drm/amd/display: Split program front end part that occur outside lock

[Why]
Eventually want to lock at a higher level in stack.
To do this, we need to be able to isolate the parts that need to be done
after pipe unlock.

[How]
Split out programming that is done post unlock.

Signed-off-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Do not set optimized_require to false after plane disable

[WHY]
The optimized_require flag is needed to set watermarks and clocks lower
in certain conditions. This flag is set to true and then set to false
while programming front end in dcn20.

[HOW]
Do not set the flag to false while disabling plane.

Signed-off-by: Sung Lee <sung.lee@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Fix psr static frames calculation

[Why]
Driver crash with psr feature enabled due to divide-by-zero error.
This is a regression after rework to calculate static screen frame
number entry time.

[How]
Correct order of operations to avoid divide-by-zero.

Signed-off-by: Roman Li <roman.li@amd.com>
Reviewed-by: Zhan Liu <Zhan.Liu@amd.com>
Acked-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: remove invalid dc_is_hw_initialized function

[why/how]
We found out that the register we read actually gets reset by SMU
after we loose power, meaning this always returns true

Signed-off-by: Joseph Gravenor <joseph.gravenor@amd.com>
Reviewed-by: Eric Yang <eric.yang2@amd.com>
Acked-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Move USB-C workaround to after parameter variables are set

[Why]
The call to dp_enable_link_phy are using default/invalid values for clock id
and link settings.

[How]
Move workaround code to after its parameter variables are determined.

Signed-off-by: George Shen <george.shen@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Add set psr version message

[Why]
Must know psr version during runtime.

[How]
Add set psr version message structures.

Signed-off-by: Wyatt Wood <wyatt.wood@amd.com>
Reviewed-by: Nicholas Kazlauskas <Nicholas.Kazlauskas@amd.com>
Acked-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: move xgmi init/fini to xgmi_add/remove_device call (v2)

For sriov, psp ip block has to be initialized before
ih block for the dynamic register programming interface
that needed for vf ih ring buffer. On the other hand,
current psp ip block hw_init function will initialize
xgmi session which actaully depends on interrupt to
return session context. This results an empty xgmi ta
session id and later failures on all the xgmi ta cmd
invoked from vf. xgmi ta session initialization has to
be done after ih ip block hw_init call.

to unify xgmi session init/fini for both bare-metal
sriov virtualization use scenario, move xgmi ta init
to xgmi_add_device call, and accordingly terminate xgmi
ta session in xgmi_remove_device call.

The existing suspend/resume sequence will not be changed.

v2: squash in return fix from Nirmoy

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

drm/amd/display: call psp set/get interfaces

Call the cmd ids for set/get srm according to the sysfs call

v2: Use define for the magic number

Signed-off-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Reviewed-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Load srm before enabling HDCP

[Why]
we need to load SRM before we start HDCP. Because for S3 case the sysfs call will be
after we have already enabled HDCP, so we might not be using the latest SRM

[How]
Set srm before starting HDCP.

Signed-off-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Reviewed-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Add sysfs interface for set/get srm

[Why]
PSP doesn't have the ability to store SRM in a non-volatile memory.  And since
the kernel cannot write to the storage directly, we need usermode to facilitate
this

As per spec the SRM needs to be persistent so this interface is to be
called by the usermode anytime the system goes down/powers on

*boot/resume: load from storage
*shutdown/suspend: save to storage

[How]
Provide a sysfs interface so that the usermode can set/get srm at the right times

save to storage: call "cat /sys/class/drm/card0/device/hdcp_srm > file" after boot and resume
-driver calls psp_get_srm() to get the stored srm and outputs it

load from storage: call "cat file > /sys/class/drm/card0/device/hdcp_srm" before shutdown and suspend
-driver reads the file from sysfs and calls psp_set_srm() to send the SRM to PSP

v2:
-update commit description
-add comment about sysfs file handling in the code

v3:
- squash in use after free fix (Dan Carpenter)

Signed-off-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Reviewed-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: update psp interface header

[Why]
We need to support SRM(System Renewability Message)
As per hdcp spec (5.Renewability) SRM needs to be storage in a non-volatile
memory.

PSP owns the checking of SRM but doesn't have the ability to store it in a
non-volatile memory. So we need the kernel driver to facilitate it using the
interface provided by PSP

[How]
Add the interface to the header file, so the driver can use them

v2: update commit description

Signed-off-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Reviewed-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Pass amdgpu_device instead of psp_context

[Why]
We need this to create sysfs (followup patch)

[How]
Change the parameter

Signed-off-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Reviewed-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: rework synchronization of VM updates v4

If provided we only sync to the BOs reservation
object and no longer to the root PD.

v2: update comment, cleanup amdgpu_bo_sync_wait_resv
v3: use correct reservation object while clearing
v4: fix typo in amdgpu_bo_sync_wait_resv

Signed-off-by: Christian König <christian.koenig@amd.com>
Tested-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: simplify and fix amdgpu_sync_resv

No matter what we always need to sync to moves.

Signed-off-by: Christian König <christian.koenig@amd.com>
Tested-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: allow higher level PD invalidations

Allow partial invalidation on unallocated PDs. This is useful when we
need to silence faults to stop interrupt floods on Vega.

Signed-off-by: Christian König <christian.koenig@amd.com>
Tested-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: return EINVAL instead of ENOENT in the VM code

That we can't find a PD above the root is expected can only happen if
we try to update a larger range than actually managed by the VM.

Signed-off-by: Christian König <christian.koenig@amd.com>
Tested-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: fix parentheses in amdgpu_vm_update_ptes

For the root PD mask can be 0xffffffff as well which would
overrun to 0 if we don't cast it before we add one.

Signed-off-by: Christian König <christian.koenig@amd.com>
Tested-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: make sure to never allocate PDs/PTs for invalidations

Make sure that we never allocate a page table for an invalidation operation.

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

drm/amdgpu: drop unnecessary restriction for huge root PDEs

The root PD can also contain huge PDEs.

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

drm/amdgpu: stop using amdgpu_bo_gpu_offset in the VM backend

We need to update page tables without any lock held.

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

drm/amdgpu: rework job synchronization v2

For unlocked page table updates we need to be able
to sync to fences of a specific VM.

v2: use SYNC_ALWAYS in the UVD code

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

drm/amdgpu: use the VM as job owner

For HMM we need to rework how VM synchronization works, so instead of the filp use VM as job owner.

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

drm/amdgpu: explicitly sync VM update to PDs/PTs

Explicitly sync VM updates to the moving fence in PDs and PTs.

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

drm/amd/dm/mst: Ignore payload update failures

Disabling a display on MST can potentially happen after the entire MST
topology has been removed, which means that we can't communicate with
the topology at all in this scenario. Likewise, this also means that we
can't properly update payloads on the topology and as such, it's a good
idea to ignore payload update failures when disabling displays.
Currently, amdgpu makes the mistake of halting the payload update
process when any payload update failures occur, resulting in leaving
DC's local copies of the payload tables out of date.

This ends up causing problems with hotplugging MST topologies, and
causes modesets on the second hotplug to fail like so:

[drm] Failed to updateMST allocation table forpipe idx:1
------------[ cut here ]------------
WARNING: CPU: 5 PID: 1511 at
drivers/gpu/drm/amd/amdgpu/../display/dc/core/dc_link.c:2677
update_mst_stream_alloc_table+0x11e/0x130 [amdgpu]
Modules linked in: cdc_ether usbnet fuse xt_conntrack nf_conntrack
nf_defrag_ipv6 libcrc32c nf_defrag_ipv4 ipt_REJECT nf_reject_ipv4
nft_counter nft_compat nf_tables nfnetlink tun bridge stp llc sunrpc
vfat fat wmi_bmof uvcvideo snd_hda_codec_realtek snd_hda_codec_generic
snd_hda_codec_hdmi videobuf2_vmalloc snd_hda_intel videobuf2_memops
videobuf2_v4l2 snd_intel_dspcfg videobuf2_common crct10dif_pclmul
snd_hda_codec videodev crc32_pclmul snd_hwdep snd_hda_core
ghash_clmulni_intel snd_seq mc joydev pcspkr snd_seq_device snd_pcm
sp5100_tco k10temp i2c_piix4 snd_timer thinkpad_acpi ledtrig_audio snd
wmi soundcore video i2c_scmi acpi_cpufreq ip_tables amdgpu(O)
rtsx_pci_sdmmc amd_iommu_v2 gpu_sched mmc_core i2c_algo_bit ttm
drm_kms_helper syscopyarea sysfillrect sysimgblt fb_sys_fops cec drm
crc32c_intel serio_raw hid_multitouch r8152 mii nvme r8169 nvme_core
rtsx_pci pinctrl_amd
CPU: 5 PID: 1511 Comm: gnome-shell Tainted: G           O      5.5.0-rc7Lyude-Test+ #4
Hardware name: LENOVO FA495SIT26/FA495SIT26, BIOS R12ET22W(0.22 ) 01/31/2019
RIP: 0010:update_mst_stream_alloc_table+0x11e/0x130 [amdgpu]
Code: 28 00 00 00 75 2b 48 8d 65 e0 5b 41 5c 41 5d 41 5e 5d c3 0f b6 06
49 89 1c 24 41 88 44 24 08 0f b6 46 01 41 88 44 24 09 eb 93 <0f> 0b e9
2f ff ff ff e8 a6 82 a3 c2 66 0f 1f 44 00 00 0f 1f 44 00
RSP: 0018:ffffac428127f5b0 EFLAGS: 00010202
RAX: 0000000000000002 RBX: ffff8d1e166eee80 RCX: 0000000000000000
RDX: ffffac428127f668 RSI: ffff8d1e166eee80 RDI: ffffac428127f610
RBP: ffffac428127f640 R08: ffffffffc03d94a8 R09: 0000000000000000
R10: ffff8d1e24b02000 R11: ffffac428127f5b0 R12: ffff8d1e1b83d000
R13: ffff8d1e1bea0b08 R14: 0000000000000002 R15: 0000000000000002
FS:  00007fab23ffcd80(0000) GS:ffff8d1e28b40000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007f151f1711e8 CR3: 00000005997c0000 CR4: 00000000003406e0
Call Trace:
 ? mutex_lock+0xe/0x30
 dc_link_allocate_mst_payload+0x9a/0x210 [amdgpu]
 ? dm_read_reg_func+0x39/0xb0 [amdgpu]
 ? core_link_enable_stream+0x656/0x730 [amdgpu]
 core_link_enable_stream+0x656/0x730 [amdgpu]
 dce110_apply_ctx_to_hw+0x58e/0x5d0 [amdgpu]
 ? dcn10_verify_allow_pstate_change_high+0x1d/0x280 [amdgpu]
 ? dcn10_wait_for_mpcc_disconnect+0x3c/0x130 [amdgpu]
 dc_commit_state+0x292/0x770 [amdgpu]
 ? add_timer+0x101/0x1f0
 ? ttm_bo_put+0x1a1/0x2f0 [ttm]
 amdgpu_dm_atomic_commit_tail+0xb59/0x1ff0 [amdgpu]
 ? amdgpu_move_blit.constprop.0+0xb8/0x1f0 [amdgpu]
 ? amdgpu_bo_move+0x16d/0x2b0 [amdgpu]
 ? ttm_bo_handle_move_mem+0x118/0x570 [ttm]
 ? ttm_bo_validate+0x134/0x150 [ttm]
 ? dm_plane_helper_prepare_fb+0x1b9/0x2a0 [amdgpu]
 ? _cond_resched+0x15/0x30
 ? wait_for_completion_timeout+0x38/0x160
 ? _cond_resched+0x15/0x30
 ? wait_for_completion_interruptible+0x33/0x190
 commit_tail+0x94/0x130 [drm_kms_helper]
 drm_atomic_helper_commit+0x113/0x140 [drm_kms_helper]
 drm_atomic_helper_set_config+0x70/0xb0 [drm_kms_helper]
 drm_mode_setcrtc+0x194/0x6a0 [drm]
 ? _cond_resched+0x15/0x30
 ? mutex_lock+0xe/0x30
 ? drm_mode_getcrtc+0x180/0x180 [drm]
 drm_ioctl_kernel+0xaa/0xf0 [drm]
 drm_ioctl+0x208/0x390 [drm]
 ? drm_mode_getcrtc+0x180/0x180 [drm]
 amdgpu_drm_ioctl+0x49/0x80 [amdgpu]
 do_vfs_ioctl+0x458/0x6d0
 ksys_ioctl+0x5e/0x90
 __x64_sys_ioctl+0x16/0x20
 do_syscall_64+0x55/0x1b0
 entry_SYSCALL_64_after_hwframe+0x44/0xa9
RIP: 0033:0x7fab2121f87b
Code: 0f 1e fa 48 8b 05 0d 96 2c 00 64 c7 00 26 00 00 00 48 c7 c0 ff ff
ff ff c3 66 0f 1f 44 00 00 f3 0f 1e fa b8 10 00 00 00 0f 05 <48> 3d 01
f0 ff ff 73 01 c3 48 8b 0d dd 95 2c 00 f7 d8 64 89 01 48
RSP: 002b:00007ffd045f9068 EFLAGS: 00000246 ORIG_RAX: 0000000000000010
RAX: ffffffffffffffda RBX: 00007ffd045f90a0 RCX: 00007fab2121f87b
RDX: 00007ffd045f90a0 RSI: 00000000c06864a2 RDI: 000000000000000b
RBP: 00007ffd045f90a0 R08: 0000000000000000 R09: 000055dbd2985d10
R10: 000055dbd2196280 R11: 0000000000000246 R12: 00000000c06864a2
R13: 000000000000000b R14: 0000000000000000 R15: 000055dbd2196280
---[ end trace 6ea888c24d2059cd ]---

Note as well, I have only been able to reproduce this on setups with 2
MST displays.

Changes since v1:
* Don't return false when part 1 or part 2 of updating the payloads
  fails, we don't want to abort at any step of the process even if
  things fail

Reviewed-by: Mikita Lipski <Mikita.Lipski@amd.com>
Signed-off-by: Lyude Paul <lyude@redhat.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Cc: stable@vger.kernel.org
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: update default voltage for boot od table for navi1x

It needed to be updated as well so it will show the proper values
if you reset to the defaults.

Bug: https://gitlab.freedesktop.org/drm/amd/issues/1020
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/smu10: fix smu10_get_clock_by_type_with_voltage

Cull out 0 clocks to avoid a warning in DC.

Bug: https://gitlab.freedesktop.org/drm/amd/issues/963
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/smu10: fix smu10_get_clock_by_type_with_latency

Only send non-0 clocks to DC for validation.  This mirrors
what the windows driver does.

Bug: https://gitlab.freedesktop.org/drm/amd/issues/963
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/display: handle multiple numbers of fclks in dcn_calcs.c (v2)

We might get different numbers of clocks from powerplay depending
on what the OEM has populated.

v2: add assert for at least one level

Bug: https://gitlab.freedesktop.org/drm/amd/issues/963
Reviewed-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: fetch default VDDC curve voltages (v2)

Ask the SMU for the default VDDC curve voltage values.  This
properly reports the VDDC values in the OD interface.

v2: only update if the original values are 0

Bug: https://gitlab.freedesktop.org/drm/amd/issues/1020
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org # 5.5.x

drm/amdgpu/smu_v11_0: Correct behavior of restoring default tables (v2)

Previously, the syfs functionality for restoring the default powerplay
table was sourcing it's information from the currently-staged powerplay
table.

This patch adds a step to cache the first overdrive table that we see on
boot, so that it can be used later to "restore" the powerplay table

v2: sqaush my original with Matt's fix

Bug: https://gitlab.freedesktop.org/drm/amd/issues/1020
Signed-off-by: Matt Coffin <mcoffin13@gmail.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org # 5.5.x

drm/amdgpu/navi10: add OD_RANGE for navi overclocking

So users can see the range of valid values.

Bug: https://gitlab.freedesktop.org/drm/amd/issues/1020
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org # 5.5.x

drm/amdgpu/navi: fix index for OD MCLK

You can only adjust the max mclk, not the min.

Bug: https://gitlab.freedesktop.org/drm/amd/issues/1020
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org # 5.5.x

drm/amd/display: Fix HW/SW state mismatch

[Why]
When we disable a connector we don't explicitly remove it from the module so the
display is still cached(SW) in the hdcp_module.

SST: no issues because we can only have 1 display per link

MST: We have x displays per link, now if we disable 1 we don't remove it from the
module so the module has x display cached(SW).

If we try to enable HDCP, psp verification will fail because we are reporting x
displays while the HW only has x-1 display enabled

[How]
Check the callback for when we disable stream and call remove display.

Signed-off-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Reviewed-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Fix a typo when computing dsc configuration

[why]
Remove a backslash symbol accidentally left in increase bpp function
when computing mst dsc configuration.

Signed-off-by: Mikita Lipski <mikita.lipski@amd.com>
Reviewed-by: Zhan Liu <zhan.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/powerplay: fix navi10 system intermittent reboot issue V2

This workaround is needed only for Navi10 12 Gbps SKUs.

V2: added SMU firmware version guard

Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Feifei Xu <Feifei.Xu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org

drm/amdkfd: Fix a bug in SDMA RLC queue counting under HWS mode

The sdma_queue_count increment should be done before
execute_queues_cpsch(), which calls pm_calc_rlib_size() where
sdma_queue_count is used to calculate whether over_subscription is
triggered.

With the previous code, when a SDMA queue is created,
compute_queue_count in pm_calc_rlib_size() is one more than the
actual compute queue number, because the queue_count has been
incremented while sdma_queue_count has not. This patch fixes that.

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

drm/amd/display: Only enable cursor on pipes that need it

[Why]
In current code we're essentially drawing the cursor on every pipe
that contains it. This only works when the planes have the same
scaling for src to dest rect, otherwise we'll get "double cursor" where
one cursor is incorrectly filtered and offset from the real position.

[How]
Without dedicated cursor planes on DCN we require at least one pipe
that matches the scaling of the current timing.

This is an optimization and workaround for the most common case where
the top-most plane is not scaled but the bottom-most plane is scaled.

Whenever a pipe has a parent pipe in the blending tree whose recout
fully contains the current pipe we can disable the pipe.

This only applies when the pipe is actually visible of course.

Signed-off-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/navi10: add mclk to navi10_get_clock_by_type_with_latency

Doesn't seem to be used, but add it just in case.

Reviewed-by: Matt Coffin <mcoffin13@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Fix implicit enum conversion in gfx_v9_4_ras_error_inject

Clang warns:

../drivers/gpu/drm/amd/amdgpu/gfx_v9_4.c:967:35: warning: implicit
conversion from enumeration type 'enum amdgpu_ras_block' to different
enumeration type 'enum ta_ras_block' [-Wenum-conversion]
        block_info.block_id = info->head.block;
                            ~ ~~~~~~~~~~~^~~~~
1 warning generated.

Use the function added in commit 8f10ff291b3d ("drm/amdgpu: Fix amdgpu
ras to ta enums conversion") that handles this conversion explicitly.

Fixes: 88600d16a817 ("drm/amdgpu: add RAS support for the gfx block of Arcturus")
Link: https://github.com/ClangBuiltLinux/linux/issues/849
Signed-off-by: Nathan Chancellor <natechancellor@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

radeon: completely remove lut leftovers

This is an oversight from

commit 9fecee3b4300357053f62f856bfc051e5964cb3e
Author: Peter Rosin <peda@axentia.se>
Date:   Thu Jul 13 18:25:36 2017 +0200

    drm: radeon: remove dead code and pointless local lut storage

v2: Also remove leftover local variable.

Cc: Peter Rosin <peda@axentia.se>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Michel Dänzer <mdaenzer@redhat.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Move drm_dp_mst_atomic_check() to the front of dc_validate_global_state()

[Why]
Need to do atomic check first, then validate global state.
If not, when connecting both MST and HDMI displays and
set a bad mode via xrandr, system will hang.

[How]
Move drm_dp_mst_atomic_check() to the front of
dc_validate_global_state().

Signed-off-by: Zhan Liu <zhan.liu@amd.com>
Reviewed-by: Mikita Lipski <mikita.lipski@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

radeon: insert 10ms sleep in dce5_crtc_load_lut

Per at least one tester this is enough magic to recover the regression
introduced for some people (but not all) in

commit 046144a856af7551fc95f309a17cc07ec45803d2
Author: Peter Rosin <peda@axentia.se>
Date:   Tue Jul 4 12:36:57 2017 +0200

    drm/fb-helper: factor out pseudo-palette

which for radeon had the side-effect of refactoring out a seemingly
redudant writing of the color palette.

10ms in a fairly slow modeset path feels like an acceptable form of
duct-tape, so maybe worth a shot and see what sticks.

Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Michel Dänzer <michel.daenzer@amd.com>
References: https://bugzilla.kernel.org/show_bug.cgi?id=198123
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: fix spelling mistake link_integiry_check -> link_integrity_check

There is a spelling mistake on the struct field name link_integiry_check,
fix this by renaming it.

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

amdgpu: using vmalloc requires includeing vmalloc.h

Fixes: d529df30bb4e ("drm/amdgpu: fix VRAM partially encroached issue in GDDR6 memory training(V2)")
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: allocate entities on demand

Currently we pre-allocate entities and fences for all the HW IPs on
context creation and some of which are might never be used.

This patch tries to resolve entity/fences wastage by creating entity
only when needed.

v2: allocate memory for entity and fences together

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

drm/amdgpu: Enable DISABLE_BARRIER_WAITCNT for Arcturus

In previous gfx9 parts, S_BARRIER shader instructions are implicitly
S_WAITCNT 0 instructions as well. This setting turns off that
mechanism in Arcturus and beyond. With this, shaders must follow the
ISA guide insofar as putting in explicit S_WAITCNT operations even
after an S_BARRIER.

v2: Fix patch title to list component

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

drm/amd/powerplay: fix spelling mistake "Attemp" -> "Attempt"

There are several spelling mistakes in PP_ASSERT_WITH_CODE messages.
Fix these.

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: fix for-loop with incorrectly sized loop counter (v2)

A for-loop is iterating from 0 up to 1000 however the loop variable count
is a u8 and hence not large enough.  Fix this by making count an int.
Also remove the redundant initialization of count since this is never used
and add { } on the loop statement make the loop block clearer.

v2: drop useless else (Walter Harms)

Addresses-Coverity: ("Operands don't affect result")
Fixes: 096b796c0e2d ("drm/amd/display: wait for update when setting dpg test pattern")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: enable GPU reset by default on renoir

Everything is in place.

Reviewed-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: enable GPU reset by default on Navi

Has been working fine for a while.

Reviewed-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: do not allocate display_mode_lib unnecessarily

This allocation isn't required and can fail when resuming from suspend.

Bug: https://gitlab.freedesktop.org/drm/amd/issues/1009
Signed-off-by: Dor Askayo <dor.askayo@gmail.com>
Reviewed-by: Leo Li <sunpeng.li@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add coreboot workaround for KV/KB

Coreboot seems to have a problem correctly setting up access to the stolen VRAM
on KV/KB. Use the direct access only when necessary.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reported-and-tested-by: Fredrik Bruhn <fredrik.bruhn@unibap.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

Revert "drm/amd/display: Don't skip link training for empty dongle"

This reverts commit f9a238ec4d7c86bc140151a88149ce6dc8df7a0a.

[WHY]
This change was working around a regression that occured in this:
commit a199d8733541 ("drm/amd/display: DP Compliance 400.1.1 failure")

With the fix to run verify_link_cap when the SINK_COUNT of
dongles becomes non-zero this change is no longer needed.

Cc: Louis Li <Ching-shih.Li@amd.com>
Cc: Wenjing Liu <Wenjing.Liu@amd.com>
Cc: Hersen Wu <hersenxs.wu@amd.com>
Cc: Eric Yang <Eric.Yang2@amd.com>
Reviewed-by: Wenjing Liu <Wenjing.Liu@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Retrain dongles when SINK_COUNT becomes non-zero

[WHY]
Two years ago the patch referenced by the Fixes tag stopped running
dp_verify_link_cap_with_retries during DP detection when the reason
for the detection was a short-pulse interrupt. This effectively meant
that we were no longer doing the verify_link_cap training on active
dongles when their SINK_COUNT changed from 0 to 1.

A year ago this was partly remedied with:
commit f9a238ec4d7c ("drm/amd/display: Don't skip link training for empty dongle")

This made sure that we trained the dongle on initial hotplug (without
connected downstream devices).

This is all fine and dandy if it weren't for the fact that there are
some dongles on the market that don't like link training when SINK_COUNT
is 0 These dongles will in fact indicate a SINK_COUNT of 0 immediately
after hotplug, even when a downstream device is connected, and then
trigger a shortpulse interrupt indicating a SINK_COUNT change to 1.

In order to play nicely we will need our policy to not link train an
active DP dongle when SINK_COUNT is 0 but ensure we train it when the
SINK_COUNT changes to 1.

[HOW]
Call dp_verify_link_cap_with_retries on detection even when the detection
is triggered from a short pulse interrupt.

With this change we can also revert this commit which we'll do in a separate
follow-up change:
commit f9a238ec4d7c ("drm/amd/display: Don't skip link training for empty dongle")

Fixes: a199d8733541 ("drm/amd/display: DP Compliance 400.1.1 failure")
Suggested-by: Louis Li <Ching-shih.Li@amd.com>
Tested-by: Louis Li <Ching-shih.Li@amd.com>
Cc: Wenjing Liu <Wenjing.Liu@amd.com>
Cc: Hersen Wu <hersenxs.wu@amd.com>
Cc: Eric Yang <Eric.Yang2@amd.com>
Reviewed-by: Wenjing Liu <Wenjing.Liu@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: original raven doesn't support full asic reset

So don't use it.

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

drm/amdgpu: attempt to enable gfxoff on more raven1 boards (v2)

Switch to a blacklist so we can disable specific boards
that are problematic.

v2: make the blacklist non-raven specific.

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

drm/amd/amdgpu: fix spelling mistake "to" -> "too"

There is a spelling mistake in a DRM_ERROR message. Fix it.

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/powerplay: use true, false for bool variable in smu7_hwmgr.c

Fixes coccicheck warning:

drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c:723:2-50: WARNING: Assignment of 0/1 to bool variable
drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c:733:3-52: WARNING: Assignment of 0/1 to bool variable
drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c:747:3-51: WARNING: Assignment of 0/1 to bool variable

Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: zhengbin <zhengbin13@huawei.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: fix doc by clarifying sched_list definition

expand sched_list definition for better understanding.
Also fix a typo atleast -> at least

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

drm/amdgpu: initialize bo_va_list when add gws to process

bo_va_list is list_head, so initialize it.

Signed-off-by: xinhui pan <xinhui.pan@amd.com>
Reviewed-by: Feifei Xu <Feifei.Xu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/vcn: use inst_idx relacing inst

Use inst_idx relacing inst in SOC15_DPG_MODE macro to avoid confusion.

Signed-off-by: James Zhu <James.Zhu@amd.com>
Reviewed-by: Leo Liu <leo.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/vcn: fix typo error

Fix typo error, should be inst_idx instead of inst.

Signed-off-by: James Zhu <James.Zhu@amd.com>
Reviewed-by: Leo Liu <leo.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/vcn: fix vcn2.5 instance issue

Fix vcn2.5 instance issue, vcn0 and vcn1 have same register offset

Signed-off-by: James Zhu <James.Zhu@amd.com>
Reviewed-by: Leo Liu <leo.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/vcn2.5: fix a bug for the 2nd vcn instance (v2)

Fix a bug for the 2nd vcn instance at start and stop.

v2: squash in unused label removal.

Signed-off-by: James Zhu <James.Zhu@amd.com>
Reviewed-by: Leo Liu <leo.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/vcn: Share vcn_v2_0_dec_ring_test_ring to vcn2.5

Share vcn_v2_0_dec_ring_test_ring to vcn2.5 to support
vcn software ring.

Signed-off-by: James Zhu <James.Zhu@amd.com>
Reviewed-by: Leo Liu <leo.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Use the correct flush_type in flush_gpu_tlb_pasid

The flush_type was incorrectly hard-coded to 0 when calling falling back
to MMIO-based invalidation in flush_gpu_tlb_pasid.

Fixes: ebc3e0b58e01 ("drm/amdgpu: export function to flush TLB via pasid")
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Oak Zeng <Oak.Zeng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Fix TLB invalidation request when using semaphore

Use a more meaningful variable name for the invalidation request
that is distinct from the tmp variable that gets overwritten when
acquiring the invalidation semaphore.

Fixes: dca15a58bcf4 ("drm/amdgpu: invalidate mmhub semaphore workaround in gmc9/gmc10")
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Yong Zhao <Yong.Zhao@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: use odm combine for YCbCr420 timing with h_active greater than 4096

[why]
FMT has limitation to support YCbCr420 with h_active greater than 4096.

[how]
Use odm combine to overcome the limitation.

Signed-off-by: Wenjing Liu <Wenjing.Liu@amd.com>
Reviewed-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Acked-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: init hw i2c speed

[Why]
Driver didn't init hw i2c speed cause hdcp hw cannot
send command, because the default value of speed register
is 0x2.

[How]
Restore the default speed when release i2c engine

Signed-off-by: Lewis Huang <Lewis.Huang@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: changed max_downscale_src_width to 4096.

[Why]
MPO isn't enabled on some 4k videos due to video source width is 4096
and the current limit is 3840.

[How]
Changed the limit to 4096.

Signed-off-by: Isabel Zhang <isabel.zhang@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: support VSC SDP update on video test pattern request

[why]
MSA will be deprecated in the future.
Need to support VSC during DP test automation.

[how]
Do not disable VSC during DP test automation.
TODO - need to add VSC update on DM side on test request.

Signed-off-by: Wenjing Liu <Wenjing.Liu@amd.com>
Reviewed-by: Nikola Cornij <Nikola.Cornij@amd.com>
Acked-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Check hw_init state when determining if DMCUB is initialized

[Why]
CW2 may already be programmed when coming back from S4. In this case
we want to unconditionally replace whatever DMCUB version is currently
enabled with the latest.

[How]
Check the hw_init flag to know whether or not we've previously executed
the initliazed routine.

Signed-off-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Reset inbox rptr/wptr when resetting DMCUB

[Why]
These logically make sense more to be set after the DMCUB has been
reset rather than when we setup the inbox.

[How]
Move them into the reset callback.

Signed-off-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Call ATOM_INIT instead of ATOM_ENABLE for DMCUB

[Why]
DMCUB command table doesn't support ATOM_ENABLE/ATOM_DISABLE anymore
so we never end up calling the DCN init path in DMCUB.

[How]
Map ATOM_ENABLE to ATOM_INIT only for DMCUB command table offloading.

Signed-off-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Add hardware reset interface for DMUB service

[Why]
We'll need this to perform a clean shutdown before unloading the driver.

[How]
It will call reset internally and set hw_init to false. It won't do
anything if the hardware isn't initialized.

Signed-off-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Fix update type for multiple planes

[Why]
determine_update_type_for_commit() uses pointers to single instance
of local variable to fill scaling/color info for all planes updates.
This is a bug, that leads to incorrect update type for commit in case
of multiple planes per crtc.
Each plane should refer to separate scaling/color data.

[How]
Use arrays for plane properties.
Bundle all properties into a single structure to simplify memory allocation.

Signed-off-by: Roman Li <roman.li@amd.com>
Reviewed-by: Nicholas Kazlauskas <Nicholas.Kazlauskas@amd.com>
Acked-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: fix rotation_angle to use enum values

[Why]
Hardcoded fixed values are not proper.

[How]
Use enum values instead of fixed numbers.

Signed-off-by: Brandon Syu <Brandon.Syu@amd.com>
Reviewed-by: Eric Yang <eric.yang2@amd.com>
Acked-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: 3.2.69

Signed-off-by: Aric Cyr <aric.cyr@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Refine i2c frequency calculating sequence

[Why]
In HG mode, vbios didn't call DispController_Init to program NV1x
XTAL_REF_DIV value when ASIC_INIT, but driver read XTAL_REF_DIV
to calculate i2c reference frequency. it cause i2c frequency change
from 100kHz to 200kHz.

[How]
remove get_speed function and calculate reference frequency at
set_speed functiton.

Signed-off-by: Lewis Huang <Lewis.Huang@amd.com>
Reviewed-by: Jun Lei <Jun.Lei@amd.com>
Acked-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Refactor to remove diags specific rgam func

[Why]
It has duplicate code for building regamma curve

[How]
Remove the duplicate code and use the same function for building regamma

Signed-off-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Reviewed-by: Krunoslav Kovac <Krunoslav.Kovac@amd.com>
Acked-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Fix DML dummyinteger types mismatch

[Why]
The types for dummyinteger1 and dummyinteger2 are unsigned
as part of the DML spec. They should not be long.

[How]
Make them unsigned int instead of long.

Signed-off-by: Jerry (Fangzhi) Zuo <Jerry.Zuo@amd.com>
Reviewed-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Do DMCUB hw_init before DC

[Why]
For DMCUB enabled hardware DC has a dependency on DMCUB already being
running.

Command table offloading will fail on first modeset if DMCUB isn't
initialized first.

[How]
Perform DMCUB hardware initialization before DC.

Signed-off-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Reviewed-by: Hersen Wu <hersenxs.wu@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Fixed comment styling

Switched to C-style comments for consistency

Signed-off-by: Haiyi Zhou <haiyi.zhou@amd.com>
Reviewed-by: Reza Amini <Reza.Amini@amd.com>
Acked-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: check pipe_ctx is split pipe or not

[Why]
Driver use pipe_ctx to reallocate payload may cause allocate
payload twice on same sink with split pipe.

[How]
Drvier must to check pipe_ctx is split pipe or not to avoid
reallocate payload twice on same sink.

Signed-off-by: Paul Hsieh <paul.hsieh@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Fallback to DMCUB when command table is missing

[Why]
If the command table isn't available then we can fallback to DMCUB
offloading if it's enabled and available.

[How]
Instead of assigning NULL for supported command table functions we can
fallback to the DMCUB when it's available.

Signed-off-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Reviewed-by: Wesley Chalmers <Wesley.Chalmers@amd.com>
Acked-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Get fb base and fb offset for DMUB from registers

[Why]
Under some hardware initialization sequences the fb base/fb offset
provided can be zero or hardwareinit can happen too late.

We want to ensure that we always have the correct fb_base/fb_offset
when performing DMCUB hardware initialization so we can do DMCUB
command table offloading during first dc hardware init.

[How]
Read from the DCN registers. VBIOS already filled these in for us.

Signed-off-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Reviewed-by: Wesley Chalmers <Wesley.Chalmers@amd.com>
Acked-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Add debug option to disable DSC support

[how]
Empty dsc enc caps when debug option is set to disable DSC.

Signed-off-by: Wenjing Liu <Wenjing.Liu@amd.com>
Reviewed-by: Nikola Cornij <Nikola.Cornij@amd.com>
Acked-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Do not send training pattern if VS Different

[Why]
The DP 1.4a Spec requires that training pattern only under certain
specific conditions. Currently driver will re-send
training pattern every time voltage swing value changes,
but that should not be the case.

[How]
Do not re-send training pattern every time VS values
are different. Only send it on the first iteration.

Signed-off-by: Sung Lee <sung.lee@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Abdoulaye Berthe <Abdoulaye.Berthe@amd.com>
Acked-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: update MSA and VSC SDP on video test pattern request

[why]
On video test pattern request we need to update MSA and VSC so
it will match the requested test pattern dynamic range field.

[how]
Update dynamic range field in MSA and disable VSC as updating VSC
info packet is complicated and not required for test pattern purpose.

Signed-off-by: Wenjing Liu <Wenjing.Liu@amd.com>
Reviewed-by: Nikola Cornij <Nikola.Cornij@amd.com>
Acked-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: individualize fence allocation per entity

Allocate fences for each entity and remove ctx->fences reference as
fences should be bound to amdgpu_ctx_entity instead amdgpu_ctx.

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

Revert "drm/amdgpu: fix modprobe failure of the secondary GPU when GDDR6 training enabled(V5)"

This reverts commit d48031b059e8c525f1fc7a29b6b961b53e53d881.

The patch will be replaced with a better solution, revert it.

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

drm/amdgpu: fix VRAM partially encroached issue in GDDR6 memory training(V2)

[why]
In GDDR6 BIST training, a certain mount of bottom VRAM will be encroached by
UMC, that causes problems(like GTT corrupted and page fault observed).

[how]
Saving the content of this bottom VRAM to system memory before training, and
restoring it after training to avoid VRAM corruption.

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Tianci.Yin <tianci.yin@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu:  remove unnecessary conversion to bool

Better clean that up before some automation starts to complain about it

Signed-off-by: Nirmoy Das <nirmoy.das@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add RAS support for the gfx block of Arcturus

Implement functions to do the RAS error injection and
query EDC counter.

Signed-off-by: Dennis Li <Dennis.Li@amd.com>
Reviewed-by: Guchun Chen <guchun.chen@amd.com>
Reviewed-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 EDC counter registers of gc for Arcturus

add reg headers to gc includes

v2: remove unused registers and fields in this patch set

Signed-off-by: Dennis Li <Dennis.Li@amd.com>
Reviewed-by: Guchun Chen <guchun.chen@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: abstract EDC counter clear to a separated function

1. Add IP prefix for the IP related codes.
2. Refactor the code to clear EDC counter.

Signed-off-by: Dennis Li <Dennis.Li@amd.com>
Reviewed-by: Guchun Chen <guchun.chen@amd.com>
Reviewed-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: refine the security check for RAS functions

To avoid calling RAS related functions when RAS feature isn't
supported in hardware. Change to check supported features, instead
of checking asic type.

v2: reuse amdgpu_ras_is_supported function, instead of introducing
a new flag for hardware ras feature.

Signed-off-by: Dennis Li <Dennis.Li@amd.com>
Reviewed-by: Guchun Chen <guchun.chen@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: enable RAS feature for more mmhub sub-blocks of Acrturus

Compared with Vg20, the size of mmhub range is changed from 2 to 8.

Signed-off-by: Dennis Li <Dennis.Li@amd.com>
Reviewed-by: Guchun Chen <guchun.chen@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: update mmhub 9.4.1 header files for Acrturus

Add mask & shift definition of MAM_D(0~3)MEM for all mmhub
ranges.

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

drm/amdgpu: read gfx register using RREG32_KIQ macro

Reading CP_MEM_SLP_CNTL register with RREG32_SOC15 macro will lead to
hang when GPU is in "gfxoff" state.
I do a uniform substitution here.

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