Merge tag 'amd-drm-fixes-6.9-2024-03-21' of https://gitlab.freedesktop.org/agd5f/linux into drm-next

amd-drm-fixes-6.9-2024-03-21:

amdgpu:
- Freesync fixes
- UAF IOCTL fixes
- Fix mmhub client ID mapping
- IH 7.0 fix
- DML2 fixes
- VCN 4.0.6 fix
- GART bind fix
- GPU reset fix
- SR-IOV fix
- OD table handling fixes
- Fix TA handling on boards without display hardware
- DML1 fix
- ABM fix
- eDP panel fix
- DPPCLK fix
- HDCP fix
- Revert incorrect error case handling in ioremap
- VPE fix
- HDMI fixes
- SDMA 4.4.2 fix
- Other misc fixes

amdkfd:
- Fix duplicate BO handling in process restore

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240321220514.1418288-1-alexander.deucher@amd.com

Merge tag 'drm-misc-next-fixes-2024-03-21' of https://gitlab.freedesktop.org/drm/misc/kernel into drm-next

Short summary of fixes pull:

core:
- fix rounding in drm_fixp2int_round()

bridge:
- fix documentation for DRM_BRIDGE_OP_EDID

nouveau:
- don't check devinit disable on GSP

sun4i:
- fix 64-bit division on 32-bit architectures

tests:
- fix dependency on DRM_KMS_HELPER

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Thomas Zimmermann <tzimmermann@suse.de>
Link: https://patchwork.freedesktop.org/patch/msgid/20240321161948.GA30430@linux.fritz.box

drm/amdgpu/pm: Don't use OD table on Arcturus

OD is not supported on Arcturus, so the OD table
should not be used.

Signed-off-by: Ma Jun <Jun.Ma2@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: drop setting buffer funcs in sdma442

To fix the entity rq NULL issue. This setting has been moved
to upper level.

Fixes: b70438004a14 ("drm/amdgpu: move buffer funcs setting up a level")
Signed-off-by: Le Ma <le.ma@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Fix noise issue on HDMI AV mute

[Why]
When mode switching is triggered there is momentary noise visible on
some HDMI TV or displays.

[How]
Wait for 2 frames to make sure we have enough time to send out AV mute
and sink receives a full frame.

Cc: Mario Limonciello <mario.limonciello@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Reviewed-by: Wenjing Liu <wenjing.liu@amd.com>
Acked-by: Wayne Lin <wayne.lin@amd.com>
Signed-off-by: Leo Ma <hanghong.ma@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Revert Remove pixle rate limit for subvp

This reverts commit 340383c734f8 ("drm/amd/display: Remove pixle rate
limit for subvp")

[why]
The original commit causes a regression when subvp is applied
on ODM required 8k60hz timing. The display shows black screen
on boot. The issue can be recovered with hotplug. It also causes
MPO to fail. We will temprarily revert this commit and investigate
the root cause further.

Cc: Mario Limonciello <mario.limonciello@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Reviewed-by: Chaitanya Dhere <chaitanya.dhere@amd.com>
Reviewed-by: Martin Leung <martin.leung@amd.com>
Acked-by: Wayne Lin <wayne.lin@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>

Revert "drm/amdgpu/vpe: don't emit cond exec command under collaborate mode"

Ready now. Remove this workaround.
This reverts commit d40f6213b52c161fd4634933acbc32103a283363.

Signed-off-by: Lang Yu <Lang.Yu@amd.com>
Tested-by: Alan Liu <haoping.liu@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

Revert "drm/amd/amdgpu: Fix potential ioremap() memory leaks in amdgpu_device_init()"

This patch causes the following iounmap erorr and calltrace
iounmap: bad address 00000000d0b3631f

The original patch was unjustified because amdgpu_device_fini_sw() will
always cleanup the rmmio mapping.

This reverts commit eb4f139888f636614dab3bcce97ff61cefc4b3a7.

Signed-off-by: Ma Jun <Jun.Ma2@amd.com>
Suggested-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Add a dc_state NULL check in dc_state_release

[How]
Check wheather state is NULL before releasing it.

Cc: Mario Limonciello <mario.limonciello@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Reviewed-by: Charlene Liu <charlene.liu@amd.com>
Acked-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Allen Pan <allen.pan@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Return the correct HDCP error code

[WHY & HOW]
If the display is null when creating an HDCP session, return a proper
error code.

Cc: Mario Limonciello <mario.limonciello@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Acked-by: Alex Hung <alex.hung@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: Implement wait_for_odm_update_pending_complete

[WHY]
Odm update is doubled buffered. We need to wait for ODM update to be
completed before optimizing bandwidth or programming new udpates.

[HOW]
implement wait_for_odm_update_pending_complete function to wait for:
1. odm configuration update is no longer pending in timing generator.
2. no pending dpg pattern update for each active OPP.

Cc: Mario Limonciello <mario.limonciello@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Reviewed-by: Alvin Lee <alvin.lee2@amd.com>
Acked-by: Alex Hung <alex.hung@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: Lock all enabled otg pipes even with no planes

[WHY]
On DCN32 we support dynamic ODM even when OTG is blanked. When ODM
configuration is dynamically changed and the OTG is on blank pattern,
we will need to reprogram OPP's test pattern based on new ODM
configuration. Therefore we need to lock the OTG pipe to avoid temporary
corruption when we are reprogramming OPP blank patterns.

[HOW]
Add a new interdependent update lock implementation to lock all enabled
OTG pipes even when there is no plane on the OTG for DCN32.

Cc: Mario Limonciello <mario.limonciello@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Reviewed-by: Alvin Lee <alvin.lee2@amd.com>
Acked-by: Alex Hung <alex.hung@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: Amend coasting vtotal for replay low hz

[WHY]
The original coasting vtotal is 2 bytes, and it need to
be amended to 4 bytes because low hz case.

[HOW]
Amend coasting vtotal from 2 bytes to 4 bytes.

Cc: Mario Limonciello <mario.limonciello@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Reviewed-by: Alvin Lee <alvin.lee2@amd.com>
Acked-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: ChunTao Tso <chuntao.tso@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Fix idle check for shared firmware state

[WHY]
We still had an instance of get_idle_state checking the PMFW scratch
register instead of the actual idle allow signal.

[HOW]
Replace it with the SW state check for whether we had allowed idle
through notify_idle.

Cc: Mario Limonciello <mario.limonciello@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Reviewed-by: Duncan Ma <duncan.ma@amd.com>
Acked-by: Alex Hung <alex.hung@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: Update odm when ODM combine is changed on an otg master pipe with no plane

[WHY]
When committing an update with ODM combine change when the plane is
removing or already removed, we fail to detect odm change in pipe
update flags. This has caused mismatch between new dc state and the
actual hardware state, because we missed odm programming.

[HOW]
- Detect odm change even for otg master pipe without a plane.
- Update odm config before calling program pipes for pipe with planes.

The commit also updates blank pattern programming when odm is changed
without plane. This is because number of OPP is changed when ODM
combine is changed. Blank pattern is per OPP so we will need to
reprogram OPP based on the new pipe topology.

Cc: Mario Limonciello <mario.limonciello@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Reviewed-by: Dillon Varone <dillon.varone@amd.com>
Acked-by: Alex Hung <alex.hung@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: Init DPPCLK from SMU on dcn32

[WHY & HOW]
DPPCLK ranges should be obtained from the SMU when available.

Cc: Mario Limonciello <mario.limonciello@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Reviewed-by: Chaitanya Dhere <chaitanya.dhere@amd.com>
Acked-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Dillon Varone <dillon.varone@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Add monitor patch for specific eDP

[WHY]
Some eDP panels' ext caps don't write initial values. The value of
dpcd_addr (0x317) can be random and the backlight control interface
will be incorrect.

[HOW]
Add new panel patches to remove sink ext caps.

Cc: Mario Limonciello <mario.limonciello@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org # 6.5.x
Cc: Tsung-hua Lin <tsung-hua.lin@amd.com>
Cc: Chris Chi <moukong.chi@amd.com>
Reviewed-by: Wayne Lin <wayne.lin@amd.com>
Acked-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Ryan Lin <tsung-hua.lin@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Allow dirty rects to be sent to dmub when abm is active

[WHY]
It's beneficial for ABM to know when new frame data are available.

[HOW]
Add new condition to allow dirty rects to be sent to DMUB when ABM is
active. ABM will use this as a signal that a new frame has arrived.

Cc: Mario Limonciello <mario.limonciello@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Reviewed-by: Anthony Koo <anthony.koo@amd.com>
Acked-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Josip Pavic <josip.pavic@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Override min required DCFCLK in dml1_validate

[WHY]:
Increasing min DCFCLK addresses underflow issues that occur when phantom
pipe is turned on for some Sub-Viewport configs

[HOW]:
dcn32_override_min_req_dcfclk is added to override DCFCLK value in
dml1_validate when subviewport is being used.

Cc: Mario Limonciello <mario.limonciello@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Reviewed-by: Alvin Lee <alvin.lee2@amd.com>
Acked-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Sohaib Nadeem <sohaib.nadeem@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Bypass display ta if display hw is not available

Do not load/invoke display TA if display hardware
is not available.

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

drm/amdgpu: correct the KGQ fallback message

Fix the KGQ fallback function name, as this will
help differentiate the failure in the KCQ enablement.

Signed-off-by: Prike Liang <Prike.Liang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/pm: Check the validity of overdiver power limit

Check the validity of overdriver power limit before using it.

Fixes: 7968e9748fbb ("drm/amdgpu/pm: Fix the power1_min_cap value")
Signed-off-by: Ma Jun <Jun.Ma2@amd.com>
Suggested-by: Lazar Lijo <lijo.lazar@amd.com>
Suggested-by: Alex Deucher <Alexander.Deucher@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org

drm/amdgpu/pm: Fix NULL pointer dereference when get power limit

Because powerplay_table initialization is skipped under
sriov case, We check and set default lower and upper OD
value if powerplay_table is NULL.

Fixes: 7968e9748fbb ("drm/amdgpu/pm: Fix the power1_min_cap value")
Signed-off-by: Ma Jun <Jun.Ma2@amd.com>
Reported-by: Yin Zhenguo <zhenguo.yin@amd.com>
Suggested-by: Lazar Lijo <lijo.lazar@amd.com>
Suggested-by: Alex Deucher <Alexander.Deucher@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org

drm/amdgpu: Skip access PF-only registers on gfx10/gfxhub2_1 under SRIOV

[Why]
RLCG interface returns "out-of-range" error under SRIOV VF when accessing
PF-only registers.

[How]
Skip access PF-only registers on gfx10/gfxhub2_1 under SRIOV.

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

drm/amdgpu: Init zone device and drm client after mode-1 reset on reload

In passthrough environment, when amdgpu is reloaded after unload, mode-1
is triggered after initializing the necessary IPs, That init does not
include KFD, and KFD init waits until the reset is completed. KFD init
is called in the reset handler, but in this case, the zone device and
drm client is not initialized, causing app to create kernel panic.

v2: Removing the init KFD condition from amdgpu_amdkfd_drm_client_create.
As the previous version has the potential of creating DRM client twice.

v3: v2 patch results in SDMA engine hung as DRM open causes VM clear to SDMA
before SDMA init. Adding the condition to in drm client creation, on top of v1,
to guard against drm client creation call multiple times.

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

drm/amdgpu: amdgpu_ttm_gart_bind set gtt bound flag

Otherwise after the GTT bo is released, the GTT and gart space is freed
but amdgpu_ttm_backend_unbind will not clear the gart page table entry
and leave valid mapping entry pointing to the stale system page. Then
if GPU access the gart address mistakely, it will read undefined value
instead page fault, harder to debug and reproduce the real issue.

Cc: stable@vger.kernel.org
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/vcn: enable vcn1 fw load for VCN 4_0_6

v1 - update the fw header for each vcn instance (Veera)

VCN1 has different FW binary in VCN v4_0_6.
Add changes to load the VCN1 fw binary

Signed-off-by: Saleemkhan Jamadar <saleemkhan.jamadar@amd.com>
Reviewed-by: Veerabadhran Gopalakrishnan <Veerabadhran.Gopalakrishnan@amd.com>
Reviewed-by: Leo Liu <leo.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Enable DML2 debug flags

[WHY & HOW]
Enable DML2 related debug config options in DM for testing purposes.

Reviewed-by: Chaitanya Dhere <chaitanya.dhere@amd.com>
Acked-by: Alex Hung <alex.hung@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: Change default size for dummy plane in DML2

[WHY & HOW]
Currently, to map dc states into dml_display_cfg,
We create a dummy plane if the stream doesn't have any planes
attached to it. This dummy plane uses max addersable width height.
This results in certain mode validations failing when they shouldn't.

Cc: Mario Limonciello <mario.limonciello@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Reviewed-by: Chaitanya Dhere <chaitanya.dhere@amd.com>
Acked-by: Alex Hung <alex.hung@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/amdgpu: Reset IH OVERFLOW_EN bit for IH 7.0

IH 7.0 support landed shortly after the original patch for resetting the
bit on all other generations, but without that patch applied.

Fixes: 12443fc53e7d ("drm/amdgpu: Add ih v7_0 ip block support")
Cc: Christian König <christian.koenig@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Friedrich Vock <friedrich.vock@gmx.de>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: fix mmhub client id out-of-bounds access

Properly handle cid 0x140.

Fixes: aba2be41470a ("drm/amdgpu: add mmhub 3.3.0 support")
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/amdgpu: fix use-after-free bug

The bug can be triggered by sending a single amdgpu_gem_userptr_ioctl
to the AMDGPU DRM driver on any ASICs with an invalid address and size.
The bug was reported by Joonkyo Jung <joonkyoj@yonsei.ac.kr>.
For example the following code:

static void Syzkaller1(int fd)
{
struct drm_amdgpu_gem_userptr arg;
int ret;

arg.addr = 0xffffffffffff0000;
arg.size = 0x80000000; /*2 Gb*/
arg.flags = 0x7;
ret = drmIoctl(fd, 0xc1186451/*amdgpu_gem_userptr_ioctl*/, &arg);
}

Due to the address and size are not valid there is a failure in
amdgpu_hmm_register->mmu_interval_notifier_insert->__mmu_interval_notifier_insert->
check_shl_overflow, but we even the amdgpu_hmm_register failure we still call
amdgpu_hmm_unregister into  amdgpu_gem_object_free which causes access to a bad address.
The following stack is below when the issue is reproduced when Kazan is enabled:

[  +0.000014] Hardware name: ASUS System Product Name/ROG STRIX B550-F GAMING (WI-FI), BIOS 1401 12/03/2020
[  +0.000009] RIP: 0010:mmu_interval_notifier_remove+0x327/0x340
[  +0.000017] Code: ff ff 49 89 44 24 08 48 b8 00 01 00 00 00 00 ad de 4c 89 f7 49 89 47 40 48 83 c0 22 49 89 47 48 e8 ce d1 2d 01 e9 32 ff ff ff <0f> 0b e9 16 ff ff ff 4c 89 ef e8 fa 14 b3 ff e9 36 ff ff ff e8 80
[  +0.000014] RSP: 0018:ffffc90002657988 EFLAGS: 00010246
[  +0.000013] RAX: 0000000000000000 RBX: 1ffff920004caf35 RCX: ffffffff8160565b
[  +0.000011] RDX: dffffc0000000000 RSI: 0000000000000004 RDI: ffff8881a9f78260
[  +0.000010] RBP: ffffc90002657a70 R08: 0000000000000001 R09: fffff520004caf25
[  +0.000010] R10: 0000000000000003 R11: ffffffff8161d1d6 R12: ffff88810e988c00
[  +0.000010] R13: ffff888126fb5a00 R14: ffff88810e988c0c R15: ffff8881a9f78260
[  +0.000011] FS:  00007ff9ec848540(0000) GS:ffff8883cc880000(0000) knlGS:0000000000000000
[  +0.000012] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[  +0.000010] CR2: 000055b3f7e14328 CR3: 00000001b5770000 CR4: 0000000000350ef0
[  +0.000010] Call Trace:
[  +0.000006]  <TASK>
[  +0.000007]  ? show_regs+0x6a/0x80
[  +0.000018]  ? __warn+0xa5/0x1b0
[  +0.000019]  ? mmu_interval_notifier_remove+0x327/0x340
[  +0.000018]  ? report_bug+0x24a/0x290
[  +0.000022]  ? handle_bug+0x46/0x90
[  +0.000015]  ? exc_invalid_op+0x19/0x50
[  +0.000016]  ? asm_exc_invalid_op+0x1b/0x20
[  +0.000017]  ? kasan_save_stack+0x26/0x50
[  +0.000017]  ? mmu_interval_notifier_remove+0x23b/0x340
[  +0.000019]  ? mmu_interval_notifier_remove+0x327/0x340
[  +0.000019]  ? mmu_interval_notifier_remove+0x23b/0x340
[  +0.000020]  ? __pfx_mmu_interval_notifier_remove+0x10/0x10
[  +0.000017]  ? kasan_save_alloc_info+0x1e/0x30
[  +0.000018]  ? srso_return_thunk+0x5/0x5f
[  +0.000014]  ? __kasan_kmalloc+0xb1/0xc0
[  +0.000018]  ? srso_return_thunk+0x5/0x5f
[  +0.000013]  ? __kasan_check_read+0x11/0x20
[  +0.000020]  amdgpu_hmm_unregister+0x34/0x50 [amdgpu]
[  +0.004695]  amdgpu_gem_object_free+0x66/0xa0 [amdgpu]
[  +0.004534]  ? __pfx_amdgpu_gem_object_free+0x10/0x10 [amdgpu]
[  +0.004291]  ? do_syscall_64+0x5f/0xe0
[  +0.000023]  ? srso_return_thunk+0x5/0x5f
[  +0.000017]  drm_gem_object_free+0x3b/0x50 [drm]
[  +0.000489]  amdgpu_gem_userptr_ioctl+0x306/0x500 [amdgpu]
[  +0.004295]  ? __pfx_amdgpu_gem_userptr_ioctl+0x10/0x10 [amdgpu]
[  +0.004270]  ? srso_return_thunk+0x5/0x5f
[  +0.000014]  ? __this_cpu_preempt_check+0x13/0x20
[  +0.000015]  ? srso_return_thunk+0x5/0x5f
[  +0.000013]  ? sysvec_apic_timer_interrupt+0x57/0xc0
[  +0.000020]  ? srso_return_thunk+0x5/0x5f
[  +0.000014]  ? asm_sysvec_apic_timer_interrupt+0x1b/0x20
[  +0.000022]  ? drm_ioctl_kernel+0x17b/0x1f0 [drm]
[  +0.000496]  ? __pfx_amdgpu_gem_userptr_ioctl+0x10/0x10 [amdgpu]
[  +0.004272]  ? drm_ioctl_kernel+0x190/0x1f0 [drm]
[  +0.000492]  drm_ioctl_kernel+0x140/0x1f0 [drm]
[  +0.000497]  ? __pfx_amdgpu_gem_userptr_ioctl+0x10/0x10 [amdgpu]
[  +0.004297]  ? __pfx_drm_ioctl_kernel+0x10/0x10 [drm]
[  +0.000489]  ? srso_return_thunk+0x5/0x5f
[  +0.000011]  ? __kasan_check_write+0x14/0x20
[  +0.000016]  drm_ioctl+0x3da/0x730 [drm]
[  +0.000475]  ? __pfx_amdgpu_gem_userptr_ioctl+0x10/0x10 [amdgpu]
[  +0.004293]  ? __pfx_drm_ioctl+0x10/0x10 [drm]
[  +0.000506]  ? __pfx_rpm_resume+0x10/0x10
[  +0.000016]  ? srso_return_thunk+0x5/0x5f
[  +0.000011]  ? __kasan_check_write+0x14/0x20
[  +0.000010]  ? srso_return_thunk+0x5/0x5f
[  +0.000011]  ? _raw_spin_lock_irqsave+0x99/0x100
[  +0.000015]  ? __pfx__raw_spin_lock_irqsave+0x10/0x10
[  +0.000014]  ? srso_return_thunk+0x5/0x5f
[  +0.000013]  ? srso_return_thunk+0x5/0x5f
[  +0.000011]  ? srso_return_thunk+0x5/0x5f
[  +0.000011]  ? preempt_count_sub+0x18/0xc0
[  +0.000013]  ? srso_return_thunk+0x5/0x5f
[  +0.000010]  ? _raw_spin_unlock_irqrestore+0x27/0x50
[  +0.000019]  amdgpu_drm_ioctl+0x7e/0xe0 [amdgpu]
[  +0.004272]  __x64_sys_ioctl+0xcd/0x110
[  +0.000020]  do_syscall_64+0x5f/0xe0
[  +0.000021]  entry_SYSCALL_64_after_hwframe+0x6e/0x76
[  +0.000015] RIP: 0033:0x7ff9ed31a94f
[  +0.000012] Code: 00 48 89 44 24 18 31 c0 48 8d 44 24 60 c7 04 24 10 00 00 00 48 89 44 24 08 48 8d 44 24 20 48 89 44 24 10 b8 10 00 00 00 0f 05 <41> 89 c0 3d 00 f0 ff ff 77 1f 48 8b 44 24 18 64 48 2b 04 25 28 00
[  +0.000013] RSP: 002b:00007fff25f66790 EFLAGS: 00000246 ORIG_RAX: 0000000000000010
[  +0.000016] RAX: ffffffffffffffda RBX: 000055b3f7e133e0 RCX: 00007ff9ed31a94f
[  +0.000012] RDX: 000055b3f7e133e0 RSI: 00000000c1186451 RDI: 0000000000000003
[  +0.000010] RBP: 00000000c1186451 R08: 0000000000000000 R09: 0000000000000000
[  +0.000009] R10: 0000000000000008 R11: 0000000000000246 R12: 00007fff25f66ca8
[  +0.000009] R13: 0000000000000003 R14: 000055b3f7021ba8 R15: 00007ff9ed7af040
[  +0.000024]  </TASK>
[  +0.000007] ---[ end trace 0000000000000000 ]---

v2: Consolidate any error handling into amdgpu_hmm_register
    which applied to kfd_bo also. (Christian)
v3: Improve syntax and comment (Christian)

Cc: Christian Koenig <christian.koenig@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Felix Kuehling <felix.kuehling@amd.com>
Cc: Joonkyo Jung <joonkyoj@yonsei.ac.kr>
Cc: Dokyung Song <dokyungs@yonsei.ac.kr>
Cc: <jisoo.jang@yonsei.ac.kr>
Cc: <yw9865@yonsei.ac.kr>
Signed-off-by: Vitaly Prosyak <vitaly.prosyak@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Handle duplicate BOs during process restore

In certain situations, some apps can import a BO multiple times
(through IPC for example). To restore such processes successfully,
we need to tell drm to ignore duplicate BOs.
While at it, also add additional logging to prevent silent failures
when process restore fails.

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/amd/display: Use freesync when `DRM_EDID_FEATURE_CONTINUOUS_FREQ` found

The monitor shipped with the Framework 16 supports VRR [1], but it's not
being advertised.

This is because the detailed timing block doesn't contain
`EDID_DETAIL_MONITOR_RANGE` which amdgpu looks for to find min and max
frequencies.  This check however is superfluous for this case because
update_display_info() calls drm_get_monitor_range() to get these ranges
already.

So if the `DRM_EDID_FEATURE_CONTINUOUS_FREQ` EDID feature is found then
turn on freesync without extra checks.

v2: squash in fix from Harry

Closes: https://www.reddit.com/r/framework/comments/1b4y2i5/no_variable_refresh_rate_on_the_framework_16_on/
Closes: https://www.reddit.com/r/framework/comments/1b6vzcy/framework_16_variable_refresh_rate/
Closes: https://community.frame.work/t/resolved-no-vrr-freesync-with-amd-version/42338
Link: https://gist.github.com/superm1/e8fbacfa4d0f53150231d3a3e0a13faf
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

nouveau/gsp: don't check devinit disable on GSP.

GSP should be handling this and I can see no evidence in opengpu
driver that this register should be touched.

Fixed acceleration on 2080 Ti GPUs.

Fixes: 15740541e8f0 ("drm/nouveau/devinit/tu102-: prepare for GSP-RM")

Signed-off-by: Dave Airlie <airlied@redhat.com>
Signed-off-by: Danilo Krummrich <dakr@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240314014521.2695233-1-airlied@gmail.com

Merge tag 'drm-misc-next-fixes-2024-03-14' of https://gitlab.freedesktop.org/drm/misc/kernel into drm-next

Short summary of fixes pull:

probe-helper:
- never return negative values from .get_modes() plus driver fixes

nouveau:
- clear bo resource bus after eviction
- documentation fixes

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Thomas Zimmermann <tzimmermann@suse.de>
Link: https://patchwork.freedesktop.org/patch/msgid/20240314082833.GA8761@localhost.localdomain

Merge tag 'drm-xe-next-fixes-2024-03-14' of https://gitlab.freedesktop.org/drm/xe/kernel into drm-next

Driver changes:

- Invalidate userptr VMA on page pin fault, allowing userspace
  to free userptr while still having bindings
- Fail early on sysfs file creation error
- Skip VMA pinning on xe_exec with num_batch_buffer == 0

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Lucas De Marchi <lucas.demarchi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/c4epi2j6anpc77z73zbgibxg7bxsmmkb522aa7tyei6oa6uunn@3oad4cgomd5a

drm: Fix drm_fixp2int_round() making it add 0.5

As well noted by Pekka[1], the rounding of drm_fixp2int_round is wrong.
To round a number, you need to add 0.5 to the number and floor that,
drm_fixp2int_round() is adding 0.0000076. Make it add 0.5.

[1]: https://lore.kernel.org/all/20240301135327.22efe0dd.pekka.paalanen@collabora.com/

Fixes: 8b25320887d7 ("drm: Add fixed-point helper to get rounded integer values")
Suggested-by: Pekka Paalanen <pekka.paalanen@collabora.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Melissa Wen <mwen@igalia.com>
Signed-off-by: Arthur Grillo <arthurgrillo@riseup.net>
Signed-off-by: Melissa Wen <melissa.srw@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240316-drm_fixed-v2-1-c1bc2665b5ed@riseup.net

drm/tests: Build KMS helpers when DRM_KUNIT_TEST_HELPERS is enabled

Commit 66671944e176 ("drm/tests: helpers: Add atomic helpers")
introduced a dependency on CRTC helpers in KUnit test helpers.
Select the former when building KUnit test helpers to avoid
linker errors.

Fixes: 66671944e176 ("drm/tests: helpers: Add atomic helpers")
Cc: Maxime Ripard <mripard@kernel.org>
Cc: Maíra Canal <mcanal@igalia.com>
Signed-off-by: Karolina Stolarek <karolina.stolarek@intel.com>
Link: https://lore.kernel.org/r/20240313142142.1318718-1-karolina.stolarek@intel.com
Signed-off-by: Maxime Ripard <mripard@kernel.org>

drm/xe: Skip VMAs pin when requesting signal to the last XE_EXEC

Doing a XE_EXEC with num_batch_buffer == 0 makes signals passed as
argument to be signaled when the last real XE_EXEC is completed.
But to do that it was first pinning all VMAs in drm_gpuvm_exec_lock(),
this patch remove this pinning as it is not required.

This change also help Mesa implementing memory over-commiting recovery
as it needs to unbind not needed VMAs when the whole VM can't fit
in GPU memory but it can only do the unbiding when the last XE_EXEC
is completed.
So with this change Mesa can get the signal it want without getting
out-of-memory errors.

Fixes: eb9702ad2986 ("drm/xe: Allow num_batch_buffer / num_binds == 0 in IOCTLs")
Cc: Thomas Hellstrom <thomas.hellstrom@linux.intel.com>
Co-developed-by: Matthew Brost <matthew.brost@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Reviewed-by: Matthew Brost <matthew.brost@intel.com>
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240313171318.121066-1-jose.souza@intel.com
(cherry picked from commit 58480c1c912ff8146d067301a0d04cca318b4a66)
Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>

drm/xe: Return if kobj creation is failed

Return after warning regarding kobj creation failure.

Fixes: 4ae3aeab32d7 ("drm/xe: Add vram frequency sysfs attributes")
Cc: Sujaritha Sundaresan <sujaritha.sundaresan@intel.com>
Cc: Tejas Upadhyay <tejas.upadhyay@intel.com>
Cc: Bommu Krishnaiah <krishnaiah.bommu@intel.com>
Reviewed-by: Tejas Upadhyay <tejas.upadhyay@intel.com>
Signed-off-by: Himal Prasad Ghimiray <himal.prasad.ghimiray@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240313150545.2830408-2-himal.prasad.ghimiray@intel.com
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
(cherry picked from commit 989d07ac6bb7d269e975f85e8f683f496faa0380)
Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>

drm/xe: Invalidate userptr VMA on page pin fault

Rather than return an error to the user or ban the VM when userptr VMA
page pin fails with -EFAULT, invalidate VMA mappings. This supports the
UMD use case of freeing userptr while still having bindings.

Now that non-faulting VMs can invalidate VMAs, drop the usm prefix for
the tile_invalidated member.

v2:
 - Fix build error (CI)
v3:
 - Don't invalidate VMA if in fault mode, rather kill VM (Thomas)
 - Update commit message with tile_invalidated name chagne (Thomas)
 - Wait VM bookkeep slots with VM resv lock (Thomas)
v4:
 - Move list_del_init(&userptr.repin_link) after error check (Thomas)
 - Assert not in fault mode (Matthew)

Fixes: dd08ebf6c352 ("drm/xe: Introduce a new DRM driver for Intel GPUs")
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240312183907.933835-1-matthew.brost@intel.com
(cherry picked from commit 521db22a1d70dbc596a07544a738416025b1b63c)
Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>

drm/bridge: correct DRM_BRIDGE_OP_EDID documentation

While the commit d807ad80d811 ("drm/bridge: add ->edid_read hook and
drm_bridge_edid_read()") and the commit 27b8f91c08d9 ("drm/bridge:
remove ->get_edid callback") replaced ->get_edid() callback with the
->edid_read(), they failed to update documentation. Fix the drm_bridge
docs to point to edid_read().

Fixes: 27b8f91c08d9 ("drm/bridge: remove ->get_edid callback")
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: Neil Armstrong <neil.armstrong@linaro.org>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20240310-drm-bridge-fix-docs-v1-1-70d3d741cb7a@linaro.org

drm/sun4i: hdmi: Fix u64 div on 32bit arch

Commit 358e76fd613a ("drm/sun4i: hdmi: Consolidate atomic_check and
mode_valid") changed the clock rate from an unsigned long to an unsigned
long long resulting in a a 64-bit division that might not be supported
on all platforms.

The resulted in compilation being broken at least for m68k, xtensa and
some arm configurations, at least.

Fixes: 358e76fd613a ("drm/sun4i: hdmi: Consolidate atomic_check and mode_valid")
Reported-by: Geert Uytterhoeven <geert@linux-m68k.org>
Reported-by: Naresh Kamboju <naresh.kamboju@linaro.org>
Closes: https://lore.kernel.org/r/CA+G9fYvG9KE15PGNoLu+SBVyShe+u5HBLQ81+kK9Zop6u=ywmw@mail.gmail.com/
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202403011839.KLiXh4wC-lkp@intel.com/
Acked-by: Thomas Zimmermann <tzimmermann@suse.de>
Link: https://lore.kernel.org/r/20240304091225.366325-1-mripard@kernel.org
Signed-off-by: Maxime Ripard <mripard@kernel.org>

drm/exynos: simplify the return value handling in exynos_dp_get_modes()

Just use 0 and 1 directly, instead of the confusing local variable
that's always set to 0.

Cc: Inki Dae <inki.dae@samsung.com>
Cc: Seung-Woo Kim <sw0312.kim@samsung.com>
Cc: Kyungmin Park <kyungmin.park@samsung.com>
Acked-by: Thomas Zimmermann <tzimmermann@suse.de>
Link: https://patchwork.freedesktop.org/patch/msgid/64cc680f14d961cedb726a6fd5c6dfd53ca9bb85.1709913674.git.jani.nikula@intel.com
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

drm/bridge: lt9611uxc: use int for holding number of modes

lt9611uxc_connector_get_modes() propagates the return value of
drm_edid_connector_add_modes() but stores the int temporarily in an
unsigned int. Use the correct type.

Cc: Andrzej Hajda <andrzej.hajda@intel.com>
Cc: Neil Armstrong <neil.armstrong@linaro.org>
Cc: Robert Foss <rfoss@kernel.org>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Neil Armstrong <neil.armstrong@linaro.org>
Acked-by: Thomas Zimmermann <tzimmermann@suse.de>
Link: https://patchwork.freedesktop.org/patch/msgid/ed97f4f036263cdc4f34330cef91214970f99a77.1709913674.git.jani.nikula@intel.com
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

drm/vc4: hdmi: do not return negative values from .get_modes()

The .get_modes() hooks aren't supposed to return negative error
codes. Return 0 for no modes, whatever the reason.

Cc: Maxime Ripard <mripard@kernel.org>
Cc: stable@vger.kernel.org
Acked-by: Maxime Ripard <mripard@kernel.org>
Acked-by: Thomas Zimmermann <tzimmermann@suse.de>
Link: https://patchwork.freedesktop.org/patch/msgid/dcda6d4003e2c6192987916b35c7304732800e08.1709913674.git.jani.nikula@intel.com
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

drm/imx/ipuv3: do not return negative values from .get_modes()

The .get_modes() hooks aren't supposed to return negative error
codes. Return 0 for no modes, whatever the reason.

Cc: Philipp Zabel <p.zabel@pengutronix.de>
Cc: stable@vger.kernel.org
Acked-by: Philipp Zabel <p.zabel@pengutronix.de>
Acked-by: Thomas Zimmermann <tzimmermann@suse.de>
Link: https://patchwork.freedesktop.org/patch/msgid/311f6eec96d47949b16a670529f4d89fcd97aefa.1709913674.git.jani.nikula@intel.com
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

drm/bridge: lt8912b: do not return negative values from .get_modes()

The .get_modes() hooks aren't supposed to return negative error
codes. Return 0 for no modes, whatever the reason.

Cc: Adrien Grassein <adrien.grassein@gmail.com>
Cc: stable@vger.kernel.org
Acked-by: Thomas Zimmermann <tzimmermann@suse.de>
Link: https://patchwork.freedesktop.org/patch/msgid/dcdddcbcb64b6f6cdc55022ee50c10dee8ddbc3d.1709913674.git.jani.nikula@intel.com
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

drm/exynos: do not return negative values from .get_modes()

The .get_modes() hooks aren't supposed to return negative error
codes. Return 0 for no modes, whatever the reason.

Cc: Inki Dae <inki.dae@samsung.com>
Cc: Seung-Woo Kim <sw0312.kim@samsung.com>
Cc: Kyungmin Park <kyungmin.park@samsung.com>
Cc: stable@vger.kernel.org
Acked-by: Thomas Zimmermann <tzimmermann@suse.de>
Link: https://patchwork.freedesktop.org/patch/msgid/d8665f620d9c252aa7d5a4811ff6b16e773903a2.1709913674.git.jani.nikula@intel.com
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

drm/panel: do not return negative error codes from drm_panel_get_modes()

None of the callers of drm_panel_get_modes() expect it to return
negative error codes. Either they propagate the return value in their
struct drm_connector_helper_funcs .get_modes() hook (which is also not
supposed to return negative codes), or add it to other counts leading to
bogus values.

On the other hand, many of the struct drm_panel_funcs .get_modes() hooks
do return negative error codes, so handle them gracefully instead of
propagating further.

Return 0 for no modes, whatever the reason.

Cc: Neil Armstrong <neil.armstrong@linaro.org>
Cc: Jessica Zhang <quic_jesszhan@quicinc.com>
Cc: Sam Ravnborg <sam@ravnborg.org>
Cc: stable@vger.kernel.org
Reviewed-by: Neil Armstrong <neil.armstrong@linaro.org>
Reviewed-by: Jessica Zhang <quic_jesszhan@quicinc.com>
Acked-by: Thomas Zimmermann <tzimmermann@suse.de>
Link: https://patchwork.freedesktop.org/patch/msgid/79f559b72d8c493940417304e222a4b04dfa19c4.1709913674.git.jani.nikula@intel.com
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

drm/probe-helper: warn about negative .get_modes()

The .get_modes() callback is supposed to return the number of modes,
never a negative error code. If a negative value is returned, it'll just
be interpreted as a negative count, and added to previous calculations.

Document the rules, but handle the negative values gracefully with an
error message.

Cc: stable@vger.kernel.org
Acked-by: Thomas Zimmermann <tzimmermann@suse.de>
Link: https://patchwork.freedesktop.org/patch/msgid/50208c866facc33226a3c77b82bb96aeef8ef310.1709913674.git.jani.nikula@intel.com
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

drm/nouveau: fix kerneldoc warnings

kernel test robot complains about missing kerneldoc entries:

  drivers-gpu-drm-nouveau-nvkm-subdev-gsp-r535.c:warning:
    Function-parameter-or-struct-member-gsp-not-described-in-nvkm_gsp_radix3_sg

Signed-off-by: Timur Tabi <ttabi@nvidia.com>
Signed-off-by: Danilo Krummrich <dakr@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240210002900.148982-1-ttabi@nvidia.com

nouveau: reset the bo resource bus info after an eviction

Later attempts to refault the bo won't happen and the whole
GPU does to lunch. I think Christian's refactoring of this
code out to the driver broke this not very well tested path.

Fixes: 141b15e59175 ("drm/nouveau: move io_reserve_lru handling into the driver v5")
Cc: Christian König <christian.koenig@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Danilo Krummrich <dakr@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240311072037.287905-1-airlied@gmail.com

Merge tag 'amd-drm-next-6.9-2024-03-08-1' of https://gitlab.freedesktop.org/agd5f/linux into drm-next

amd-drm-next-6.9-2024-03-08-1:

amdgpu:
- DCN 3.5.1 support
- Fixes for IOMMUv2 removal
- UAF fix
- Misc small fixes and cleanups
- SR-IOV fixes
- MCBP cleanup
- devcoredump update
- NBIF 6.3.1 support
- VPE 6.1.1 support

amdkfd:
- Misc fixes and cleanups
- GFX10.1 trap fixes

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240308170741.3691166-1-alexander.deucher@amd.com

Merge tag 'drm-msm-next-2024-03-07' of https://gitlab.freedesktop.org/drm/msm into drm-next

Late updates for v6.9, the main part is CDM (YUV over DP) which was
waiting for drm-misc-next-2024-02-29.

DPU:
- Add support for YUV420 over DP
- Patchset to ease debugging of vblank timeouts
- Small cleanup

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Rob Clark <robdclark@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/CAF6AEGvedk6OCOZ-NNtGf_pNiGuK9uvWj1MCDZLX9Jo2nHS=Zg@mail.gmail.com

Merge tag 'drm-etnaviv-next-2024-03-07' of https://git.pengutronix.de/git/lst/linux into drm-next

- various code cleanups
- enhancements for NPU and MRT support

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Lucas Stach <l.stach@pengutronix.de>
Link: https://patchwork.freedesktop.org/patch/msgid/72a783cd98d60f6ebb43b90a6b453eea87224409.camel@pengutronix.de

Merge tag 'drm-xe-next-fixes-2024-03-04' of ssh://gitlab.freedesktop.org/drm/xe/kernel into drm-next

Driver Changes:

- Fix kunit link failure with built-in xe
- Fix one more 32-bit build failure with ARM compiler
- Fix initialization order of topology struct
- Cleanup unused fields in struct xe_vm
- Fix xe_vm leak when handling page fault on a VM not in fault mode
- Drop use of "grouped target" feature in Makefile since that's
  only available in make >= 4.3

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Lucas De Marchi <lucas.demarchi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/kaypobelrl7u7rtnu6hg5czs3vptbhs4rp24vnwuo2ajoxysto@l5u7377hz4es

Merge tag 'drm-misc-next-fixes-2024-03-07' of https://anongit.freedesktop.org/git/drm/drm-misc into drm-next

Short summary of fixes pull:

- i915: Fix applying placement flags
- fbdev: Fix build on PowerMacs after header cleanup

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Thomas Zimmermann <tzimmermann@suse.de>
Link: https://patchwork.freedesktop.org/patch/msgid/20240307124640.GA18593@localhost.localdomain

Merge tag 'amd-drm-next-6.9-2024-03-01' of https://gitlab.freedesktop.org/agd5f/linux into drm-next

amd-drm-next-6.9-2024-03-01:

amdgpu:
- GC 11.5.1 updates
- Misc display cleanups
- NBIO 7.9 updates
- Backlight fixes
- DMUB fixes
- MPO fixes
- atomfirmware table updates
- SR-IOV fixes
- VCN 4.x updates
- use RMW accessors for pci config registers
- PSR fixes
- Suspend/resume fixes
- RAS fixes
- ABM fixes
- Misc code cleanups
- SI DPM fix
- Revert freesync video

amdkfd:
- Misc cleanups
- Error handling fixes

radeon:
- use RMW accessors for pci config registers

From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240301204857.13960-1-alexander.deucher@amd.com
Signed-off-by: Dave Airlie <airlied@redhat.com>

drm/amd/display: Removed redundant @ symbol to fix kernel-doc warnings in -next repo

For linux-next repository.

./drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h:1: warning: no structured comments found
./drivers/gpu/drm/amd/display/dc/inc/hw/mpc.h:132: warning: Incorrect use of kernel-doc format:          * @@overlap_only: Whether overlapping of different planes is allowed.
./drivers/gpu/drm/amd/display/dc/inc/hw/mpc.h:132: warning: Incorrect use of kernel-doc format:          * @@overlap_only: Whether overlapping of different planes is allowed.
./drivers/gpu/drm/amd/display/dc/inc/hw/mpc.h:1: warning: no structured comments found
./drivers/gpu/drm/amd/display/dc/inc/hw/mpc.h:132: warning: Incorrect use of kernel-doc format:          * @@overlap_only: Whether overlapping of different planes is allowed.
./drivers/gpu/drm/amd/display/dc/inc/hw/mpc.h:162: warning: Function parameter or struct member 'pre_multiplied_alpha' not described in 'mpcc_blnd_cfg'

Signed-off-by: R SUNDAR <prosunofficial@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: wait for completion of the EnableGfxImu message

Wait for completion of sending the EnableGfxImu message
when using the PSP FW loading for SMU ip v14.0.

Signed-off-by: Tim Huang <Tim.Huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/soc21: add mode2 asic reset for SMU IP v14.0.1

Set the default reset method to mode2 for SMU IP v14.0.1

Signed-off-by: lima1002 <li.ma@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add smu 14.0.1 support

This patch to add smu 14.0.1 support.

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

drm/amdgpu: add VPE 6.1.1 discovery support

Enable VPE 6.1.1.

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

drm/amdgpu/vpe: add VPE 6.1.1 support

Add initial support for VPE 6.1.1.

v2: squash in updates (Alex)

Signed-off-by: Lang Yu <Lang.Yu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/vpe: don't emit cond exec command under collaborate mode

Not ready now.

Signed-off-by: Lang Yu <Lang.Yu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/vpe: add collaborate mode support for VPE

Under clollaborate mode, multiple VPE instances share a ring buferr
and work together to finish a job.

Signed-off-by: Lang Yu <Lang.Yu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/vpe: add PRED_EXE and COLLAB_SYNC OPCODE

To support multi VPE collaborate mode.

Signed-off-by: Lang Yu <Lang.Yu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/vpe: add multi instance VPE support

Add support for multi instance VPE processing.

Signed-off-by: Lang Yu <Lang.Yu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/discovery: add nbif v6_3_1 ip block

Add nbif v6_3_1 ip block.

Signed-off-by: Likun Gao <Likun.Gao@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Add nbif v6_3_1 ip block support

Add nbif v6_3_1 ip block support.

Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Likun Gao <Likun.Gao@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Add pcie v6_1_0 ip headers (v5)

v1: Add pcie v6_1_0 register offset and shift masks
    header files. (Hawking)
v2: Update pcie v6_1_0 register offset and shift masks
    header files to RE2. (Likun)
v3: Update pcie v6_1_0 register offset and shift masks
    header files to RE2.5. (Likun)
v4: Update pcie v6_1_0 register offset and shift masks
    header files to RE3. (Likun)
v5: Updates (Alex)

Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Likun Gao <Likun.Gao@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Likun Gao <Likun.Gao@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Add nbif v6_3_1 ip headers (v5)

v1: Add nbif v6_3_1 register offset and shift masks
    header files. (Hawking)
v2: Update nbif v6_3_1 register offset and shift masks
    header files to RE2. (Likun)
v3: Update nbif v6_3_1 register offset and shift masks
    header files to RE2.5. (Likun)
v4: Update nbif v6_3_1 register offset and shift masks
    header files to RE3. (Likun)
v5: Updates (Alex)

Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Likun Gao <Likun.Gao@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Likun Gao <Likun.Gao@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

arch/powerpc: Remove <linux/fb.h> from backlight code

Replace <linux/fb.h> with a forward declaration in <asm/backlight.h> to
resolve an unnecessary dependency. Remove pmac_backlight_curve_lookup()
and struct fb_info from source and header files. The function and the
framebuffer struct are unused. No functional changes.

v3:
* Add Fixes tag (Christophe)
* fix typos in commit message (Jani)

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Fixes: d565dd3b0824 ("[PATCH] powerpc: More via-pmu backlight fixes")
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Acked-by: Michael Ellerman <mpe@ellerman.id.au> # (powerpc)
Link: https://patchwork.freedesktop.org/patch/msgid/20240306122935.10626-4-tzimmermann@suse.de

macintosh/via-pmu-backlight: Include <linux/backlight.h>

Fix builds with CONFIG_PMAC_BACKLIGHT=y. The include statement for the
backlight header has recently been removed from <linux/fb.h>. Add it to
via-pmu-backlight.c to get the necessary symbols.

Reported-by: Naresh Kamboju <naresh.kamboju@linaro.org>
Closes: https://lore.kernel.org/dri-devel/CA+G9fYsAk5TbqqxFC2W4oHLGA0CbTHMxbeq8QayFXTU75YiueA@mail.gmail.com/
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Fixes: 11b4eedfc87d ("fbdev: Do not include <linux/backlight.h> in header")
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Cc: Thomas Zimmermann <tzimmermann@suse.de>
Cc: Jani Nikula <jani.nikula@intel.com>
Cc: Helge Deller <deller@gmx.de>
Cc: linux-fbdev@vger.kernel.org
Cc: dri-devel@lists.freedesktop.org
Link: https://patchwork.freedesktop.org/patch/msgid/20240306122935.10626-3-tzimmermann@suse.de

fbdev/chipsfb: Include <linux/backlight.h>

Fix builds with CONFIG_PMAC_BACKLIGHT=y. The include statement for
the backlight header has recently been removed from <linux/fb.h>.

Reported-by: Naresh Kamboju <naresh.kamboju@linaro.org>
Closes: https://lore.kernel.org/dri-devel/CA+G9fYsAk5TbqqxFC2W4oHLGA0CbTHMxbeq8QayFXTU75YiueA@mail.gmail.com/
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Fixes: 11b4eedfc87d ("fbdev: Do not include <linux/backlight.h> in header")
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Cc: Thomas Zimmermann <tzimmermann@suse.de>
Cc: Jani Nikula <jani.nikula@intel.com>
Cc: Helge Deller <deller@gmx.de>
Cc: linux-fbdev@vger.kernel.org
Cc: dri-devel@lists.freedesktop.org
Link: https://patchwork.freedesktop.org/patch/msgid/20240306122935.10626-2-tzimmermann@suse.de

Merge drm/drm-next into drm-misc-next-fixes

Backmerging to get a few more commits that came from drm-misc-next.
See [1]

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Link: https://lore.kernel.org/dri-devel/20240229084806.GA21616@localhost.localdomain/

drm/etnaviv: Restore some id values

The hwdb selection logic as a feature that allows it to mark some fields
as 'don't care'. If we match with such a field we memcpy(..)
the current etnaviv_chip_identity into ident.

This step can overwrite some id values read from the GPU with the
'don't care' value.

Fix this issue by restoring the affected values after the memcpy(..).

As this is crucial for user space to know when this feature works as
expected increment the minor version too.

Fixes: 4078a1186dd3 ("drm/etnaviv: update hwdb selection logic")
Cc: stable@vger.kernel.org
Signed-off-by: Christian Gmeiner <cgmeiner@igalia.com>
Reviewed-by: Tomeu Vizoso <tomeu@tomeuvizoso.net>
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>

drm/amdkfd: make kfd_class constant

Since commit 43a7206b0963 ("driver core: class: make class_register() take
a const *"), the driver core allows for struct class to be in read-only
memory, so move the kfd_class structure to be declared at build time
placing it into read-only memory, instead of having to be dynamically
allocated at boot time.

Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Suggested-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Ricardo B. Marliere <ricardo@marliere.net>
Signed-off-by: Felix Kuehling <felix.kuehling@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add ring timeout information in devcoredump

Add ring timeout related information in the amdgpu
devcoredump file for debugging purposes.

During the gpu recovery process the registered call
is triggered and add the debug information in data
file created by devcoredump framework under the
directory /sys/class/devcoredump/devcdx/

Signed-off-by: Sunil Khatri <sunil.khatri@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add dcn3.5.1 support

This patch to add dcn3.5.1 support.

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

drm/amdkfd: Use SQC when TCP would fail in gfx10.1 context save

Similarly to gfx9, gfx10.1 drops vector stores when an xnack error is
raised. To work around this issue, use scalar stores instead of vector
stores when trapsts.xnack_error == 1.

Signed-off-by: Laurent Morichetti <laurent.morichetti@amd.com>
Reviewed-by: Jay Cornwall <jay.cornwall@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdkfd: Increase the size of the memory reserved for the TBA

In a future commit, the cwsr trap handler code size for gfx10.1 will
increase to slightly above the one page mark. Since the TMA does not
need to be page aligned, and only 2 pointers are stored in it, push
the TMA offset by 2 KiB and keep the TBA+TMA reserved memory size
to two pages.

Signed-off-by: Laurent Morichetti <laurent.morichetti@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: disable ring_muxer if mcbp is off

Using the ring_muxer without preemption adds overhead for no
reason since mcbp cannot be triggered.

Moving back to a single queue in this case also helps when
high priority app are used: in this case the gpu_scheduler
priority handling will work as expected - much better than
ring_muxer with its 2 independant schedulers competing for
the same hardware queue.

This change requires moving amdgpu_device_set_mcbp above
amdgpu_device_ip_early_init because we use adev->gfx.mcbp.

Signed-off-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Acked-by: Jiadong Zhu <Jiadong.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: disable pp_dpm_dcefclk node for gfx 11.0.3 sriov

the dce dpm is not available on gfx 11.0.3 sriov device.

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

drm/amdgpu: remove unused code

Remove the unused function - amdgpu_vm_pt_is_root_clean
and remove the impossible condition

v1: entries == 0 is not possible any more,
    so this condition could probably be removed (Felix)

Signed-off-by: Jesse Zhang <jesse.zhang@amd.com>
Suggested-by：Felix Kuehling <felix.kuehling@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: workaround to avoid SET_Q_MODE packets v2

It turned out that executing the SET_Q_MODE packet on every submission
creates to much overhead.

Implement a workaround which allows skipping the SET_Q_MODE packet if
subsequent submissions all use the same parameters.

v2: add a NULL check for ring_obj

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

drm/amdgpu: cleanup conditional execution

First of all calculating the number of dw to patch into a
conditional execution is not something HW generation specific.
This is just standard ring buffer calculations. While at it also
reduce the BUG_ON() into WARN_ON().

Then instead of a random bit pattern use 0 as default value for
the number of dw skipped, this way it's not mandatory any more
to patch the conditional execution.

And last make the address to check a parameter of the
conditional execution instead of getting this from the ring.

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

drm/amdgpu: Use rpm_mode flag instead of checking it again for rpm

Because the rpm_mode flag is already set when the driver
is initialized, we use it directly for runtime suspend/resume
instead of checking it again

Signed-off-by: Ma Jun <Jun.Ma2@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/pm: Fix the error of pwm1_enable setting

Fix the pwm_mode value error which used for
pwm1_enable setting

Signed-off-by: Ma Jun <Jun.Ma2@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: change vm->task_info handling

This patch changes the handling and lifecycle of vm->task_info object.
The major changes are:
- vm->task_info is a dynamically allocated ptr now, and its uasge is
  reference counted.
- introducing two new helper funcs for task_info lifecycle management
    - amdgpu_vm_get_task_info: reference counts up task_info before
      returning this info
    - amdgpu_vm_put_task_info: reference counts down task_info
- last put to task_info() frees task_info from the vm.

This patch also does logistical changes required for existing usage
of vm->task_info.

V2: Do not block all the prints when task_info not found (Felix)

V3: Fixed review comments from Felix
   - Fix wrong indentation
   - No debug message for -ENOMEM
   - Add NULL check for task_info
   - Do not duplicate the debug messages (ti vs no ti)
   - Get first reference of task_info in vm_init(), put last
     in vm_fini()

V4: Fixed review comments from Felix
   - fix double reference increment in create_task_info
   - change amdgpu_vm_get_task_info_pasid
   - additional changes in amdgpu_gem.c while porting

Cc: Christian Koenig <christian.koenig@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Shashank Sharma <shashank.sharma@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: handle range offsets in VRR ranges

Need to check the offset bits for values greater than 255.

v2: also update amdgpu_dm_connector values.

Suggested-by: Mano Ségransan <mano.segransan@protonmail.com>
Tested-by: Mano Ségransan <mano.segransan@protonmail.com>
Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/3203
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: add amdgpu_dm support for DCN351

Add Display Manager specific changes for DCN3.5.1.

Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: add DC changes for DCN351

Add DC support for DCN 3.5.1.

Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: add DCN351 IRQ changes

Add DCN3.5.1 interrupt support.

Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: add DMUB source files and changes for DCN351

DMUB support is required to light-up displays.

Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd: add register headers for DCN351

Add register headers for DCN 3.5.1.

Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: add DCN351 version identifiers

Add DCN3.5.1 ASIC identifiers.

Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

Revert "drm/amdgpu: remove vm sanity check from amdgpu_vm_make_compute" for Raven

fix the issue:
"amdgpu: Failed to create process VM object".

[Why]when amdgpu initialized, seq64 do mampping and update bo mapping in vm page table.
But when clifo run. It also initializes a vm for a process device through the function kfd_process_device_init_vm and ensure the root PD is clean through the function amdgpu_vm_pt_is_root_clean.
So they have a conflict, and clinfo  always failed.

v1:
  - remove all the pte_supports_ats stuff from the amdgpu_vm code (Felix)

Signed-off-by: Jesse Zhang <Jesse.Zhang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>