= appletbdrm_suspend,
+	.resume		= appletbdrm_resume,
 	.id_table	= appletbdrm_usb_id_table,
 };
 module_usb_driver(appletbdrm_usb_driver);
-- 
2.53.0

Re: [PATCH v2] drm/appletbdrm: Add suspend and resume support

[Aditya Garg]

> On 4 Oct 2026, at 5:36 PM, Aditya Garg <aditya.garg@linux.dev> wrote:
> 
> From: Vasilij Markin <vasilij.kernel@tototo.si>
> 
> The driver has no USB power management callbacks, so the USB core
> unbinds it before system suspend and binds it again on resume. The DRM
> device is unregistered and registered again on every suspend and resume
> cycle, userspace driving the Touch Bar through it (such as tiny-dfr)
> loses its device, and the Touch Bar stays dark after resume.
> 
> Implement suspend and resume with drm_mode_config_helper_suspend() and
> drm_mode_config_helper_resume(), so the device stays registered and the
> last committed state is restored on resume. If the device was reset
> while suspended, repeat the probe-time handshake first.
> 
> Fixes: 0670c2f56e45 ("drm/tiny: add driver for Apple Touch Bars in x86 Macs")
> Assisted-by: Claude:claude-opus-5-5
> Signed-off-by: Vasilij Markin <vasilij.kernel@tototo.si>
> Signed-off-by: Aditya Garg <aditya.garg@linux.dev>

Could anyone please merge it if they are ok with it? Thanks.
> ---
> v2: Remove reset_resume
> 
> drivers/gpu/drm/tiny/appletbdrm.c | 17 +++++++++++++++++
> 1 file changed, 17 insertions(+)
> 
> diff --git a/drivers/gpu/drm/tiny/appletbdrm.c b/drivers/gpu/drm/tiny/appletbdrm.c
> index ef71b9957..c94ed1068 100644
> --- a/drivers/gpu/drm/tiny/appletbdrm.c
> +++ b/drivers/gpu/drm/tiny/appletbdrm.c
> @@ -30,6 +30,7 @@
> #include <drm/drm_gem_atomic_helper.h>
> #include <drm/drm_gem_framebuffer_helper.h>
> #include <drm/drm_gem_shmem_helper.h>
> +#include <drm/drm_modeset_helper.h>
> #include <drm/drm_plane.h>
> #include <drm/drm_print.h>
> #include <drm/drm_probe_helper.h>
> @@ -821,6 +822,20 @@ static void appletbdrm_shutdown(struct usb_interface *intf)
>    drm_atomic_helper_shutdown(&adev->drm);
> }
> 
> +static int appletbdrm_suspend(struct usb_interface *intf, pm_message_t message)
> +{
> +    struct appletbdrm_device *adev = usb_get_intfdata(intf);
> +
> +    return drm_mode_config_helper_suspend(&adev->drm);
> +}
> +
> +static int appletbdrm_resume(struct usb_interface *intf)
> +{
> +    struct appletbdrm_device *adev = usb_get_intfdata(intf);
> +
> +    return drm_mode_config_helper_resume(&adev->drm);
> +}
> +
> static const struct usb_device_id appletbdrm_usb_id_table[] = {
>    { USB_DEVICE_INTERFACE_CLASS(0x05ac, 0x8302, USB_CLASS_AUDIO_VIDEO) },
>    {}
> @@ -832,6 +847,8 @@ static struct usb_driver appletbdrm_usb_driver = {
>    .probe        = appletbdrm_probe,
>    .disconnect    = appletbdrm_disconnect,
>    .shutdown    = appletbdrm_shutdown,
> +    .suspend    = appletbdrm_suspend,
> +    .resume        = appletbdrm_resume,
>    .id_table    = appletbdrm_usb_id_table,
> };
> module_usb_driver(appletbdrm_usb_driver);
> --
> 2.53.0

Re: [PATCH v2] drm/appletbdrm: Add suspend and resume support

[Lukas Wunner]

On Wed, Oct 07, 2026 at 07:50:38PM +0530, Aditya Garg wrote:
> > On 4 Oct 2026, at 5:36???PM, Aditya Garg <aditya.garg@linux.dev> wrote:
> > The driver has no USB power management callbacks, so the USB core
> > unbinds it before system suspend and binds it again on resume. The DRM
> > device is unregistered and registered again on every suspend and resume
> > cycle, userspace driving the Touch Bar through it (such as tiny-dfr)
> > loses its device, and the Touch Bar stays dark after resume.
> > 
> > Implement suspend and resume with drm_mode_config_helper_suspend() and
> > drm_mode_config_helper_resume(), so the device stays registered and the
> > last committed state is restored on resume. If the device was reset
> > while suspended, repeat the probe-time handshake first.
[...]
> Could anyone please merge it if they are ok with it? Thanks.

I'd love to, but both v1 and v2 have sashiko findings that nobody bothered
to debunk or fix.  (The fix for v1 consisted of removing a callback, which
triggered other sashiko findings.)  I'm neither an expert on this driver
nor the tiny drm library, but I'd like to help improve Linux Mac support.
Please help me understand whether the sashiko findings are valid.

Thanks,

Lukas

Re: [PATCH v2] drm/appletbdrm: Add suspend and resume support

[Vasilij]

On Thu, Oct 15, 2026 at 10:26:15PM +0200, Lukas Wunner wrote:
> I'd love to, but both v1 and v2 have sashiko findings that nobody bothered
> to debunk or fix.  (The fix for v1 consisted of removing a callback, which
> triggered other sashiko findings.)  I'm neither an expert on this driver
> nor the tiny drm library, but I'd like to help improve Linux Mac support.
> Please help me understand whether the sashiko findings are valid.

Thanks for looking at it, and sorry for leaving the findings
unanswered. I went through all of them; in short, the v1 finding and
the first two v2 findings are valid. v3 brings back reset_resume with
the error path fixed:
https://lore.kernel.org/all/20261015-appletbdrm-pm-v3-1-47377176619a@tototo.si/

v1, "memory leak of suspend_state on reset_resume error": valid. If
appletbdrm_get_information() or appletbdrm_signal_readiness() fails,
reset_resume returns before drm_mode_config_helper_resume(), so the
state saved by drm_mode_config_helper_suspend() is never put, and
fbdev/polling stay suspended. The USB core only logs a reset_resume
error and keeps the driver bound, so nothing else cleans it up. v3
logs the handshake error and still calls
drm_mode_config_helper_resume(), the same way udl and gm12u320 do it.

v2, "missing reset_resume breaks reset recovery": valid, and not only
in theory. v2 dropped reset_resume because the device was not reset
on resume, but with the current t2linux 7.2.11 kernel (which carries
v2) the Touch Bar is reset on every resume here (MacBookPro16,2, S3):

  usb 5-6: reset high-speed USB device number 7 using t2bce_core
  [drm] Initialized appletbdrm 1.0.0 for 5-6:2.1 on minor 2

Without reset_resume the USB core unbinds and probes the driver
again, tiny-dfr loses its device and the Touch Bar stays dark, which
is the bug the patch is meant to fix. With v1 (reset_resume doing the
probe handshake) on the same kernel the reset still happens, but the
driver is not probed again and the Touch Bar works after resume.

v2, "suspend_state leaked on unbind while suspended": with
reset_resume back, a reset no longer unbinds the driver, which was
the scenario sashiko described. What is left is the device going
away while the system sleeps: then disconnect runs without a resume
and the saved state is not freed. As far as I can see udl and
gm12u320 have the same gap, and drm_mode_config_cleanup() does not
free suspend_state either. v3 leaves it out; I can drop the saved
state in appletbdrm_disconnect() in a follow-up, or it could be fixed
once in the helpers if the DRM maintainers prefer that.

v3 is tested on the t2linux 7.2.11 kernel: 7 suspend/resume cycles,
the Touch Bar was reset on each resume, the driver was not probed
again, tiny-dfr kept running and the Touch Bar worked every time.
(I went through the findings with the help of an AI assistant,
Claude; the logs and tests are from my machine.)

Thanks,
Vasilij