drm/amd/display: Create dc_sink when EDID fail

[Why]
While reading remote EDID via Startech 1-to-4 hub, occasionally we
won't get response in time and won't light up corresponding monitor.

Ideally, we can still add generic modes for userspace to choose to try
to light up the monitor and which is done in
drm_helper_probe_single_connector_modes(). So the main problem here is
that we fail .mode_valid since we don't create remote dc_sink for this
case.

[How]
Also add default dc_sink if we can't get the EDID.

Reviewed-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Signed-off-by: Wayne Lin <Wayne.Lin@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: correct the address of Arcturus fan related registers

These registers have different address from other SMU V11 ASICs.

Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: drop unnecessary manual mode check

As the fan control was guarded under manual mode before fan speed
RPM/PWM setting. Thus the extra check is totally redundant.

Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: drop the unnecessary intermediate percent-based transition

Currently, the readout of fan speed pwm is transited into percent-based
and then pwm-based. However, the transition into percent-based is totally
unnecessary and make the final output less accurate.

Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: correct the fan speed RPM retrieving

The relationship "PWM = RPM / smu->fan_max_rpm" between fan speed
PWM and RPM is not true for SMU11 ASICs. So, we need a new way to
retrieving the fan speed RPM.

Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: correct the fan speed PWM retrieving

The relationship "PWM = RPM / smu->fan_max_rpm" between fan speed
PWM and RPM is not true for SMU11 ASICs. So, we need a new way to
retrieving the fan speed PWM.

Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: record the RPM and PWM based fan speed settings

As the relationship "PWM = RPM / smu->fan_max_rpm" between fan speed
PWM and RPM is not true for SMU11 ASICs. So, both the RPM and PWM
settings need to be saved.

Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: correct the fan speed RPM setting

The relationship "PWM = RPM / smu->fan_max_rpm" between fan speed
PWM and RPM is not true for SMU11 ASICs. So, we need a new way to
perform the fan speed RPM setting.

Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/amdgpu: remove unnecessary RAS context field

Delete ras_if->name in the RAS ctx structure and remove related lines.

Signed-off-by: Candice Li <candice.li@amd.com>
Reviewed-by: John Clements <john.clements@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdkfd: fix random KFDSVMRangeTest.SetGetAttributesTest test failure

KFDSVMRangeTest.SetGetAttributesTest randomly fails in stress test.

Note: Google Test filter = KFDSVMRangeTest.*
[==========] Running 18 tests from 1 test case.
[----------] Global test environment set-up.
[----------] 18 tests from KFDSVMRangeTest
[ RUN      ] KFDSVMRangeTest.BasicSystemMemTest
[       OK ] KFDSVMRangeTest.BasicSystemMemTest (30 ms)
[ RUN      ] KFDSVMRangeTest.SetGetAttributesTest
[          ] Get default atrributes
/home/yifan/brahma/libhsakmt/tests/kfdtest/src/KFDSVMRangeTest.cpp:154: Failure
Value of: expectedDefaultResults[i]
  Actual: 4294967295
Expected: outputAttributes[i].value
Which is: 0
/home/yifan/brahma/libhsakmt/tests/kfdtest/src/KFDSVMRangeTest.cpp:154: Failure
Value of: expectedDefaultResults[i]
  Actual: 4294967295
Expected: outputAttributes[i].value
Which is: 0
/home/yifan/brahma/libhsakmt/tests/kfdtest/src/KFDSVMRangeTest.cpp:152: Failure
Value of: expectedDefaultResults[i]
  Actual: 4
Expected: outputAttributes[i].type
Which is: 2
[          ] Setting/Getting atrributes
[  FAILED  ]

the root cause is that svm work queue has not finished when svm_range_get_attr is called, thus
some garbage svm interval tree data make svm_range_get_attr get wrong result. Flush work queue before
iterate svm interval tree.

Signed-off-by: Yifan Zhang <yifan1.zhang@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: change the workload type for some cards

change the workload type for some cards as it is needed.

Signed-off-by: Kenneth Feng <kenneth.feng@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

Revert "drm/amd/pm: fix workload mismatch on vega10"

This reverts commit 0979d43259e13846d86ba17e451e17fec185d240.
Revert this because it does not apply to all the cards.

Signed-off-by: Kenneth Feng <kenneth.feng@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/amdgpu: consolidate PSP TA context

Signed-off-by: Candice Li <candice.li@amd.com>
Reviewed-by: John Clements <john.clements@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Add MB_REQ_MSG_READY_TO_RESET response when VF get FLR notification.

When guest received FLR notification from host, it would
lock adapter into reset state. There will be no more
job submission and hardware access after that.

Then it should send a response to host that it has prepared
for host reset.

Signed-off-by: Jiange Zhao <Jiange.Zhao@amd.com>
Signed-off-by: Peng Ju Zhou <PengJu.Zhou@amd.com>
Reviewed-by: Emily.Deng <Emily.Deng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: change pp_dpm_sclk/mclk/fclk attribute is RO for aldebaran

the following clock is only support voltage DPM, change attribute to RO:
1. pp_dpm_sclk
2. pp_dpm_mclk
3. pp_dpm_fclk

Signed-off-by: Kevin Wang <kevin1.wang@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: change smu msg's attribute to allow working under sriov

the following message is allowed in sriov mode:
1. GetEnabledSmuFeaturesLow
2. GetEnabledSmuFeaturesHigh

Signed-off-by: Kevin Wang <kevin1.wang@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: change return value in aldebaran_get_power_limit()

v1:
1. change return value to avoid smu driver probe fails when FEATURE_PPT is
not enabled.
2. if FEATURE_PPT is not enabled, set power limit value to 0.

v2:
instead dev_err with dev_warn

Signed-off-by: Kevin Wang <kevin1.wang@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: skip to load smu microcode on sriov for aldebaran

v1:
1. skip to load smu firmware in sriov mode for aldebaran chip
2. using vbios pptable if in sriov mode.

v2:
clean up smu driver code in sriov code path

Signed-off-by: Kevin Wang <kevin1.wang@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: correct DPM_XGMI/VCN_DPM feature name

the following feature is wrong, it will cause sysnode of pp_features show error:
1. DPM_XGMI
2. VCN_DPM

Signed-off-by: Kevin Wang <kevin1.wang@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/amdgpu embed hw_fence into amdgpu_job

Why: Previously hw fence is alloced separately with job.
It caused historical lifetime issues and corner cases.
The ideal situation is to take fence to manage both job
and fence's lifetime, and simplify the design of gpu-scheduler.

How:
We propose to embed hw_fence into amdgpu_job.
1. We cover the normal job submission by this method.
2. For ib_test, and submit without a parent job keep the
legacy way to create a hw fence separately.
v2:
use AMDGPU_FENCE_FLAG_EMBED_IN_JOB_BIT to show that the fence is
embedded in a job.
v3:
remove redundant variable ring in amdgpu_job
v4:
add tdr sequence support for this feature. Add a job_run_counter to
indicate whether this job is a resubmit job.
v5
add missing handling in amdgpu_fence_enable_signaling

Signed-off-by: Jingwen Chen <Jingwen.Chen2@amd.com>
Signed-off-by: Jack Zhang <Jack.Zhang7@hotmail.com>
Reviewed-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Reviewed by: Monk Liu <monk.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/ttm: Include pagemap.h from ttm_tt.h

It's needed for pgprot_t which is used in the header.

Signed-off-by: Jason Ekstrand <jason@jlekstrand.net>
Cc: Christian König <christian.koenig@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210812203443.1725307-2-jason@jlekstrand.net
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Christian König <christian.koenig@amd.com>

drm/ttm: ttm_bo_device is now ttm_device

These names were changed in

commit 8af8a109b34fa88b8b91f25d11485b37d37549c3
Author: Christian König <christian.koenig@amd.com>
Date:   Thu Oct 1 14:51:40 2020 +0200

    drm/ttm: device naming cleanup

But he missed a couple of them.

Signed-off-by: Jason Ekstrand <jason@jlekstrand.net>
Cc: Christian König <christian.koenig@amd.com>
Fixes: 8af8a109b34f ("drm/ttm: device naming cleanup")
Link: https://patchwork.freedesktop.org/patch/msgid/20210812203443.1725307-1-jason@jlekstrand.net
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Christian König <christian.koenig@amd.com>

Merge tag 'drm-misc-next-2021-08-12' of git://anongit.freedesktop.org/drm/drm-misc into drm-next

drm-misc-next for v5.15:

UAPI Changes:

Cross-subsystem Changes:
- Add lockdep_assert(once) helpers.

Core Changes:
- Add lockdep assert to drm_is_current_master_locked.
- Fix typos in dma-buf documentation.
- Mark drm irq midlayer as legacy only.
- Fix GPF in udmabuf_create.
- Rename member to correct value in drm_edid.h

Driver Changes:
- Build fix to make nouveau build with NOUVEAU_BACKLIGHT.
- Add MI101AIT-ICP1, LTTD800480070-L6WWH-RT panels.
- Assorted fixes to bridge/it66121, anx7625.
- Add custom crtc_state to simple helpers, and use it to
  convert pll handling in mgag200 to atomic.
- Convert drivers to use offset-adjusted framebuffer bo mappings.
- Assorted small fixes and fix for a use-after-free in vmwgfx.
- Convert remaining callers of non-legacy drivers to use linux irqs directly.
- Small cleanup in ingenic.
- Small fixes to virtio and ti-sn65dsi86.

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1cf2d7fc-402d-1852-574a-21cbbd2eaebf@linux.intel.com

gpu: host1x: debug: Dump DMASTART and DMAEND register

Show the values of the DMASTART and DMAEND registers when dumping status
to help with failure analysis.

Signed-off-by: Thierry Reding <treding@nvidia.com>

gpu: host1x: debug: Dump only relevant parts of CDMA push buffer

Dumping the full CDMA push buffer takes a long time and isn't very
useful since most of the contents are not relevant. Instead only show
the CDMA push buffer entries associated with current jobs.

While at it, tweak the indentation a bit to make the output more
readable.

Signed-off-by: Thierry Reding <treding@nvidia.com>

gpu: host1x: debug: Use dma_addr_t more consistently

The host1x debug code uses a mix of phys_addr_t, dma_addr_t and u32 to
represent addresses. However, these addresses are always DMA addresses
so use the appropriate type.

This fixes some issues with how these addresses are displayed, because
they could be truncated in some cases and not show the full address.

Signed-off-by: Thierry Reding <treding@nvidia.com>

drm/tegra: dc: Extend debug stats with total number of events

It's useful to know the total number of underflow events and currently
the debug stats are getting reset each time CRTC is being disabled. Let's
account the overall number of events that doesn't get a reset.

Reviewed-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Signed-off-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>

drm/tegra: dc: Support memory bandwidth management

Display controller (DC) performs isochronous memory transfers, and thus,
has a requirement for a minimum memory bandwidth that shall be fulfilled,
otherwise framebuffer data can't be fetched fast enough and this results
in a DC's data-FIFO underflow that follows by a visual corruption.

The Memory Controller drivers provide facility for memory bandwidth
management via interconnect API. Let's wire up the interconnect API
support to the DC driver in order to fix the distorted display output
on T30 Ouya, T124 TK1 and other Tegra devices.

Tested-by: Peter Geis <pgwipeout@gmail.com> # Ouya T30
Tested-by: Matt Merhar <mattmerhar@protonmail.com> # Ouya T30
Tested-by: Nicolas Chauvet <kwizart@gmail.com> # PAZ00 T20 and TK1 T124
Signed-off-by: Dmitry Osipenko <digetx@gmail.com>
[treding@nvidia.com: unbreak Tegra186+ display support]
Signed-off-by: Thierry Reding <treding@nvidia.com>

drm/bridge: ti-sn65dsi86: Avoid creating multiple connectors

If we created our own connector because the driver does not support the
NO_CONNECTOR flag, we don't want the downstream bridge to *also* create
a connector.  And if this driver did pass the NO_CONNECTOR flag (and we
supported that mode) this would change nothing.

Fixes: 4e5763f03e10 ("drm/bridge: ti-sn65dsi86: Wrap panel with panel-bridge")
Reported-by: Stephen Boyd <swboyd@chromium.org>
Signed-off-by: Rob Clark <robdclark@chromium.org>
Tested-by: Stephen Boyd <swboyd@chromium.org>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Douglas Anderson <dianders@chromium.org>
Tested-by: Douglas Anderson <dianders@chromium.org>
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20210811235253.924867-2-robdclark@gmail.com

drm/edid: fix edid field name

Byte 26 in a edid struct is supposed to be "Blue and white
least-significant 2 bits", not "black and white". Rename the field
accordingly. This field is not used anywhere, so just renaming it here
for correctness.

Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
Reviewed-by: Simon Ser <contact@emersion.fr>
Signed-off-by: Simon Ser <contact@emersion.fr>
Link: https://patchwork.freedesktop.org/patch/msgid/20210811205818.156100-1-lucas.demarchi@intel.com

drm/virtio: set non-cross device blob uuid_state

Blob resources without the cross device flag don't have a uuid to share
with other virtio devices. When exporting such blobs, set uuid_state to
STATE_ERR so that virtgpu_virtio_get_uuid doesn't hang.

Signed-off-by: David Stevens <stevensd@chromium.org>
Link: http://patchwork.freedesktop.org/patch/msgid/20210811040401.1264234-1-stevensd@chromium.org
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>

udmabuf: fix general protection fault in udmabuf_create

Syzbot reported general protection fault in udmabuf_create. The problem
was in wrong error handling.

In commit 16c243e99d33 ("udmabuf: Add support for mapping hugepages (v4)")
shmem_read_mapping_page() call was replaced with find_get_page_flags(),
but find_get_page_flags() returns NULL on failure instead PTR_ERR().

Wrong error checking was causing GPF in get_page(), since passed page
was equal to NULL. Fix it by changing if (IS_ER(!hpage)) to if (!hpage)

Reported-by: syzbot+e9cd3122a37c5d6c51e8@syzkaller.appspotmail.com
Fixes: 16c243e99d33 ("udmabuf: Add support for mapping hugepages (v4)")
Signed-off-by: Pavel Skripkin <paskripkin@gmail.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20210811175052.21254-1-paskripkin@gmail.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>

Merge tag 'drm-intel-next-2021-08-10-1' of git://anongit.freedesktop.org/drm/drm-intel into drm-next

drm/i915 changes for v5.15:

Features:
- Basic DG2 platform enabling (Matt, Animesh, Gwan-gyeong, José)
- Add PSF GV point support for display bandwidth calculation (Stan)
- Add platform release id version support (Lucas)
- Add support for forcing DSC BPP for testing (Vandita, Patnana)

Refactoring and cleanups:
- Remove CNL support completely (Lucas)
- Revid/stepping cleanup (Matt, Anusha)
- Make display stepping check upper bounds exclusive (Matt)
- Remove old GEN macros (Lucas)
- Refactor DG1 interrupt handler (Paulo)
- Refactor DMC stepping info (Anusha)

Fixes:
- Fix XELPD color capability reporting; it's not yet enabled (Uma)
- Fix DG1 memory bandwidth computation (Clint)
- Fix mux on certain HP laptops (Kai-Heng)
- Various display workarounds (José, Matt, Imre)
- Fix error state dumps wrt SFC_DONE (Matt)
- Fix DG1 and XEPLD audio power domains (Anshuman)
- Fix ADL-P and ADL-S ddi buf translation tables (Matt)
- Fix DP/HDMI modeset sequences causing issues on ADL-P (José)
- PSR2 fixes (José)
- Fix DP MST modeset with FEC on TGL+
- Fix MBUS DBOX A credits on ADL-P (José)
- Fix DP PHY test training set programming (Khaled)
- Fix dgfx pcode uncore init done wait (Badal)
- Fix DSC disable fuse check on GLK (Lucas)
- Fix shared dpll mismatch for bigjoiner secondary pipe (Manasi)
- Fix ADL-P underrun recovery (Matt)
- Fix permissions on FEC support debugfs file (Vandita)

Misc:
- Backmerge drm-next (Rodrigo)
- Bump RKL and TGL DMC firmware version (Anusha)

Signed-off-by: Dave Airlie <airlied@redhat.com>
# Conflicts:
# drivers/gpu/drm/i915/i915_drv.h
# drivers/gpu/drm/i915/intel_device_info.h
From: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/87a6lpo1a9.fsf@intel.com

Merge tag 'drm-xilinx-dpsub-20210809' of git://linuxtv.org/pinchartl/media into drm-next

- Miscellaneous fixes in ZynqMP DPSUB driver

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Link: https://patchwork.freedesktop.org/patch/msgid/YRCSRZZV1HZYPvaG@pendragon.ideasonboard.com

Merge tag 'drm-intel-gt-next-2021-08-06-1' of ssh://git.freedesktop.org/git/drm/drm-intel into drm-next

UAPI Changes:

- Add I915_MMAP_OFFSET_FIXED

  On devices with local memory `I915_MMAP_OFFSET_FIXED` is the only valid
  type. On devices without local memory, this caching mode is invalid.

  As caching mode when specifying `I915_MMAP_OFFSET_FIXED`, WC or WB will
  be used, depending on the object placement on creation. WB will be used
  when the object can only exist in system memory, WC otherwise.

  Userspace: https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11888

- Reinstate the mmap ioctl for (already released) integrated Gen12 platforms

  Rationale: Otherwise media driver breaks eg. for ADL-P. Long term goal is
  still to sunset the IOCTL even for integrated and require using mmap_offset.

- Reject caching/set_domain IOCTLs on discrete

  Expected to become immutable property of the BO

- Disallow changing context parameters after first use on Gen12 and earlier
- Require setting context parameters at creation on platforms after Gen12

  Rationale (for both): Allow less dynamic changes to the context to simplify
  the implementation and avoid user shooting theirselves in the foot.

- Drop I915_CONTEXT_PARAM_RINGSIZE

  Userspace PR for compute-driver has not been merged

- Drop I915_CONTEXT_PARAM_NO_ZEROMAP

  Userspace PR for libdrm / Beignet was never landed

- Drop CONTEXT_CLONE API

  Userspace PR for Mesa was never landed

- Drop getparam support for I915_CONTEXT_PARAM_ENGINES

  Only existed for symmetry wrt. setparam, never used.

- Disallow bonding of virtual engines

  Drop the prep work, no hardware has been released needing it.

- (Implicit) Disable gpu relocations

  Media userspace was the last userspace to still use them. They
  have converted so performance can be regained with an update.

Core Changes:

- Merge topic branch 'topic/i915-ttm-2021-06-11' (from Maarten)
- Merge topic branch 'topic/revid_steppings' (from Matt R)
- Merge topic branch 'topic/xehp-dg2-definitions-2021-07-21' (from Matt R)
- Backmerges drm-next (Rodrigo)

Driver Changes:

- Initial workarounds for ADL-P (Clint)
- Preliminary code for XeHP/DG2 (Stuart, Umesh, Matt R, Prathap, Ram,
  Venkata, Akeem, Tvrtko, John, Lucas)
- Fix ADL-S DMA mask size to 39 bits (Tejas)
- Remove code for CNL (Lucas)
- Add ADL-P GuC/HuC firmwares (John)
- Update HuC to 7.9.3 for TGL/ADL-S/RKL (John)
- Fix -EDEADLK handling regression (Ville)
- Implement Wa_1508744258 for DG1 and Gen12 iGFX (Jose)
- Extend Wa_1406941453 to ADL-S (Jose)
- Drop unnecessary workarounds per stepping for SKL/BXT/ICL (Matt R)
- Use fuse info to enable SFC on Gen12 (Venkata)
- Unconditionally flush the pages on acquire on EHL/JSL (Matt A)
- Probe existence of backing struct pages upon userptr creation (Chris, Matt A)

- Add an intermediate GEM proto-context to delay real context creation (Jason)
- Implement SINGLE_TIMELINE with a syncobj (Jason)
- Set the watchdog timeout directly in intel_context_set_gem (Jason)
- Disallow userspace from creating contexts with too many engines (Jason)
- Revert "drm/i915/gem: Asynchronous cmdparser" (Jason)
- Revert "drm/i915: Propagate errors on awaiting already signaled fences" (Jason)
- Revert "drm/i915: Skip over MI_NOOP when parsing" (Jason)
- Revert "drm/i915: Shrink the GEM kmem_caches upon idling" (Daniel)
- Always let TTM handle object migration (Jason)
- Correct the locking and pin pattern for dma-buf (Thomas H, Michael R, Jason)
- Migrate to system at dma-buf attach time (Thomas, Michael R)

- MAJOR refactoring of the GuC backend code to allow for enabling on Gen11+
  (Matt B, John, Michal Wa., Fernando, Daniele, Vinay)
- Update GuC firmware interface to v62.0.0 (John, Michal Wa., Matt B)
- Add GuCRC feature to hand over the control of HW RC6 to the GuC on
  Gen12+ when GuC submission is enabled (Vinay, Sujaritha, Daniele,
  John, Tvrtko)
- Use the correct IRQ during resume and eliminate DRM IRQ midlayer (Thomas Z)
- Add pipelined page migration and clearing (Chris, Thomas H)
- Use TTM for system memory on discrete (Thomas H)
- Implement object migration for display vs. dma-buf (Thomas H)
- Perform execbuffer object locking as a separate step (Thomas H)
- Add support for explicit L3BANK steering (Matt, Daniele)
- Remove duplicated call to ops->pread (Daniel)
- Fix pagefault disabling in the first execbuf slowpath (Daniel)
- Simplify userptr locking (Thomas H)
- Improvements to the GuC CTB code (Matt B, John)
- Make GT workaround upper bounds exclusive (Matt R)
- Check for nomodeset in i915_init() first (Daniel)
- Delete now unused gpu reloc code (Daniel)

- Document RFC plans for GuC submission, DRM scheduler and new parallel
  submit uAPI (Matt B)
- Reintroduce buddy allocator this time with TTM (Matt A)
- Support forcing page size with LMEM (Matt A)
- Add i915_sched_engine to abstract a submission queue between backends (Matt B)
- Use accelerated move in TTM (Ram)
- Fix memory leaks from TTM backend (Thomas H)
- Introduce WW transaction helper (Thomas H)
- Improve debug Kconfig texts a bit (Daniel)
- Unify user object creation code (Jason)
- Use a table for i915_init/exit (Jason)
- Move slabs to module init/exit (Daniel)
- Remove now unused i915_globals (Daniel)
- Extract i915_module.c (Daniel)

- Consistently use adl-p/adl-s in WA comments (Jose)
- Finish INTEL_GEN and friends conversion (Lucas)
- Correct variable/function namings (Lucas)
- Code checker fixes (Wan, Matt A)
- Tracepoint improvements (Matt B)
- Kerneldoc improvements (Tvrtko, Jason, Matt A, Maarten)
- Selftest improvements (Chris, Matt A, Tejas, Thomas H, John, Matt B,
  Rahul, Vinay)

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/YQ0JmYiXhGskNcrI@jlahtine-mobl.ger.corp.intel.com

drm/display: fix possible null-pointer dereference in dcn10_set_clock()

The variable dc->clk_mgr is checked in:
  if (dc->clk_mgr && dc->clk_mgr->funcs->get_clock)

This indicates dc->clk_mgr can be NULL.
However, it is dereferenced in:
    if (!dc->clk_mgr->funcs->get_clock)

To fix this null-pointer dereference, check dc->clk_mgr and the function
pointer dc->clk_mgr->funcs->get_clock earlier, and return if one of them
is NULL.

Reported-by: TOTE Robot <oslab@tsinghua.edu.cn>
Signed-off-by: Tuo Li <islituo@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

gpu: drm: amd: amdgpu: amdgpu_i2c: fix possible uninitialized-variable access in amdgpu_i2c_router_select_ddc_port()

The variable val is declared without initialization, and its address is
passed to amdgpu_i2c_get_byte(). In this function, the value of val is
accessed in:
  DRM_DEBUG("i2c 0x%02x 0x%02x read failed\n",
       addr, *val);

Also, when amdgpu_i2c_get_byte() returns, val may remain uninitialized,
but it is accessed in:
  val &= ~amdgpu_connector->router.ddc_mux_control_pin;

To fix this possible uninitialized-variable access, initialize val to 0 in
amdgpu_i2c_router_select_ddc_port().

Reported-by: TOTE Robot <oslab@tsinghua.edu.cn>
Signed-off-by: Tuo Li <islituo@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdkfd: CWSR with software scheduler

This patch adds support to program trap handler settings
when loading driver with software scheduler (sched_policy=2).

Signed-off-by: Mukul Joshi <mukul.joshi@amd.com>
Suggested-by: Jay Cornwall <Jay.Cornwall@amd.com>
Reviewed-by: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdkfd: AIP mGPUs best prefetch location for xnack on

For xnack on, if range ACCESS or ACCESS_IN_PLACE (AIP) by single GPU, or
range is ACCESS_IN_PLACE by mGPUs and all mGPUs connection on XGMI same
hive, the best prefetch location is prefetch_loc GPU. Otherwise, the best
prefetch location is always CPU because GPU does not have coherent
mapping VRAM of other GPUs even with large-BAR PCIe connection.

Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: graceful exit on restore fan mode failure (v2)

Attempt od settings restore and disable restore flag on restore fan mode
failure.

v2: Update fan mode to auto and fan speed to zero (Lijo)

Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Ryan Taylor <Ryan.Taylor@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: restore fan_mode AMD_FAN_CTRL_NONE on resume (v2)

Adds missing edge case to smu_restore_dpm_user_profile.

v2: Don't restore fan mode auto (Lijo)

Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Ryan Taylor <Ryan.Taylor@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/msi/mdp4: populate priv->kms in mdp4_kms_init

Without this fix boot throws NULL ptr exception at msm_dsi_manager_setup_encoder
on devices like Nexus 7 2013 (MDP4 v4.4).

Fixes: 03436e3ec69c ("drm/msm/dsi: Move setup_encoder to modeset_init")

Cc: <stable@vger.kernel.org>
Signed-off-by: David Heidelberg <david@ixit.cz>
Link: https://lore.kernel.org/r/20210811170631.39296-1-david@ixit.cz
Signed-off-by: Rob Clark <robdclark@chromium.org>

drm/ingenic: Use standard drm_atomic_helper_commit_tail

By making the CRTC's .vblank_enable() function return an error when it
is known that the hardware won't deliver a VBLANK, we can drop the
ingenic_drm_atomic_helper_commit_tail() function and use the standard
drm_atomic_helper_commit_tail() function instead.

Signed-off-by: Paul Cercueil <paul@crapouillou.net>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20210808134526.119198-4-paul@crapouillou.net

drm/ingenic: Remove dead code

The priv->ipu_plane would get a different value further down the code,
without the first assigned value being read first; so the first
assignation can be dropped.

Signed-off-by: Paul Cercueil <paul@crapouillou.net>
Acked-by: Thomas Zimmermann <tzimmermann@suse.de>
Link: https://patchwork.freedesktop.org/patch/msgid/20210808134526.119198-2-paul@crapouillou.net

Merge tag 'amd-drm-next-5.15-2021-08-06' of https://gitlab.freedesktop.org/agd5f/linux into drm-next

amd-drm-next-5.15-2021-08-06:

amdgpu:
- Aldebaran fixes
- Powergating fix for Renoir
- Switch virtual DCE over to vkms based atomic modesetting
- Misc typo fixes
- PSP handling cleanups
- DC FP cleanups
- RAS fixes
- Wave debug improvements
- Freesync fix
- BACO/BOCO fixes
- Misc fixes

amdkfd:
- Expose gfx version in sysfs
- Aldebaran fixes

radeon:
- Coding style fix
- Typo fixes
- Pageflip fix

UAPI:
- amdkfd: SVM address range query
  Proposed userspace: https://github.com/RadeonOpenCompute/ROCR-Runtime/tree/memory_model_queries

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210806205248.3864-1-alexander.deucher@amd.com

Merge tag 'bus_remove_return_void-5.15' of git://git./linux/kernel/git/gregkh/driver-core into drm-next

Bus: Make remove callback return void tag

Tag for other trees/branches to pull from in order to have a stable
place to build off of if they want to add new busses for 5.15.

Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Dave Airlie <airlied@redhat.com>
[airlied: fixed up merge conflict in drm]
From:   Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Link: https://patchwork.freedesktop.org/patch/msgid/YPkwQwf0dUKnGA7L@kroah.com

drm/msm/dsi: add continuous clock support for 7nm PHY

Unlike previous generations, 7nm PHYs are required to collaborate with
the host for continuos clock mode. Add changes neccessary to enable
continuous clock mode in the 7nm DSI PHYs.

Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20210805170817.3337665-1-dmitry.baryshkov@linaro.org
[Fix merge conflict, and $description typo]
Signed-off-by: Rob Clark <robdclark@chromium.org>

drm/msm/dp: add drm debug logs to dp_pm_resume/suspend

Changes in V2:
-- correct Fixes text
-- drop commit text

Fixes: 601f0479c583 ("drm/msm/dp: add logs across DP driver for ease of debugging")
Signed-off-by: Kuogee Hsieh <khsieh@codeaurora.org>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Link: https://lore.kernel.org/r/1628609362-2109-1-git-send-email-khsieh@codeaurora.org
Signed-off-by: Rob Clark <robdclark@chromium.org>

drm/msm/dp: do not end dp link training until video is ready

Initialize both pre-emphasis and voltage swing level to 0 before
start link training and do not end link training until video is
ready to reduce the period between end of link training and video
start to meet Link Layer CTS requirement.  Some dongle main link
symbol may become unlocked again if host did not end link training
soon enough after completion of link training 2. Host have to re
train main link if loss of symbol locked detected before end link
training so that the coming video stream can be transmitted to sink
properly. This fixes Link Layer CTS cases 4.3.2.1, 4.3.2.2, 4.3.2.3
and 4.3.2.4.

Changes in v3:
-- merge retrain link if loss of symbol locked happen into this patch
-- replace dp_ctrl_loss_symbol_lock() with dp_ctrl_channel_eq_ok()

Signed-off-by: Kuogee Hsieh <khsieh@codeaurora.org>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Link: https://lore.kernel.org/r/1628196295-7382-7-git-send-email-khsieh@codeaurora.org
Signed-off-by: Rob Clark <robdclark@chromium.org>

drm/msm/dp: return correct edid checksum after corrupted edid checksum read

Response with correct edid checksum saved at connector after corrupted edid
checksum read. This fixes Link Layer CTS cases 4.2.2.3, 4.2.2.6.

Signed-off-by: Kuogee Hsieh <khsieh@codeaurora.org>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Link: https://lore.kernel.org/r/1628196295-7382-6-git-send-email-khsieh@codeaurora.org
Signed-off-by: Rob Clark <robdclark@chromium.org>

drm/msm/dp: replug event is converted into an unplug followed by an plug events

Remove special handling of replug interrupt and instead treat replug event
as a sequential unplug followed by a plugin event. This is needed to meet
the requirements of DP Link Layer CTS test case 4.2.1.3.

Changes in V2:
-- add fixes statement

Changes in V3:
-- delete EV_HPD_REPLUG_INT

Fixes: f21c8a276c2d ("drm/msm/dp: handle irq_hpd with sink_count = 0 correctly")

Signed-off-by: Kuogee Hsieh <khsieh@codeaurora.org>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Link: https://lore.kernel.org/r/1628196295-7382-5-git-send-email-khsieh@codeaurora.org
Signed-off-by: Rob Clark <robdclark@chromium.org>

drm/msm/dp: reset aux controller after dp_aux_cmd_fifo_tx() failed.

Aux hardware calibration sequence requires resetting the aux controller
in order for the new setting to take effect. However resetting the AUX
controller will also clear HPD interrupt status which may accidentally
cause pending unplug interrupt to get lost. Therefore reset aux
controller only when link is in connection state when dp_aux_cmd_fifo_tx()
fail. This fixes Link Layer CTS cases 4.2.1.1 and 4.2.1.2.

Signed-off-by: Kuogee Hsieh <khsieh@codeaurora.org>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Link: https://lore.kernel.org/r/1628196295-7382-4-git-send-email-khsieh@codeaurora.org
Signed-off-by: Rob Clark <robdclark@chromium.org>

drm/msm/dp: reduce link rate if failed at link training 1

Reduce link rate and re start link training if link training 1
failed due to loss of clock recovery done to fix Link Layer
CTS case 4.3.1.7.  Also only update voltage and pre-emphasis
swing level after link training started to fix Link Layer CTS
case 4.3.1.6.

Changes in V2:
-- replaced cr_status with link_status[DP_LINK_STATUS_SIZE]
-- replaced dp_ctrl_any_lane_cr_done() with dp_ctrl_colco_recovery_any_ok()
-- replaced dp_ctrl_any_ane_cr_lose() with !drm_dp_clock_recovery_ok()

Changes in V3:
-- return failed if lane_count <= 1

Signed-off-by: Kuogee Hsieh <khsieh@codeaurora.org>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Link: https://lore.kernel.org/r/1628196295-7382-3-git-send-email-khsieh@codeaurora.org
[remove unused cr_status variable]
Signed-off-by: Rob Clark <robdclark@chromium.org>

drm/msm/dp: use dp_ctrl_off_link_stream during PHY compliance test run

DP cable should always connect to DPU during the entire PHY compliance
testing run. Since DP PHY compliance test is executed at irq_hpd event
context, dp_ctrl_off_link_stream() should be used instead of dp_ctrl_off().
dp_ctrl_off() is used for unplug event which is triggered when DP cable is
dis connected.

Changes in V2:
-- add fixes statement

Fixes: f21c8a276c2d ("drm/msm/dp: handle irq_hpd with sink_count = 0 correctly")

Signed-off-by: Kuogee Hsieh <khsieh@codeaurora.org>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Link: https://lore.kernel.org/r/1628196295-7382-2-git-send-email-khsieh@codeaurora.org
Signed-off-by: Rob Clark <robdclark@chromium.org>

drm/msm/dsi: Fix some reference counted resource leaks

'of_find_device_by_node()' takes a reference that must be released when
not needed anymore.
This is expected to be done in 'dsi_destroy()'.

However, there are 2 issues in 'dsi_get_phy()'.

First, if 'of_find_device_by_node()' succeeds but 'platform_get_drvdata()'
returns NULL, 'msm_dsi->phy_dev' will still be NULL, and the reference
won't be released in 'dsi_destroy()'.

Secondly, as 'of_find_device_by_node()' already takes a reference, there is
no need for an additional 'get_device()'.

Move the assignment to 'msm_dsi->phy_dev' a few lines above and remove the
unneeded 'get_device()' to solve both issues.

Fixes: ec31abf6684e ("drm/msm/dsi: Separate PHY to another platform device")
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Link: https://lore.kernel.org/r/f15bc57648a00e7c99f943903468a04639d50596.1628241097.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Rob Clark <robdclark@chromium.org>

drm/msm: Rework SQE version check

This check is really about which SQE firmware, rather than which GPU.
Rework to match minimum version based on firmware name, so it doesn't
need to be updated when adding additional GPUs using the same fw.

Signed-off-by: Rob Clark <robdclark@chromium.org>
Link: https://lore.kernel.org/r/20210807163019.379003-2-robdclark@gmail.com
Signed-off-by: Rob Clark <robdclark@chromium.org>

drm/msm: Add adreno_is_a640_family()

Combine adreno_is_a640() and adreno_is_a680().

Signed-off-by: Rob Clark <robdclark@chromium.org>
Link: https://lore.kernel.org/r/20210807163019.379003-1-robdclark@gmail.com
Signed-off-by: Rob Clark <robdclark@chromium.org>

drm: IRQ midlayer is now legacy

Hide the DRM midlayer behind CONFIG_DRM_LEGACY, make functions use
the prefix drm_legacy_, and move declarations to drm_legacy.h.
In struct drm_device, move the fields irq and irq_enabled behind
CONFIG_DRM_LEGACY.

All callers have been updated.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20210803090704.32152-15-tzimmermann@suse.de

drm: Remove unused devm_drm_irq_install()

DRM IRQ helpers will become legacy. The function devm_drm_irq_install()
is unused and won't be required later.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20210803090704.32152-14-tzimmermann@suse.de

drm/vc4: Convert to Linux IRQ interfaces

Drop the DRM IRQ midlayer in favor of Linux IRQ interfaces. DRM's
IRQ helpers are mostly useful for UMS drivers. Modern KMS drivers
don't benefit from using it.

DRM IRQ callbacks are now being called directly or inlined.

Calls to platform_get_irq() can fail with a negative errno code.
Abort initialization in this case. The DRM IRQ midlayer does not
handle this case correctly.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20210803090704.32152-13-tzimmermann@suse.de

drm/tilcdc: Convert to Linux IRQ interfaces

Drop the DRM IRQ midlayer in favor of Linux IRQ interfaces. DRM's
IRQ helpers are mostly useful for UMS drivers. Modern KMS drivers
don't benefit from using it.

DRM IRQ callbacks are now being called directly or inlined.

Calls to platform_get_irq() can fail with a negative errno code.
Abort initialization in this case. The DRM IRQ midlayer does not
handle this case correctly.

For most drivers, only the DRM IRQ helpers use irq_enabled from
struct drm_device. Tilcdc also uses irq_enabled to make its error
rollback work correctly. As the field will become legacy, duplicated
the state in the driver's local private structure.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20210803090704.32152-12-tzimmermann@suse.de

drm/tidss: Convert to Linux IRQ interfaces

Drop the DRM IRQ midlayer in favor of Linux IRQ interfaces. DRM's
IRQ helpers are mostly useful for UMS drivers. Modern KMS drivers
don't benefit from using it.

DRM IRQ callbacks are now being called directly or inlined.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
Acked-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20210803090704.32152-11-tzimmermann@suse.de

drm/radeon: Convert to Linux IRQ interfaces

Drop the DRM IRQ midlayer in favor of Linux IRQ interfaces. DRM's
IRQ helpers are mostly useful for UMS drivers. Modern KMS drivers
don't benefit from using it.

DRM IRQ callbacks are now being called directly or inlined.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210803090704.32152-10-tzimmermann@suse.de

drm/mxsfb: Convert to Linux IRQ interfaces

Drop the DRM IRQ midlayer in favor of Linux IRQ interfaces. DRM's
IRQ helpers are mostly useful for UMS drivers. Modern KMS drivers
don't benefit from using it.

DRM IRQ callbacks are now being called directly or inlined.

Calls to platform_get_irq() can fail with a negative errno code.
Abort initialization in this case. The DRM IRQ midlayer does not
handle this case correctly.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20210803090704.32152-9-tzimmermann@suse.de

drm/msm: Convert to Linux IRQ interfaces

Drop the DRM IRQ midlayer in favor of Linux IRQ interfaces. DRM's
IRQ helpers are mostly useful for UMS drivers. Modern KMS drivers
don't benefit from using it.

DRM IRQ callbacks are now being called directly or inlined.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Abhinav Kumar <abhinavk@codeaurora.org>
Acked-by: Rob Clark <robdclark@chromium.org>
Acked-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20210803090704.32152-8-tzimmermann@suse.de

drm/kmb: Convert to Linux IRQ interfaces

Drop the DRM IRQ midlayer in favor of Linux IRQ interfaces. DRM's
IRQ helpers are mostly useful for UMS drivers. Modern KMS drivers
don't benefit from using it.

DRM IRQ callbacks are now being called directly or inlined.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Sam Ravnborg <sam@ravnborg.org>
Tested-by: Anitha Chrisanthus <anitha.chrisanthus@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210803090704.32152-7-tzimmermann@suse.de

drm/gma500: Convert to Linux IRQ interfaces

Drop the DRM IRQ midlayer in favor of Linux IRQ interfaces. DRM's
IRQ helpers are mostly useful for UMS drivers. Modern KMS drivers
don't benefit from using it. DRM IRQ callbacks are now being called
directly or inlined.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20210803090704.32152-6-tzimmermann@suse.de

drm/fsl-dcu: Convert to Linux IRQ interfaces

Drop the DRM IRQ midlayer in favor of Linux IRQ interfaces. DRM's
IRQ helpers are mostly useful for UMS drivers. Modern KMS drivers
don't benefit from using it. DRM IRQ callbacks are now being called
directly or inlined.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20210803090704.32152-5-tzimmermann@suse.de

drm/atmel-hlcdc: Convert to Linux IRQ interfaces

Drop the DRM IRQ midlayer in favor of Linux IRQ interfaces. DRM's
IRQ helpers are mostly useful for UMS drivers. Modern KMS drivers
don't benefit from using it. DRM IRQ callbacks are now being called
directly or inlined.

v2:
* use managed release via devm_request_irq() (Sam)
* drop extra test for irq != IRQ_NOTCONNECTED (Sam)

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Sam Ravnborg <sam@ravnborg.org>
Tested-by: Dan Sneddon <Dan.Sneddon@microchip.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210803090704.32152-4-tzimmermann@suse.de

drm/arm/hdlcd: Convert to Linux IRQ interfaces

Drop the DRM IRQ midlayer in favor of Linux IRQ interfaces. DRM's
IRQ helpers are mostly useful for UMS drivers. Modern KMS drivers
don't benefit from using it.

DRM IRQ callbacks are now being called directly or inlined.

Calls to platform_get_irq() can fail with a negative errno code.
Abort initialization in this case. The DRM IRQ midlayer does not
handle this case correctly.

v2:
* name struct drm_device variables 'drm' (Sam)

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20210803090704.32152-3-tzimmermann@suse.de

drm/amdgpu: Convert to Linux IRQ interfaces

Drop the DRM IRQ midlayer in favor of Linux IRQ interfaces. DRM's
IRQ helpers are mostly useful for UMS drivers. Modern KMS drivers
don't benefit from using it.

DRM IRQ callbacks are now being called directly or inlined.

The interrupt number returned by pci_msi_vector() is now stored
in struct amdgpu_irq. Calls to pci_msi_vector() can fail and return
a negative errno code. Abort initlaizaton in thi case. The DRM IRQ
midlayer does not handle this correctly.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210803090704.32152-2-tzimmermann@suse.de

drm/tegra: Bump driver version

Bump driver version to 1.0.0 to allow userspace to detect
availability of new interfaces.

Signed-off-by: Mikko Perttunen <mperttunen@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>

drm/tegra: Add job firewall

Add a firewall that validates jobs before submission to ensure
they don't do anything they aren't allowed to do, like accessing
memory they should not access.

The firewall is functionality-wise a copy of the firewall already
implemented in gpu/host1x. It is copied here as it makes more
sense for it to live on the DRM side, as it is only needed for
userspace job submissions, and generally the data it needs to
do its job is easier to access here.

In the future, the other implementation will be removed.

Signed-off-by: Mikko Perttunen <mperttunen@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>

drm/tegra: Implement job submission part of new UAPI

Implement the job submission IOCTL with a minimum feature set.

Signed-off-by: Mikko Perttunen <mperttunen@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>

drm/tegra: Implement syncpoint wait UAPI

Implement new syncpoint wait UAPI. This is different from the
legacy one in taking an absolute timestamp in line with modern
DRM conventions.

Signed-off-by: Mikko Perttunen <mperttunen@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>

drm/tegra: Implement syncpoint management UAPI

Implement TegraDRM IOCTLs for allocating and freeing syncpoints.

Signed-off-by: Mikko Perttunen <mperttunen@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>

drm/tegra: Implement new UAPI

Implement the non-submission parts of the new UAPI, including
channel management and memory mapping. The UAPI is under the
CONFIG_DRM_TEGRA_STAGING config flag for now.

Signed-off-by: Mikko Perttunen <mperttunen@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>

amdgpu/pm: Replace amdgpu_pm usage of sprintf with sysfs_emit

 replacing printfs with sysfs_emit

=== Test ===
AMDGPU_PCI_ADDR=`lspci -nn | grep "VGA\|Display" | cut -d " " -f 1`
AMDGPU_HWMON=`ls -la /sys/class/hwmon | grep $AMDGPU_PCI_ADDR | awk '{print $9}'`
HWMON_DIR=/sys/class/hwmon/${AMDGPU_HWMON}
LOGFILE=pp_printf.test.log

lspci -nn | grep "VGA\|Display"  > $LOGFILE
FILES="pwm1_enable
pwm1_min
pwm1_max
pwm1
fan1_input
fan1_target
fan1_enable
power1_cap_min "

for f in $FILES
do
  echo === $f === >> $LOGFILE
  cat $HWMON_DIR/device/$f >> $LOGFILE
done
cat $LOGFILE

Signed-off-by: Darren Powell <darren.powell@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

amdgpu/pm: Replace hwmgr smu usage of sprintf with sysfs_emit

 replacing printfs with sysfs_emit
 minor smu7 change to remove compiler warning comparison of int and uint32_t
 minor smu8 change to remove compiler warning comparison of int and uint32_t

=== Test ===
AMDGPU_PCI_ADDR=`lspci -nn | grep "VGA\|Display" | cut -d " " -f 1`
AMDGPU_HWMON=`ls -la /sys/class/hwmon | grep $AMDGPU_PCI_ADDR | awk '{print $9}'`
HWMON_DIR=/sys/class/hwmon/${AMDGPU_HWMON}
LOGFILE=pp_printf.test.log

lspci -nn | grep "VGA\|Display"  > $LOGFILE
FILES="pp_dpm_sclk
pp_power_profile_mode "

for f in $FILES
do
  echo === $f === >> $LOGFILE
  cat $HWMON_DIR/device/$f >> $LOGFILE
done
cat $LOGFILE

Signed-off-by: Darren Powell <darren.powell@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

amdgpu/pm: Replace vega12,20 usage of sprintf with sysfs_emit

=== Test ===
AMDGPU_PCI_ADDR=`lspci -nn | grep "VGA\|Display" | cut -d " " -f 1`
AMDGPU_HWMON=`ls -la /sys/class/hwmon | grep $AMDGPU_PCI_ADDR | awk '{print $9}'`
HWMON_DIR=/sys/class/hwmon/${AMDGPU_HWMON}
LOGFILE=pp_printf.test.log

lspci -nn | grep "VGA\|Display"  > $LOGFILE
FILES="pp_dpm_sclk
pp_features
pp_power_profile_mode "

for f in $FILES
do
  echo === $f === >> $LOGFILE
  cat $HWMON_DIR/device/$f >> $LOGFILE
done
cat $LOGFILE

Signed-off-by: Darren Powell <darren.powell@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

amdgpu/pm: Replace vega10 usage of sprintf with sysfs_emit

 initial modification of files
  vega10_hwmgr.c

=== Test ===
AMDGPU_PCI_ADDR=`lspci -nn | grep "VGA\|Display" | cut -d " " -f 1`
AMDGPU_HWMON=`ls -la /sys/class/hwmon | grep $AMDGPU_PCI_ADDR | awk '{print $9}'`
HWMON_DIR=/sys/class/hwmon/${AMDGPU_HWMON}
LOGFILE=pp_printf.test.log

lspci -nn | grep "VGA\|Display"  > $LOGFILE
FILES="pp_dpm_sclk
pp_features
pp_power_profile_mode "

for f in $FILES
do
  echo === $f === >> $LOGFILE
  cat $HWMON_DIR/device/$f >> $LOGFILE
done
cat $LOGFILE

Signed-off-by: Darren Powell <darren.powell@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

amdgpu/pm: Replace smu12/13 usage of sprintf with sysfs_emit

 initial modification of files
  renoir_ppt.c
  aldebaran_ppt.c
  yellow_carp_ppt.c

=== Test ===
AMDGPU_PCI_ADDR=`lspci -nn | grep "VGA\|Display" | cut -d " " -f 1`
AMDGPU_HWMON=`ls -la /sys/class/hwmon | grep $AMDGPU_PCI_ADDR | awk '{print $9}'`
HWMON_DIR=/sys/class/hwmon/${AMDGPU_HWMON}
LOGFILE=pp_printf.test.log

lspci -nn | grep "VGA\|Display"  > $LOGFILE
FILES="pp_dpm_sclk
pp_power_profile_mode "

for f in $FILES
do
  echo === $f === >> $LOGFILE
  cat $HWMON_DIR/device/$f >> $LOGFILE
done
cat $LOGFILE

Signed-off-by: Darren Powell <darren.powell@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

amdgpu/pm: Replace smu11 usage of sprintf with sysfs_emit

 modification of smu11 files
  arcturus_ppt.c
  sienna_cichlid_ppt.c
  vangogh_ppt.c

=== Test ===
AMDGPU_PCI_ADDR=`lspci -nn | grep "VGA\|Display" | cut -d " " -f 1`
AMDGPU_HWMON=`ls -la /sys/class/hwmon | grep $AMDGPU_PCI_ADDR | awk '{print $9}'`
HWMON_DIR=/sys/class/hwmon/${AMDGPU_HWMON}
LOGFILE=pp_printf.test.log

lspci -nn | grep "VGA\|Display"  > $LOGFILE
FILES="pp_dpm_sclk
pp_power_profile_mode "

for f in $FILES
do
  echo === $f === >> $LOGFILE
  cat $HWMON_DIR/device/$f >> $LOGFILE
done
cat $LOGFILE

Signed-off-by: Darren Powell <darren.powell@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

amdgpu/pm: Replace navi10 usage of sprintf with sysfs_emit

 initial modification of files
  smu_cmn.c
  navi10_ppt.c

=== Test ===
AMDGPU_PCI_ADDR=`lspci -nn | grep "VGA\|Display" | cut -d " " -f 1`
AMDGPU_HWMON=`ls -la /sys/class/hwmon | grep $AMDGPU_PCI_ADDR | awk '{print $9}'`
HWMON_DIR=/sys/class/hwmon/${AMDGPU_HWMON}
LOGFILE=pp_printf.test.log

lspci -nn | grep "VGA\|Display"  > $LOGFILE
FILES="pp_dpm_sclk
pp_sclk_od
pp_mclk_od
pp_dpm_pcie
pp_od_clk_voltage
pp_power_profile_mode "

for f in $FILES
do
  echo === $f === >> $LOGFILE
  cat $HWMON_DIR/device/$f >> $LOGFILE
done
cat $LOGFILE

Signed-off-by: Darren Powell <darren.powell@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: handle VCN instances when harvesting (v2)

There may be multiple instances and only one is harvested.

v2: fix typo in commit message

Fixes: 83a0b8639185 ("drm/amdgpu: add judgement when add ip blocks (v2)")
Bug: https://gitlab.freedesktop.org/drm/amd/-/issues/1673
Reviewed-by: Guchun Chen <guchun.chen@amd.com>
Reviewed-by: James Zhu <James.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/tegra: Allocate per-engine channel in core code

To avoid code duplication, allocate the per-engine shared channel in
the core code instead. This is the usual channel that all jobs are
submitted to when MLOCKing is not in use. Once MLOCKs are implemented
on Host1x side, we can also update this to avoid allocating a shared
channel when MLOCKs are enabled.

Signed-off-by: Mikko Perttunen <mperttunen@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>

drm/tegra: Boot VIC during runtime PM resume

With the new UAPI implementation, engines are powered on and off
when there are active jobs, and the core code handles channel
allocation. To accommodate that, boot the engine as part of
runtime PM instead of using the open_channel callback, which is
not used by the new submit path.

Signed-off-by: Mikko Perttunen <mperttunen@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>

drm/tegra: Add new UAPI to header

Update the tegra_drm.h UAPI header, adding the new proposed UAPI.
The old staging UAPI is left in for now, with minor modification
to avoid name collisions.

Signed-off-by: Mikko Perttunen <mperttunen@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>

drm/tegra: Extract tegra_gem_lookup()

The static function host1x_bo_lookup() in drm.c is also useful
elsewhere. Extract it as tegra_gem_lookup() in gem.c.

Signed-off-by: Mikko Perttunen <mperttunen@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>

gpu: host1x: Add option to skip firewall for a job

The new UAPI will have its own firewall, and we don't want to run
the firewall in the Host1x driver for those jobs. As such, add a
parameter to host1x_job_alloc to specify if we want to skip the
firewall in the Host1x driver.

Signed-off-by: Mikko Perttunen <mperttunen@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>

gpu: host1x: Add support for syncpoint waits in CDMA pushbuffer

Add support for inserting syncpoint waits in the CDMA pushbuffer.
These waits need to be done in HOST1X class, while gather submitted
by the application execute in engine class.

Support is added by converting the gather list of job into a command
list that can include both gathers and waits. When the job is
submitted, these commands are pushed as the appropriate opcodes
on the CDMA pushbuffer.

Also supported are waits relative to the start of the job,
which are useful for jobs doing multiple things with an engine
that doesn't natively support pipelining.

While at it, use 32-bit waits on chips that support them.

Signed-off-by: Mikko Perttunen <mperttunen@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>

gpu: host1x: Add job release callback

Add a callback field to the job structure, to be called just before
the job is to be freed. This allows the job's submitter to clean
up any of its own state, like decrement runtime PM refcounts.

Signed-off-by: Mikko Perttunen <mperttunen@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>

gpu: host1x: Add no-recovery mode

Add a new property for jobs to enable or disable recovery i.e.
CPU increments of syncpoints to max value on job timeout. This
allows for a more solid model for hanged jobs, where userspace
doesn't need to guess if a syncpoint increment happened because
the job completed, or because job timeout was triggered.

On job timeout, we stop the channel, NOP all future jobs on the
channel using the same syncpoint, mark the syncpoint as locked
and resume the channel from the next job, if any.

The future jobs are NOPed, since because we don't do the CPU
increments, the value of the syncpoint is no longer synchronized,
and any waiters would become confused if a future job incremented
the syncpoint. The syncpoint is marked locked to ensure that any
future jobs cannot increment the syncpoint either, until the
application has recognized the situation and reallocated the
syncpoint.

Signed-off-by: Mikko Perttunen <mperttunen@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>

gpu: host1x: Add DMA fence implementation

Add an implementation of dma_fences based on syncpoints. Syncpoint
interrupts are used to signal fences. Additionally, after
software signaling has been enabled, a 30 second timeout is started.
If the syncpoint threshold is not reached within this period,
the fence is signalled with an -ETIMEDOUT error code. This is to
allow fences that would never reach their syncpoint threshold to
be cleaned up. The timeout can potentially be removed in the future
after job tracking code has been refactored.

Signed-off-by: Mikko Perttunen <mperttunen@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>

dma-buf: Fix a few typos in dma-buf documentation

Fix a few typos in the documentation:
- Remove an extraneous 'or'
- 'unpins' -> 'unpin'
- 'braket' -> 'bracket'
- 'mappinsg' -> 'mappings'
- 'fullfills' -> 'fulfills'

Reviewed-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Gal Pressman <galpress@amazon.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20210809122247.15869-1-galpress@amazon.com

drm/vmwgfx: Replace "vmw_num_pages" with "PFN_UP"

we counld use PFN_UP instead of vmw_num_pages()

Signed-off-by: Cai Huoqing <caihuoqing@baidu.com>
Signed-off-by: Zack Rusin <zackr@vmware.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210802033552.990-1-caihuoqing@baidu.com

drm/vmwgfx: Make use of PFN_ALIGN/PFN_UP helper macro

it's a refactor to make use of PFN_ALIGN/PFN_UP helper macro

Signed-off-by: Cai Huoqing <caihuoqing@baidu.com>
Signed-off-by: Zack Rusin <zackr@vmware.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210802033308.927-1-caihuoqing@baidu.com

drm/vmwgfx: fix potential UAF in vmwgfx_surface.c

drm_file.master should be protected by either drm_device.master_mutex
or drm_file.master_lookup_lock when being dereferenced. However,
drm_master_get is called on unprotected file_priv->master pointers in
vmw_surface_define_ioctl and vmw_gb_surface_define_internal.

This is fixed by replacing drm_master_get with drm_file_get_master.

Signed-off-by: Desmond Cheong Zhi Xi <desmondcheongzx@gmail.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Zack Rusin <zackr@vmware.com>
Signed-off-by: Zack Rusin <zackr@vmware.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210724111824.59266-4-desmondcheongzx@gmail.com

drm/vmwgfx: Use list_move_tail instead of list_del/list_add_tail in vmwgfx_cmdbuf_res.c

Using list_move_tail() instead of list_del() + list_add_tail() in vmwgfx_cmdbuf_res.c.

Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Baokun Li <libaokun1@huawei.com>
Signed-off-by: Zack Rusin <zackr@vmware.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210609072248.1353421-1-libaokun1@huawei.com

drm/vmwgfx: Use list_move_tail instead of list_del/list_add_tail in vmwgfx_cmdbuf.c

Using list_move_tail() instead of list_del() + list_add_tail() in vmwgfx_cmdbuf.c.

Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Baokun Li <libaokun1@huawei.com>
Signed-off-by: Zack Rusin <zackr@vmware.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210609071803.1347254-1-libaokun1@huawei.com