drm/amdgpu: once more fix the call oder in amdgpu_ttm_move() v2

This reverts drm/amdgpu: fix ftrace event amdgpu_bo_move always move
on same heap. The basic problem here is that after the move the o

drm/amdgpu: once more fix the call oder in amdgpu_ttm_move() v2

This reverts drm/amdgpu: fix ftrace event amdgpu_bo_move always move
on same heap. The basic problem here is that after the move the old
location is simply not available any more.

Some fixes were suggested, but essentially we should call the move
notification before actually moving things because only this way we have
the correct order for DMA-buf and VM move notifications as well.

Also rework the statistic handling so that we don't update the eviction
counter before the move.

v2: add missing NULL check

Signed-off-by: Christian König <christian.koenig@amd.com>
Fixes: 94aeb4117343 ("drm/amdgpu: fix ftrace event amdgpu_bo_move always move on same heap")
Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/3171
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
CC: stable@vger.kernel.org

show more ...

drm/amdgpu: Modify the contiguous flags behaviour

Now we have two flags for contiguous VRAM buffer allocation.
If the application request for AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS,
it would set the ttm

drm/amdgpu: Modify the contiguous flags behaviour

Now we have two flags for contiguous VRAM buffer allocation.
If the application request for AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS,
it would set the ttm place TTM_PL_FLAG_CONTIGUOUS flag in the
buffer's placement function.

This patch will change the default behaviour of the two flags.

When we set AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS
- This means contiguous is not mandatory.
- we will try to allocate the contiguous buffer. Say if the
allocation fails, we fallback to allocate the individual pages.

When we setTTM_PL_FLAG_CONTIGUOUS
- This means contiguous allocation is mandatory.
- we are setting this in amdgpu_bo_pin_restricted() before bo validation
and check this flag in the vram manager file.
- if this is set, we should allocate the buffer pages contiguously.
the allocation fails, we return -ENOSPC.

v2:
- keep the mem_flags and bo->flags check as is(Christian)
- place the TTM_PL_FLAG_CONTIGUOUS flag setting into the
amdgpu_bo_pin_restricted function placement range iteration
loop(Christian)
- rename find_pages with amdgpu_vram_mgr_calculate_pages_per_block
(Christian)
- Keep the kernel BO allocation as is(Christain)
- If BO pin vram allocation failed, we need to return -ENOSPC as
RDMA cannot work with scattered VRAM pages(Philip)

v3(Christian):
- keep contiguous flag handling outside of pages_per_block
calculation
- remove the hacky implementation in contiguous flag error
handling code

v4(Christian):
- use any variable and return value for non-contiguous
fallback

v5: rebase to amd-staging-drm-next branch

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

show more ...

drm/amdgpu: replace tmz flag into buffer flag

Replace tmz flag into buffer flag to make it easier to understand
and extend

Signed-off-by: Likun Gao <Likun.Gao@amd.com>
Signed-off-by: Frank Min <Fra

drm/amdgpu: replace tmz flag into buffer flag

Replace tmz flag into buffer flag to make it easier to understand
and extend

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

show more ...

drm/amdgpu: Update BO eviction priorities

Make SVM BOs more likely to get evicted than other BOs. These BOs
opportunistically use available VRAM, but can fall back relatively
seamlessly to system me

drm/amdgpu: Update BO eviction priorities

Make SVM BOs more likely to get evicted than other BOs. These BOs
opportunistically use available VRAM, but can fall back relatively
seamlessly to system memory. It also avoids SVM migrations evicting
other, more important BOs as they will evict other SVM allocations
first.

Signed-off-by: Felix Kuehling <felix.kuehling@amd.com>
Acked-by: Mukul Joshi <mukul.joshi@amd.com>
Tested-by: Mukul Joshi <mukul.joshi@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

show more ...

drm/amdgpu: Update BO eviction priorities

Make SVM BOs more likely to get evicted than other BOs. These BOs
opportunistically use available VRAM, but can fall back relatively
seamlessly to system me

drm/amdgpu: Update BO eviction priorities

Make SVM BOs more likely to get evicted than other BOs. These BOs
opportunistically use available VRAM, but can fall back relatively
seamlessly to system memory. It also avoids SVM migrations evicting
other, more important BOs as they will evict other SVM allocations
first.

Signed-off-by: Felix Kuehling <felix.kuehling@amd.com>
Acked-by: Mukul Joshi <mukul.joshi@amd.com>
Tested-by: Mukul Joshi <mukul.joshi@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

show more ...

drm/amdgpu: Enable clear page functionality

Add clear page support in vram memory region.

v1(Christian):
- Dont handle clear page as TTM flag since when moving the BO back
in from GTT again w

drm/amdgpu: Enable clear page functionality

Add clear page support in vram memory region.

v1(Christian):
- Dont handle clear page as TTM flag since when moving the BO back
in from GTT again we don't need that.
- Make a specialized version of amdgpu_fill_buffer() which only
clears the VRAM areas which are not already cleared
- Drop the TTM_PL_FLAG_WIPE_ON_RELEASE check in
amdgpu_object.c

v2:
- Modify the function name amdgpu_ttm_* (Alex)
- Drop the delayed parameter (Christian)
- handle amdgpu_res_cleared(&cursor) just above the size
calculation (Christian)
- Use AMDGPU_GEM_CREATE_VRAM_WIPE_ON_RELEASE for clearing the buffers
in the free path to properly wait for fences etc.. (Christian)

v3(Christian):
- Remove buffer clear code in VRAM manager instead change the
AMDGPU_GEM_CREATE_VRAM_WIPE_ON_RELEASE handling to set
the DRM_BUDDY_CLEARED flag.
- Remove ! from amdgpu_res_cleared(&cursor) check.

v4(Christian):
- vres flag setting move to vram manager file
- use dma_fence_get_stub in amdgpu_ttm_clear_buffer function
- make fence a mandatory parameter and drop the if and the get/put dance

Signed-off-by: Arunpravin Paneer Selvam <Arunpravin.PaneerSelvam@amd.com>
Suggested-by: Christian König <christian.koenig@amd.com>
Acked-by: Felix Kuehling <felix.kuehling@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240419063538.11957-2-Arunpravin.PaneerSelvam@amd.com
Signed-off-by: Christian König <christian.koenig@amd.com>

show more ...

drm/amdgpu: fix visible VRAM handling during faults

When we removed the hacky start code check we actually didn't took into
account that *all* VRAM pages needs to be CPU accessible.

Clean up the co

drm/amdgpu: fix visible VRAM handling during faults

When we removed the hacky start code check we actually didn't took into
account that *all* VRAM pages needs to be CPU accessible.

Clean up the code and unify the handling into a single helper which
checks if the whole resource is CPU accessible.

The only place where a partial check would make sense is during
eviction, but that is neglitible.

Signed-off-by: Christian König <christian.koenig@amd.com>
Fixes: aed01a68047b ("drm/amdgpu: Remove TTM resource->start visible VRAM condition v2")
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
CC: stable@vger.kernel.org

show more ...

drm/amdgpu: fix visible VRAM handling during faults

When we removed the hacky start code check we actually didn't took into
account that *all* VRAM pages needs to be CPU accessible.

Clean up the co

drm/amdgpu: fix visible VRAM handling during faults

When we removed the hacky start code check we actually didn't took into
account that *all* VRAM pages needs to be CPU accessible.

Clean up the code and unify the handling into a single helper which
checks if the whole resource is CPU accessible.

The only place where a partial check would make sense is during
eviction, but that is neglitible.

Signed-off-by: Christian König <christian.koenig@amd.com>
Fixes: aed01a68047b ("drm/amdgpu: Remove TTM resource->start visible VRAM condition v2")
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
CC: stable@vger.kernel.org

show more ...

drm/amdgpu: use GTT only as fallback for VRAM|GTT

Try to fill up VRAM as well by setting the busy flag on GTT allocations.

This fixes the issue that when VRAM was evacuated for suspend it's never
f

drm/amdgpu: use GTT only as fallback for VRAM|GTT

Try to fill up VRAM as well by setting the busy flag on GTT allocations.

This fixes the issue that when VRAM was evacuated for suspend it's never
filled up again unless the application is restarted.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Zack Rusin <zack.rusin@broadcom.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240229134003.3688-2-christian.koenig@amd.com

show more ...

drm/amdgpu: add shared fdinfo stats

Add shared stats. Useful for seeing shared memory.

v2: take dma-buf into account as well
v3: use the new gem helper

Link: https://lore.kernel.org/all/202312071

drm/amdgpu: add shared fdinfo stats

Add shared stats. Useful for seeing shared memory.

v2: take dma-buf into account as well
v3: use the new gem helper

Link: https://lore.kernel.org/all/20231207180225.439482-1-alexander.deucher@amd.com/
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: Rob Clark <robdclark@gmail.com>
Reviewed-by: Christian König <christian.keonig@amd.com>
Signed-off-by: Christian König <christian.koenig@amd.com>

show more ...

drm/ttm: replace busy placement with flags v6

Instead of a list of separate busy placement add flags which indicate
that a placement should only be used when there is room or if we need to
evict.

v

drm/ttm: replace busy placement with flags v6

Instead of a list of separate busy placement add flags which indicate
that a placement should only be used when there is room or if we need to
evict.

v2: add missing TTM_PL_FLAG_IDLE for i915
v3: fix auto build test ERROR on drm-tip/drm-tip
v4: fix some typos pointed out by checkpatch
v5: cleanup some rebase problems with VMWGFX
v6: implement some missing VMWGFX functionality pointed out by Zack,
rename the flags as suggested by Michel, rebase on drm-tip and
adjust XE as well

Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Somalapuram Amaranath <Amaranath.Somalapuram@amd.com>
Reviewed-by: Zack Rusin <zack.rusin@broadcom.com>
Reviewed-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240112125158.2748-4-christian.koenig@amd.com

show more ...

drm/amdgpu: fix ftrace event amdgpu_bo_move always move on same heap

Issue: during evict or validate happened on amdgpu_bo, the 'from' and
'to' is always same in ftrace event of amdgpu_bo_move

wher

drm/amdgpu: fix ftrace event amdgpu_bo_move always move on same heap

Issue: during evict or validate happened on amdgpu_bo, the 'from' and
'to' is always same in ftrace event of amdgpu_bo_move

where calling the 'trace_amdgpu_bo_move', the comment says move_notify
is called before move happens, but actually it is called after move
happens, here the new_mem is same as bo->resource

Fix: move trace_amdgpu_bo_move from move_notify to amdgpu_bo_move

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

show more ...

drm/amdgpu: warn when there are still mappings when a BO is destroyed v2

This can only happen when there is a reference counting bug.

v2: fix typo

Signed-off-by: Christian König <christian.koenig@

drm/amdgpu: warn when there are still mappings when a BO is destroyed v2

This can only happen when there is a reference counting bug.

v2: fix typo

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

show more ...

drm/amdgpu: fix AGP addressing when GART is not at 0

This worked by luck if the GART aperture ended up at 0. When
we ended up moving GART on some chips, the GART aperture ended
up offsetting the AG

drm/amdgpu: fix AGP addressing when GART is not at 0

This worked by luck if the GART aperture ended up at 0. When
we ended up moving GART on some chips, the GART aperture ended
up offsetting the AGP address since the resource->start is
a GART offset, not an MC address. Fix this by moving the AGP
address setup into amdgpu_bo_gpu_offset_no_check().

v2: check mem_type before checking agp
v3: check if the ttm bo has a ttm_tt allocated yet

Fixes: 67318cb84341 ("drm/amdgpu/gmc11: set gart placement GC11")
Tested-by: Mario Limonciello <mario.limonciello@amd.com>
Reported-by: Jesse Zhang <Jesse.Zhang@amd.com>
Reported-by: Yifan Zhang <yifan1.zhang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: christian.koenig@amd.com
Cc: mario.limonciello@amd.com

show more ...

drm/amdgpu: fix AGP init order

The default AGP settings were overwriting the IP selected
ones since the default was getting set after the IP ones
were selected.

Fixes: de59b69932e6 ("drm/amdgpu/gmc

drm/amdgpu: fix AGP init order

The default AGP settings were overwriting the IP selected
ones since the default was getting set after the IP ones
were selected.

Fixes: de59b69932e6 ("drm/amdgpu/gmc: set a default disable value for AGP")
Link: https://lists.freedesktop.org/archives/amd-gfx/2023-November/100966.html
Tested-by: Mikhail Gavrilov <mikhail.v.gavrilov@gmail.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: Mikhail Gavrilov <mikhail.v.gavrilov@gmail.com>

show more ...

drm/amdgpu/gmc: set a default disable value for AGP

To disable AGP, the start needs to be set to a higher
value than the end. Set a default disable value for
the AGP aperture and allow the IP speci

drm/amdgpu/gmc: set a default disable value for AGP

To disable AGP, the start needs to be set to a higher
value than the end. Set a default disable value for
the AGP aperture and allow the IP specific GMC code
to enable it selectively be calling amdgpu_gmc_agp_location().

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

show more ...

drm/amd: Simplify the bo size check funciton

Simplify the code logic of size check function amdgpu_bo_validate_size

Signed-off-by: Ma Jun <Jun.Ma2@amd.com>
Reviewed-by: Christian König <christian.k

drm/amd: Simplify the bo size check funciton

Simplify the code logic of size check function amdgpu_bo_validate_size

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

show more ...

drm/amdgpu: accommodate DOMAIN/PL_DOORBELL

This patch adds changes:
- to accommodate the new GEM domain for DOORBELLs
- to accommodate the new TTM PL for DOORBELLs

in order to manage doorbell pages

drm/amdgpu: accommodate DOMAIN/PL_DOORBELL

This patch adds changes:
- to accommodate the new GEM domain for DOORBELLs
- to accommodate the new TTM PL for DOORBELLs

in order to manage doorbell pages as GEM object.

V2: Addressed reviwe comments from Christian
- drop the doorbell changes for pinning/unpinning
- drop the doorbell changes for dma-buf map
- drop the doorbell changes for sgt
- no need to handle TTM_PL_FLAG_CONTIGUOUS for doorbell
- add caching type for doorbell

V3: - Removed unrelated empty line (Christian)
- Add PL_DOORBELL in mem_type_to_domain() as well (Alex)

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

show more ...

drm/amdgpu: add VISIBLE info in amdgpu_bo_print_info

This allows tools to distinguish between VRAM and visible VRAM.

Use the opportunity to fix locking before accessing bo.

v2: squash in unused va

drm/amdgpu: add VISIBLE info in amdgpu_bo_print_info

This allows tools to distinguish between VRAM and visible VRAM.

Use the opportunity to fix locking before accessing bo.

v2: squash in unused variable fix

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

show more ...

drm/amdgpu: make sure that BOs have a backing store

It's perfectly possible that the BO is about to be destroyed and doesn't
have a backing store associated with it.

Signed-off-by: Christian König

drm/amdgpu: make sure that BOs have a backing store

It's perfectly possible that the BO is about to be destroyed and doesn't
have a backing store associated with it.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Guchun Chen <guchun.chen@amd.com>
Tested-by: Mikhail Gavrilov <mikhail.v.gavrilov@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

show more ...

Revert "drm/amdgpu: remove TOPDOWN flags when allocating VRAM in large bar system"

This reverts commit c105518679b6e87232874ffc989ec403bee59664.

This patch disables the TOPDOWN flag for APU and few

Revert "drm/amdgpu: remove TOPDOWN flags when allocating VRAM in large bar system"

This reverts commit c105518679b6e87232874ffc989ec403bee59664.

This patch disables the TOPDOWN flag for APU and few dGPU cards
which has the VRAM size equal to the BAR size.

When we enable the TOPDOWN flag, we get the free blocks at
the highest available memory region and we don't split the
lower order blocks. This change is required to keep off
the fragmentation related issues particularly in ASIC
which has VRAM space <= 500MiB

Hence, we are reverting this patch.

Link: https://gitlab.freedesktop.org/drm/amd/-/issues/2270
Signed-off-by: Arunpravin Paneer Selvam <Arunpravin.PaneerSelvam@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

show more ...

drm/amdgpu: fix Null pointer dereference error in amdgpu_device_recover_vram

Use the function of amdgpu_bo_vm_destroy to handle the resource release
of shadow bo. During the amdgpu_mes_self_test, sh

drm/amdgpu: fix Null pointer dereference error in amdgpu_device_recover_vram

Use the function of amdgpu_bo_vm_destroy to handle the resource release
of shadow bo. During the amdgpu_mes_self_test, shadow bo released, but
vmbo->shadow_list was not, which caused a null pointer reference error
in amdgpu_device_recover_vram when GPU reset.

Fixes: 6c032c37ac3e ("drm/amdgpu: Fix vram recover doesn't work after whole GPU reset (v2)")
Signed-off-by: xinhui pan <xinhui.pan@amd.com>
Signed-off-by: Horatio Zhang <Hongkun.Zhang@amd.com>
Acked-by: Feifei Xu <Feifei.Xu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

show more ...

drm/amdgpu: Add a low priority scheduler for VRAM clearing

Add a low priority DRM scheduler for VRAM clearing instead of using
the exisiting high priority scheduler. Use the high priority scheduler

drm/amdgpu: Add a low priority scheduler for VRAM clearing

Add a low priority DRM scheduler for VRAM clearing instead of using
the exisiting high priority scheduler. Use the high priority scheduler
for migrations and evictions.

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

show more ...

drm/amdkfd: Store xcp partition id to amdgpu bo

For memory accounting per compute partition and export drm amdgpu bo and
then import to KFD, we need the xcp id to account the memory usage or
find th

drm/amdkfd: Store xcp partition id to amdgpu bo

For memory accounting per compute partition and export drm amdgpu bo and
then import to KFD, we need the xcp id to account the memory usage or
find the KFD node of the original amdgpu bo to create the KFD bo on the
correct adev KFD node.

Set xcp_id_plus1 of amdgpu_bo_param to create bo and store xcp_id to
amddgpu bo. Add helper macro to get the mem_id from adev and xcp_id.

v2: squash in fix ("drm/amdgpu: Fix BO creation failure on GFX 9.4.3 dGPU")

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

show more ...

drm/amdgpu: dGPU mode set VRAM range lpfn as exclusive

TTM place lpfn is exclusive used as end (start + size) in drm and buddy
allocator, adev->gmc memory partition range lpfn is inclusive (start +

drm/amdgpu: dGPU mode set VRAM range lpfn as exclusive

TTM place lpfn is exclusive used as end (start + size) in drm and buddy
allocator, adev->gmc memory partition range lpfn is inclusive (start +
size - 1), should plus 1 to set TTM place lpfn.

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

show more ...