apple-drm: quiesce inactive DCPs on suspend, fix resume crash#510
Open
areofyl wants to merge 48 commits into
Open
apple-drm: quiesce inactive DCPs on suspend, fix resume crash#510areofyl wants to merge 48 commits into
areofyl wants to merge 48 commits into
Conversation
areofyl
force-pushed
the
fix/drm-dcp-suspend
branch
from
d33be4e to
0b2b907
Compare
areofyl
force-pushed
the
fix/drm-dcp-suspend
branch
from
0b2b907 to
6e8660e
Compare
Certain Broadcom bluetooth chips (bcm4377/bcm4378/bcm438) need ACL streams carrying audio to be set as "high priority" using a vendor specific command to prevent 10-ish second-long dropouts whenever something does a device scan. This patch sends the command when the socket priority is set to TC_PRIO_INTERACTIVE, as BlueZ does for audio. From experimenting with the hardware - this command is not suitable for per-skb priority switching, as prioritization is done on the handle level, with this command reconfiguring certain radio timings, and dropping to low priority in order to send a low packet on the same handle as an audio stream is being played on causes the same kind of dropout it is supposed to avoid. In addition, the hardware is rather picky about when this command can be sent, as sending it during connection open results in a timeout. The vendor stacks solve it by having high-level visibility into what a connection is used for and sending it from userspace when it is known that an audio stream is about to start. As we can't have that visibility without introducing a new ioctl, the socket priority is used as proxy. Reviewed-by: Neal Gompa <neal@gompa.dev> Signed-off-by: Sasha Finkelstein <k@chaosmail.tech>
The current approach of silently disabling all rust drivers if the toolchain is missing results in users that try to compile their own kernels getting a "successful" build and then being confused about where did their drivers go. In comparison, missing openssl results in a build failure, not a disappearance of everything that depends on it. This also means that allyesconfig will depend on rust, but since the rust experiment concluded with "rust is here to stay", i believe that allyesconfig should be building rust drivers too. Signed-off-by: Sasha Finkelstein <k@chaosmail.tech>
Signed-off-by: Janne Grunau <j@jannau.net>
Apple M3 Pro and Max devices are using 'gp00' keys for GPIO in addition to 'gP00' keys. Add a second compatible to handle this keys with an additional macsmc-gpio instance. Signed-off-by: Janne Grunau <j@jannau.net>
Add support for SMC GPIO keys with a lower letter 'p' via the "apple,smc-low-gpio" compatible. This adds support for a second macsmc-gpio controller using 'gp00' keys. These keys are used on Apple M3 Pro and Max MacBooks in the controller for keyboard and trackpad and for the built-in DisplayPort to HDMI converter. Signed-off-by: Janne Grunau <j@jannau.net>
Apple M3 Pro and Max devices are using 'gp00' keys for GPIO in addition to 'gP00' keys. These keys are handled by an additional macsmc-gpio instance using the "apple,smc-low-gpio" compatible. Signed-off-by: Janne Grunau <j@jannau.net>
- WLAN/BT (SMC PMU GPIO AsahiLinux#13) (all devices) - ASM3142 (SMC PMU GPIO AsahiLinux#14) (j434, iMac with 4 USB-C ports) - SD card reader (SMC PMU GPIO AsahiLinux#23) (j504, 14-inch MacBook Pro) Signed-off-by: Janne Grunau <j@jannau.net>
Signed-off-by: Yureka <yuka@yuka.dev>
The internal keyboard and trackpad HID on MacBook variants of the Apple M3 (t8122) SoC are connected through a Apple -developed protocol called DockChannel and mediated by a coprocessor known as the Multi-Touch Processor (MTP). This commit adds the nessecary device tree nodes to the M3's device tree for internal HID to work. It is disabled by default, to be enabled only in MacBook board files where it is tested and confirmed to work. Co-developed-by: Alyssa Milburn <amilburn@zall.org> Signed-off-by: Alyssa Milburn <amilburn@zall.org> Signed-off-by: Michael Reeves <michael.reeves077@gmail.com>
Add mtp device nodes for t8122 (M3) based MacBooks. Signed-off-by: Michael Reeves <michael.reeves077@gmail.com>
Signed-off-by: Janne Grunau <j@jannau.net>
Signed-off-by: Janne Grunau <j@jannau.net>
List trackpad firmware files and activate MTP devices nodes on all t6030, t6031 and t6034 based MacBooks. Signed-off-by: Janne Grunau <j@jannau.net>
HDP status for DisplayPort alt-mode is signaled data_status. Track changes to have a debounced HPD to forward to the DRM KMS driver. Signed-off-by: Janne Grunau <j@jannau.net>
This is not how dp-altmode support should be implemented but it works for new. Requires a "displayport" property in the connector node with a phandle of the connector. Signed-off-by: Janne Grunau <j@jannau.net>
Enable DP alt mode for all M1 devices: - Mac Mini (M1): USB-C port next to the HDMI port - Macbook Pro (M1, 13-inch): front left USB-C port - Macbook Air (M1, 13-inch): front left USB-C port - iMac (M1, 2 USB-C ports): back left USB-C port - iMac (M1, 4 USB-C ports): back right middle USB-C port Signed-off-by: Janne Grunau <j@jannau.net>
Works around missing suspend/resume handling in ATC phy for DP-altmode. Signed-off-by: Janne Grunau <j@jannau.net>
Enable DP alt mode for the front left USB-C port of Macbook Air 13 (M2, 13/15-inch) and Macbook Pro (M2, 13-inch). Can't easily enabled on on the M2 Mac Mini since dcpext is used for the HDMI port and dcp bringup is troublesome. Signed-off-by: Janne Grunau <j@jannau.net>
Works around missing suspend/resume handling in ATC phy for DP-altmode. Signed-off-by: Janne Grunau <j@jannau.net>
Needs more testing, maybe a little unstable and somehow limits the HDMI out to 1280x720 (to be verified). Using dcp as display coproc since dcpext is used for the HDMI port. Signed-off-by: Janne Grunau <j@jannau.net>
Works around missing suspend/resume handling in ATC phy for DP-altmode. Signed-off-by: Janne Grunau <j@jannau.net>
Blessed dp-altmode port is front left port on j314/j316/j414/j416. Signed-off-by: Janne Grunau <j@jannau.net>
…ways on Works around missing suspend/resume handling in ATC phy for DP-altmode. Signed-off-by: Janne Grunau <j@jannau.net>
DP alt mode for Mac Studio, the blessed port is the back left middle port (second closest USB-C port to the power connector). Signed-off-by: Janne Grunau <j@jannau.net>
…s on Works around missing suspend/resume handling in ATC phy for DP-altmode. Signed-off-by: Janne Grunau <j@jannau.net>
DP alt mode for Mac Mini M2 Pro, the blessed port is the back right middle port (second closest USB-C port to the power connector). Signed-off-by: Janne Grunau <j@jannau.net>
areofyl
force-pushed
the
fix/drm-dcp-suspend
branch
from
6e8660e to
8faa278
Compare
drm_mode_config_helper_suspend returns -EINVAL when a secondary DCP (external display) is disconnected, which blocks the entire PM suspend path. Even if suspend returns 0 anyway, the DCP RTKit firmware keeps running and generates mailbox IRQs that immediately wake from s2idle. In suspend_noirq, iterate CRTCs and call apple_rtkit_quiesce on disconnected DCPs so they stop generating IRQs right before s2idle entry. In resume_noirq, re-boot them with apple_rtkit_boot. Also fix the resume path: drm_mode_config_helper_resume crashes on NULL suspend_state when suspend failed, so check for that and fall back to a hotplug event instead. Tested on M1 MacBook Air (J313) with USB-C display, fairydust 6.18.10. Signed-off-by: areofyl <areofyl@users.noreply.github.com>
areofyl
force-pushed
the
fix/drm-dcp-suspend
branch
from
8faa278 to
9a0b5c3
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
drm_mode_config_helper_suspend returns -EINVAL when a secondary DCP (external display) is disconnected, which blocks the entire PM suspend path. Even if you make suspend return 0 anyway, the DCP RTKit firmware keeps running and generates mailbox IRQs that immediately wake from s2idle.
When the drm suspend helper fails, iterate CRTCs and call apple_rtkit_quiesce on inactive DCPs so they stop generating IRQs. On resume, re-boot them with apple_rtkit_boot and fire a hotplug event so the DRM state gets reconciled.
Also fixed the resume path. drm_mode_config_helper_resume crashes on NULL suspend_state if suspend failed, so check for that and fall back to a hotplug event instead.
Tested on M1 MacBook Air (J313) with USB-C display, fairydust 6.18.10. Still needs more stress testing with repeated plug/unplug + suspend cycles before taking this out of draft.