msm-next pull request has a baseline with stuff from -fixes, roll
forward first.
Some simple conflicts in amdgpu, ttm and one in i915 where git gets
confused and tries to add the same function twice.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
[Why]
Previous tdr design treats the first job in job_timeout as the bad job.
But sometimes a later bad compute job can block a good gfx job and
cause an unexpected gfx job timeout because gfx and compute ring share
internal GC HW mutually.
[How]
This patch implements an advanced tdr mode.It involves an additinal
synchronous pre-resubmit step(Step0 Resubmit) before normal resubmit
step in order to find the real bad job.
1. At Step0 Resubmit stage, it synchronously submits and pends for the
first job being signaled. If it gets timeout, we identify it as guilty
and do hw reset. After that, we would do the normal resubmit step to
resubmit left jobs.
2. For whole gpu reset(vram lost), do resubmit as the old way.
v2: squash in build fix (Alex)
Signed-off-by: Jack Zhang <Jack.Zhang1@amd.com>
Reviewed-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Since we're about to move drm_dp_helper.c over to drm_dbg_*(), we'll want
to make sure that we can also add ratelimited versions of these macros in
order to retain some of the previous debugging output behavior we had.
However, as I was preparing to do this I noticed that the current
rate limited macros we have are kind of bogus. It looks like when I wrote
these, I didn't notice that we'd always be calling __ratelimit() even if
the debugging message we'd be printing would normally be filtered out due
to the relevant DRM debugging category being disabled.
So, let's fix this by making sure to check drm_debug_enabled() in our
ratelimited macros before calling __ratelimit(), and start using
drm_dev_printk() in order to print debugging messages since that will save
us from doing a redundant drm_debug_enabled() check. And while we're at it,
let's move the code for this into another macro that we can reuse for
defining new ratelimited DRM debug macros more easily.
v2:
* Make sure to use tabs where possible in __DRM_DEFINE_DBG_RATELIMITED()
Signed-off-by: Lyude Paul <lyude@redhat.com>
Cc: Robert Foss <robert.foss@linaro.org>
Reviewed-by: Robert Foss <robert.foss@linaro.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20210326203807.105754-8-lyude@redhat.com
* Make sure that struct members are referred to using @, otherwise they
won't be formatted as such
* Make sure to refer to other struct types using & so they link back to
each struct's definition
* Make sure to precede constant values with % so they're formatted
correctly
Signed-off-by: Lyude Paul <lyude@redhat.com>
Acked-by: Randy Dunlap <rdunlap@infradead.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20210326203807.105754-2-lyude@redhat.com
Iterating DisplayID blocks across sections (in EDID extensions) is
unnecessarily complicated for the caller. Implement DisplayID iterators
to go through all blocks in all sections.
Usage example:
const struct displayid_block *block;
struct displayid_iter iter;
displayid_iter_edid_begin(edid, &iter);
displayid_iter_for_each(block, &iter) {
/* operate on block */
}
displayid_iter_end(&iter);
When DisplayID is stored in EDID extensions, the DisplayID sections map
to extensions as described in VESA DisplayID v1.3 Appendix B: DisplayID
as an EDID Extension. This is implemented here.
When DisplayID is stored in its dedicated DDC device 0xA4, according to
VESA E-DDC v1.3, different rules apply for the structure. This is not
implemented here, as we don't currently use it, but the idea is you'd
have a different call for beginning the iteration, for example simply:
displayid_iter_begin(displayid, &iter);
instead of displayid_iter_edid_begin(), and everything else would be
hidden away in the iterator functions.
v2:
- sizeof(struct displayid_block) -> sizeof(*block) (Ville)
- remove __ prefix from displayid_iter_block
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/da3dead1752ab16c061f7bd248ac1a4268f7fefb.1617024940.git.jani.nikula@intel.com
We'll be adding more DisplayID specific functions going forward, so
start off by splitting out a few functions to a separate file.
We don't bother with exporting the functions; at least for now they
should be needed solely within drm.ko.
No functional changes.
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/07942d5011891b8e8f77245c78b34f4af97a9315.1617024940.git.jani.nikula@intel.com
As documented in HDCP 2.2 DP Errata spec transmitter should abort the
authentication protocol in case transmitter has not received the
entire {AKE_Send_Cert, AKE_Send_H_prime, AKE_Send_Paring_Info} msg
within {110,7,5} miliseconds.
Adding above msg timeout values and aborting the HDCP authentication
in case it timedout to read entire msg.
https://www.digital-cp.com/sites/default/files/HDCP%202_2_DisplayPort_Errata_v3_0.pdf
v2:
- Removed redundant variable msg_can_timedout. [Ankit]
Cc: Ramalingam C <ramalingam.c@intel.com>
Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com>
Reviewed-by: Ankit Nautiyal <ankit.k.nautiyal@intel.com>
Acked-by: Maxime Ripard <mripard@kernel.org>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210324113012.7564-2-anshuman.gupta@intel.com
Currently the FRL training mode (Concurrent, Sequential) and
training type (Normal, Extended) are not defined properly and
are passed as bool values in drm_helpers for pcon
configuration for FRL training.
This patch:
-Add register masks for Sequential and Normal FRL training options.
-Fixes the drm_helpers for FRL Training configuration to use the
appropriate masks.
-Modifies the calls to the above drm_helpers in i915/intel_dp as per
the above change.
v2: Re-used the register masks for these options, instead of enum. (Ville)
Signed-off-by: Ankit Nautiyal <ankit.k.nautiyal@intel.com>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Acked-by: Maxime Ripard <mripard@kernel.org>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210323112422.1211-2-ankit.k.nautiyal@intel.com
This performs the same operation as drmm_encoder_alloc(), but
only allocates and returns a struct drm_encoder instance.
v4: Rename macro drmm_plain_encoder_alloc() and move to
<drm/drm_encoder.h>. Since it's not "simple" anymore it
will now take funcs/name arguments as well.
Signed-off-by: Paul Cercueil <paul@crapouillou.net>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210327115742.18986-3-paul@crapouillou.net
The shrinker based approach still has some flaws. Especially that we need
temporary pages to free up the pages allocated to the driver is problematic
in a shrinker.
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20210324134845.2338-1-christian.koenig@amd.com
amd-drm-next-5.13-2021-03-23:
amdgpu:
- Debugfs cleanup
- Various cleanups and spelling fixes
- Flexible array cleanups
- Initial AMD Freesync HDMI
- Display fixes
- 10bpc dithering improvements
- Display ASSR support
- Clean up and unify powerplay and swsmu interfaces
- Vangogh fixes
- Add SMU gfx busy queues for RV/PCO
- PCIE DPM fixes
- S0ix fixes
- GPU metrics data fixes
- DCN secure display support
- Backlight type override
- Add initial support for Aldebaran
- RAS fixes
- Prime fixes for A+A systems
- Reset fixes
- Initial resource cursor support
- Drop legacy IO BAR requirements
- Various power fixes
amdkfd:
- MMU notifier fixes
- APU fixes
radeon:
- Debugfs cleanups
- Flexible array cleanups
UAPI:
- amdgpu: Add a new INFO ioctl interface to query video capabilities
rather than hardcoding them in userspace. This allows us to provide
fine grained asic capabilities (e.g., if a particular part is
bandwidth limited, we can limit the capabilities). Proposed userspace:
https://gitlab.freedesktop.org/leoliu/drm/-/commits/info_video_capshttps://gitlab.freedesktop.org/leoliu/mesa/-/commits/info_video_caps
- amdkfd: bump the driver version. There was a problem with reporting
some RAS features on older versions of the driver. Proposed userspace:
7cdd63475c
Danvet: A bunch of conflicts all over, but it seems to compile ... I
did put the call to dc_allow_idle_optimizations() on a single line
since it looked a bit too jarring to be left alone.
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210324040147.1990338-1-alexander.deucher@amd.com
Fix the following typos:
1. When mentioning a list of functions, the function
drm_atomic_helper_disable_plane is mentioned twice.
2. drop the word 'afterwards':
s/afterwards after that/after that/'
3. drop extra 'the':
s/but do not the support the full/but do not support the full/
Signed-off-by: Dafna Hirschfeld <dafna.hirschfeld@collabora.com>
Reviewed-by: Simon Ser <contact@emersion.fr>
Signed-off-by: Simon Ser <contact@emersion.fr>
Link: https://patchwork.freedesktop.org/patch/msgid/20210326103216.7918-1-dafna.hirschfeld@collabora.com
Instead of having a global lock for potentially less contention.
Signed-off-by: Christian König <christian.koenig@amd.com>
Tested-by: Nirmoy Das <nirmoy.das@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/424010/
Instead evict round robin from each devices SYSTEM and TT domain.
v2: reorder num_pages access reported by Dan's script
v3: fix rebase fallout, num_pages should be 32bit
Signed-off-by: Christian König <christian.koenig@amd.com>
Tested-by: Nirmoy Das <nirmoy.das@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/424009/
Move the iteration of the global lru into the new function
ttm_global_swapout() and use that instead in drivers.
v2: consistently return int
v3: fix build fail
Signed-off-by: Christian König <christian.koenig@amd.com>
Tested-by: Nirmoy Das <nirmoy.das@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/424008/
Highlights:
- Alderlake S enabling, via topic branch (Aditya, Anusha, Caz, José, Lucas, Matt, Tejas)
- Refactor display code to shrink intel_display.c etc. (Dave)
- Support more gen 9 and Tigerlake PCH combinations (Lyude, Tejas)
- Add eDP MSO support (Jani)
Display:
- Refactor to support multiple PSR instances (Gwan-gyeong)
- Link training debug logging updates (Sean)
- Updates to eDP fixed mode handling (Jani)
- Disable PSR2 on JSL/EHL (Edmund)
- Support DDR5 and LPDDR5 for bandwidth computation (Clint, José)
- Update VBT DP max link rate table (Shawn)
- Disable the QSES check for HDCP2.2 over MST (Juston)
- PSR updates, refactoring, selective fetch (José, Gwan-gyeong)
- Display init sequence refactoring (Lucas)
- Limit LSPCON to gen 9 and 10 platforms (Ankit)
- Fix DDI lane polarity per VBT info (Uma)
- Fix HDMI vswing programming location in mode set (Ville)
- Various display improvements and refactorings and cleanups (Ville)
- Clean up DDI clock routing and readout (Ville)
- Workaround async flip + VT-d corruption on HSW/BDW (Ville)
- SAGV watermark fixes and cleanups (Ville)
- Silence pipe tracepoint WARNs (Ville)
Other:
- Remove require_force_probe protection from RKL, may need to be revisited (Tejas)
- Detect loss of MMIO access (Matt)
- GVT display improvements
- drm/i915: Disable runtime power management during shutdown (Imre)
- Perf/OA updates (Umesh)
- Remove references to struct drm_device.pdev, via topic branch (Thomas)
- Backmerge (Jani)
Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/87v99rnk1g.fsf@intel.com
This adds a USB display driver with the intention that it can be
used with future USB interfaced low end displays/adapters. The Linux
gadget device driver will serve as the canonical device implementation.
The following DRM properties are supported:
- Plane rotation
- Connector TV properties
There is also support for backlight brightness exposed as a backlight
device.
Display modes can be made available to the host driver either as DRM
display modes or through EDID. If both are present, EDID is just passed
on to userspace.
Performance is preferred over color depth, so if the device supports
RGB565, DRM_CAP_DUMB_PREFERRED_DEPTH will return 16.
If the device transfer buffer can't fit an uncompressed framebuffer
update, the update is split up into parts that do fit.
Optimal user experience is achieved by providing damage reports either by
setting FB_DAMAGE_CLIPS on pageflips or calling DRM_IOCTL_MODE_DIRTYFB.
LZ4 compression is used if the device supports it.
The driver supports a one bit monochrome transfer format: R1. This is not
implemented in the gadget driver. It is added in preparation for future
monochrome e-ink displays.
The driver is MIT licensed to smooth the path for any BSD port of the
driver.
v2:
- Use devm_drm_dev_alloc() and drmm_mode_config_init()
- drm_fbdev_generic_setup: Use preferred_bpp=0, 16 was a copy paste error
- The drm_backlight_helper is dropped, copy in the code
- Support protocol version backwards compatibility for device
v3:
- Use donated Openmoko USB pid
- Use direct compression from framebuffer when pitch matches, not only on
full frames, so split updates can benefit
- Use __le16 in struct gud_drm_req_get_connector_status
- Set edid property when the device only provides edid
- Clear compression fields in struct gud_drm_req_set_buffer
- Fix protocol version negotiation
- Remove mode->vrefresh, it's calculated
v4:
- Drop the status req polling which was a workaround for something that
turned out to be a dwc2 udc driver problem
- Add a flag for the Linux gadget to require a status request on
SET operations. Other devices will only get status req on STALL errors
- Use protocol specific error codes (Peter)
- Add a flag for devices that want to receive the entire framebuffer on
each flush (Lubomir)
- Retry a failed framebuffer flush
- If mode has changed wait for worker and clear pending damage before
queuing up new damage, fb width/height might have changed
- Increase error counter on bulk transfer failures
- Use DRM_MODE_CONNECTOR_USB
- Handle R1 kmalloc error (Peter)
- Don't try and replicate the USB get descriptor request standard for the
display descriptor (Peter)
- Make max_buffer_size optional (Peter), drop the pow2 requirement since
it's not necessary anymore.
- Don't pre-alloc a control request buffer, it was only 4k
- Let gud.h describe the whole protocol explicitly and don't let DRM
leak into it (Peter)
- Drop display mode .hskew and .vscan from the protocol
- Shorten names: s/GUD_DRM_/GUD_/ s/gud_drm_/gud_/ (Peter)
- Fix gud_pipe_check() connector picking when switching connector
- Drop gud_drm_driver_gem_create_object() cached is default now
- Retrieve USB device from struct drm_device.dev instead of keeping a
pointer
- Honour fb->offsets[0]
- Fix mode fetching when connector status is forced
- Check EDID length reported by the device
- Use drm_do_get_edid() so userspace can overrride EDID
- Set epoch counter to signal connector status change
- gud_drm_driver can be const now
v5:
- GUD_DRM_FORMAT_R1: Use non-human ascii values (Daniel)
- Change name to: GUD USB Display (Thomas, Simon)
- Change one __u32 -> __le32 in protocol header
- Always log fb flush errors, unless the previous one failed
- Run backlight update in a worker to avoid upsetting lockdep (Daniel)
- Drop backlight_ops.get_brightness, there's no readback from the device
so it doesn't really add anything.
- Set dma mask, needed by dma-buf importers
v6:
- Use obj-y in Makefile (Peter)
- Fix missing le32_to_cpu() when using GUD_DISPLAY_MAGIC (Peter)
- Set initial brightness on backlight device
v7:
- LZ4_compress_default() can return zero, check for that
- Fix memory leak in gud_pipe_check() error path (Peter)
- Improve debug and error messages (Peter)
- Don't pass length in protocol structs (Peter)
- Pass USB interface to gud_usb_control_msg() et al. (Peter)
- Improve gud_connector_fill_properties() (Peter)
- Add GUD_PIXEL_FORMAT_RGB111 (Peter)
- Remove GUD_REQ_SET_VERSION (Peter)
- Fix DRM_IOCTL_MODE_OBJ_SETPROPERTY and the rotation property
- Fix dma-buf import (Thomas)
v8:
- Forgot to filter RGB111 from reaching userspace
- Handle a device that only returns unknown device properties (Peter)
- s/GUD_PIXEL_FORMAT_RGB111/GUD_PIXEL_FORMAT_XRGB1111/ (Peter)
- Fix R1 and XRGB1111 format conversion
- Add FIXME about Big Endian being broken (Peter, Ilia)
Cc: Lubomir Rintel <lkundrak@v3.sk>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Peter Stuge <peter@stuge.se>
Tested-by: Peter Stuge <peter@stuge.se>
Signed-off-by: Noralf Trønnes <noralf@tronnes.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20210313112545.37527-4-noralf@tronnes.org
Add aldebaran in amdgpu_asic_name array and amdgpu_asic_type enum
Signed-off-by: Le Ma <le.ma@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Acked-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The current atomic helpers have either their object state being passed as
an argument or the full atomic state.
The former is the pattern that was done at first, before switching to the
latter for new hooks or when it was needed.
Let's convert all the remaining helpers to provide a consistent
interface, starting with the planes atomic_check.
The conversion was done using the coccinelle script below plus some
manual changes for vmwgfx, built tested on all the drivers.
@@
identifier plane, plane_state;
symbol state;
@@
struct drm_plane_helper_funcs {
...
int (*atomic_check)(struct drm_plane *plane,
- struct drm_plane_state *plane_state);
+ struct drm_atomic_state *state);
...
}
@ plane_atomic_func @
identifier helpers;
identifier func;
@@
static const struct drm_plane_helper_funcs helpers = {
...,
.atomic_check = func,
...,
};
@@
struct drm_plane_helper_funcs *FUNCS;
identifier f;
identifier dev;
identifier plane, plane_state, state;
@@
f(struct drm_device *dev, struct drm_atomic_state *state)
{
<+...
- FUNCS->atomic_check(plane, plane_state)
+ FUNCS->atomic_check(plane, state)
...+>
}
@ ignores_new_state @
identifier plane_atomic_func.func;
identifier plane, new_plane_state;
@@
func(struct drm_plane *plane, struct drm_plane_state *new_plane_state)
{
... when != new_plane_state
}
@ adds_new_state depends on plane_atomic_func && !ignores_new_state @
identifier plane_atomic_func.func;
identifier plane, new_plane_state;
@@
func(struct drm_plane *plane, struct drm_plane_state *new_plane_state)
{
+ struct drm_plane_state *new_plane_state = drm_atomic_get_new_plane_state(state, plane);
...
}
@ depends on plane_atomic_func @
identifier plane_atomic_func.func;
identifier plane, new_plane_state;
@@
func(struct drm_plane *plane,
- struct drm_plane_state *new_plane_state
+ struct drm_atomic_state *state
)
{ ... }
@ include depends on adds_new_state @
@@
#include <drm/drm_atomic.h>
@ no_include depends on !include && adds_new_state @
@@
+ #include <drm/drm_atomic.h>
#include <drm/...>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Acked-by: Thomas Zimmermann <tzimmermann@suse.de>
Link: https://patchwork.freedesktop.org/patch/msgid/20210219120032.260676-4-maxime@cerno.tech
The explicit out-fences in crtc are signaled as part of vblank event,
indicating all framebuffers present on the Atomic Commit request are
scanned out on the screen. Though the fence signal and the vblank event
notification happens at the same time, triggered by the same hardware
vsync event, the timestamp set in both are different. With drivers
supporting precise vblank timestamp the difference between the two
timestamps would be even higher. This might have an impact on use-mode
frameworks using these fence timestamps for purposes other than simple
buffer usage. For instance, the Android framework [1] uses the
retire-fences as an alternative to vblank when frame-updates are in
progress. Set the fence timestamp during send vblank event using a new
drm_send_event_timestamp_locked variant to avoid discrepancies.
[1] https://android.googlesource.com/platform/frameworks/native/+/master/
services/surfaceflinger/Scheduler/Scheduler.cpp#397
Changes in v2:
- Use drm_send_event_timestamp_locked to update fence timestamp
- add more information to commit text
Changes in v3:
- use same backend helper function for variants of drm_send_event to
avoid code duplications
Changes in v4:
- remove WARN_ON from drm_send_event_timestamp_locked
Signed-off-by: Veera Sundaram Sankaran <veeras@codeaurora.org>
Reviewed-by: John Stultz <john.stultz@linaro.org>
Signed-off-by: Sumit Semwal <sumit.semwal@linaro.org>
[sumits: minor parenthesis alignment correction]
Link: https://patchwork.freedesktop.org/patch/msgid/1610757107-11892-2-git-send-email-veeras@codeaurora.org
(cherry picked from commit a78e7a51d2)
Signed-off-by: Sumit Semwal <sumit.semwal@linaro.org>
The function drm_gem_fb_prepare_fb() is a helper for atomic modesetting,
but currently located next to framebuffer helpers. Move it to GEM atomic
helpers, rename it slightly and adopt the drivers. Same for the rsp
simple-pipe helper.
Compile-tested with x86-64, aarch64 and arm. The patch is fairly large,
but there are no functional changes.
v3:
* remove out-comented line in drm_gem_framebuffer_helper.h
(Maxime)
v2:
* rename to drm_gem_plane_helper_prepare_fb() (Daniel)
* add tutorial-style documentation
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Maxime Ripard <mripard@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20210222141756.7864-1-tzimmermann@suse.de
This reverts commit c10983e14e.
This commit is not meant for drm-misc-next-fixes, and was accidentally
cherry picked over.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
I got real badly confused when trying to review a fix from Ville for
this. Let's try to document better what's required for this, and check
the minimal settings at runtime - we can't check ofc that there's
indeed no races in the driver callback.
Also noticed that the drm_vblank_restore version is unused, so lets
unexport that while at it.
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Cc: Ville Syrjala <ville.syrjala@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210209101523.2954281-1-daniel.vetter@ffwll.ch
This is just another feature which is only used by VMWGFX, so move
it into the driver instead.
I've tried to add the accounting sysfs file to the kobject of the drm
minor, but I'm not 100% sure if this works as expected.
v2: fix typo in KFD and avoid 64bit divide
v3: fix init order in VMWGFX
v4: use pdev sysfs reference instead of drm
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Zack Rusin <zackr@vmware.com> (v3)
Tested-by: Nirmoy Das <nirmoy.das@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210208133226.36955-2-christian.koenig@amd.com
TTM implements a rather extensive accounting of allocated memory.
There are two reasons for this:
1. It tries to block userspace allocating a huge number of very small
BOs without accounting for the kmalloced memory.
2. Make sure we don't over allocate and run into an OOM situation
during swapout while trying to handle the memory shortage.
This is only partially a good idea. First of all it is perfectly
valid for an application to use all of system memory, limiting it to
50% is not really acceptable.
What we need to take care of is that the application is held
accountable for the memory it allocated. This is what control
mechanisms like memcg and the normal Linux page accounting already do.
Making sure that we don't run into an OOM situation while trying to
cope with a memory shortage is still a good idea, but this is also
not very well implemented since it means another opportunity of
recursion from the driver back into TTM.
So start to rework all of this by implementing a shrinker callback which
allows for TT object to be swapped out if necessary.
v2: Switch from limit to shrinker callback.
v3: fix gfp mask handling, use atomic for swapable_pages, add debugfs
v4: drop the extra gfp_mask checks
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20210208133226.36955-1-christian.koenig@amd.com
This patch does not change current behaviour.
The driver's job timeout handler now returns
status indicating back to the DRM layer whether
the device (GPU) is no longer available, such as
after it's been unplugged, or whether all is
normal, i.e. current behaviour.
All drivers which make use of the
drm_sched_backend_ops' .timedout_job() callback
have been accordingly renamed and return the
would've-been default value of
DRM_GPU_SCHED_STAT_NOMINAL to restart the task's
timeout timer--this is the old behaviour, and is
preserved by this patch.
v2: Use enum as the status of a driver's job
timeout callback method.
v3: Return scheduler/device information, rather
than task information.
Cc: Alexander Deucher <Alexander.Deucher@amd.com>
Cc: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Cc: Christian König <christian.koenig@amd.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Lucas Stach <l.stach@pengutronix.de>
Cc: Russell King <linux+etnaviv@armlinux.org.uk>
Cc: Christian Gmeiner <christian.gmeiner@gmail.com>
Cc: Qiang Yu <yuq825@gmail.com>
Cc: Rob Herring <robh@kernel.org>
Cc: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Cc: Steven Price <steven.price@arm.com>
Cc: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Cc: Eric Anholt <eric@anholt.net>
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Luben Tuikov <luben.tuikov@amd.com>
Acked-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Acked-by: Christian König <christian.koenig@amd.com>
Acked-by: Steven Price <steven.price@arm.com>
Signed-off-by: Christian König <christian.koenig@amd.com>
Link: https://patchwork.freedesktop.org/patch/415095/
(cherry picked from commit a6a1f036c7)
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Export the helpers for shadow-buffered planes. These will be used by
several drivers.
v3:
* fix documentation typos and formatting (Daniel)
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Tested-by: Hans de Goede <hdegoede@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210209121042.24098-2-tzimmermann@suse.de
Several drivers use GEM buffer objects as shadow buffers for the actual
framebuffer memory. Right now, drivers do these vmap operations in their
commit tail, which is actually not allowed by the locking rules for
the dma-buf reservation lock. The involved BO has to be vmapped in the
plane's prepare_fb callback and vunmapped in cleanup_fb.
This patch introduces atomic helpers for such shadow planes. Plane
functions manage the plane state for shadow planes. The provided
implementations for prepare_fb and cleanup_fb vmap and vunmap all BOs of
struct drm_plane_state.fb. The mappings are afterwards available in the
plane's commit-tail functions.
For now, all rsp drivers use the simple KMS helpers, so we add the plane
callbacks and wrappers for simple KMS. The internal plane functions can
later be exported as needed.
v3:
* documentation fixes
v2:
* make duplicate_state interface compatible with
struct drm_plane_funcs
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Tested-by: Gerd Hoffmann <kraxel@redhat.com>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210208115538.6430-4-tzimmermann@suse.de
Just like regular plane-state helpers, drivers can use these new
callbacks to create and destroy private plane state.
v2:
* make duplicate_state interface compatible with
struct drm_plane_funcs
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Tested-by: Gerd Hoffmann <kraxel@redhat.com>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210208115538.6430-2-tzimmermann@suse.de
- Prevent double YUV range correction on HDR planes (Andres)
- DP MST related Fixes (Sean, Imre)
- More clean-up around DRAM detection code (Jose)
- Actually async flips enable for all ilk+ platforms (Ville)
-----BEGIN PGP SIGNATURE-----
iQEzBAABCAAdFiEEbSBwaO7dZQkcLOKj+mJfZA7rE8oFAmAUkUIACgkQ+mJfZA7r
E8rdqAgArg9X7LaZ9pnrPNDK2CSIZ7OSFmLjqYePDlKM+13TstM0NgMzMMCad7oM
0687oOaoT8NTaTvs5mQOG1TMWm8EfFsm1hzY8SRBDqHIjROUQWjYJRU202yWB2Ik
Od5rsJ7U+RiVmdTCU0vPahSCSgqLS7s9/Cj0uxjP5q7Fpg8ab7R5pQL3a/XFqRFY
VEynVt9GyGV/0bD7S8n6sJTKKMtdjWif1t6OiIdYZeU2UQY9juaIs+WuC6HaL+Zr
mq2p1zUJfpEK1r7GupWUHQm/bhwhaBVbuqvhWZFFyyv5gyyHY8dLTXmAgZioMEJT
F0TZR5zMCuBFJ0BSo7IuhJnJsBSHtQ==
=VcVL
-----END PGP SIGNATURE-----
Merge tag 'drm-intel-next-2021-01-29' of git://anongit.freedesktop.org/drm/drm-intel into drm-next
- WARN if plane src coords are too big (Ville)
- Prevent double YUV range correction on HDR planes (Andres)
- DP MST related Fixes (Sean, Imre)
- More clean-up around DRAM detection code (Jose)
- Actually async flips enable for all ilk+ platforms (Ville)
Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210129225328.GA1041349@intel.com
This patch does not change current behaviour.
The driver's job timeout handler now returns
status indicating back to the DRM layer whether
the device (GPU) is no longer available, such as
after it's been unplugged, or whether all is
normal, i.e. current behaviour.
All drivers which make use of the
drm_sched_backend_ops' .timedout_job() callback
have been accordingly renamed and return the
would've-been default value of
DRM_GPU_SCHED_STAT_NOMINAL to restart the task's
timeout timer--this is the old behaviour, and is
preserved by this patch.
v2: Use enum as the status of a driver's job
timeout callback method.
v3: Return scheduler/device information, rather
than task information.
Cc: Alexander Deucher <Alexander.Deucher@amd.com>
Cc: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Cc: Christian König <christian.koenig@amd.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Lucas Stach <l.stach@pengutronix.de>
Cc: Russell King <linux+etnaviv@armlinux.org.uk>
Cc: Christian Gmeiner <christian.gmeiner@gmail.com>
Cc: Qiang Yu <yuq825@gmail.com>
Cc: Rob Herring <robh@kernel.org>
Cc: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Cc: Steven Price <steven.price@arm.com>
Cc: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Cc: Eric Anholt <eric@anholt.net>
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Luben Tuikov <luben.tuikov@amd.com>
Acked-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Acked-by: Christian König <christian.koenig@amd.com>
Acked-by: Steven Price <steven.price@arm.com>
Signed-off-by: Christian König <christian.koenig@amd.com>
Link: https://patchwork.freedesktop.org/patch/415095/
- Fix DP vswing settings and handling (Imre, Ville)
- Various display code clean-up (Jani, Ville)
- Various display refactoring, including split out of pps, aux, and fdi (Ja\
ni, Dave)
- Add DG1 missing workarounds (Jose)
- Fix display color conversion (Chris, Ville)
- Try to guess PCH type even without ISA bridge (Zhenyu)
- More backlight refactor (Lyude)
- Support two CSC module on gen11 and later (Lee)
- Async flips for all ilk+ platforms (Ville)
- Clear color support for TGL (RK)
- Add a helper to read data from a GEM object page (Imre)
- VRR/Adaptive Sync Enabling on DP/eDP for TGL+ (Manasi, Ville Aditya)
-----BEGIN PGP SIGNATURE-----
iQEzBAABCAAdFiEEbSBwaO7dZQkcLOKj+mJfZA7rE8oFAmARc1oACgkQ+mJfZA7r
E8ovFAgAhBlnhrquGrgtuRmS5tkpLQShqiwrPv7ogDBkcayJ6Is+AO+FoE8kuzZO
q/TW2enkcRKFHhQnm9RA0Ayk+rwYqkEOcBU5gvzusdwbEjkygJfH/yA2ukAmaw32
QxP9A2pW+iWsEcqtMuOEu1ahu0Z8gnkCe8DakuBMf/0y5D9DairwaCC8nyZQdEJq
YFTxp+pILE57PhOVbKS3EZPbnl5t28NgCtMc8hjlPTWZxRSVV0eDrY2tx9vyhFhU
Eq8aLDxPKCRm18sg+J+lDytOvFoljXcglAEwPODVMmDNolG8R6aRQ9T6tG6qfxyI
UO1KyeYUHHEIJ4IBJKK+KWeUWLLBhA==
=3Jc8
-----END PGP SIGNATURE-----
Merge tag 'drm-intel-next-2021-01-27' of git://anongit.freedesktop.org/drm/drm-intel into drm-next
- HDCP 2.2 and HDCP 1.4 Gen12 DP MST support (Anshuman)
- Fix DP vswing settings and handling (Imre, Ville)
- Various display code clean-up (Jani, Ville)
- Various display refactoring, including split out of pps, aux, and fdi (Ja\
ni, Dave)
- Add DG1 missing workarounds (Jose)
- Fix display color conversion (Chris, Ville)
- Try to guess PCH type even without ISA bridge (Zhenyu)
- More backlight refactor (Lyude)
- Support two CSC module on gen11 and later (Lee)
- Async flips for all ilk+ platforms (Ville)
- Clear color support for TGL (RK)
- Add a helper to read data from a GEM object page (Imre)
- VRR/Adaptive Sync Enabling on DP/eDP for TGL+ (Manasi, Ville Aditya)
Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210127140822.GA711686@intel.com
This function will be needed by the next patch where the driver
calculates the BW based on driver specific parameters, so export it.
At the same time sanitize the function params, passing the more natural
link rate instead of the encoding of the same rate.
v2:
- Fix function documentation. (Lyude)
Cc: Lyude Paul <lyude@redhat.com>
Cc: Ville Syrjala <ville.syrjala@intel.com>
Cc: <stable@vger.kernel.org>
Cc: dri-devel@lists.freedesktop.org
Signed-off-by: Imre Deak <imre.deak@intel.com>
Reviewed-by: Lyude Paul <lyude@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210125173636.1733812-1-imre.deak@intel.com
Daniel Vetter
Jack Zhang
Dave Airlie
Lyude Paul
Christian König
Jani Nikula
Anshuman Gupta
Ankit Nautiyal
Paul Cercueil
Dafna Hirschfeld
Noralf Trønnes
Le Ma
Maxime Ripard
Veera Sundaram Sankaran
Thomas Zimmermann
Maarten Lankhorst
Luben Tuikov
Imre Deak