Our default behavior continues to match the vanilla kernel.

The Nvidia proprietary driver has some bugs that leads to issues if used
with the simpledrm driver. The most noticeable is that does not register
an emulated fbdev device.

It just relies on a fbdev to be registered by another driver, that could
be that could be attached to the framebuffer console. On UEFI machines,
this is the efifb driver.

This means that disabling the efifb driver will cause virtual consoles to
not be present in the system when using the Nvidia driver. Legacy BIOS is
not affected just because fbcon is not used there, but instead vgacon.

Unless a VGA mode is specified using the vga= kernel command line option,
in that case the vesafb driver is used instead and its fbdev attached to
the fbcon.

This is a problem because with CONFIG_SYSFB_SIMPLEFB=y, the sysfb platform
code attempts to register a "simple-framebuffer" platform device (that is
matched against simpledrm) and only registers either an "efi-framebuffer"
or "vesa-framebuffer" if this fails to be registered due the video modes
not being compatible.

The Nvidia driver relying on another driver to register the fbdev is quite
fragile, since it can't really assume those will stick around. For example
there are patches posted to remove the EFI and VESA platform devices once
a real DRM or fbdev driver probes.

But in any case, moving to a simpledrm + emulated fbdev only breaks this
assumption and causes users to not have VT if the Nvidia driver is used.

So to prevent this, let's add a workaround and make the sysfb to skip the
"simple-framebuffer" registration when nvidia-drm.modeset=1 option is set.

This is quite horrible, but honestly I can't think of any other approach.

For this to work, the CONFIG_FB_EFI and CONFIG_FB_VESA config options must
be enabled besides CONFIG_DRM_SIMPLEDRM.

Signed-off-by: Javier Martinez Canillas <[email protected]>
Cherry-picked-for: https://bugs.archlinux.org/task/73720

Add quirk orientation for AYA NEO 2. The name appears without spaces in
dmi strings. That made it difficult to reuse the 2021 match and the
display is greater in resolution.

Tested by the JELOS team that has been patching their own kernel for a
while now and confirmed by users in the AYA NEO and ChimeraOS discord
servers.

Signed-off-by: Joaquín Ignacio Aramendía <[email protected]>

Add quirk orientation for AYA NEO Founder. The name appears with spaces in
dmi strings as other devices of the brand. The panel is the same as the
NEXT and 2021 models. Those could not be reused as the former has VENDOR
name as "AYANEO" without spaces and the latter has "AYADEVICE".

Tested by the JELOS team that has been patching their own kernel for a
while now and confirmed by users in the AYA NEO and ChimeraOS discord
servers.

Signed-off-by: Joaquín Ignacio Aramendía <[email protected]>

Add quirk orientation for AYA NEO GEEK. One of the more recent devices by
the brand. The name appears without spaces in dmi strings. The board
name is completely different to the previous models making it difficult
to reuse their quirks despite being the same resolution and mounting.

Tested by the JELOS team that has been patching their own kernel for a
while now and confirmed by users in the AYA NEO and ChimeraOS discord
servers.

Signed-off-by: Joaquín Ignacio Aramendía <[email protected]>

Support adding the missing DSD properties required  for ASUS ROG 2023
laptops and other ASUS laptops to properly utilise the cs35l41.

The currently added laptops are:
- ASUS GS650P,   i2c
- ASUS GA402X,   i2c
- ASUS GU604V,   spi
- ASUS GU603V,   spi
- ASUS GV601V,   spi
- ASUS GZ301V,   spi
- ASUS ROG ALLY, i2c
- ASUS G614J,    spi
- ASUS G634J,    spi
- ASUS G614JI,   spi
- ASUS G713P,    i2c

The SPI connected amps may be required to use an external DSD patch
to fix or add the "cs-gpios" property.

Co-developed-by: Jonathan LoBue <[email protected]>
Signed-off-by: Jonathan LoBue <[email protected]>
Co-developed-by: Luke D. Jones <[email protected]>
Signed-off-by: Luke D. Jones <[email protected]>

This patch adds a new sysfs event, which will indicate
the userland about a GPU reset, and can also provide
some information like:
- process ID of the process involved with the GPU reset
- process name of the involved process
- the GPU status info (using flags)

This patch also introduces the first flag of the flags
bitmap, which can be appended as and when required.

V2: Addressed review comments from Christian and Amar
   - move the reset information structure to DRM layer
   - drop _ctx from struct name
   - make pid 32 bit(than 64)
   - set flag when VRAM invalid (than valid)
   - add process name as well (Amar)

Cc: Alexandar Deucher <[email protected]>
Cc: Christian Koenig <[email protected]>
Cc: Amaranath Somalapuram <[email protected]>
Signed-off-by: Shashank Sharma <[email protected]>
(cherry picked from commit 90230bd9d9c7d979038547460c9a2cbbeff8d6b9)
[Forward port to 6.0]
Signed-off-by: Cristian Ciocaltea <[email protected]>

This patch adds a work function, which sends a GPU reset
uevent and some contextual infomration, like the PID and
some status flags. This work should be scheduled during
a GPU reset.

The userspace can do some recovery and post-processing work
based on this event and information.

V2: Addressed review comments from Christian
- Changed the name of the work to gpu_reset_event_work
- Added a structure to accommodate some additional information
  (like a PID and some flags)
- Do not add new structure in amdgpu.h

Cc: Alexander Deucher <[email protected]>
Cc: Christian Koenig <[email protected]>
Cc: Amaranath Somalapuram <[email protected]>
Signed-off-by: Shashank Sharma <[email protected]>
(cherry picked from commit f63b09e78126f7da67b69409e2cce1d3ab2d7f46)
[Forward port to 6.0]
Signed-off-by: Cristian Ciocaltea <[email protected]>

Schedule work function with valid PID, process name,
and vram lost status during a GPU reset/ recovery.

Signed-off-by: Somalapuram Amaranath <[email protected]>
(cherry picked from commit 293c019a84c6402b08db9579819b555b01cd613b)
[Forward ported to 6.0]
Signed-off-by: Cristian Ciocaltea <[email protected]>

While the HID gadget implementation has been sufficient for devices that only
use INTERRUPT transfers, the USB HID standard includes provisions for Set- and
Get-Feature report CONTROL transfers that go over endpoint 0. These were
previously impossible with the existing implementation, and would either send
an empty reply, or stall out.

As the feature is a standard part of USB HID, it stands to reason that devices
would use it, and that the HID gadget should support it. This patch adds
support for (polled) device-to-host Get-Feature reports through a new ioctl
interface to the hidg class dev nodes.

Signed-off-by: Vicki Pfau <[email protected]>
(cherry picked from commit 8437fa3861c7198a3e286f393c8637c4fc08d2bc)
Signed-off-by: Cristian Ciocaltea <[email protected]>

While the HID gadget implementation has been sufficient for devices that only
use INTERRUPT transfers, the USB HID standard includes provisions for Set- and
Get-Feature report CONTROL transfers that go over endpoint 0. These were
previously impossible with the existing implementation, and would either send
an empty reply, or stall out.

As the feature is a standard part of USB HID, it stands to reason that devices
would use it, and that the HID gadget should support it. This patch adds
support for host-to-device Set-Feature reports through a new ioctl
interface to the hidg class dev nodes.

Signed-off-by: Vicki Pfau <[email protected]>
(cherry picked from commit 3d82be0ec3aa3b947d9c927d7b06c433de15be8b)
Signed-off-by: Cristian Ciocaltea <[email protected]>

SDL includes a list of settings (registers), reports (cmds), and various other
identifiers that were provided by Valve. This commit imports a significant
chunk of that list as well as updating the guessed names and replacing a
handful of magic constants. It also replaces bitmask definitions that used hex
with the BIT macro.

Signed-off-by: Vicki Pfau <[email protected]>

Signed-off-by: Vicki Pfau <[email protected]>

This cleans up the locking logic so that the spinlock is consistently used for
access to a small handful of struct variables, and the mutex is exclusively and
consistently used for ensuring that mutliple threads aren't trying to
send/receive reports at the same time. Previously, only some report
transactions were guarded by this mutex, potentially breaking atomicity. The
mutex has been renamed to reflect this usage.

Signed-off-by: Vicki Pfau <[email protected]>

The client_opened variable was used to track if the hidraw was opened by any
clients to silence keyboard/mouse events while opened. However, there was no
counting of how many clients were opened, so opening two at the same time and
then closing one would fool the driver into thinking it had no remaining opened
clients.

Signed-off-by: Vicki Pfau <[email protected]>

The second byte of the GET_STRING_ATTRIB report is a length, so we should set
the size of the buffer to be the size we're actually requesting, and only
reject the reply if the length out is nonsensical.

Signed-off-by: Vicki Pfau <[email protected]>

Add quirk orientation for the Ayn Loki Zero.

This also has been tested/used by the JELOS team.

Signed-off-by: Bouke Sybren Haarsma <[email protected]>

Add quirk orientation for Ayn Loki Max model.

This has been tested by JELOS team that uses their
own patched kernel for a while now and confirmed by
users in the ChimeraOS discord servers.

Signed-off-by: Bouke Sybren Haarsma <[email protected]>

Fixing no patch applied exit code