drm/amdgpu: add gmc support for dimgrey_cavefish

Same as navy_flounder.

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

drm/amdgpu: add support for dimgrey_cavefish firmware

Add support for dimgrey_cavefish cp/rlc firmware.

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

drm/amdgpu: set asic family and ip blocks for dimgrey_cavefish

Same as navi series.

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

drm/amdgpu: set fw load type for dimgrey_cavefish

Use direct load for dimgrey_cavefish.

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

drm/amdgpu: add dimgrey_cavefish gpu info firmware

Load gpu info firmware for dimgrey_cavefish.

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

drm/amdgpu: add dimgrey_cavefish asic type

Add chip type for dimgrey_cavefish.

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

drm/amd/display: Fix module load hangs when connected to an eDP

It was recently introduced a change that enables driver to disable
streams if pixel clock changes. Consequently, the code path executed in
the disable vbios function expanded to an encoder verification part.
The encoder loop is nested inside the pipe count loop, and both loops
share the 'i' variable in control of their flow. This situation may lead
to an infinite loop because the encoder loop constantly updates the `i`
variable, making the first loop always positive. As a result, we can see
a soft hang during the module load (modprobe amdgpu) and a series of
dmesg log that looks like this:

kernel:[  124.538727] watchdog: BUG: soft lockup - CPU#2 stuck for 22s!
[modprobe:1000]

RSP: 0018:ffffabbf419bf0e8 EFLAGS: 00000282
RAX: ffffffffc0809de0 RBX: ffff93b35ccc0000 RCX: ffff93b366c21800
RDX: 0000000000000000 RSI: 0000000000000141 RDI: ffff93b35ccc0000
RBP: ffffabbf419bf108 R08: ffffabbf419bf164 R09: 0000000000000001
R10: 0000000000000003 R11: 0000000000000003 R12: 0000000008677d40
R13: 0000000000000141 R14: ffff93b35cfc0000 R15: ffff93b35abc0000
FS:  00007f1400717540(0000) GS:ffff93b37f680000(0000)
     knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00005649b66b0968 CR3: 00000003e0fec000 CR4: 0000000000350ee0
Call Trace:
 amdgpu_device_rreg+0x17/0x20 [amdgpu]
 amdgpu_cgs_read_register+0x14/0x20 [amdgpu]
 dm_read_reg_func+0x3a/0xb0 [amdgpu]
 get_pixel_clk_frequency_100hz+0x30/0x50 [amdgpu]
 dc_commit_state+0x8f1/0xae0 [amdgpu]
 ? drm_calc_timestamping_constants+0x101/0x160 [drm]
 amdgpu_dm_atomic_commit_tail+0x39d/0x21a0 [amdgpu]
 ? dcn21_validate_bandwidth+0xe5/0x290 [amdgpu]
 ? kfree+0xc3/0x390
 ? dcn21_validate_bandwidth+0xe5/0x290 [amdgpu]
...
RSP: 002b:00007fff26009bd8 EFLAGS: 00000246 ORIG_RAX: 0000000000000139
RAX: ffffffffffffffda RBX: 000055a8025bea50 RCX: 00007f140085c89d
RDX: 0000000000000000 RSI: 000055a8025b8290 RDI: 000000000000000c
RBP: 0000000000040000 R08: 0000000000000000 R09: 0000000000000000
R10: 000000000000000c R11: 0000000000000246 R12: 000055a8025b8290
R13: 0000000000000000 R14: 000055a8025bead0 R15: 000055a8025bea50

This issue was fixed by introducing a second variable for the internal
loop.

Fixes: 8353d30e747f4e ("drm/amd/display: disable stream if pixel clock changed with link active")
Reviewed-by: Roman Li <Roman.Li@amd.com>
Reviewed-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Signed-off-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Add missing function pointers for dcn3

These function pointers are missing from dcn30_init

.calc_vupdate_position : Used to help avoid cursor stuttering
.set_pipe : Needed for setting ABM

So add them

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

drm/amdgpu: Fix inconsistent of format with argument type in amdgpu_xgmi.c

Fix follow warning:
[drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c:249]: (warning) %d in format
string (no. 1) requires 'int' but the argument type is 'unsigned int'.

Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Ye Bin <yebin10@huawei.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Fix invalid number of character '{' in amdgpu_acpi_init

Fix follow warning:
Checking drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c...
[drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c:770]: (error) Invalid number
of character '{' when these macros are defined: ''.
Checking drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c: CONFIG_ACPI...
[drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c:770]: (error) Invalid number
of character '{' when these macros are defined: 'CONFIG_ACPI'.
......
Checking drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c: CONFIG_X86...
[drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c:770]: (error) Invalid number
of character '{' when these macros are defined: 'CONFIG_X86'.
Checking drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c: _X86_...
[drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c:770]: (error) Invalid number
of character '{' when these macros are defined: '_X86_'.
Checking drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c: __linux__...
[drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c:770]: (error) Invalid number
of character '{' when these macros are defined: '__linux__'.

Fixes: 97d798b276e9 ("drm/amdgpu: simplify ATIF backlight handling")
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Ye Bin <yebin10@huawei.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/swsmu: clean up a bunch of stale interfaces

These were leftover from the initial implementation, but
never used.  Drop them.

Reviewed-by: Evan Quan <evan.quan@amd.com>
Noticed-by: Ryan Taylor <ryan.taylor@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: setup APU dpm clock table in SMU HW initialization

As the dpm clock table is needed during DC HW initialization.
And that (DC HW initialization) comes before smu_late_init()
where current APU dpm clock table setup is performed. So, NULL
pointer dereference will be triggered. By moving APU dpm clock
table setup to smu_hw_init(), this can be avoided.

Fixes: 02cf91c113ea ("drm/amd/powerplay: postpone operations not required for hw setup to late_init")
Signed-off-by: Evan Quan <evan.quan@amd.com>
Reported-by: Dirk Gouders <dirk@gouders.net>
Acked-by: Nirmoy Das <nirmoy.das@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: modify cp_flags to pg_flags in gfx_v10_cntl_power_gating

It needs to use adev->pg_flags other than adev->cg_glags in
gfx_v10_cntl_power_gating

Signed-off-by: Changfeng <Changfeng.Zhu@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Remove warning for virtual_display

Remove the virtual_display warning in drm_crtc_vblank_off when
dev->num_crtcs is null.

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

drm/amdgpu: kfd_initialized can be static

Fixes: c7651b73586600 ("drm/amdgpu: Fix handling of KFD initialization failures")
Signed-off-by: kernel test robot <lkp@intel.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

amd/amdgpu_ctx: Use struct_size() helper and kmalloc() (v2)

Make use of the new struct_size() helper instead of the offsetof() idiom.
Also, use kmalloc() instead of kcalloc().

v2: squash in kzalloc fix

Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Use struct_size() helper in kmalloc()

Make use of the new struct_size() helper instead of the offsetof() idiom.

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: Replace one-element array with flexible-array in struct phm_ppt_v1_pcie_table

There is a regular need in the kernel to provide a way to declare having
a dynamically sized set of trailing elements in a structure. Kernel code
should always use “flexible array members”[1] for these cases. The older
style of one-element or zero-length arrays should no longer be used[2].

Refactor the code according to the use of a flexible-array member in
struct phm_ppt_v1_pcie_table, instead of a one-element array, and use
the struct_size() helper to calculate the size for the allocation.

[1] https://en.wikipedia.org/wiki/Flexible_array_member
[2] https://www.kernel.org/doc/html/v5.9-rc1/process/deprecated.html#zero-length-and-one-element-arrays

Build-tested-by: kernel test robot <lkp@intel.com>
Link: https://lore.kernel.org/lkml/5f7db0bc.7Xivn4K83f7XW0ug%25lkp@intel.com/
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: Replace one-element array with flexible-array in struct phm_ppt_v1_voltage_lookup_table

There is a regular need in the kernel to provide a way to declare having
a dynamically sized set of trailing elements in a structure. Kernel code
should always use “flexible array members”[1] for these cases. The older
style of one-element or zero-length arrays should no longer be used[2].

Refactor the code according to the use of a flexible-array member in
struct phm_ppt_v1_voltage_lookup_table, instead of a one-element array,
and use the struct_size() helper to calculate the size for the allocation.

[1] https://en.wikipedia.org/wiki/Flexible_array_member
[2] https://www.kernel.org/doc/html/v5.9-rc1/process/deprecated.html#zero-length-and-one-element-arrays

Build-tested-by: kernel test robot <lkp@intel.com>
Link: https://lore.kernel.org/lkml/5f7d61df.jWrFfnjxGbjSkPOp%25lkp@intel.com/
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: Replace one-element array with flexible-array in struct phm_ppt_v1_mm_clock_voltage_dependency_table

There is a regular need in the kernel to provide a way to declare having
a dynamically sized set of trailing elements in a structure. Kernel code
should always use “flexible array members”[1] for these cases. The older
style of one-element or zero-length arrays should no longer be used[2].

Refactor the code according to the use of a flexible-array member in
struct phm_ppt_v1_mm_clock_voltage_dependency_table, instead of a
one-element array, and use the struct_size() helper to calculate the
size for the allocation.

[1] https://en.wikipedia.org/wiki/Flexible_array_member
[2] https://www.kernel.org/doc/html/v5.9-rc1/process/deprecated.html#zero-length-and-one-element-arrays

Build-tested-by: kernel test robot <lkp@intel.com>
Link: https://lore.kernel.org/lkml/5f7d61e2.qiTVTyG2pVoG8bb0%25lkp@intel.com/
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: Replace one-element array with flexible-array in struct phm_ppt_v1_clock_voltage_dependency_table

There is a regular need in the kernel to provide a way to declare having
a dynamically sized set of trailing elements in a structure. Kernel code
should always use “flexible array members”[1] for these cases. The older
style of one-element or zero-length arrays should no longer be used[2].

Refactor the code according to the use of a flexible-array member in
struct phm_ppt_v1_clock_voltage_dependency_table, instead of a one-element
array, and use the struct_size() helper to calculate the size for the
allocation.

[1] https://en.wikipedia.org/wiki/Flexible_array_member
[2] https://www.kernel.org/doc/html/v5.9-rc1/process/deprecated.html#zero-length-and-one-element-arrays

Build-tested-by: kernel test robot <lkp@intel.com>
Link: https://lore.kernel.org/lkml/5f7c433c.TTk9rnA+F58kyDUy%25lkp@intel.com/
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: Replace one-element array with flexible-array in struct phm_samu_clock_voltage_dependency_table

There is a regular need in the kernel to provide a way to declare having
a dynamically sized set of trailing elements in a structure. Kernel code
should always use “flexible array members”[1] for these cases. The older
style of one-element or zero-length arrays should no longer be used[2].

Refactor the code according to the use of a flexible-array member in
struct phm_samu_clock_voltage_dependency_table, instead of a one-element array,
and use the struct_size() helper to calculate the size for the allocation.

Also, save some heap space as the original code is multiplying
table->numEntries by sizeof(struct phm_samu_clock_voltage_dependency_table)
when it should have been multiplied it by
sizeof(struct phm_samu_clock_voltage_dependency_record) instead.

[1] https://en.wikipedia.org/wiki/Flexible_array_member
[2] https://www.kernel.org/doc/html/v5.9-rc1/process/deprecated.html#zero-length-and-one-element-arrays

Build-tested-by: kernel test robot <lkp@intel.com>
Link: https://lore.kernel.org/lkml/5f7c5d3a.ryM4GmZr3e0JeZy+%25lkp@intel.com/
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: Replace one-element array with flexible-array in struct phm_cac_leakage_table

There is a regular need in the kernel to provide a way to declare having
a dynamically sized set of trailing elements in a structure. Kernel code
should always use “flexible array members”[1] for these cases. The older
style of one-element or zero-length arrays should no longer be used[2].

Refactor the code according to the use of a flexible-array member in
struct phm_cac_leakage_table, instead of a one-element array,
and use the struct_size() helper to calculate the size for the allocation.

Also, save some heap space as the original code is multiplying
table->ucNumEntries by sizeof(struct phm_cac_leakage_table) when it
should have been multiplied it by sizeof(struct phm_cac_leakage_record)
instead.

[1] https://en.wikipedia.org/wiki/Flexible_array_member
[2] https://www.kernel.org/doc/html/v5.9-rc1/process/deprecated.html#zero-length-and-one-element-arrays

Build-tested-by: kernel test robot <lkp@intel.com>
Link: https://lore.kernel.org/lkml/5f7c5d38.iT%2FQTjN+659XUDo5%25lkp@intel.com/
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: Replace one-element array with flexible-array in struct phm_vce_clock_voltage_dependency_table

There is a regular need in the kernel to provide a way to declare having
a dynamically sized set of trailing elements in a structure. Kernel code
should always use “flexible array members”[1] for these cases. The older
style of one-element or zero-length arrays should no longer be used[2].

Refactor the code according to the use of a flexible-array member in
struct phm_vce_clock_voltage_dependency_table, instead of a one-element array,
and use the struct_size() helper to calculate the size for the allocation.

Also, save some heap space as the original code is multiplying
table->numEntries by sizeof(struct phm_vce_clock_voltage_dependency_table)
when it should have multiplied it by sizeof(struct phm_vce_clock_voltage_dependency_record)
instead.

[1] https://en.wikipedia.org/wiki/Flexible_array_member
[2] https://www.kernel.org/doc/html/v5.9-rc1/process/deprecated.html#zero-length-and-one-element-arrays

Build-tested-by: kernel test robot <lkp@intel.com>
Link: https://lore.kernel.org/lkml/5f7c5d35.pJToGs3H9khZK6ws%25lkp@intel.com/
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: Replace one-element array with flexible-array in struct phm_phase_shedding_limits_table

There is a regular need in the kernel to provide a way to declare having
a dynamically sized set of trailing elements in a structure. Kernel code
should always use “flexible array members”[1] for these cases. The older
style of one-element or zero-length arrays should no longer be used[2].

Refactor the code according to the use of a flexible-array member in
struct phm_phase_shedding_limits_table, instead of a one-element array,
and use the struct_size() helper to calculate the size for the allocation.

Also, save some heap space as the original code is multiplying
ptable->ucNumEntries by sizeof(struct phm_phase_shedding_limits_table)
when it should have multiplied it by sizeof(struct phm_phase_shedding_limits_record)
instead.

[1] https://en.wikipedia.org/wiki/Flexible_array_member
[2] https://www.kernel.org/doc/html/v5.9-rc1/process/deprecated.html#zero-length-and-one-element-arrays

Build-tested-by: kernel test robot <lkp@intel.com>
Link: https://lore.kernel.org/lkml/5f7c5d36.6PStUZp2HRxAz7IM%25lkp@intel.com/
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: Replace one-element array with flexible-array in struct phm_acp_clock_voltage_dependency_table

There is a regular need in the kernel to provide a way to declare having
a dynamically sized set of trailing elements in a structure. Kernel code
should always use “flexible array members”[1] for these cases. The older
style of one-element or zero-length arrays should no longer be used[2].

Refactor the code according to the use of a flexible-array member in
struct phm_acp_clock_voltage_dependency_table, instead of a one-element
array, and use the struct_size() helper to calculate the size for the
allocation.

Also, save some heap space as the original code is multiplying
table->numEntries by sizeof(struct phm_acp_clock_voltage_dependency_table)
when it should have multiplied it by sizeof(phm_acp_clock_voltage_dependency_record)
instead.

[1] https://en.wikipedia.org/wiki/Flexible_array_member
[2] https://www.kernel.org/doc/html/v5.9-rc1/process/deprecated.html#zero-length-and-one-element-arrays

Build-tested-by: kernel test robot <lkp@intel.com>
Link: https://lore.kernel.org/lkml/5f7c5d3c.TyfOhg%2FA6JycL6ZN%25lkp@intel.com/
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: Replace one-element array with flexible-array in struct phm_uvd_clock_voltage_dependency_table

There is a regular need in the kernel to provide a way to declare having
a dynamically sized set of trailing elements in a structure. Kernel code
should always use “flexible array members”[1] for these cases. The older
style of one-element or zero-length arrays should no longer be used[2].

Refactor the code according to the use of a flexible-array member in
struct phm_uvd_clock_voltage_dependency_table, instead of a one-element
array, and use the struct_size() helper to calculate the size for the
allocation.

Also, save some heap space as the original code is multiplying
table->numEntries by sizeof(struct phm_uvd_clock_voltage_dependency_table)
when it should have multiplied it by sizeof(phm_uvd_clock_voltage_dependency_record)
instead.

[1] https://en.wikipedia.org/wiki/Flexible_array_member
[2] https://www.kernel.org/doc/html/v5.9-rc1/process/deprecated.html#zero-length-and-one-element-arrays

Build-tested-by: kernel test robot <lkp@intel.com>
Link: https://lore.kernel.org/lkml/5f7c433e.pXkC6KsN6HN%2FLdhj%25lkp@intel.com/
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: Replace one-element array with flexible-array in struct phm_clock_array

There is a regular need in the kernel to provide a way to declare having
a dynamically sized set of trailing elements in a structure. Kernel code
should always use “flexible array members”[1] for these cases. The older
style of one-element or zero-length arrays should no longer be used[2].

Refactor the code according to the use of a flexible-array member in
struct phm_clock_array, instead of a one-element array, and use the
struct_size() helper to calculate the size for the allocation.

[1] https://en.wikipedia.org/wiki/Flexible_array_member
[2] https://www.kernel.org/doc/html/v5.9-rc1/process/deprecated.html#zero-length-and-one-element-arrays

Build-tested-by: kernel test robot <lkp@intel.com>
Link: https://lore.kernel.org/lkml/5f7c433f.ZyMD+YUIVAwiHGVe%25lkp@intel.com/
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: Replace one-element array with flexible-array member in struct vi_dpm_table

There is a regular need in the kernel to provide a way to declare having
a dynamically sized set of trailing elements in a structure. Kernel code
should always use “flexible array members”[1] for these cases. The older
style of one-element or zero-length arrays should no longer be used[2].

Use a flexible-array member in struct vi_dpm_table instead of a
one-element array.

[1] https://en.wikipedia.org/wiki/Flexible_array_member
[2] https://www.kernel.org/doc/html/v5.9-rc1/process/deprecated.html#zero-length-and-one-element-arrays

Build-tested-by: kernel test robot <lkp@intel.com>
Link: https://lore.kernel.org/lkml/5f7c433c.TTk9rnA+F58kyDUy%25lkp@intel.com/
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: Replace one-element array with flexible-array member

There is a regular need in the kernel to provide a way to declare having
a dynamically sized set of trailing elements in a structure. Kernel code
should always use “flexible array members”[1] for these cases. The older
style of one-element or zero-length arrays should no longer be used[2].

Refactor the code according to the use of a flexible-array member in
struct phm_clock_voltage_dependency_table, instead of a one-element
array, and use the struct_size() helper to calculate the size for the
allocation.

[1] https://en.wikipedia.org/wiki/Flexible_array_member
[2] https://www.kernel.org/doc/html/v5.9-rc1/process/deprecated.html#zero-length-and-one-element-arrays

Build-tested-by: kernel test robot <lkp@intel.com>
Link: https://lore.kernel.org/lkml/5f7c295c.8iqp1Ifc6oiVDq%2F%2F%25lkp@intel.com/
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: enable vcn support for green_sardine (v2)

Enable Green_Sardine VCN support and VCN firmware loading

v2: use apu flags

Signed-off-by: Thong Thai <thong.thai@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: enable green_sardine_asd.bin loading (v2)

This patch enable green_sardine_asd.bin loading.

v2: use apu flags

Signed-off-by: Aaron Liu <aaron.liu@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Prike Liang <Prike.Liang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/sdma: add sdma engine support for green_sardine (v2)

Initialize the SDMA IP for green_sardine.

v2: use apu flags

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

drm/amdgpu: add gfx support for green_sardine (v2)

Enable the gfx base HW function of green_sardine.

v2: use apu flags

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

drm/amdgpu: add soc15 common ip block support for green_sardine (v3)

This patch adds common ip support for green_sardine.

v2: use apu flags, squash in CG/PG enablement
v3: rebase

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

drm/amdgpu: add green_sardine support for gpu_info and ip block setting (v2)

This patch adds green_sardine support for gpu_info firmware and ip block setting.

v2: use apu flag

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

drm/amdgpu: add Green_Sardine APU flag

Will be used for Green_Sardine which is a new APU.

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

drm/amdgpu: prevent spurious warning

The default auto setting for kcq should not generate
a warning.

Fixes: a300de40f66b ("drm/amdgpu: introduce a new parameter to configure how many KCQ we want(v5)")
Reviewed-by: Kent Russell <kent.russell@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/swsmu: fix ARC build errors

We want to use the dev_* functions here rather than the pr_* variants.
Switch to using dev_warn() which mirrors what we do on other asics.

Fixes the following build errors on ARC:

../drivers/gpu/drm/amd/amdgpu/../powerplay/navi10_ppt.c: In function 'navi10_fill_i2c_req':
../arch/arc/include/asm/bug.h:24:2: error: implicit declaration of function 'pr_warn'; did you mean 'drm_warn'? [-Werror=implicit-function-declaration]

../drivers/gpu/drm/amd/amdgpu/../powerplay/sienna_cichlid_ppt.c: In function 'sienna_cichlid_fill_i2c_req':
../arch/arc/include/asm/bug.h:24:2: error: implicit declaration of function 'pr_warn'; did you mean 'drm_warn'? [-Werror=implicit-function-declaration]

Reported-by: kernel test robot <lkp@intel.com>
Cc: Randy Dunlap <rdunlap@infradead.org>
Cc: Evan Quan <evan.quan@amd.com>
Cc: Vineet Gupta <vgupta@synopsys.com>
Cc: linux-snps-arc@lists.infradead.org
Acked-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add xgmi perfmons for arcturus

Add xgmi perfmons for Arcturus.

v4: Align with patch 2 with struct improvements
v3: Align with patch 2 streamlining perf types versus event config types.
v2: Resend for re-review with alignment for v3 in patch 2.

Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Tested-by: Chris Freehill <chris.freehill@amd.com>
Reviewed-by: Harish Kasiviswanathan <harish.kasiviswanathan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add per device user friendly xgmi events for vega20

Non-outbound data metrics are non useful so mark them as legacy.
Bucket new perf counters into device and not device ip.
Bind events to chip instead of IP.
Report available event counters and not number of hw counter banks.
Move DF public macros to private since not needed outside of IP version.

v5: cleanup by moving per chip configs into structs

v4: After more discussion, replace *_LEGACY references with IP references
to indicate concept of pmu-typed versus event-config-typed event
registration.

v3: attr groups const array is global but attr groups are allocated per
device which doesn't work and causes problems on memory allocation and
de-allocation for pmu unregister. Switch to building const attr groups
per pmu instead to simplify solution.

v2: add comments on sysfs structure and formatting.

Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Reviewed-by: Harish Kasiviswanathan <harish.kasiviswanathan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: fix xgmi perfmon a-b-a problem

Mapping hw counters per event config will cause ABA problems so map per
event instead.

v2: Discontinue starting perf counters if add fails.  Make it clear what's
happening with pmc_start.

Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Reviewed-by: Harish Kasiviswanathan <harish.kasiviswanathan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: 3.2.106

Signed-off-by: Aric Cyr <aric.cyr@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: [FW Promotion] Release 0.0.36

Signed-off-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Only flush inst_fb if backdoor loading

[Why]
DAL resume from BACO time is longer if we always flush inst_fb

[How]
Check if backdoor loading to flush inst_fb

Signed-off-by: Alvin Lee <alvin.lee2@amd.com>
Reviewed-by: Nicholas Kazlauskas <Nicholas.Kazlauskas@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Program meta addresses correctly

[Why]
When forcing 3D mode in DAL, we set the right address to be the same as the
left address. We need to do the same for the meta addresses.

[How]
Program right meta to be same as left meta.

Signed-off-by: Alvin Lee <alvin.lee2@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Hardcode 45W Bounding Box for DCN2.1 Diags

[WHY]
Currently construction of clock limits gets skipped for diags.
This logic would be useful to get tested in diags.

[HOW]
Copy existing states to clk_table such that update_bw_bounding_box
logic gets used to fill the table.

Signed-off-by: Sung Lee <sung.lee@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: FreeSync not active near lower bound of non-LFC monitor range

[Why]
On narrow range monitors without LFC, a margin prevents good utilization
of the available range.

[How]
Decrease the margin for exiting fixed mode and fix the frame counter to
reset if a non-consecutive render is found.

Signed-off-by: Aric Cyr <aric.cyr@amd.com>
Reviewed-by: Anthony Koo <Anthony.Koo@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Fix OPTC_DATA_FORMAT programming

This should be programmed with timing rather than with odm.

Signed-off-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Add i2c speed arbitration for dc_i2c and hdcp_i2c

[why]
HDCP 1.4 failed on SL8800 SW w/a test driver use.

[how]
Slow down the HW i2c speed when used by HW i2c.
This request: each acquired_i2c_engine setup the i2c speed needed
and sets the I2c engine for HDCP use at release_engine.

This covers SW using HW i2c engine and HDCP using HW i2c engine.
for dmcu using HW i2c engine, needs add similar logic in dmcu fw.

Signed-off-by: Charlene Liu <Charlene.Liu@amd.com>
Reviewed-by: Chris Park <Chris.Park@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Block ABM in case of eDP ODM

[Why]
enable ODM on eDP panel with ABM will result in color difference
on the panel due to only one ABM module to set one pipe.

[How]
Block ABM in case of ODM enabled on eDP.

Signed-off-by: Yongqiang Sun <yongqiang.sun@amd.com>
Reviewed-by: Eric Yang <eric.yang2@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Copy WM values from set A to other sets in hw_init

[Why]
When we transfer the WM range table to SMU, they can perform a watermark
switch right away. This can be a problem if we're in not in accelerated mode
during hw_init as SMU may initiate a dummy p-state change before the rest
of the watermarks are programmed. Watermark set A is defined to be
sufficient for all cases, so we can copy the values from set A to all other
sets, avoiding any issues from SMU doing WM switches.

[How]
 - new hubbub func init_watermarks
 - copy register values from set A to all other sets
 - call init_watermarks before calling notify_wm_ranges

Signed-off-by: Joshua Aberback <joshua.aberback@amd.com>
Reviewed-by: Jun Lei <Jun.Lei@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Force enable pstate on driver unload

[Why]
During driver unload, it is expected that p-state switching is supported.
If it's not supported, PMFW will hang due to a forced p-state switch. Even
if the current timing does not support p-state normally, we still want to
force allow because the worst that can happen is underflow. This will
match Navi10 behaviour.

[How]
 - new hubbub func to control the force pstate register
 - force allow when releasing display ownership
   - registers are inaccessible after due to m_cgs.hwNotAvailable
 - explicitly disable force signal during hw_init
   - if driver is disabled and re-enabled, register not cleared otherwise

Also, remove DCN3 part of dcn10_init_hw, we will not be going back to it.

Signed-off-by: Joshua Aberback <joshua.aberback@amd.com>
Reviewed-by: Jun Lei <Jun.Lei@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Fixed comments (uniform style)

[WHY]
This change was implemented because the comment style was not uniform
across the file. In some lines comments were initiated with // and in
others they were in between /* ... */.
Additionally, the style for multi-line comments was also not uniform and
some comment lines were missing the space between the opening /* and the
first word of the comment.

[HOW]
All comments are now in between /*.../*, multi line comments also use
/*...*/ and for every comment there is now a space between the opening
/* and the first word of the comment.

Signed-off-by: Felipe <Felipe.Clark@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Check for flip pending before locking pipes

[Why]
When running games or benchmarking with v-sync disabled, disabling
a plane (which is v-sync) can cause underflow. This is caused by
flips pending before pipe locking being applied after locks are
released and pipes could have been re-arranged or disconnected. This
could potentially apply a flip on incorrect pipe. Also, previous logic
of always locking pipes was unnecessary.

[How]
Only lock the pipes when there is a pipe being disabled to increase
efficiency. Before the pipes are locked, check that any pending flips
are cleared to ensure the flips are applied to the correct pipe.

Signed-off-by: Aric Cyr <aric.cyr@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Don't allow pstate if no support in blank

[Why]
We will hang if we report switch in VACTIVE but not in VBLANK and DPG_EN = 1

[How]
Block switch in ACTIVE if not supported in BLANK

Signed-off-by: Alvin Lee <alvin.lee2@amd.com>
Reviewed-by: Jun Lei <Jun.Lei@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: 3.2.105

Signed-off-by: Aric Cyr <aric.cyr@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add van gogh pci id

Add Van Gogh PCI id support.

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Add dcn3.01 support to DM

Update dm for vangogh support.

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

drm/amd/display: Add dcn3.01 support to DC (v2)

Update dc for vangogh support.

v2: fix compilation without DCN 301 set.

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

drm/amdgpu: enable gfx clock gating and power gating for vangogh

This patch is to enable the gfx cg and pg for vangogh.

Signed-off-by: Huang Rui <ray.huang@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add gfx power gating for gfx10

This patch adds power gating handler for gfx10.

v2: simplify function

Signed-off-by: Huang Rui <ray.huang@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/mmhub2.3: print client id string for mmhub

Print the name of the client rather than the number.  This
makes it easier to debug what block is causing the fault.

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

drm/amdgpu: IP discovery table is not ready yet for VG

Fallback to legacy path for now.

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

drm/amdgpu: disable gfxoff on vangogh for the moment (v2)

GFXOFF will be enabled once it's verified on real asic.

v2: move check into gfx10 module.

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Aaron Liu <aaron.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: enable psp support for vangogh

This patch is to enable psp support for vangogh

Signed-off-by: Huang Rui <ray.huang@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add TOC firmware support for apu (v3)

APU needs load toc firmware for gfx10 series on psp front door loading.

v2: rebase against latest code
v3: clarify error message

Signed-off-by: Huang Rui <ray.huang@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add TOC firmware definition

This patch is to add TOC firmware definition on uapi.

Signed-off-by: Huang Rui <ray.huang@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add smu ip block for vangogh

This patch is to add ip block for vangogh.

Signed-off-by: Huang Rui <ray.huang@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Aaron Liu <aaron.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/powerplay: add vangogh ppt into swSMU

This patch is to add vangogh ppt funcions into swSMU block.

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/powerplay: partially enable swsmu for vangogh

This patch is to partially enable swSMU for vangogh for the moment.

Signed-off-by: Huang Rui <ray.huang@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Aaron Liu <aaron.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/powerplay: add smu initialize funcitons for vangogh (v4)

This patch is to add smu initialize functions for vangogh.

v2: squash in updates
v3: drop duplicate table entries
v4: rebase fixes

Signed-off-by: Xiaojian Du <xiaojian.du@amd.com>
Reviewed-by: Kevin Wang <kevin1.wang@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/powerplay: add vangogh asic name in smu v11 (v2)

This patch is to add vangogh asic name in smu v11.

v2: drop smu firmware name (N/A for VG)

Signed-off-by: Xiaojian Du <xiaojian.du@amd.com>
Reviewed-by: Kevin Wang <kevin1.wang@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/powerplay: add smu v11.5 smc header for vangogh

This patch is to add smu v11.5 smc header for vangogh.

Signed-off-by: Xiaojian Du <xiaojian.du@amd.com>
Reviewed-by: Kevin Wang <kevin1.wang@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/powerplay: add smu v11.5 firmware header for vangogh (v2)

This patch is to add smu v11.5 firmware header for vangogh

v2: squash in updates

Signed-off-by: Xiaojian Du <xiaojian.du@amd.com>
Reviewed-by: Kevin Wang <kevin1.wang@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/admgpu/powerplay: add smu v11.5 driver interface header for vangogh

This patch is to add smu v11.5 driver interface header for vangogh.

Signed-off-by: Xiaojian Du <xiaojian.du@amd.com>
Reviewed-by: Kevin Wang <kevin1.wang@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/powerplay: add new smu messages and feature masks for vangogh (v2)

This patch is to add new smu messages and feature masks for vangogh.

v2: squash in updates and typo fixes

Signed-off-by: Xiaojian Du <xiaojian.du@amd.com>
Reviewed-by: Kevin Wang <kevin1.wang@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add nbio v7.2 for vangogh (v2)

VanGogh uses nbio v7.2, and a couple of offsets are changed since nbio
v2.3 for navi series, so add new nbio v7.2 block.

v2: squash in fix for sdma and vcn instances

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add pcie port indirect read and write on nv

This patch is to add pcie port indirect read/write callback for nv
series. They will be used for new asic.

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: enable vcn3.0 for van gogh

Same as other VCN 3.0 asics.

Signed-off-by: Thong Thai <thong.thai@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add mmhub v2.3 for vangogh (v4)

There are too many register offset mismatch between mmhub v2.0 and v2.3.

E.X:

mmMM_ATC_L2_MISC_CG:  0x064a(v2.0)  0x06cd(v2.3)
mmMMVM_L2_PROTECTION_FAULT_CNTL: 0x0688(v2.0) 0x0708(v2.3)
mmMMVM_CONTEXT0_PAGE_TABLE_BASE_ADDR_LO32: 0x072b(v2.0) 0x0940(v2.3)
mmMMVM_CONTEXT0_PAGE_TABLE_BASE_ADDR_HI32: 0x072c(v2.0) 0x0941(v2.3)
mmMMVM_INVALIDATE_ENG0_REQ: 0x06e3(v2.0) 0x0a01(v2.3)
mmMMVM_INVALIDATE_ENG0_ACK: 0x06f5(v2.0) 0x0a02(v2.3)
mmMMVM_CONTEXT0_CNTL: 0x06c0(v2.0) 0x0740(v2.3)
mmMMVM_L2_PROTECTION_FAULT_STATUS: 0x068c(v2.0) 0x070c(v2.3)
mmMMVM_L2_PROTECTION_FAULT_CNTL: 0x0688(v2.0) 0x0708(v2.3)
mmMM_ATC_L2_MISC_CG: 0x064a(v2.0) 0x06cd(v2.3)
mmDAGB0_CNTL_MISC2: 0x0071(v2.0) 0x0096(v2.3)
...

Continuing using the same file mmhub v2.0 is not good choice, it will
introduce a lot of checking with ASIC types. And also easy to introduce the
issues that offset not align, this kind of issues are really hard to find. Van
Gogh's mmhub vm invalidation is actually caused by the offset mismatch as well.

So it would like to create a new file rather than stick to re-use orignal mmhub
v2.0 here.

v2: add missed translate_further programming.
v3: sync with latest code
v4: add missing callbacks

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdkfd: add Van Gogh KFD support

This patch is to add GFX10 based APU Van Gogh KFD support. We will treat Van
Gogh as "dgpu" (bypass IOMMU v2).

Signed-off-by: Huang Rui <ray.huang@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Yong Zhao <Yong.Zhao@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: set ip blocks for van gogh

Enable ip blocks for van gogh.

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add sdma support for van gogh

This patch adds the sdma v5.2 support for van gogh.

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/gfx10: add updated register offsets for VGH

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

drm/amdgpu: add gfx golden settings for vangogh (v3)

This patch is to add gfx golden settings for vangogh post si.

v2: squash in updates
v3: fix SPI register offset

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add gfx support for van gogh (v3)

Add van gogh checks to gfx10 code.

v2: squash in fixes
v3: fix mode

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: set fw load type for van gogh

This patch sets fw load type as direct for van gogh for the moment.
Will switch to psp when psp is ready.

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add gmc v10 supports for van gogh (v4)

Add gfx memory controller support for van gogh.

v2: don't use dynamic invalidate eng allocation for van gogh.
v3: squash in other fixes
v4: rebase

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: get the correct vram type for van gogh

This patch is to get the correct vram type from atombios for van gogh.

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

drm/amdgpu/atomfirmware: Add edp and integrated info v2.1 tables

Required for vangogh.

Signed-off-by: Roman Li <Roman.Li@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: update new memory types in atomfirmware header

Add new nemory types in atomfirmware header.

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

drm/amdgpu: add uapi to define van gogh memory type

This patch adds van gogh memory type as DDR5.

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

drm/amdgpu: use gpu virtual address for interrupt packet write space for vangogh

The interrupts are not stable while uses guest physical address (GPA)
for interrupt packet write space even on direct loading case.

v2: make condition more readable

Signed-off-by: Huang Rui <ray.huang@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add van gogh support for ih block

This patch adds the support for van gogh ih block.

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: skip sdma1 in nv_allowed_read_registers list for van gogh (v2)

Van gogh only has one sdma.

v2: use num_instances rather than APU flag

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

drm/amdgpu: add nv common ip block support for van gogh

This patch adds common ip support for van gogh.

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add vangogh_reg_base_init function for van gogh

This patch adds vangogh_reg_base_init function to init the register base for
van gogh.

v2: make vangogh_reg_base_init void, align equality sign

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add van gogh support for gpu_info and ip block setting

This patch adds van gogh support for gpu_info firmware and ip block setting.

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add uapi to define van gogh series

Add a flag to define van gogh series.

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>