The HVS can change AXI request mode based on how full the COB
FIFOs are.
Until now the vc4 driver has been relying on the firmware to
have set these to sensible values.
With HVS channel 2 now being used for live video, change the
panic mode for all channels to be explicitly set by the driver,
and the same for all channels.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Support displaying DRM_FORMAT_YUV444 and DRM_FORMAT_YVU444 formats.
Tested with kmstest and kodi. e.g.
kmstest -r 1920x1080@60 -f 400x300-YU24
Note: without the shift of width, only half the chroma is fetched,
resulting in correct left half of image and corrupt colours on right half.
The increase in width shouldn't affect fetching of Y data,
as the hardware will clamp at dest width.
Signed-off-by: Dom Cobley <popcornmix@gmail.com>
When running a kunit test, the driver runs with a mock device. As such,
any attempt to read or write to a hardware register will fail the
current test immediately.
The dlist allocation management recently introduced will read the
current frame count from the HVS to defer its destruction until a
subsequent frame has been output. This obviously involves a register
read that fails the Kunit tests.
Change the destruction deferral function to destroy the allocation
immediately if we run under kunit. This is essentially harmless since
the main reason for that deferall is to prevent any access to the
hardware dlist while a frame described by that list is rendered. On our
mock driver, we have neither a hardware dlist nor a rendering, so it
doesn't matter.
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
We'll need to destroy a dlist allocation in multiple code paths, so
let's move it to a separate function.
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
The vc4_hvs_dlist_allocation structure has a list that we don't
initialize when we allocate a new instance.
This makes any call reading the list structure (such as list_empty) fail
with a NULL pointer dereference.
Let's make sure our list is always initialized.
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
During normal operations, the cursor position update is done through an
asynchronous plane update, which on the vc4 driver basically just
modifies the right dlist word to move the plane to the new coordinates.
However, when we have the overscan margins setup, we fall back to a
regular commit when we are next to the edges. And since that commit
happens to be on a cursor plane, it's considered a legacy cursor update
by KMS.
The main difference it makes is that it won't wait for its completion
(ie, next vblank) before returning. This means if we have multiple
commits happening in rapid succession, we can have several of them
happening before the next vblank.
In parallel, our dlist allocation is tied to a CRTC state, and each time
we do a commit we end up with a new CRTC state, with the previous one
being freed. This means that we free our previous dlist entry (but don't
clear it though) every time a new one is being committed.
Now, if we were to have two commits happening before the next vblank, we
could end up freeing reusing the same dlist entries before the next
vblank.
Indeed, we would start from an initial state taking, for example, the
dlist entries 10 to 20, then start a commit taking the entries 20 to 30
and setting the dlist pointer to 20, and freeing the dlist entries 10 to
20. However, since we haven't reach vblank yet, the HVS is still using
the entries 10 to 20.
If we were to make a new commit now, chances are the allocator are going
to give the 10 to 20 entries back, and we would change their content to
match the new state. If vblank hasn't happened yet, we just corrupted
the active dlist entries.
A first attempt to solve this was made by creating an intermediate dlist
buffer to store the current (ie, as of the last commit) dlist content,
that we would update each time the HVS is done with a frame. However, if
the interrupt handler missed the vblank window, we would end up copying
our intermediate dlist to the hardware one during the composition,
essentially creating the same issue.
Since making sure that our interrupt handler runs within a fixed,
constrained, time window would require to make Linux a real-time kernel,
this seems a bit out of scope.
Instead, we can work around our original issue by keeping the dlist
slots allocation longer. That way, we won't reuse a dlist slot while
it's still in flight. In order to achieve this, instead of freeing the
dlist slot when its associated CRTC state is destroyed, we'll queue it
in a list.
A naive implementation would free the buffers in that queue when we get
our end of frame interrupt. However, there's still a race since, just
like in the shadow dlist case, we don't control when the handler for
that interrupt is going to run. Thus, we can end up with a commit adding
an old dlist allocation to our queue during the window between our
actual interrupt and when our handler will run. And since that buffer is
still being used for the composition of the current frame, we can't free
it right away, exposing us to the original bug.
Fortunately for us, the hardware provides a frame counter that is
increased each time the first line of a frame is being generated.
Associating the frame counter the image is supposed to go away to the
allocation, and then only deallocate buffers that have a counter below
or equal to the one we see when the deallocation code should prevent the
above race from occuring.
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
We regularly get dmesg error reports of:
[ 18.184066] hdmi-audio-codec hdmi-audio-codec.3.auto: ASoC: error at snd_soc_dai_startup on i2s-hifi: -19
[ 18.184098] MAI: soc_pcm_open() failed (-19)
Currently I get 30 of these when booting to desktop.
We always say, ignore they are harmless, but removing them would be good.
A bit of investigation shows, for me, the errors are all generated by second, unused hdmi interface.
It shows as an alsa device, and pulseaudio attempts to open it (numerous times), generating a kernel
error message each time.
systemctl --user restart pulseaudio.service generates 6 additional error messages.
The error messages all come through:
a009a9c0d7/sound/soc/soc-pcm.c (L39)
which suggests returning ENOTSUPP, rather that ENODEV will be quiet. And indeed it is.
Note: earlier kernels do not have the quiet ENOTSUPP, so additional cherry-picks will be needed to backport
Signed-off-by: Dom Cobley <popcornmix@gmail.com>
See: https://forum.libreelec.tv/thread/24783-tv-avr-turns-back-on-right-after-turning-them-off
While the kernel provides a :D flag for assuming device is connected,
it doesn't stop this function from being called and generating a cec_phys_addr_invalidate
message when hotplug is deasserted.
That message provokes a flurry of CEC messages which for many users results in the TV
switching back on again and it's very hard to get it to stay switched off.
It seems to only occur with an AVR and TV connected but has been observed across a
number of manufacturers.
The issue started with https://github.com/raspberrypi/linux/pull/4371
and this provides an optional way of getting back the old behaviour
Signed-off-by: Dom Cobley <popcornmix@gmail.com>
FKMS doesn't have an HVS and it's expected. Return from the debugfs init
function immediately if we're running with fkms.
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Margins may be implemented by scaling the planes, but as there
is no way of intercepting the set_property for a standard property,
and all planes are checked in drm_atomic_check_only before the
connectors, there's now way to add the planes into the state
from the driver.
If the margin properties change, add all corresponding planes to
the state.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
The stuff never really worked, and leads to lots of fun because it
out-of-order frees atomic states. Which upsets KASAN, among other
things.
For async updates we now have a more solid solution with the
->atomic_async_check and ->atomic_async_commit hooks. Support for that
for msm and vc4 landed. nouveau and i915 have their own commit
routines, doing something similar.
For everyone else it's probably better to remove the use-after-free
bug, and encourage folks to use the async support instead. The
affected drivers which register a legacy cursor plane and don't either
use the new async stuff or their own commit routine are: amdgpu,
atmel, mediatek, qxl, rockchip, sti, sun4i, tegra, virtio, and vmwgfx.
Inspired by an amdgpu bug report.
v2: Drop RFC, I think with amdgpu converted over to use
atomic_async_check/commit done in
commit 674e78acae
Author: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Date: Wed Dec 5 14:59:07 2018 -0500
drm/amd/display: Add fast path for cursor plane updates
we don't have any driver anymore where we have userspace expecting
solid legacy cursor support _and_ they are using the atomic helpers in
their fully glory. So we can retire this.
v3: Paper over msm and i915 regression. The complete_all is the only
thing missing afaict.
v4: Rebased on recent kernel, added extra link for vc4 bug.
Link: https://bugzilla.kernel.org/show_bug.cgi?id=199425
Link: https://lore.kernel.org/all/20220221134155.125447-9-maxime@cerno.tech/
Cc: mikita.lipski@amd.com
Cc: Michel Dänzer <michel@daenzer.net>
Cc: harry.wentland@amd.com
Cc: Rob Clark <robdclark@gmail.com>
Cc: "Kazlauskas, Nicholas" <nicholas.kazlauskas@amd.com>
Tested-by: Maxime Ripard <maxime@cerno.tech>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
When the margins are changed, the dlist needs to be regenerated
with the changed updated dest regions for each of the planes.
Setting the zpos_changed flag is sufficient to trigger that
without doing a full modeset, therefore set it should the
margins be changed.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
mipi_dsi_attach can fail due to resources not being available
yet, therefore do not log error messages should they occur.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Still under investigation, but the conditions under which the HVS
will accept values written to the gamma PWL are not straightforward.
Disable gamma on HVS5 again until it can be resolved to avoid
gamma being enabled with an incorrect table.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Add a check to vc4_hvs_gamma_check to ensure a new non-empty
gamma LUT is of the correct length before accepting it.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Two calls were made to drm_crtc_enable_color_mgmt to add gamma
and CTM, however they were both set to add the gamma properties,
so they ended up added twice.
Fixes: 766cc6b1f7 "drm/vc4: Add CTM support"
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
With HVS5 the gamma block is now only reprogrammed with
a disable/enable. Loading the table from vc4_hvs_init_channel
(called from vc4_hvs_atomic_enable) appears to be at an
invalid point in time and so isn't applied.
Switch to enabling and disabling the gamma table instead. This
isn't safe if the pipeline is running, but it isn't now.
For HVS4 it is safe to enable and disable dynamically, so
adopt that approach there too.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
drm_crtc_legacy_gamma_set updates the gamma_lut blob unconditionally,
which leads to unnecessary reprogramming of hardware.
Check whether the blob contents has actually changed before
signalling that it has been updated.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Add predefined modelines for the 240p (NTSC) and 288p (PAL) progressive
modes, and report them through vc4_vec_connector_get_modes().
Signed-off-by: Mateusz Kwiatkowski <kfyatek+publicgit@gmail.com>
The HVS Gamma block can only be updated when idle, so we need to disable
the HVS channel when the gamma property is set in an atomic commit.
Since the pixelvalve cannot have its assigned channel halted without
stalling unless it's disabled as well, in our case that means forcing a
full disable / enable cycle on the pipeline.
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
BCM2711 changes from a 256 entry lookup table to a 16 point
piecewise linear function as the pipeline bitdepth has increased
to make a LUT unwieldy.
Implement a simple conversion from a 256 entry LUT that userspace
is likely to expect to 16 evenly spread points in the PWL. This
could be improved with curve fitting at a later date.
Co-developed-by: Juerg Haefliger <juergh@canonical.com>
Signed-off-by: Juerg Haefliger <juergh@canonical.com>
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
This is a squash of all firmware-kms related patches from previous
branches, up to and including
"drm/vc4: Set the possible crtcs mask correctly for planes with FKMS"
plus a couple of minor fixups for the 5.9 branch.
Please refer to earlier branches for full history.
This patch includes work by Eric Anholt, James Hughes, Phil Elwell,
Dave Stevenson, Dom Cobley, and Jonathon Bell.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
drm/vc4: Fixup firmware-kms after "drm/atomic: Pass the full state to CRTC atomic enable/disable"
Prototype for those calls changed, so amend fkms (which isn't
upstream) to match.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
drm/vc4: Fixup fkms for API change
Atomic flush and check changed API, so fix up the downstream-only
FKMS driver.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
drm/vc4: Make normalize_zpos conditional on using fkms
Eric's view was that there was no point in having zpos
support on vc4 as all the planes had the same functionality.
Can be later squashed into (and fixes):
drm/vc4: Add firmware-kms mode
Signed-off-by: Dom Cobley <popcornmix@gmail.com>
drm/vc4: FKMS: Change of Broadcast RGB mode needs a mode change
The Broadcast RGB (aka HDMI limited/full range) property is only
notified to the firmware on mode change, so this needs to be
signalled when set.
https://github.com/raspberrypi/firmware/issues/1580
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
vc4/drv: Only notify firmware of display done with kms
fkms driver still wants firmware display to be active
Signed-off-by: Dom Cobley <popcornmix@gmail.com>
ydrm/vc4: fkms: Fix margin calculations for the right/bottom edges
The calculations clipped the right/bottom edge of the clipped
range based on the left/top margins.
https://github.com/raspberrypi/linux/issues/4447
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
drm/vc4: fkms: Use new devm_rpi_firmware_get api
drm/kms: Add allow_fb_modifiers
Signed-off-by: Dom Cobley <popcornmix@gmail.com>
drm/vc4: Add async update support for cursor planes
Now that cursors are implemented as regular planes, all cursor
movements result in atomic updates. As the firmware-kms driver
doesn't support asynchronous updates, these are synchronous, which
limits the update rate to the screen refresh rate. Xorg seems unaware
of this (or at least of the effect of this), because if the mouse is
configured with a higher update rate than the screen then continuous
mouse movement results in an increasing backlog of mouse events -
cue extreme lag.
Add minimal support for asynchronous updates - limited to cursor
planes - to eliminate the lag.
See: https://github.com/raspberrypi/linux/pull/4971https://github.com/raspberrypi/linux/issues/4988
Signed-off-by: Phil Elwell <phil@raspberrypi.com>
drivers/gpu/drm/vc4: Add missing 32-bit RGB formats
The missing 32-bit per pixel ABGR and various "RGB with an X value"
formats are added. Change sent by Dave Stevenson.
Signed-off-by: David Plowman <david.plowman@raspberrypi.com>
drm: vc4: Fixup duplicated macro definition in vc4_firmware_kms
Both vc4_drv.h and vc4_firmware_kms.c had definitions for
to_vc4_crtc.
Rename the fkms one to make it unique, and drop the magic
define vc4_crtc vc4_kms_crtc
define to_vc4_crtc to_vc4_kms_crtc
that renamed half the variable and function names in a slightly
unexpected way.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
drm/vc4: Fix FKMS for when the YUV chroma planes are different buffers
The code was assuming that it was a single buffer with offsets,
when kmstest uses separate buffers and 0 offsets for each plane.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
drm/vc4: fkms: Rename plane related functions
The name collide with the Full KMS functions that are going to be made
public.
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
drm/vc4_fkms: Fix up interrupt handler for both 2835/2711 and 2712
2712 has switched from using the SMI peripheral to another interrupt
source for the vsync interrupt, so handle both sources cleanly.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Similar to the ch7006 and nouveau drivers, introduce a "tv_mode" module
parameter that allow setting the TV norm by specifying vc4.tv_norm= on
the kernel command line.
If that is not specified, try inferring one of the most popular norms
(PAL or NTSC) from the video mode specified on the command line. On
Raspberry Pis, this causes the most common cases of the sdtv_mode
setting in config.txt to be respected.
Signed-off-by: Mateusz Kwiatkowski <kfyatek+publicgit@gmail.com>
drm/vc4: Do not reset tv mode as this is already handled by framework
In vc4_vec_connector_reset, the tv mode is already reset to the
property default by drm_atomic_helper_connector_tv_reset, so there
is no need for a local fixup to potentially some other default.
Fixes: 96922af144 ("drm/vc4: Allow setting the TV norm via module parameter")
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Under FKMS, the firmware (via FKMS) also requires the VideoCore cache
aliases for image planes, as defined by the dma-ranges under /soc.
Add rpi-firmware-kms to the list of acceptable nodes to look for
to copy dma config from.
Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Starting from LNL, CCS has moved over to flat CCS model where there is
now dedicated memory reserved for storing compression state. On
platforms like LNL this reserved memory lives inside graphics stolen
memory, which is not treated like normal RAM and is therefore skipped by
the core kernel when creating the hibernation image. Currently if
something was compressed and we enter hibernation all the corresponding
CCS state is lost on such HW, resulting in corrupted memory. To fix this
evict user buffers from TT -> SYSTEM to ensure we take a snapshot of the
raw CCS state when entering hibernation, where upon resuming we can
restore the raw CCS state back when next validating the buffer. This has
been confirmed to fix display corruption on LNL when coming back from
hibernation.
Fixes: cbdc52c11c ("drm/xe/xe2: Support flat ccs")
Link: https://gitlab.freedesktop.org/drm/xe/kernel/-/issues/3409
Signed-off-by: Matthew Auld <matthew.auld@intel.com>
Cc: Matthew Brost <matthew.brost@intel.com>
Cc: <stable@vger.kernel.org> # v6.8+
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241112162827.116523-2-matthew.auld@intel.com
(cherry picked from commit c8b3c6db94)
Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
The GGTT looks to be stored inside stolen memory on igpu which is not
treated as normal RAM. The core kernel skips this memory range when
creating the hibernation image, therefore when coming back from
hibernation the GGTT programming is lost. This seems to cause issues
with broken resume where GuC FW fails to load:
[drm] *ERROR* GT0: load failed: status = 0x400000A0, time = 10ms, freq = 1250MHz (req 1300MHz), done = -1
[drm] *ERROR* GT0: load failed: status: Reset = 0, BootROM = 0x50, UKernel = 0x00, MIA = 0x00, Auth = 0x01
[drm] *ERROR* GT0: firmware signature verification failed
[drm] *ERROR* CRITICAL: Xe has declared device 0000:00:02.0 as wedged.
Current GGTT users are kernel internal and tracked as pinned, so it
should be possible to hook into the existing save/restore logic that we
use for dgpu, where the actual evict is skipped but on restore we
importantly restore the GGTT programming. This has been confirmed to
fix hibernation on at least ADL and MTL, though likely all igpu
platforms are affected.
This also means we have a hole in our testing, where the existing s4
tests only really test the driver hooks, and don't go as far as actually
rebooting and restoring from the hibernation image and in turn powering
down RAM (and therefore losing the contents of stolen).
v2 (Brost)
- Remove extra newline and drop unnecessary parentheses.
Fixes: dd08ebf6c3 ("drm/xe: Introduce a new DRM driver for Intel GPUs")
Link: https://gitlab.freedesktop.org/drm/xe/kernel/-/issues/3275
Signed-off-by: Matthew Auld <matthew.auld@intel.com>
Cc: Matthew Brost <matthew.brost@intel.com>
Cc: <stable@vger.kernel.org> # v6.8+
Reviewed-by: Matthew Brost <matthew.brost@intel.com>
Reviewed-by: Lucas De Marchi <lucas.demarchi@intel.com>
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241101170156.213490-2-matthew.auld@intel.com
(cherry picked from commit f2a6b8e396)
Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
This commit fixes the bug in the handling of partial mapping of the
buffer objects to the GPU, which caused kernel warnings.
Panthor didn't correctly handle the case where the partial mapping
spanned multiple scatterlists and the mapping offset didn't point
to the 1st page of starting scatterlist. The offset variable was
not cleared after reaching the starting scatterlist.
Following warning messages were seen.
WARNING: CPU: 1 PID: 650 at drivers/iommu/io-pgtable-arm.c:659 __arm_lpae_unmap+0x254/0x5a0
<snip>
pc : __arm_lpae_unmap+0x254/0x5a0
lr : __arm_lpae_unmap+0x2cc/0x5a0
<snip>
Call trace:
__arm_lpae_unmap+0x254/0x5a0
__arm_lpae_unmap+0x108/0x5a0
__arm_lpae_unmap+0x108/0x5a0
__arm_lpae_unmap+0x108/0x5a0
arm_lpae_unmap_pages+0x80/0xa0
panthor_vm_unmap_pages+0xac/0x1c8 [panthor]
panthor_gpuva_sm_step_unmap+0x4c/0xc8 [panthor]
op_unmap_cb.isra.23.constprop.30+0x54/0x80
__drm_gpuvm_sm_unmap+0x184/0x1c8
drm_gpuvm_sm_unmap+0x40/0x60
panthor_vm_exec_op+0xa8/0x120 [panthor]
panthor_vm_bind_exec_sync_op+0xc4/0xe8 [panthor]
panthor_ioctl_vm_bind+0x10c/0x170 [panthor]
drm_ioctl_kernel+0xbc/0x138
drm_ioctl+0x210/0x4b0
__arm64_sys_ioctl+0xb0/0xf8
invoke_syscall+0x4c/0x110
el0_svc_common.constprop.1+0x98/0xf8
do_el0_svc+0x24/0x38
el0_svc+0x34/0xc8
el0t_64_sync_handler+0xa0/0xc8
el0t_64_sync+0x174/0x178
<snip>
panthor : [drm] drm_WARN_ON(unmapped_sz != pgsize * pgcount)
WARNING: CPU: 1 PID: 650 at drivers/gpu/drm/panthor/panthor_mmu.c:922 panthor_vm_unmap_pages+0x124/0x1c8 [panthor]
<snip>
pc : panthor_vm_unmap_pages+0x124/0x1c8 [panthor]
lr : panthor_vm_unmap_pages+0x124/0x1c8 [panthor]
<snip>
panthor : [drm] *ERROR* failed to unmap range ffffa388f000-ffffa3890000 (requested range ffffa388c000-ffffa3890000)
Fixes: 647810ec24 ("drm/panthor: Add the MMU/VM logical block")
Signed-off-by: Akash Goel <akash.goel@arm.com>
Reviewed-by: Liviu Dudau <liviu.dudau@arm.com>
Reviewed-by: Steven Price <steven.price@arm.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241111134720.780403-1-akash.goel@arm.com
Signed-off-by: Liviu Dudau <liviu.dudau@arm.com>
