drm/amd/display: Increase frame warning limit with KASAN or KCSAN in dml2

When building ARCH=x86_64 allmodconfig with clang, which will typically
have sanitizers enabled, there is a warning about a large stack frame.

  drivers/gpu/drm/amd/amdgpu/../display/dc/dml2/display_mode_core.c:6265:13: error: stack frame size (2520) exceeds limit (2048) in 'dml_prefetch_check' [-Werror,-Wframe-larger-than]
   6265 | static void dml_prefetch_check(struct display_mode_lib_st *mode_lib)
        |             ^
  1 error generated.

Notably, GCC 13.2.0 does not do too much of a better job, as it is right
at the current limit of 2048 (and others have reported being over with
older GCC versions):

  drivers/gpu/drm/amd/amdgpu/../display/dc/dml2/display_mode_core.c: In function 'dml_prefetch_check':
  drivers/gpu/drm/amd/amdgpu/../display/dc/dml2/display_mode_core.c:6705:1: error: the frame size of 2048 bytes is larger than 1800 bytes [-Werror=frame-larger-than=]
   6705 | }
        | ^

In the past, these warnings have been avoided by reducing the number of
parameters to various functions so that not as many arguments need to be
passed on the stack. However, these patches take a good amount of effort
to write despite being mechanical due to code structure and complexity
and they are never carried forward to new generations of the code so
that effort has to be expended every new hardware generation, which
becomes harder to justify as time goes on.

To avoid having a noticeable or lengthy breakage in all{mod,yes}config,
which are easy testing targets that have -Werror enabled, increase the
limit for configurations that have KASAN or KCSAN enabled by 50% so that
cases of extremely poor code generation can still be caught while not
breaking the majority of builds. CONFIG_KMSAN also causes high stack
usage but the frame limit is already set to zero when it is enabled,
which is accounted for by the check for CONFIG_FRAME_WARN=0 in the dml2
Makefile.

Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Avoid NULL dereference of timing generator

[Why & How]
Check whether assigned timing generator is NULL or not before
accessing its funcs to prevent NULL dereference.

Reviewed-by: Jun Lei <jun.lei@amd.com>
Acked-by: Hersen Wu <hersenxs.wu@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/amdkfd: Update cache info for GFX 9.4.3

Update cache info reporting based on compute and
memory partitioning modes.

Signed-off-by: Mukul Joshi <mukul.joshi@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdkfd: Populate cache info for GFX 9.4.3

GFX 9.4.3 uses a new version of the GC info table which
contains the cache info. This patch adds a new function
to populate the cache info from IP discovery for GFX 9.4.3.

Signed-off-by: Mukul Joshi <mukul.joshi@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: don't put MQDs in VRAM on ARM | ARM64

Issues were reported with commit 1cfb4d612127
("drm/amdgpu: put MQDs in VRAM") on an ADLINK Ampere
Altra Developer Platform (AVA developer platform).

Various ARM systems seem to have problems related
to PCIe and MMIO access.  In this case, I'm not sure
if this is specific to the ADLINK platform or ARM
in general.  Seems to be some coherency issue with
VRAM.  For now, just don't put MQDs in VRAM on ARM.

Link: https://lists.freedesktop.org/archives/amd-gfx/2023-October/100453.html
Fixes: 1cfb4d612127 ("drm/amdgpu: put MQDs in VRAM")
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: alexey.klimov@linaro.org

drm/amdgpu/smu13: drop compute workload workaround

This was fixed in PMFW before launch and is no longer
required.

Reviewed-by: Yang Wang <kevinyang.wang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org # 6.1.x

drm/amdgpu: add a retry for IP discovery init

AMD dGPUs have integrated FW that runs as soon as the
device gets power and initializes the board (determines
the amount of memory, provides configuration details to
the driver, etc.).  For direct PCIe attached cards this
happens as soon as power is applied and normally completes
well before the OS has even started loading.  However, with
hotpluggable ports like USB4, the driver needs to wait for
this to complete before initializing the device.

This normally takes 60-100ms, but could take longer on
some older boards periodically due to memory training.

Retry for up to a second.  In the non-hotplug case, there
should be no change in behavior and this should complete
on the first try.

v2: adjust test criteria
v3: adjust checks for the masks, only enable on removable devices
v4: skip bif_fb_en check

Link: https://gitlab.freedesktop.org/drm/amd/-/issues/2925
Reviewed-by: Mario Limonciello <mario.limonciello@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org

drm/amdgpu: ungate power gating when system suspend

[Why] During suspend, if GFX DPM is enabled and GFXOFF feature is
enabled the system may get hung. So, it is suggested to disable
GFXOFF feature during suspend and enable it after resume.

[How] Update the code to disable GFXOFF feature during suspend and enable
it after resume.

[  311.396526] amdgpu 0000:03:00.0: amdgpu: SMU: I'm not done with your previous command: SMN_C2PMSG_66:0x0000001E SMN_C2PMSG_82:0x00000000
[  311.396530] amdgpu 0000:03:00.0: amdgpu: Fail to disable dpm features!
[  311.396531] [drm:amdgpu_device_ip_suspend_phase2 [amdgpu]] *ERROR* suspend of IP block <smu> failed -62

Acked-by: Yang Wang <kevinyang.wang@amd.com>
Reviewed-by: Kenneth Feng <kenneth.feng@amd.com>
Signed-off-by: Perry Yuan <perry.yuan@amd.com>
Signed-off-by: Kun Liu <kun.liu2@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org

drm/radeon: replace 1-element arrays with flexible-array members

Reported by coccinelle, the following patch will move the
following 1 element arrays to flexible arrays.

drivers/gpu/drm/radeon/atombios.h:5523:32-48: WARNING use flexible-array member instead (https://www.kernel.org/doc/html/latest/process/deprecated.html#zero-length-and-one-element-arrays)
drivers/gpu/drm/radeon/atombios.h:5545:32-48: WARNING use flexible-array member instead (https://www.kernel.org/doc/html/latest/process/deprecated.html#zero-length-and-one-element-arrays)
drivers/gpu/drm/radeon/atombios.h:5461:34-44: WARNING use flexible-array member instead (https://www.kernel.org/doc/html/latest/process/deprecated.html#zero-length-and-one-element-arrays)
drivers/gpu/drm/radeon/atombios.h:4447:30-40: WARNING use flexible-array member instead (https://www.kernel.org/doc/html/latest/process/deprecated.html#zero-length-and-one-element-arrays)
drivers/gpu/drm/radeon/atombios.h:4236:30-41: WARNING use flexible-array member instead (https://www.kernel.org/doc/html/latest/process/deprecated.html#zero-length-and-one-element-arrays)
drivers/gpu/drm/radeon/atombios.h:7095:28-45: WARNING use flexible-array member instead (https://www.kernel.org/doc/html/latest/process/deprecated.html#zero-length-and-one-element-arrays)
drivers/gpu/drm/radeon/atombios.h:3896:27-37: WARNING use flexible-array member instead (https://www.kernel.org/doc/html/latest/process/deprecated.html#zero-length-and-one-element-arrays)
drivers/gpu/drm/radeon/atombios.h:5443:16-25: WARNING use flexible-array member instead (https://www.kernel.org/doc/html/latest/process/deprecated.html#zero-length-and-one-element-arrays)
drivers/gpu/drm/radeon/atombios.h:5454:34-43: WARNING use flexible-array member instead (https://www.kernel.org/doc/html/latest/process/deprecated.html#zero-length-and-one-element-arrays)
drivers/gpu/drm/radeon/atombios.h:4603:21-32: WARNING use flexible-array member instead (https://www.kernel.org/doc/html/latest/process/deprecated.html#zero-length-and-one-element-arrays)
drivers/gpu/drm/radeon/atombios.h:4628:32-46: WARNING use flexible-array member instead (https://www.kernel.org/doc/html/latest/process/deprecated.html#zero-length-and-one-element-arrays)
drivers/gpu/drm/radeon/atombios.h:6285:29-39: WARNING use flexible-array member instead (https://www.kernel.org/doc/html/latest/process/deprecated.html#zero-length-and-one-element-arrays)
drivers/gpu/drm/radeon/atombios.h:4296:30-36: WARNING use flexible-array member instead (https://www.kernel.org/doc/html/latest/process/deprecated.html#zero-length-and-one-element-arrays)
drivers/gpu/drm/radeon/atombios.h:4756:28-36: WARNING use flexible-array member instead (https://www.kernel.org/doc/html/latest/process/deprecated.html#zero-length-and-one-element-arrays)
drivers/gpu/drm/radeon/atombios.h:4064:22-35: WARNING use flexible-array member instead (https://www.kernel.org/doc/html/latest/process/deprecated.html#zero-length-and-one-element-arrays)
drivers/gpu/drm/radeon/atombios.h:7327:9-24: WARNING use flexible-array member instead (https://www.kernel.org/doc/html/latest/process/deprecated.html#zero-length-and-one-element-arrays)
drivers/gpu/drm/radeon/atombios.h:7332:32-53: WARNING use flexible-array member instead (https://www.kernel.org/doc/html/latest/process/deprecated.html#zero-length-and-one-element-arrays)
drivers/gpu/drm/radeon/atombios.h:7362:26-41: WARNING use flexible-array member instead (https://www.kernel.org/doc/html/latest/process/deprecated.html#zero-length-and-one-element-arrays)
drivers/gpu/drm/radeon/atombios.h:7369:29-44: WARNING use flexible-array member instead (https://www.kernel.org/doc/html/latest/process/deprecated.html#zero-length-and-one-element-arrays)
drivers/gpu/drm/radeon/atombios.h:7349:24-32: WARNING use flexible-array member instead (https://www.kernel.org/doc/html/latest/process/deprecated.html#zero-length-and-one-element-arrays)
drivers/gpu/drm/radeon/atombios.h:7355:27-35: WARNING use flexible-array member instead (https://www.kernel.org/doc/html/latest/process/deprecated.html#zero-length-and-one-element-arrays)

Signed-off-by: José Pekkarinen <jose.pekkarinen@foxhound.fi>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd: Fix UBSAN array-index-out-of-bounds for Powerplay headers

For pptable structs that use flexible array sizes, use flexible arrays.

Link: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/2039926
Reviewed-by: Mario Limonciello <mario.limonciello@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: don't use pci_is_thunderbolt_attached()

It's only valid on Intel systems with the Intel VSEC.
Use dev_is_removable() instead.  This should do the right
thing regardless of the platform.

Link: https://gitlab.freedesktop.org/drm/amd/-/issues/2925
Reviewed-by: Mario Limonciello <mario.limonciello@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org

drm/amdgpu: don't use ATRM for external devices

The ATRM ACPI method is for fetching the dGPU vbios rom
image on laptops and all-in-one systems.  It should not be
used for external add in cards.  If the dGPU is thunderbolt
connected, don't try ATRM.

v2: pci_is_thunderbolt_attached only works for Intel.  Use
    pdev->external_facing instead.
v3: dev_is_removable() seems to be what we want

Link: https://gitlab.freedesktop.org/drm/amd/-/issues/2925
Reviewed-by: Mario Limonciello <mario.limonciello@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org

drm/amdgpu/gfx10,11: use memcpy_to/fromio for MQDs

Since they were moved to VRAM, we need to use the IO
variants of memcpy.

Fixes: 1cfb4d612127 ("drm/amdgpu: put MQDs in VRAM")
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: use mode-2 reset for RAS poison consumption

Switch from mode-1 reset to mode-2 for poison consumption.

Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Stanley.Yang <Stanley.Yang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu doorbell range should be set when gpu recovery

GFX doorbell range should be set after flr otherwise the gfx doorbell
range will be overlap with MEC.

v2: remove "amdgpu_sriov_vf" and "amdgpu_in_reset" check, and add grbm
select for the case of 2 gfx rings.

Signed-off-by: Lin.Cao <lincao12@amd.com>
Acked-by: ZhenGuo Yin <zhenguo.yin@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: Return 0 as default min power limit for legacy asics

Return 0 as the default min power limit for the asics use
powerplay.

Signed-off-by: Ma Jun <Jun.Ma2@amd.com>
Reviewed-by: Kenneth Feng <kenneth.feng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: remove unused macro HW_REV

remove unused macro HW_REV

Signed-off-by: Yang Wang <kevinyang.wang@amd.com>
Reviewed-by: Kenneth Feng <kenneth.feng@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: Fix error of MACO flag setting code

MACO only works if BACO is supported

Signed-off-by: Ma Jun <Jun.Ma2@amd.com>
Reviewed-by: Kenneth Feng <kenneth.feng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org # 6.1.x

drm/amdgpu: Fix the vram base start address

If the size returned by drm buddy allocator is higher than
the required size, we take the higher size to calculate
the buffer start address. This is required if we couldn't
trim the buffer to the requested size. This will fix the
display corruption issue on APU's which has limited VRAM
size.

Link: https://gitlab.freedesktop.org/drm/amd/-/issues/2859
Fixes: 0a1844bf0b53 ("drm/buddy: Improve contiguous memory allocation")
Signed-off-by: Arunpravin Paneer Selvam <Arunpravin.PaneerSelvam@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: set XGMI IP version manually for v6_4

The version can't be queried from discovery table.

Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add unmap latency when gfx11 set kiq resources

[why]
If driver does not set unmap latency for KIQ, the default value of KIQ
unmap latency is zero. When do unmap queue, KIQ will return that almost
immediately after receiving unmap command. So, the queue status will be
saved to MQD incorrectly or lost in some chance.

[how]
Set unmap latency when do kiq set resources. The unmap latency is set to
be 1 second that is synchronized with Windows driver.

Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Tong Liu01 <Tong.Liu01@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: fix the high voltage and temperature issue

fix the high voltage and temperature issue after the driver is unloaded on smu 13.0.0,
smu 13.0.7 and smu 13.0.10
v2 - fix the code format and make sure it is used on the unload case only.

Signed-off-by: Kenneth Feng <kenneth.feng@amd.com>
Reviewed-by: Yang Wang <kevinyang.wang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: remove amdgpu_mes_self_test in gpu recover

gpu tlb flush is skipped if reset sem is held, it makes
mes_self_test fail since it involves add_hw_queue/remove_hw_queue
which needs tlb flush functional. Remove mes_self_test in gpu
recover sequence.

This patch is to fix the recover failure in gfx11.

[ 1831.768292] [drm] ring sdma_32769.3.3 was added
[ 1831.768313] [drm] ring gfx_32769.1.1 ib test pass
[ 1831.768337] [drm] ring compute_32769.2.2 ib test pass
[ 1831.768399] amdgpu 0000:c2:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:24 vmid:8 pasid:32769, for process  pid 0 thread  pid 0)
[ 1831.768434] amdgpu 0000:c2:00.0: amdgpu:   in page starting at address 0x0000aec200000000 from client 10
[ 1831.768456] amdgpu 0000:c2:00.0: amdgpu: GCVM_L2_PROTECTION_FAULT_STATUS:0x00800A30
[ 1831.768473] amdgpu 0000:c2:00.0: amdgpu:      Faulty UTCL2 client ID: CPC (0x5)
[ 1831.768489] amdgpu 0000:c2:00.0: amdgpu:      MORE_FAULTS: 0x0
[ 1831.768501] amdgpu 0000:c2:00.0: amdgpu:      WALKER_ERROR: 0x0
[ 1831.768513] amdgpu 0000:c2:00.0: amdgpu:      PERMISSION_FAULTS: 0x3
[ 1831.768521] amdgpu 0000:c2:00.0: amdgpu:      MAPPING_ERROR: 0x0
[ 1831.768529] amdgpu 0000:c2:00.0: amdgpu:      RW: 0x0
[ 1831.931229] amdgpu 0000:c2:00.0: [drm:amdgpu_ring_test_helper [amdgpu]] *ERROR* ring sdma_32769.3.3 test failed (-110)
[ 1832.062917] [drm:mes_v11_0_submit_pkt_and_poll_completion.constprop.0 [amdgpu]] *ERROR* MES failed to response msg=3
[ 1832.063107] [drm:amdgpu_mes_remove_hw_queue [amdgpu]] *ERROR* failed to remove hardware queue, queue id = 3

Fixes: e2e3788850b9 ("drm/amdgpu: rework lock handling for flush_tlb v2")
Reported-by: Li Ma <li.ma@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Yifan Zhang <yifan1.zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Drop deferred error in uncorrectable error check

Drop checking deferred error which can be handled by poison
consumption.

Signed-off-by: Candice Li <candice.li@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: Fix warnings

Fixes warnings:

drivers/gpu/drm/amd/amdgpu/../pm/swsmu/smu13/smu_v13_0_6_ppt.c:286:45:
warning: '%s' directive output may be truncated writing up to 29 bytes
into a region of size 23 [-Wformat-truncation=]
drivers/gpu/drm/amd/amdgpu/../pm/swsmu/smu13/smu_v13_0_6_ppt.c:286:52:
warning: '%s' directive output may be truncated writing up to 29 bytes
into a region of size 23 [-Wformat-truncation=]
drivers/gpu/drm/amd/amdgpu/../pm/swsmu/smu14/smu_v14_0.c:72:45: warning:
'%s' directive output may be truncated writing up to 29 bytes into a
region of size 23 [-Wformat-truncation=]
drivers/gpu/drm/amd/amdgpu/../pm/swsmu/smu14/smu_v14_0.c:72:52: warning:
'%s' directive output may be truncated writing up to 29 bytes into a
region of size 23 [-Wformat-truncation=]

Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Yang Wang <kevinyang.wang@amd.com>
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: check RAS supported first in ras_reset_error_count

Not all platforms support RAS.

Fixes: 73582be11ac8 ("drm/amdgpu: bypass RAS error reset in some conditions")
Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Yang Wang <kevinyang.wang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

Merge tag 'amd-drm-next-6.7-2023-10-27' of https://gitlab.freedesktop.org/agd5f/linux into drm-next

amd-drm-next-6.7-2023-10-27:

amdgpu:
- RAS fixes
- Seamless boot fixes
- NBIO 7.7 fix
- SMU 14.0 fixes
- GC 11.5 fixes
- DML2 fixes
- ASPM fixes
- VPE fixes
- Misc code cleanups
- SRIOV fixes
- Add some missing copyright notices
- DCN 3.5 fixes
- FAMS fixes
- Backlight fix
- S/G display fix
- fdinfo cleanups
- EXT_COHERENT fixes for APU and NUMA systems

amdkfd:
- Misc fixes
- Misc code cleanups
- SVM fixes

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20231027200343.57132-1-alexander.deucher@amd.com

Merge tag 'drm-misc-next-2023-10-27' of git://anongit.freedesktop.org/drm/drm-misc into drm-next

drm-misc-next for v6.7-rc1:

drm-misc-next-2023-10-19 + following:

UAPI Changes:

Cross-subsystem Changes:
- Convert fbdev drivers to use fbdev i/o mem helpers.

Core Changes:
- Use cross-references for macros in docs.
- Make drm_client_buffer_addb use addfb2.
- Add NV20 and NV30 YUV formats.
- Documentation updates for create_dumb ioctl.
- CI fixes.
- Allow variable number of run-queues in scheduler.

Driver Changes:
- Rename drm/ast constants.
- Make ili9882t its own driver.
- Assorted fixes in ivpu, vc4, bridge/synopsis, amdgpu.
- Add planar formats to rockchip.

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/3d92fae8-9b1b-4165-9ca8-5fda11ee146b@linux.intel.com

drm/amdgpu: Remove unused variables from amdgpu_show_fdinfo

Remove unused variables from amdgpu_show_fdinfo

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Umio Yasuno <coelacanth_dream@protonmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Remove duplicate fdinfo fields

Some of the fields that are handled by drm_show_fdinfo() crept back in
when rebasing the patch.  Remove them again.

Fixes: 376c25f8ca47 ("drm/amdgpu: Switch to fdinfo helper")
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Rob Clark <robdclark@chromium.org>
Reviewed-by: <alexander.deucher@amd.com>
Co-developed-by: Umio Yasuno <coelacanth_dream@protonmail.com>
Signed-off-by: Umio Yasuno <coelacanth_dream@protonmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/amdgpu: avoid to disable gfxhub interrupt when driver is unloaded

avoid to disable gfxhub interrupt when driver is unloaded on gmc 11

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

drm/amdgpu: Add EXT_COHERENT support for APU and NUMA systems

On gfx943 APU, EXT_COHERENT should give MTYPE_CC for local and
MTYPE_UC for nonlocal memory.

On NUMA systems, local memory gets the local mtype, set by an
override callback. If EXT_COHERENT is set, memory will be set as
MTYPE_UC by default, with local memory MTYPE_CC.

Add an option in the override function for this case, and
add a check to ensure it is not used on UNCACHED memory.

V2: Combined APU and NUMA code into one patch
V3: Fixed a potential nullptr in amdgpu_vm_bo_update

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

drm/amdgpu: Retrieve CE count from ce_count_lo_chip in EccInfo table

Retrieve correctable error count from ce_count_lo_chip instead of
mca_umc_status.

Signed-off-by: Candice Li <candice.li@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Identify data parity error corrected in replay mode

Use ErrorCodeExt field to identify data parity error in replay mode.

Signed-off-by: Candice Li <candice.li@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Yang Wang <kevinyang.wang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Fix typo in IP discovery parsing

Fix a typo in parsing of the GC info table header when
reading the IP discovery table.

Fixes: 0e64c9aad031 ("drm/amdgpu: add type conversion for gc info")
Signed-off-by: Mukul Joshi <mukul.joshi@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: fix S/G display enablement

An assignment statement was reversed during a refactor which effectively
disabled S/G display outright. Since, we use
adev->mode_info.gpu_vm_support to indicate to the rest of the driver
that S/G display should be enabled and currently it is always set to
false. So, to fix this set adev->mode_info.gpu_vm_support's value to
that of init_data.flags.gpu_vm_support (and not vice versa).

Fixes: 098c13079c6f ("drm/amd/display: enable S/G display for for recent APUs by default")
Reported-by: Mark Broadworth <mark.broadworth@amd.com>
Tested-by: Mark Broadworth <mark.broadworth@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Yifan Zhang <yifan1.zhang@amd.com>
Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdxcp: fix amdxcp unloads incompletely

amdxcp unloads incompletely, and below error will be seen during load/unload,
sysfs: cannot create duplicate filename '/devices/platform/amdgpu_xcp.0'

devres_release_group will free xcp device at first, platform device will be
unregistered later in platform_device_unregister.

Signed-off-by: James Zhu <James.Zhu@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/amdgpu: fix the GPU power print error in pm info

Modify the print format of the fractional part to avoid display error.

Signed-off-by: Li Ma <li.ma@amd.com>
Reviewed-by: Yifan Zhang <yifan1.zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Use pcie domain of xcc acpi objects

PCI domain/segment information of xccs is available through ACPI DSM
methods. Consider that also while looking for devices.

Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Rajneesh Bhardwaj <rajneesh.bhardwaj@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd: check num of link levels when update pcie param

In SR-IOV environment, the value of pcie_table->num_of_link_levels will
be 0, and num_of_levels - 1 will cause array index out of bounds

Signed-off-by: Lin.Cao <lincao12@amd.com>
Acked-by: Jingwen Chen <Jingwen.Chen2@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Add a read to GFX v9.4.3 ring test

Issue a read to confirm the register write before ringing doorbell. With
multiple XCCs there is chance for race condition.

Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Acked-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Asad Kamal <asad.kamal@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: call smu_cmn_get_smc_version in is_mode1_reset_supported.

is_mode1_reset_supported may be called before smu init, when smu_context
is unitialized in driver load/unload test. Call smu_cmn_get_smc_version
explicitly in is_mode1_reset_supported.

v2: apply to aldebaran in case is_mode1_reset_supported will be
uncommented (Candice Li)

Fixes: 710d9caec70c ("drm/amd/pm: drop most smu_cmn_get_smc_version in smu")
Signed-off-by: Yifan Zhang <yifan1.zhang@amd.com>
Reviewed-by: Candice Li <candice.li@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: get RAS poison status from DF v4_6_2

Add DF block and RAS poison mode query for DF v4_6_2.

Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Stanley.Yang <Stanley.Yang@amd.com>
Acked-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Use discovery table's subrevision

Use subrevision of IP version in discovery table to identify SOC
revision id for NBIO v7.9 SOCs. Only newer bootloaders update
subrevision field.

Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Le Ma <le.ma@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: 3.2.256

DC v3.2.256
Summary:
* Fixes null-deref regression after
  "drm/amd/display: Update OPP counter from new interface"
* Fixes display flashing when VSR and HDR enabled on dcn32
* Fixes dcn3x intermittent hangs due to FPO
* Fixes MST Multi-Stream light up on dcn35
* Fixes green screen on DCN31x when DVI and HDMI monitors attached
* Adds DML2 improvements
* Adds idle power optimization improvements
* Accommodates panels with lower nit backlight
* Updates SDP VSC colorimetry from DP test automation request
* Reverts "drm/amd/display: allow edp updates for virtual signal"

Acked-by: Roman Li <roman.li@amd.com>
Signed-off-by: Aric Cyr <aric.cyr@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: add interface to query SubVP status

[Why&How]
To enable automated testing through IGT, expose an API that is
accessible through debugfs to query current status of SubVP feature.

Reviewed-by: Alvin Lee <alvin.lee2@amd.com>
Acked-by: Roman Li <roman.li@amd.com>
Signed-off-by: Aurabindo Pillai <aurabindo.pillai@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Read before writing Backlight Mode Set Register

[HOW&WHY]
Reading the value from
DP_EDP_BACKLIGHT_MODE_SET_REGISTER, DPCD 0x721
before setting the
BP_EDP_PANEL_LUMINANC_CONTROL_ENABLE bit
to ensure there are no accidental overwrites.

Reviewed-by: Sreeja Golui <sreeja.golui@amd.com>
Reviewed-by: Harry Vanzylldejong <harry.vanzylldejong@amd.com>
Acked-by: Roman Li <roman.li@amd.com>
Signed-off-by: Iswara Nagulendran <iswara.nagulendran@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Disable SYMCLK32_SE RCO on DCN314

[WHY]
Currently causes some DP link layer failures, backing out until
the failures are root caused.

Reviewed-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Acked-by: Roman Li <roman.li@amd.com>
Signed-off-by: Michael Strauss <michael.strauss@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Fix shaper using bad LUT params

[Why]
LUT params are not cleared after setting blend TF, which can lead to
same params being used for the shaper, if the shaper func is bypassed.

[How]
Set lut_params to NULL after program_1dlut.

Reviewed-by: Krunoslav Kovac <krunoslav.kovac@amd.com>
Acked-by: Roman Li <roman.li@amd.com>
Signed-off-by: Ilya Bakoulin <ilya.bakoulin@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: add null check for invalid opps

[Why]
In cases where number of pipes available is less
 than num_opp, there will opp instances that are
 null

[How]
Add null check to skip over these opp instances

Fixes: 40de8403b998 ("drm/amd/display: Update OPP counter from new interface")
Reviewed-by: Alvin Lee <alvin.lee2@amd.com>
Acked-by: Roman Li <roman.li@amd.com>
Signed-off-by: Samson Tam <samson.tam@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Update FAMS sequence for DCN30 & DCN32

Provide DCN32 specific sequence and update DCN30 sequence

Reviewed-by: Samson Tam <samson.tam@amd.com>
Acked-by: Roman Li <roman.li@amd.com>
Signed-off-by: Alvin Lee <alvin.lee2@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: fix num_ways overflow error

[Why]
Helper function calculates num_ways using 32-bit.  But is
 returned as 8-bit.  If num_ways exceeds 8-bit, then it
 reports back the incorrect num_ways and erroneously
 uses MALL when it should not

[How]
Make returned value 32-bit and convert after it checks
 against caps.cache_num_ways, which is under 8-bit

Reviewed-by: Alvin Lee <alvin.lee2@amd.com>
Acked-by: Roman Li <roman.li@amd.com>
Signed-off-by: Samson Tam <samson.tam@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Add prefix for plane functions

This commit adds the amdgpu_dm_plane_ prefix for all functions in the
amdgpu_dm_plane.c. This change enables an easy way to filter code paths
via ftrace.

Reviewed-by: Aurabindo Pillai <aurabindo.pillai@amd.com>
Acked-by: Roman Li <roman.li@amd.com>
Signed-off-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Add prefix to amdgpu crtc functions

The ftrace debug feature allows filtering functions based on a prefix,
which can be helpful in some complex debug scenarios. The driver can
benefit more from this feature if the function name follows some
patterns; for this reason, this commit adds the prefix amdgpu_dm_crtc_
to all the functions that do not have it in the amdgpu_dm_crtc.c file.

Reviewed-by: Aurabindo Pillai <aurabindo.pillai@amd.com>
Acked-by: Roman Li <roman.li@amd.com>
Signed-off-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Correct enum typo

This commit just replaces dc_interrupt_po*r*larity with its correct
name, which is dc_interrupt_polarity.

Reviewed-by: Aurabindo Pillai <aurabindo.pillai@amd.com>
Acked-by: Roman Li <roman.li@amd.com>
Signed-off-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Set emulated sink type to HDMI accordingly.

[WHY & HOW]
Virtual sink is not audio-capable and this causes kms_hdmi_inject's
inject-audio to fail. Set it to HDMI according to EDID.

Reviewed-by: Chao-kai Wang <stylon.wang@amd.com>
Acked-by: Roman Li <roman.li@amd.com>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Revert "drm/amd/display: allow edp updates for virtual signal"

This reverts commit 4ad3ee5ccc77aa3f9d702f7b9ad4d9cfeca6c443.

[WHY & HOW]
Virtual signal is not supported as audio capable by DC.

Reviewed-by: Chao-kai Wang <stylon.wang@amd.com>
Acked-by: Roman Li <roman.li@amd.com>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Fix HDMI framepack 3D test issue

[why]
Bandwidth validation failure on framepack tests.
Need to double pixel clock when 3D format is
framepack. Also for HDMI displays, we need to
keep the ITC flag to 1 by default.

[how]
Double the pixel clock when using framepack 3D format.
Set hdmi ITC bit to 1.

Reviewed-by: Charlene Liu <charlene.liu@amd.com>
Acked-by: Roman Li <roman.li@amd.com>
Signed-off-by: Sung Joon Kim <sungkim@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Fix IPS handshake for idle optimizations

[Why]
Intermittent reboot hangs are observed introduced by
"Improve x86 and dmub ips handshake".

[How]
Bring back the commit but fix the polling.

Avoid hanging in place forever by bounding the delay and ensure that
we still message DMCUB on IPS2 exit to notify driver idle has been
cleared.

Reviewed-by: Duncan Ma <duncan.ma@amd.com>
Reviewed-by: Jun Lei <jun.lei@amd.com>
Acked-by: Roman Li <roman.li@amd.com>
Signed-off-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: implement map dc pipe with callback in DML2

[why]
Unify pipe resource management logic in dc resource layer.

V2:
Add default case for switch.

CC: Hamza Mahfooz <hamza.mahfooz@amd.com>
Reviewed-by: Chaitanya Dhere <chaitanya.dhere@amd.com>
Signed-off-by: Wenjing Liu <wenjing.liu@amd.com>
Reviewed-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com>
Reviewed-by: Jun Lei <jun.lei@amd.com>
Acked-by: Roman Li <roman.li@amd.com>
Signed-off-by: Qingqing Zhuo <qingqing.zhuo@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: add pipe resource management callbacks to DML2

[why]
Need DML2 to support new pipe resource management APIs.

Reviewed-by: Chaitanya Dhere <chaitanya.dhere@amd.com>
Acked-by: Roman Li <roman.li@amd.com>
Signed-off-by: Wenjing Liu <wenjing.liu@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Reduce default backlight min from 5 nits to 1 nits

[Why & How]
Currently set_default_brightness_aux function uses 5 nits as lower limit
to check for valid default_backlight setting. However some newer panels
can support even lower default settings

Reviewed-by: Agustin Gutierrez <agustin.gutierrez@amd.com>
Acked-by: Roman Li <roman.li@amd.com>
Signed-off-by: Swapnil Patel <swapnil.patel@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Update SDP VSC colorimetry from DP test automation request

[Why]
Certain test equipment vendors check the SDP VSC for colorimetry against
the value from the test request during certain DP link layer tests for
YCbCr test cases.

[How]
Update SDP VSC with colorimetry from test automation request.

Reviewed-by: Wenjing Liu <wenjing.liu@amd.com>
Acked-by: Roman Li <roman.li@amd.com>
Signed-off-by: George Shen <george.shen@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Add a check for idle power optimization

[why]
Need a helper function to check idle power is allowed
so that dc doesn't access any registers that are power-gated.

[how]
Implement helper function to check idle power optimization.
Enable a hook to check if detection is allowed.

V2:
Add function hooks for set and get idle states.
Check if function hook was properly initialized.

Reviewed-by: Aric Cyr <aric.cyr@amd.com>
Reviewed-by: Nicholas Choi <nicholas.choi@amd.com>
Acked-by: Roman Li <roman.li@amd.com>
Signed-off-by: Sung Joon Kim <sungkim@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Revert "Improve x86 and dmub ips handshake"

This reverts commit 1288d702080949f87688d49dfeeacc99f40adc9b.

Causes intermittent hangs during reboot stress testing.

Reviewed-by: Duncan Ma <duncan.ma@amd.com>
Acked-by: Roman Li <roman.li@amd.com>
Signed-off-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Fix MST Multi-Stream Not Lighting Up on dcn35

dcn35 misses .enable_symclk_se hook that makes MST DSC
not functional when having multiple FE clk to be enabled.

Reviewed-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com>
Acked-by: Roman Li <roman.li@amd.com>
Signed-off-by: Fangzhi Zuo <jerry.zuo@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd: Explicitly disable ASPM when dynamic switching disabled

Currently there are separate but related checks:
* amdgpu_device_should_use_aspm()
* amdgpu_device_aspm_support_quirk()
* amdgpu_device_pcie_dynamic_switching_supported()

Simplify into checking whether DPM was enabled or not in the auto
case.  This works because amdgpu_device_pcie_dynamic_switching_supported()
populates that value.

Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd: Move AMD_IS_APU check for ASPM into top level function

There is no need for every ASIC driver to perform the same check.
Move the duplicated code into amdgpu_device_should_use_aspm().

Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd: Disable PP_PCIE_DPM_MASK when dynamic speed switching not supported

Rather than individual ASICs checking for the quirk, set the quirk at the
driver level.

Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

Revert "drm/amdkfd: Use partial migrations in GPU page faults"

This reverts commit dc427a473e5d119232ddb27530920d9796cdea70.

The change prevents migrating the entire range to VRAM because retry
fault restore_pages map the remaining system memory range to GPUs. It
will work correctly to submit together with partial mapping to GPU
patch later.

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>

Revert "drm/amdkfd:remove unused code"

This reverts commit f9caf6cdd5cc1f4006fd7b6b113658c0b0159f23.

Needed for the next revert patch.

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/display: Fix copyright notice in DC code

[Why & How]
Fix incomplete copyright notice in DC code.

Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Stylon Wang <stylon.wang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Fix copyright notice in DML2 code

[Why & How]
Fix incomplete copyright notice in DML2 code.

Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Stylon Wang <stylon.wang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Add missing copyright notice in DMUB

[Why & How]
Add missing/incomplete copyright notice in DMUB files

Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Stylon Wang <stylon.wang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu remove restriction of sriov max_pfn on Vega10

Remove restriction of sriov max_pfn so that TBA and TMA can move to high
47 bits address.

Regression test: change range alloc flag of libdrm as
AMDGPU_VA_RANGE_HIGH and there is no flr occur when testing amdgpu_test
of drm.

Signed-off-by: Lin.Cao <lincao12@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdkfd: Address 'remap_list' not described in 'svm_range_add'

Fixes the below:

drivers/gpu/drm/amd/amdgpu/../amdkfd/kfd_svm.c:2073: warning: Function parameter or member 'remap_list' not described in 'svm_range_add'

Cc: Felix Kuehling <Felix.Kuehling@amd.com>
Cc: Christian König <christian.koenig@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: "Pan, Xinhui" <Xinhui.Pan@amd.com>
Signed-off-by: Srinivasan Shanmugam <srinivasan.shanmugam@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Fix a null pointer access when the smc_rreg pointer is NULL

In certain types of chips, such as VEGA20, reading the amdgpu_regs_smc file could result in an abnormal null pointer access when the smc_rreg pointer is NULL. Below are the steps to reproduce this issue and the corresponding exception log:

1. Navigate to the directory: /sys/kernel/debug/dri/0
2. Execute command: cat amdgpu_regs_smc
3. Exception Log::
[4005007.702554] BUG: kernel NULL pointer dereference, address: 0000000000000000
[4005007.702562] #PF: supervisor instruction fetch in kernel mode
[4005007.702567] #PF: error_code(0x0010) - not-present page
[4005007.702570] PGD 0 P4D 0
[4005007.702576] Oops: 0010 [#1] SMP NOPTI
[4005007.702581] CPU: 4 PID: 62563 Comm: cat Tainted: G           OE     5.15.0-43-generic #46-Ubunt       u
[4005007.702590] RIP: 0010:0x0
[4005007.702598] Code: Unable to access opcode bytes at RIP 0xffffffffffffffd6.
[4005007.702600] RSP: 0018:ffffa82b46d27da0 EFLAGS: 00010206
[4005007.702605] RAX: 0000000000000000 RBX: 0000000000000000 RCX: ffffa82b46d27e68
[4005007.702609] RDX: 0000000000000001 RSI: 0000000000000000 RDI: ffff9940656e0000
[4005007.702612] RBP: ffffa82b46d27dd8 R08: 0000000000000000 R09: ffff994060c07980
[4005007.702615] R10: 0000000000020000 R11: 0000000000000000 R12: 00007f5e06753000
[4005007.702618] R13: ffff9940656e0000 R14: ffffa82b46d27e68 R15: 00007f5e06753000
[4005007.702622] FS:  00007f5e0755b740(0000) GS:ffff99479d300000(0000) knlGS:0000000000000000
[4005007.702626] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[4005007.702629] CR2: ffffffffffffffd6 CR3: 00000003253fc000 CR4: 00000000003506e0
[4005007.702633] Call Trace:
[4005007.702636]  <TASK>
[4005007.702640]  amdgpu_debugfs_regs_smc_read+0xb0/0x120 [amdgpu]
[4005007.703002]  full_proxy_read+0x5c/0x80
[4005007.703011]  vfs_read+0x9f/0x1a0
[4005007.703019]  ksys_read+0x67/0xe0
[4005007.703023]  __x64_sys_read+0x19/0x20
[4005007.703028]  do_syscall_64+0x5c/0xc0
[4005007.703034]  ? do_user_addr_fault+0x1e3/0x670
[4005007.703040]  ? exit_to_user_mode_prepare+0x37/0xb0
[4005007.703047]  ? irqentry_exit_to_user_mode+0x9/0x20
[4005007.703052]  ? irqentry_exit+0x19/0x30
[4005007.703057]  ? exc_page_fault+0x89/0x160
[4005007.703062]  ? asm_exc_page_fault+0x8/0x30
[4005007.703068]  entry_SYSCALL_64_after_hwframe+0x44/0xae
[4005007.703075] RIP: 0033:0x7f5e07672992
[4005007.703079] Code: c0 e9 b2 fe ff ff 50 48 8d 3d fa b2 0c 00 e8 c5 1d 02 00 0f 1f 44 00 00 f3 0f        1e fa 64 8b 04 25 18 00 00 00 85 c0 75 10 0f 05 <48> 3d 00 f0 ff ff 77 56 c3 0f 1f 44 00 00 48 83 e       c 28 48 89 54 24
[4005007.703083] RSP: 002b:00007ffe03097898 EFLAGS: 00000246 ORIG_RAX: 0000000000000000
[4005007.703088] RAX: ffffffffffffffda RBX: 0000000000020000 RCX: 00007f5e07672992
[4005007.703091] RDX: 0000000000020000 RSI: 00007f5e06753000 RDI: 0000000000000003
[4005007.703094] RBP: 00007f5e06753000 R08: 00007f5e06752010 R09: 00007f5e06752010
[4005007.703096] R10: 0000000000000022 R11: 0000000000000246 R12: 0000000000022000
[4005007.703099] R13: 0000000000000003 R14: 0000000000020000 R15: 0000000000020000
[4005007.703105]  </TASK>
[4005007.703107] Modules linked in: nf_tables libcrc32c nfnetlink algif_hash af_alg binfmt_misc nls_       iso8859_1 ipmi_ssif ast intel_rapl_msr intel_rapl_common drm_vram_helper drm_ttm_helper amd64_edac t       tm edac_mce_amd kvm_amd ccp mac_hid k10temp kvm acpi_ipmi ipmi_si rapl sch_fq_codel ipmi_devintf ipm       i_msghandler msr parport_pc ppdev lp parport mtd pstore_blk efi_pstore ramoops pstore_zone reed_solo       mon ip_tables x_tables autofs4 ib_uverbs ib_core amdgpu(OE) amddrm_ttm_helper(OE) amdttm(OE) iommu_v       2 amd_sched(OE) amdkcl(OE) drm_kms_helper syscopyarea sysfillrect sysimgblt fb_sys_fops cec rc_core        drm igb ahci xhci_pci libahci i2c_piix4 i2c_algo_bit xhci_pci_renesas dca
[4005007.703184] CR2: 0000000000000000
[4005007.703188] ---[ end trace ac65a538d240da39 ]---
[4005007.800865] RIP: 0010:0x0
[4005007.800871] Code: Unable to access opcode bytes at RIP 0xffffffffffffffd6.
[4005007.800874] RSP: 0018:ffffa82b46d27da0 EFLAGS: 00010206
[4005007.800878] RAX: 0000000000000000 RBX: 0000000000000000 RCX: ffffa82b46d27e68
[4005007.800881] RDX: 0000000000000001 RSI: 0000000000000000 RDI: ffff9940656e0000
[4005007.800883] RBP: ffffa82b46d27dd8 R08: 0000000000000000 R09: ffff994060c07980
[4005007.800886] R10: 0000000000020000 R11: 0000000000000000 R12: 00007f5e06753000
[4005007.800888] R13: ffff9940656e0000 R14: ffffa82b46d27e68 R15: 00007f5e06753000
[4005007.800891] FS:  00007f5e0755b740(0000) GS:ffff99479d300000(0000) knlGS:0000000000000000
[4005007.800895] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[4005007.800898] CR2: ffffffffffffffd6 CR3: 00000003253fc000 CR4: 00000000003506e0

Signed-off-by: Qu Huang <qu.huang@linux.dev>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: bypass RAS error reset in some conditions

PMFW is responsible for RAS error reset in some conditions, driver can
skip the operation.

v2: add check for ras->in_recovery, it's set earlier than
amdgpu_in_reset.

v3: fix error in gpu reset check.

Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: enable RAS poison mode for APU

Enable it by default on APU platform.

Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Stanley.Yang <Stanley.Yang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/vpe: correct queue stop programing

Otherwise IB test would fail during GPU reset.

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

drm/amd/display: Fix DMUB errors introduced by DML2

When DML 2 was introduced, it changed part of the generic sequence of
DC, which caused issues on previous DCNs with DMUB support. This commit
ensures the new sequence only works for new DCNs from 3.5 and above.

Changes since V1:
- Harry: Use the attribute using_dml2 instead of check the DCN version.

Cc: Vitaly Prosyak <vprosyak@amd.com>
Cc: Roman Li <roman.li@amd.com>
Cc: Qingqing Zhuo <Qingqing.Zhuo@amd.com>
Cc: Daniel Wheeler <daniel.wheeler@amd.com>
Cc: Alex Deucher <Alexander.Deucher@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Fixes: 7966f319c66d ("drm/amd/display: Introduce DML2")
Signed-off-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd: Disable ASPM for VI w/ all Intel systems

Originally we were quirking ASPM disabled specifically for VI when
used with Alder Lake, but it appears to have problems with Rocket
Lake as well.

Like we've done in the case of dpm for newer platforms, disable
ASPM for all Intel systems.

Cc: stable@vger.kernel.org # 5.15+
Fixes: 0064b0ce85bb ("drm/amd/pm: enable ASPM by default")
Reported-and-tested-by: Paolo Gentili <paolo.gentili@canonical.com>
Closes: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/2036742
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Remove power sequencing check

[Why]
Some ASICs keep backlight powered on after dpms off
command has been issued.

[How]
The check for no edp power sequencing was never going to pass.
The value is never changed from what it is set by design.

Cc: stable@vger.kernel.org # 6.1+
Link: https://gitlab.freedesktop.org/drm/amd/-/issues/2765
Reviewed-by: Swapnil Patel <swapnil.patel@amd.com>
Acked-by: Roman Li <roman.li@amd.com>
Signed-off-by: Agustin Gutierrez <agustin.gutierrez@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Set the DML2 attribute to false in all DCNs older than version 3.5

When DML2 was introduced, it targeted only new DCN versions. For
controlling which ASIC should use this new version of DML, it was
introduced the using_dml2 attribute. To avoid ambiguities, this commit
explicitly sets using_dml2 to false in all ASICs that do not support
DML2.

Cc: Vitaly Prosyak <vprosyak@amd.com>
Cc: Roman Li <roman.li@amd.com>
Cc: Qingqing Zhuo <Qingqing.Zhuo@amd.com>
Cc: Daniel Wheeler <daniel.wheeler@amd.com>
Cc: Alex Deucher <Alexander.Deucher@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: Fix the return value in default case

Fix the return value in default case and drop
redundant 'break'.

Signed-off-by: Ma Jun <Jun.Ma2@amd.com>
Reviewed-by: Kenneth Feng <kenneth.feng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Add API to get full IP version

Fetch the full version of IP including variant and subrevision.

Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Le Ma <le.ma@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add tmz support for GC IP v11.5.0

Add tmz support for GC 11.5.0.

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

drm/amd/pm: drop unneeded dpm features disablement for SMU 14.0.0

PMFW will handle the features disablement properly for gpu reset case,
driver involvement may cause some unexpected issues.

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

drm/amdgpu: modify if condition in nbio_v7_7.c

remove unnecessary "enable" in if condition.

Signed-off-by: Li Ma <li.ma@amd.com>
Reviewed-by: Tim Huang <Tim.Huang@amd.com>
Reviewed-by: Yifan Zhang <yifan1.zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdkfd: Fix shift out-of-bounds issue

[  567.613292] shift exponent 255 is too large for 64-bit type 'long unsigned int'
[  567.614498] CPU: 5 PID: 238 Comm: kworker/5:1 Tainted: G           OE      6.2.0-34-generic #34~22.04.1-Ubuntu
[  567.614502] Hardware name: AMD Splinter/Splinter-RPL, BIOS WS43927N_871 09/25/2023
[  567.614504] Workqueue: events send_exception_work_handler [amdgpu]
[  567.614748] Call Trace:
[  567.614750]  <TASK>
[  567.614753]  dump_stack_lvl+0x48/0x70
[  567.614761]  dump_stack+0x10/0x20
[  567.614763]  __ubsan_handle_shift_out_of_bounds+0x156/0x310
[  567.614769]  ? srso_alias_return_thunk+0x5/0x7f
[  567.614773]  ? update_sd_lb_stats.constprop.0+0xf2/0x3c0
[  567.614780]  svm_range_split_by_granularity.cold+0x2b/0x34 [amdgpu]
[  567.615047]  ? srso_alias_return_thunk+0x5/0x7f
[  567.615052]  svm_migrate_to_ram+0x185/0x4d0 [amdgpu]
[  567.615286]  do_swap_page+0x7b6/0xa30
[  567.615291]  ? srso_alias_return_thunk+0x5/0x7f
[  567.615294]  ? __free_pages+0x119/0x130
[  567.615299]  handle_pte_fault+0x227/0x280
[  567.615303]  __handle_mm_fault+0x3c0/0x720
[  567.615311]  handle_mm_fault+0x119/0x330
[  567.615314]  ? lock_mm_and_find_vma+0x44/0x250
[  567.615318]  do_user_addr_fault+0x1a9/0x640
[  567.615323]  exc_page_fault+0x81/0x1b0
[  567.615328]  asm_exc_page_fault+0x27/0x30
[  567.615332] RIP: 0010:__get_user_8+0x1c/0x30

Signed-off-by: Jesse Zhang <jesse.zhang@amd.com>
Suggested-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Yifan Zhang <yifan1.zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: refine ras error kernel log print

refine ras error kernel log to avoid user-ridden ambiguity.

Signed-off-by: Yang Wang <kevinyang.wang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: fix find ras error node error

the origin function might return the wrong node.

Fixes: 5b1270beb380 ("drm/amdgpu: add ras_err_info to identify RAS error source")
Signed-off-by: Yang Wang <kevinyang.wang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: reprogram det size while seamless boot

[Why]
During system boot in second screen only mode on a seamless boot system,
there is a chance that the pipe's det size might not be reset.

[How]
Reset the det size while resetting the pipe during seamless boot.

Reviewed-by: Dmytro Laktyushkin <dmytro.laktyushkin@amd.com>
Acked-by: Roman Li <roman.li@amd.com>
Signed-off-by: Hugo Hu <hugo.hu@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: record mca debug mode in RAS

Call amdgpu_ras_set_mca_debug_mode when we set mca debug mode in smu
v13_0_6.

Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

usb: typec: altmodes/displayport: fixup drm internal api change vs new user.

usb: typec: altmodes/displayport: Signal hpd low when exiting mode
and
drm: Add HPD state to drm_connector_oob_hotplug_event()

sideswiped each other.

Signal disconnected always.

Signed-off-by: Dave Airlie <airlied@redhat.com>

drm/amdgpu: move buffer funcs setting up a level

Rather than doing this in the IP code for the SDMA paging
engine, move it up to the core device level init level.
This should fix the scheduler init ordering.

v2: drop extra parens
v3: drop SDMA helpers
v4: Added a Fixes tag because amdgpu dereferences an uninitialized
    scheduler without this patch, and this patch fixes this. (Luben)

Tested-by: Luben Tuikov <luben.tuikov@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Link: https://lore.kernel.org/r/20231025171928.3318505-1-alexander.deucher@amd.com
Acked-by: Christian König <christian.koenig@amd.com>
Fixes: 56e449603f0ac5 ("drm/sched: Convert the GPU scheduler to variable number of run-queues")
Signed-off-by: Luben Tuikov <ltuikov89@gmail.com>

MAINTAINERS: Update the GPU Scheduler email

Update the GPU Scheduler maintainer email.

Cc: Alex Deucher <Alexander.Deucher@amd.com>
Cc: Christian König <christian.koenig@amd.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Dave Airlie <airlied@gmail.com>
Cc: AMD Graphics <amd-gfx@lists.freedesktop.org>
Cc: Direct Rendering Infrastructure - Development <dri-devel@lists.freedesktop.org>
Signed-off-by: Luben Tuikov <ltuikov89@gmail.com>
Acked-by: Alex Deucher <Alexander.Deucher@amd.com>
Link: https://lore.kernel.org/r/20231026174438.18427-2-ltuikov89@gmail.com

drm/sched: Convert the GPU scheduler to variable number of run-queues

The GPU scheduler has now a variable number of run-queues, which are set up at
drm_sched_init() time. This way, each driver announces how many run-queues it
requires (supports) per each GPU scheduler it creates. Note, that run-queues
correspond to scheduler "priorities", thus if the number of run-queues is set
to 1 at drm_sched_init(), then that scheduler supports a single run-queue,
i.e. single "priority". If a driver further sets a single entity per
run-queue, then this creates a 1-to-1 correspondence between a scheduler and
a scheduled entity.

Cc: Lucas Stach <l.stach@pengutronix.de>
Cc: Russell King <linux+etnaviv@armlinux.org.uk>
Cc: Qiang Yu <yuq825@gmail.com>
Cc: Rob Clark <robdclark@gmail.com>
Cc: Abhinav Kumar <quic_abhinavk@quicinc.com>
Cc: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Cc: Danilo Krummrich <dakr@redhat.com>
Cc: Matthew Brost <matthew.brost@intel.com>
Cc: Boris Brezillon <boris.brezillon@collabora.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Christian König <christian.koenig@amd.com>
Cc: Emma Anholt <emma@anholt.net>
Cc: etnaviv@lists.freedesktop.org
Cc: lima@lists.freedesktop.org
Cc: linux-arm-msm@vger.kernel.org
Cc: freedreno@lists.freedesktop.org
Cc: nouveau@lists.freedesktop.org
Cc: dri-devel@lists.freedesktop.org
Signed-off-by: Luben Tuikov <luben.tuikov@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Link: https://lore.kernel.org/r/20231023032251.164775-1-luben.tuikov@amd.com

MAINTAINERS: drm/ci: add entries for xfail files

DRM CI keeps track of which tests are failing, flaking or being skipped
by the ci in the expectations files. Add entries for those files to the
corresponding driver maintainer, so they can be notified when they
change.

Signed-off-by: Helen Koike <helen.koike@collabora.com>
Reviewed-by: Neil Armstrong <neil.armstrong@linaro.org>
Link: https://lore.kernel.org/r/20230919182249.153499-1-helen.koike@collabora.com
Signed-off-by: Maxime Ripard <mripard@kernel.org>

drm/ci: docs: add step about how to request privileges

Clarify the procedure developer must follow to request privileges to
run tests on Freedesktop gitlab CI.

This measure was added to avoid untrusted people to misuse the
infrastructure.

Signed-off-by: Helen Koike <helen.koike@collabora.com>
Reviewed-by: David Heidelberg <david.heidelberg@collabora.com>
Link: https://lore.kernel.org/r/20231024004525.169002-11-helen.koike@collabora.com
Signed-off-by: Maxime Ripard <mripard@kernel.org>