Re: [PATCH v7 1/6] drm/amdgpu: Fix init ordering in amdgpu_vram_mgr_init()

On 7/21/2026 5:56 PM, Maarten Lankhorst wrote:
> Hey,
>
> On 7/21/26 14:10, Thomas Hellström wrote:
>> Hi, Maarten,
>>
>> On Tue, 2026-07-21 at 13:46 +0200, Maarten Lankhorst wrote:
>>> Hey,
>>>
>>> Can I merge this through drm-misc-next together with the rest of the
>>> series?
>>>
>>> Kind regards
>>> ~Maarten
>> For the xe patches, Ack from me.
>>
>> I do have a v8 that fixes a sashiko comment on the last patch, though,
>> but I was holding that awaiting a review on the other patches.
> Yeah seems about right. Patches themselves look good to me now, so would be nice if amd can ack here or in v8.
I have already sent my Reviewed-by for this patch.
Reviewed-by: Arunpravin Paneer Selvam <[email protected]>

Thanks,
Arun.
>
> Series themselves look good, so feel free to add my r-b to entire series: (and Thadeu's t-b)
>
> Reviewed-By: Maarten Lankhorst <[email protected]>
>
>> Thanks,
>> Thomas
>>
>>
>>
>>> On 7/7/26 20:05, Arunpravin Paneer Selvam wrote:
>>>>
>>>> On 7/3/2026 6:38 PM, Christian König wrote:
>>>>> Arun please take a look at this.
>>>> Sure Christian. This fix looks correct to me.
>>>> Reviewed-by: Arunpravin Paneer Selvam
>>>> <[email protected]>
>>>>
>>>> Thanks,
>>>> Arun.
>>>>> Thanks,
>>>>> Christian.
>>>>>
>>>>> On 7/3/26 15:05, Thomas Hellström wrote:
>>>>>> drmm_cgroup_register_region() is called before INIT_LIST_HEAD()
>>>>>> and
>>>>>> gpu_buddy_init() in amdgpu_vram_mgr_init(). If it fails, the
>>>>>> function
>>>>>> returns early and bypasses those initializations.
>>>>>>
>>>>>> Since adev->mman.initialized is set to true before
>>>>>> amdgpu_vram_mgr_init()
>>>>>> is called, a failure triggers amdgpu_ttm_fini(), which calls
>>>>>> amdgpu_vram_mgr_fini(), which then:
>>>>>>
>>>>>>    - Calls list_for_each_entry_safe() on reservations_pending
>>>>>> and
>>>>>>      reserved_pages, whose list_head::next pointers are zero-
>>>>>> initialized
>>>>>>      (NULL). The loop does not recognize them as empty and
>>>>>> dereferences NULL.
>>>>>>
>>>>>>    - Calls gpu_buddy_fini(), which iterates free_trees[]
>>>>>> unconditionally
>>>>>>      via for_each_free_tree(). Since mm->free_trees is NULL
>>>>>>      (never allocated), this dereferences NULL.
>>>>>>
>>>>>> Both result in a kernel panic on the module load error path.
>>>>>>
>>>>>> Fix by moving drmm_cgroup_register_region() to after the list
>>>>>> and buddy
>>>>>> allocator are fully initialized, so the teardown path is safe
>>>>>> to run.
>>>>>>
>>>>>> Reported-by: Sashiko-bot <[email protected]>
>>>>>> Closes:
>>>>>> https://sashiko.dev/#/patchset/[email protected]?part=4
>>>>>> Fixes: 2b624a2c1865 ("drm/ttm: Handle cgroup based eviction in
>>>>>> TTM")
>>>>>> Cc: Friedrich Vock <[email protected]>
>>>>>> Cc: Maarten Lankhorst <[email protected]>
>>>>>> Cc: Tejun Heo <[email protected]>
>>>>>> Cc: Maxime Ripard <[email protected]>
>>>>>> Cc: Christian König <[email protected]>
>>>>>> Cc: Alex Deucher <[email protected]>
>>>>>> Cc: [email protected]
>>>>>> Cc: [email protected]
>>>>>> Cc: <[email protected]> # v6.14+
>>>>>> Assisted-by: GitHub_Copilot:claude-sonnet-4.6
>>>>>> Signed-off-by: Thomas Hellström
>>>>>> <[email protected]>
>>>>>> ---
>>>>>>    drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c | 7 ++++---
>>>>>>    1 file changed, 4 insertions(+), 3 deletions(-)
>>>>>>
>>>>>> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
>>>>>> b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
>>>>>> index 2a241a5b12c4..ac3f71d77140 100644
>>>>>> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
>>>>>> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
>>>>>> @@ -918,9 +918,6 @@ int amdgpu_vram_mgr_init(struct
>>>>>> amdgpu_device *adev)
>>>>>>        struct ttm_resource_manager *man = &mgr->manager;
>>>>>>        int err;
>>>>>>    -    man->cg = drmm_cgroup_register_region(adev_to_drm(adev),
>>>>>> "vram", adev->gmc.real_vram_size);
>>>>>> -    if (IS_ERR(man->cg))
>>>>>> -        return PTR_ERR(man->cg);
>>>>>>        ttm_resource_manager_init(man, &adev->mman.bdev,
>>>>>>                      adev->gmc.real_vram_size);
>>>>>>    @@ -935,6 +932,10 @@ int amdgpu_vram_mgr_init(struct
>>>>>> amdgpu_device *adev)
>>>>>>        if (err)
>>>>>>            return err;
>>>>>>    +    man->cg = drmm_cgroup_register_region(adev_to_drm(adev),
>>>>>> "vram", adev->gmc.real_vram_size);
>>>>>> +    if (IS_ERR(man->cg))
>>>>>> +        return PTR_ERR(man->cg);
>>>>>> +
>>>>>>        ttm_set_driver_manager(&adev->mman.bdev, TTM_PL_VRAM,
>>>>>> &mgr->manager);
>>>>>>        ttm_resource_manager_set_used(man, true);
>>>>>>        return 0;