[PATCH] drm/amdgpu: move task_info to amdgpu_fpriv

Shashank Sharma shashank.sharma at amd.com
Tue Oct 17 07:25:26 UTC 2023


Hello Christian, Felix,

Thanks for your comments, mine inline.

On 17/10/2023 07:55, Christian König wrote:
> Am 17.10.23 um 00:15 schrieb Felix Kuehling:
>> On 2023-10-16 13:08, Shashank Sharma wrote:
>>> This patch does the following:
>>> - moves vm->task_info struct to fpriv->task_info.
>>> - makes task_info allocation dynamic.
>>> - adds reference counting support for task_info structure.
>>> - adds some new helper functions to find and put task_info.
>>> - adds respective supporting changes for existing get_task_info 
>>> consumers.
>>>
>>> Cc: Christian Koenig <christian.koenig at amd.com>
>>> Cc: Alex Deucher <alexander.deucher at amd.com>
>>> Signed-off-by: Shashank Sharma <shashank.sharma at amd.com>
>>> ---
>>>   drivers/gpu/drm/amd/amdgpu/amdgpu.h         |  2 +
>>>   drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c |  4 +-
>>>   drivers/gpu/drm/amd/amdgpu/amdgpu_device.c  | 28 +++++--
>>>   drivers/gpu/drm/amd/amdgpu/amdgpu_job.c     | 16 ++--
>>>   drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c      | 87 
>>> ++++++++++++++++-----
>>>   drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h      |  9 +--
>>>   drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c   |  5 +-
>>>   drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c      | 19 +++--
>>>   drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c      | 18 +++--
>>>   drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c       | 17 ++--
>>>   drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c       | 19 +++--
>>>   drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c      | 19 ++---
>>>   drivers/gpu/drm/amd/amdgpu/sdma_v4_4_2.c    | 20 ++---
>>>   drivers/gpu/drm/amd/amdkfd/kfd_smi_events.c | 14 ++--
>>>   14 files changed, 186 insertions(+), 91 deletions(-)
>>>
>>> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h 
>>> b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
>>> index dc2d53081e80..a90780d38725 100644
>>> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h
>>> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
>>> @@ -469,6 +469,8 @@ struct amdgpu_fpriv {
>>>       struct mutex        bo_list_lock;
>>>       struct idr        bo_list_handles;
>>>       struct amdgpu_ctx_mgr    ctx_mgr;
>>> +    struct amdgpu_task_info *task_info;
>>> +
>>>       /** GPU partition selection */
>>>       uint32_t        xcp_id;
>>>   };
>>> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c 
>>> b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
>>> index a4faea4fa0b5..6e9dcd13ee34 100644
>>> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
>>> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
>>> @@ -1763,9 +1763,11 @@ static int amdgpu_debugfs_vm_info_show(struct 
>>> seq_file *m, void *unused)
>>>       list_for_each_entry(file, &dev->filelist, lhead) {
>>>           struct amdgpu_fpriv *fpriv = file->driver_priv;
>>>           struct amdgpu_vm *vm = &fpriv->vm;
>>> +        struct amdgpu_task_info *task_info = fpriv->task_info;
>>>             seq_printf(m, "pid:%d\tProcess:%s ----------\n",
>>> -                vm->task_info.pid, vm->task_info.process_name);
>>> +                task_info ? task_info->pid : 0,
>>> +                task_info ? task_info->process_name : "");
>>>           r = amdgpu_bo_reserve(vm->root.bo, true);
>>>           if (r)
>>>               break;
>>> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c 
>>> b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
>>> index 3f001a50b34a..b372a87b9b77 100644
>>> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
>>> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
>>> @@ -4872,6 +4872,27 @@ static void 
>>> amdgpu_reset_capture_coredumpm(struct amdgpu_device *adev)
>>>       dev_coredumpm(dev->dev, THIS_MODULE, adev, 0, GFP_NOWAIT,
>>>                 amdgpu_devcoredump_read, amdgpu_devcoredump_free);
>>>   }
>>> +
>>> +static void
>>> +amdgpu_reset_dev_coredump(struct amdgpu_device *adev, struct 
>>> amdgpu_reset_context *reset_context)
>>> +{
>>> +    struct amdgpu_task_info *ti;
>>> +    struct amdgpu_vm *vm;
>>> +
>>> +    if (!reset_context->job || !reset_context->job->vm)
>>> +        return;
>>> +
>>> +    vm = reset_context->job->vm;
>>> +
>>> +    /* Get reset task info and save a copy of data to be consumed 
>>> later */
>>> +    ti = amdgpu_vm_get_task_info(adev, vm->pasid);
>>> +    if (ti) {
>>> +        adev->reset_task_info = *ti;
>>> +        amdgpu_reset_capture_coredumpm(adev);
>>> +    }
>>> +
>>> +    amdgpu_vm_put_task_info(adev, vm->pasid);
>>> +}
>>>   #endif
>>>     int amdgpu_do_asic_reset(struct list_head *device_list_handle,
>>> @@ -4976,12 +4997,7 @@ int amdgpu_do_asic_reset(struct list_head 
>>> *device_list_handle,
>>>                   vram_lost = amdgpu_device_check_vram_lost(tmp_adev);
>>>   #ifdef CONFIG_DEV_COREDUMP
>>>                   tmp_adev->reset_vram_lost = vram_lost;
>>> -                memset(&tmp_adev->reset_task_info, 0,
>>> - sizeof(tmp_adev->reset_task_info));
>>> -                if (reset_context->job && reset_context->job->vm)
>>> -                    tmp_adev->reset_task_info =
>>> - reset_context->job->vm->task_info;
>>> -                amdgpu_reset_capture_coredumpm(tmp_adev);
>>> +                amdgpu_reset_dev_coredump(tmp_adev, reset_context);
>>>   #endif
>>>                   if (vram_lost) {
>>>                       DRM_INFO("VRAM is lost due to GPU reset!\n");
>>> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c 
>>> b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c
>>> index 78476bc75b4e..99cf30c0bce6 100644
>>> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c
>>> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c
>>> @@ -35,7 +35,7 @@ static enum drm_gpu_sched_stat 
>>> amdgpu_job_timedout(struct drm_sched_job *s_job)
>>>   {
>>>       struct amdgpu_ring *ring = to_amdgpu_ring(s_job->sched);
>>>       struct amdgpu_job *job = to_amdgpu_job(s_job);
>>> -    struct amdgpu_task_info ti;
>>> +    struct amdgpu_task_info *ti;
>>>       struct amdgpu_device *adev = ring->adev;
>>>       int idx;
>>>       int r;
>>> @@ -48,7 +48,6 @@ static enum drm_gpu_sched_stat 
>>> amdgpu_job_timedout(struct drm_sched_job *s_job)
>>>           return DRM_GPU_SCHED_STAT_ENODEV;
>>>       }
>>>   -    memset(&ti, 0, sizeof(struct amdgpu_task_info));
>>>       adev->job_hang = true;
>>>         if (amdgpu_gpu_recovery &&
>>> @@ -58,12 +57,15 @@ static enum drm_gpu_sched_stat 
>>> amdgpu_job_timedout(struct drm_sched_job *s_job)
>>>           goto exit;
>>>       }
>>>   -    amdgpu_vm_get_task_info(ring->adev, job->pasid, &ti);
>>>       DRM_ERROR("ring %s timeout, signaled seq=%u, emitted seq=%u\n",
>>> -          job->base.sched->name, 
>>> atomic_read(&ring->fence_drv.last_seq),
>>> -          ring->fence_drv.sync_seq);
>>> -    DRM_ERROR("Process information: process %s pid %d thread %s pid 
>>> %d\n",
>>> -          ti.process_name, ti.tgid, ti.task_name, ti.pid);
>>> +           job->base.sched->name, 
>>> atomic_read(&ring->fence_drv.last_seq),
>>> +           ring->fence_drv.sync_seq);
>>> +
>>> +    ti = amdgpu_vm_get_task_info(ring->adev, job->pasid);
>>> +    if (ti)
>>> +        DRM_ERROR("Process information: process %s pid %d thread %s 
>>> pid %d\n",
>>> +               ti->process_name, ti->tgid, ti->task_name, ti->pid);
>>> +    amdgpu_vm_put_task_info(ring->adev, job->pasid);
>>> dma_fence_set_error(&s_job->s_fence->finished, -ETIME);
>>>   diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c 
>>> b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
>>> index f5daadcec865..aabfbb8edb2b 100644
>>> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
>>> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
>>> @@ -2335,6 +2335,7 @@ void amdgpu_vm_fini(struct amdgpu_device 
>>> *adev, struct amdgpu_vm *vm)
>>>         root = amdgpu_bo_ref(vm->root.bo);
>>>       amdgpu_bo_reserve(root, true);
>>> +    amdgpu_vm_put_task_info(adev, vm->pasid);
>>>       amdgpu_vm_set_pasid(adev, vm, 0);
>>>       dma_fence_wait(vm->last_unlocked, false);
>>>       dma_fence_put(vm->last_unlocked);
>>> @@ -2491,26 +2492,44 @@ int amdgpu_vm_ioctl(struct drm_device *dev, 
>>> void *data, struct drm_file *filp)
>>>       return 0;
>>>   }
>>>   +static struct
>>> +amdgpu_task_info *amdgpu_vm_find_task_info(struct amdgpu_device 
>>> *adev, u32 pasid)
>>> +{
>>> +    unsigned long flags;
>>> +    struct amdgpu_vm *vm;
>>> +    struct amdgpu_fpriv *fpriv;
>>> +    struct amdgpu_task_info *task_info = NULL;
>>> +
>>> +    xa_lock_irqsave(&adev->vm_manager.pasids, flags);
>>> +
>>> +    vm = xa_load(&adev->vm_manager.pasids, pasid);
>>> +    if (vm) {
>>> +        fpriv = container_of(vm, struct amdgpu_fpriv, vm);
>>> +        task_info = fpriv->task_info;
>>> +    }
>>> +
>>> +    xa_unlock_irqrestore(&adev->vm_manager.pasids, flags);
>>> +    return task_info;
>>> +}
>>> +
>>>   /**
>>>    * amdgpu_vm_get_task_info - Extracts task info for a PASID.
>>>    *
>>>    * @adev: drm device pointer
>>>    * @pasid: PASID identifier for VM
>>> - * @task_info: task_info to fill.
>>> + *
>>> + * returns the task_info* (refrence counted) set under the vm_pasid
>>> + * user must call amdgpu_vm_put_task_info when done with the 
>>> task_info ptr
>>>    */
>>> -void amdgpu_vm_get_task_info(struct amdgpu_device *adev, u32 pasid,
>>> -             struct amdgpu_task_info *task_info)
>>> +struct amdgpu_task_info *amdgpu_vm_get_task_info(struct 
>>> amdgpu_device *adev, u32 pasid)
>>>   {
>>> -    struct amdgpu_vm *vm;
>>> -    unsigned long flags;
>>> -
>>> -    xa_lock_irqsave(&adev->vm_manager.pasids, flags);
>>> +    struct amdgpu_task_info *ti;
>>>   -    vm = xa_load(&adev->vm_manager.pasids, pasid);
>>> -    if (vm)
>>> -        *task_info = vm->task_info;
>>> +    ti = amdgpu_vm_find_task_info(adev, pasid);
>>> +    if (ti)
>>> +        kref_get(&ti->refcount);
>>>   -    xa_unlock_irqrestore(&adev->vm_manager.pasids, flags);
>>> +    return ti;
>>>   }
>>>     /**
>>> @@ -2520,17 +2539,49 @@ void amdgpu_vm_get_task_info(struct 
>>> amdgpu_device *adev, u32 pasid,
>>>    */
>>>   void amdgpu_vm_set_task_info(struct amdgpu_vm *vm)
>>>   {
>>> -    if (vm->task_info.pid)
>>> -        return;
>>> +    struct amdgpu_fpriv *fpriv = container_of(vm, struct 
>>> amdgpu_fpriv, vm);
>
> Big NAK to that, no idea why everybody wants to upcast the VM 
> structure to fpriv.
>
> Instead move this into the the fpriv handling in amdgpu_kms.c
If you see the existing consumers of amdgpu_vm_get_task_info(), in most 
of the cases they either have only the vm or the pasid available (Please 
see gmc_v*_0_process_interrupts, sdma_v4_*_print_iv_entry or 
kfd_smi_event_update_vmfault). Now if we want to move the task_info from 
vm to fpriv, the only way to do it without breaking these existing 
consumers, is to get fpriv from VM. If we move this handling into 
amdgpu_kms.c , then how to handle these consumers who do not have fpriv 
available with them ?
>
>>> +    struct amdgpu_task_info *task_info;
>>>   -    vm->task_info.pid = current->pid;
>>> -    get_task_comm(vm->task_info.task_name, current);
>>> +    if (fpriv->task_info)
>>> +        return;
>>>   -    if (current->group_leader->mm != current->mm)
>>> +    task_info = kzalloc(sizeof(*task_info), GFP_KERNEL);
>>> +    if (!task_info) {
>>> +        DRM_ERROR("OOM while task_info creation\n");
>>>           return;
>>> +    }
>>> +
>>> +    kref_init(&task_info->refcount);
>>> +    task_info->pid = current->pid;
>>> +    get_task_comm(task_info->task_name, current);
>>> +
>>> +    if (current->group_leader->mm != current->mm) {
>>> +        task_info->tgid = current->group_leader->pid;
>>> +        get_task_comm(task_info->process_name, current->group_leader);
>>> +    }
>>> +
>>> +    kref_get(&task_info->refcount);
>>> +    fpriv->task_info = task_info;
>>> +}
>>> +
>>> +static void amdgpu_vm_free_task_info(struct kref *kref)
>>> +{
>>> +    kfree(container_of(kref, struct amdgpu_task_info, refcount));
>>> +}
>>> +
>>> +/**
>>> + * amdgpu_vm_put_task_info - reference down the task_info ptr.
>>> + *
>>> + * @adev: drm device pointer
>>> + * @pasid: PASID identifier for VM
>>> + */
>>> +void amdgpu_vm_put_task_info(struct amdgpu_device *adev, u32 pasid)
>>> +{
>>> +    struct amdgpu_task_info *ti;
>>>   -    vm->task_info.tgid = current->group_leader->pid;
>>> -    get_task_comm(vm->task_info.process_name, current->group_leader);
>>> +    ti = amdgpu_vm_find_task_info(adev, pasid);
>> This doesn't make sense. Anyone trying to drop a reference should 
>> already have a ti * from a previous amdgpu_vm_get_task_info call. You 
>> don't need to find it again. Just pass the ti * from the caller to 
>> this function.
Agree, I will modify it.
>
> Yeah, agree completely. Additional to that one goal here was to 
> decouple the VM from the task_info.
>
> Probably best if you put the task_info structure into amdgpu_reset.h 
> and the related handling into amdgpu_kms.c or amdgpu_reset.c instead.
>
Same question as above.

- Shashank

> Regards,
> Christian.
>
>>
>> Regards,
>>   Felix
>>
>>
>>> +    if (ti)
>>> +        kref_put(&ti->refcount, amdgpu_vm_free_task_info);
>>>   }
>>>     /**
>>> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h 
>>> b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
>>> index 204ab13184ed..40a8c532a5ed 100644
>>> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
>>> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
>>> @@ -187,6 +187,7 @@ struct amdgpu_task_info {
>>>       char    task_name[TASK_COMM_LEN];
>>>       pid_t    pid;
>>>       pid_t    tgid;
>>> +    struct kref refcount;
>>>   };
>>>     /**
>>> @@ -333,9 +334,6 @@ struct amdgpu_vm {
>>>       /* Valid while the PD is reserved or fenced */
>>>       uint64_t        pd_phys_addr;
>>>   -    /* Some basic info about the task */
>>> -    struct amdgpu_task_info task_info;
>>> -
>>>       /* Store positions of group of BOs */
>>>       struct ttm_lru_bulk_move lru_bulk_move;
>>>       /* Flag to indicate if VM is used for compute */
>>> @@ -466,8 +464,9 @@ bool amdgpu_vm_need_pipeline_sync(struct 
>>> amdgpu_ring *ring,
>>>                     struct amdgpu_job *job);
>>>   void amdgpu_vm_check_compute_bug(struct amdgpu_device *adev);
>>>   -void amdgpu_vm_get_task_info(struct amdgpu_device *adev, u32 pasid,
>>> -                 struct amdgpu_task_info *task_info);
>>> +struct amdgpu_task_info *
>>> +amdgpu_vm_get_task_info(struct amdgpu_device *adev, u32 pasid);
>>> +void amdgpu_vm_put_task_info(struct amdgpu_device *adev, u32 pasid);
>>>   bool amdgpu_vm_handle_fault(struct amdgpu_device *adev, u32 pasid,
>>>                   u32 vmid, u32 node_id, uint64_t addr,
>>>                   bool write_fault);
>>> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c 
>>> b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c
>>> index 96d601e209b8..3d7a9ad963a8 100644
>>> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c
>>> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c
>>> @@ -1023,21 +1023,24 @@ int amdgpu_vm_ptes_update(struct 
>>> amdgpu_vm_update_params *params,
>>>               uint64_t upd_end = min(entry_end, frag_end);
>>>               unsigned int nptes = (upd_end - frag_start) >> shift;
>>>               uint64_t upd_flags = flags | AMDGPU_PTE_FRAG(frag);
>>> +            struct amdgpu_task_info *task_info;
>>>                 /* This can happen when we set higher level PDs to
>>>                * silent to stop fault floods.
>>>                */
>>>               nptes = max(nptes, 1u);
>>>   +            task_info = amdgpu_vm_get_task_info(adev, vm->pasid);
>>>               trace_amdgpu_vm_update_ptes(params, frag_start, upd_end,
>>>                               min(nptes, 32u), dst, incr,
>>>                               upd_flags,
>>> -                            vm->task_info.tgid,
>>> +                            task_info ? task_info->tgid : 0,
>>>                               vm->immediate.fence_context);
>>>               amdgpu_vm_pte_update_flags(params, to_amdgpu_bo_vm(pt),
>>>                              cursor.level, pe_start, dst,
>>>                              nptes, incr, upd_flags);
>>>   +            amdgpu_vm_put_task_info(adev, vm->pasid);
>>>               pe_start += nptes * 8;
>>>               dst += nptes * incr;
>>>   diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c 
>>> b/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c
>>> index fa87a85e1017..14ded13c8b09 100644
>>> --- a/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c
>>> +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c
>>> @@ -107,7 +107,7 @@ static int gmc_v10_0_process_interrupt(struct 
>>> amdgpu_device *adev,
>>>       struct amdgpu_vmhub *hub = &adev->vmhub[vmhub_index];
>>>       bool retry_fault = !!(entry->src_data[1] & 0x80);
>>>       bool write_fault = !!(entry->src_data[1] & 0x20);
>>> -    struct amdgpu_task_info task_info;
>>> +    struct amdgpu_task_info *task_info;
>>>       uint32_t status = 0;
>>>       u64 addr;
>>>   @@ -155,15 +155,18 @@ static int 
>>> gmc_v10_0_process_interrupt(struct amdgpu_device *adev,
>>>       if (!printk_ratelimit())
>>>           return 0;
>>>   -    memset(&task_info, 0, sizeof(struct amdgpu_task_info));
>>> -    amdgpu_vm_get_task_info(adev, entry->pasid, &task_info);
>>> -
>>> -    dev_err(adev->dev,
>>> -        "[%s] page fault (src_id:%u ring:%u vmid:%u pasid:%u, for 
>>> process %s pid %d thread %s pid %d)\n",
>>> +    dev_err(adev->dev, "[%s] page fault (src_id:%u ring:%u vmid:%u 
>>> pasid:%u)\n",
>>>           entry->vmid_src ? "mmhub" : "gfxhub",
>>>           entry->src_id, entry->ring_id, entry->vmid,
>>> -        entry->pasid, task_info.process_name, task_info.tgid,
>>> -        task_info.task_name, task_info.pid);
>>> +        entry->pasid);
>>> +
>>> +    task_info = amdgpu_vm_get_task_info(adev, entry->pasid);
>>> +    if (task_info)
>>> +        dev_err(adev->dev, "for process %s pid %d thread %s pid %d\n",
>>> +                   task_info->process_name, task_info->tgid,
>>> +                   task_info->task_name, task_info->pid);
>>> +    amdgpu_vm_put_task_info(adev, entry->pasid);
>>> +
>>>       dev_err(adev->dev, "  in page starting at address 0x%016llx 
>>> from client 0x%x (%s)\n",
>>>           addr, entry->client_id,
>>>           soc15_ih_clientid_name[entry->client_id]);
>>> diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c 
>>> b/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c
>>> index e3b76fd28d15..2d96567171bc 100644
>>> --- a/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c
>>> +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c
>>> @@ -120,17 +120,21 @@ static int gmc_v11_0_process_interrupt(struct 
>>> amdgpu_device *adev,
>>>       }
>>>         if (printk_ratelimit()) {
>>> -        struct amdgpu_task_info task_info;
>>> -
>>> -        memset(&task_info, 0, sizeof(struct amdgpu_task_info));
>>> -        amdgpu_vm_get_task_info(adev, entry->pasid, &task_info);
>>> +        struct amdgpu_task_info *task_info;
>>>             dev_err(adev->dev,
>>> -            "[%s] page fault (src_id:%u ring:%u vmid:%u pasid:%u, 
>>> for process %s pid %d thread %s pid %d)\n",
>>> +            "[%s] page fault (src_id:%u ring:%u vmid:%u pasid:%u)\n",
>>>               entry->vmid_src ? "mmhub" : "gfxhub",
>>>               entry->src_id, entry->ring_id, entry->vmid,
>>> -            entry->pasid, task_info.process_name, task_info.tgid,
>>> -            task_info.task_name, task_info.pid);
>>> +            entry->pasid);
>>> +
>>> +        task_info = amdgpu_vm_get_task_info(adev, entry->pasid);
>>> +        if (task_info)
>>> +            dev_err(adev->dev, " for process %s pid %d thread %s 
>>> pid %d\n",
>>> +                task_info->process_name, task_info->tgid,
>>> +                task_info->task_name, task_info->pid);
>>> +        amdgpu_vm_put_task_info(adev, entry->pasid);
>>> +
>>>           dev_err(adev->dev, "  in page starting at address 
>>> 0x%016llx from client %d\n",
>>>               addr, entry->client_id);
>>>           if (!amdgpu_sriov_vf(adev))
>>> diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c 
>>> b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
>>> index 5af235202513..8759ef1c5ea5 100644
>>> --- a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
>>> +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
>>> @@ -1446,14 +1446,19 @@ static int gmc_v8_0_process_interrupt(struct 
>>> amdgpu_device *adev,
>>>           gmc_v8_0_set_fault_enable_default(adev, false);
>>>         if (printk_ratelimit()) {
>>> -        struct amdgpu_task_info task_info;
>>> +        struct amdgpu_task_info *task_info;
>>>   -        memset(&task_info, 0, sizeof(struct amdgpu_task_info));
>>> -        amdgpu_vm_get_task_info(adev, entry->pasid, &task_info);
>>> +        dev_err(adev->dev, "GPU fault detected: %d 0x%08x\n",
>>> +            entry->src_id, entry->src_data[0]);
>>> +
>>> +        task_info = amdgpu_vm_get_task_info(adev, entry->pasid);
>>> +        if (task_info) {
>>> +            dev_err(adev->dev, "for process %s pid %d thread %s pid 
>>> %d\n\n",
>>> +                task_info->process_name, task_info->tgid,
>>> +                task_info->task_name, task_info->pid);
>>> +        }
>>> +        amdgpu_vm_put_task_info(adev, entry->pasid);
>>>   -        dev_err(adev->dev, "GPU fault detected: %d 0x%08x for 
>>> process %s pid %d thread %s pid %d\n",
>>> -            entry->src_id, entry->src_data[0], task_info.process_name,
>>> -            task_info.tgid, task_info.task_name, task_info.pid);
>>>           dev_err(adev->dev, " VM_CONTEXT1_PROTECTION_FAULT_ADDR   
>>> 0x%08X\n",
>>>               addr);
>>>           dev_err(adev->dev, " VM_CONTEXT1_PROTECTION_FAULT_STATUS 
>>> 0x%08X\n",
>>> diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c 
>>> b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
>>> index f9a5a2c0573e..75b849d69875 100644
>>> --- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
>>> +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
>>> @@ -550,7 +550,7 @@ static int gmc_v9_0_process_interrupt(struct 
>>> amdgpu_device *adev,
>>>       bool retry_fault = !!(entry->src_data[1] & 0x80);
>>>       bool write_fault = !!(entry->src_data[1] & 0x20);
>>>       uint32_t status = 0, cid = 0, rw = 0;
>>> -    struct amdgpu_task_info task_info;
>>> +    struct amdgpu_task_info *task_info;
>>>       struct amdgpu_vmhub *hub;
>>>       const char *mmhub_cid;
>>>       const char *hub_name;
>>> @@ -625,16 +625,19 @@ static int gmc_v9_0_process_interrupt(struct 
>>> amdgpu_device *adev,
>>>       if (!printk_ratelimit())
>>>           return 0;
>>>   -
>>> -    memset(&task_info, 0, sizeof(struct amdgpu_task_info));
>>> -    amdgpu_vm_get_task_info(adev, entry->pasid, &task_info);
>>> -
>>>       dev_err(adev->dev,
>>> -        "[%s] %s page fault (src_id:%u ring:%u vmid:%u pasid:%u, 
>>> for process %s pid %d thread %s pid %d)\n",
>>> +        "[%s] %s page fault (src_id:%u ring:%u vmid:%u pasid:%u)\n",
>>>           hub_name, retry_fault ? "retry" : "no-retry",
>>>           entry->src_id, entry->ring_id, entry->vmid,
>>> -        entry->pasid, task_info.process_name, task_info.tgid,
>>> -        task_info.task_name, task_info.pid);
>>> +        entry->pasid);
>>> +
>>> +    task_info = amdgpu_vm_get_task_info(adev, entry->pasid);
>>> +    if (task_info)
>>> +        dev_err(adev->dev, " process %s pid %d thread %s pid %d\n",
>>> +            task_info->process_name, task_info->tgid,
>>> +            task_info->task_name, task_info->pid);
>>> +    amdgpu_vm_put_task_info(adev, entry->pasid);
>>> +
>>>       dev_err(adev->dev, "  in page starting at address 0x%016llx 
>>> from IH client 0x%x (%s)\n",
>>>           addr, entry->client_id,
>>>           soc15_ih_clientid_name[entry->client_id]);
>>> diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c 
>>> b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
>>> index cd37f45e01a1..d3b9fe74332a 100644
>>> --- a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
>>> +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
>>> @@ -2112,7 +2112,7 @@ static int sdma_v4_0_print_iv_entry(struct 
>>> amdgpu_device *adev,
>>>                             struct amdgpu_iv_entry *entry)
>>>   {
>>>       int instance;
>>> -    struct amdgpu_task_info task_info;
>>> +    struct amdgpu_task_info *task_info;
>>>       u64 addr;
>>>         instance = sdma_v4_0_irq_id_to_seq(entry->client_id);
>>> @@ -2124,15 +2124,16 @@ static int sdma_v4_0_print_iv_entry(struct 
>>> amdgpu_device *adev,
>>>       addr = (u64)entry->src_data[0] << 12;
>>>       addr |= ((u64)entry->src_data[1] & 0xf) << 44;
>>>   -    memset(&task_info, 0, sizeof(struct amdgpu_task_info));
>>> -    amdgpu_vm_get_task_info(adev, entry->pasid, &task_info);
>>> -
>>>       dev_dbg_ratelimited(adev->dev,
>>> -           "[sdma%d] address:0x%016llx src_id:%u ring:%u vmid:%u "
>>> -           "pasid:%u, for process %s pid %d thread %s pid %d\n",
>>> -           instance, addr, entry->src_id, entry->ring_id, entry->vmid,
>>> -           entry->pasid, task_info.process_name, task_info.tgid,
>>> -           task_info.task_name, task_info.pid);
>>> +           "[sdma%d] address:0x%016llx src_id:%u ring:%u vmid:%u 
>>> pasid:%u\n",
>>> +           instance, addr, entry->src_id, entry->ring_id, 
>>> entry->vmid, entry->pasid);
>>> +
>>> +    task_info = amdgpu_vm_get_task_info(adev, entry->pasid);
>>> +    if (task_info)
>>> +        dev_dbg_ratelimited(adev->dev, "for process %s pid %d 
>>> thread %s pid %d\n",
>>> +                    task_info->process_name, task_info->tgid,
>>> +                    task_info->task_name, task_info->pid);
>>> +    amdgpu_vm_put_task_info(adev, entry->pasid);
>>>       return 0;
>>>   }
>>>   diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v4_4_2.c 
>>> b/drivers/gpu/drm/amd/amdgpu/sdma_v4_4_2.c
>>> index f413898dda37..56c2f744d64e 100644
>>> --- a/drivers/gpu/drm/amd/amdgpu/sdma_v4_4_2.c
>>> +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v4_4_2.c
>>> @@ -1633,7 +1633,7 @@ static int sdma_v4_4_2_print_iv_entry(struct 
>>> amdgpu_device *adev,
>>>                             struct amdgpu_iv_entry *entry)
>>>   {
>>>       int instance;
>>> -    struct amdgpu_task_info task_info;
>>> +    struct amdgpu_task_info *task_info;
>>>       u64 addr;
>>>         instance = sdma_v4_4_2_irq_id_to_seq(entry->client_id);
>>> @@ -1645,15 +1645,17 @@ static int sdma_v4_4_2_print_iv_entry(struct 
>>> amdgpu_device *adev,
>>>       addr = (u64)entry->src_data[0] << 12;
>>>       addr |= ((u64)entry->src_data[1] & 0xf) << 44;
>>>   -    memset(&task_info, 0, sizeof(struct amdgpu_task_info));
>>> -    amdgpu_vm_get_task_info(adev, entry->pasid, &task_info);
>>> -
>>>       dev_dbg_ratelimited(adev->dev,
>>> -           "[sdma%d] address:0x%016llx src_id:%u ring:%u vmid:%u "
>>> -           "pasid:%u, for process %s pid %d thread %s pid %d\n",
>>> -           instance, addr, entry->src_id, entry->ring_id, entry->vmid,
>>> -           entry->pasid, task_info.process_name, task_info.tgid,
>>> -           task_info.task_name, task_info.pid);
>>> +           "[sdma%d] address:0x%016llx src_id:%u ring:%u vmid:%u 
>>> pasid:%u\n",
>>> +           instance, addr, entry->src_id, entry->ring_id, 
>>> entry->vmid, entry->pasid);
>>> +
>>> +    task_info = amdgpu_vm_get_task_info(adev, entry->pasid);
>>> +    if (task_info)
>>> +        dev_dbg_ratelimited(adev->dev, "for process %s pid %d 
>>> thread %s pid %d\n",
>>> +                    task_info->process_name, task_info->tgid,
>>> +                    task_info->task_name, task_info->pid);
>>> +    amdgpu_vm_put_task_info(adev, entry->pasid);
>>> +
>>>       return 0;
>>>   }
>>>   diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.c 
>>> b/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.c
>>> index d9953c2b2661..6b51262811f6 100644
>>> --- a/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.c
>>> +++ b/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.c
>>> @@ -238,16 +238,18 @@ void 
>>> kfd_smi_event_update_thermal_throttling(struct kfd_node *dev,
>>>     void kfd_smi_event_update_vmfault(struct kfd_node *dev, uint16_t 
>>> pasid)
>>>   {
>>> -    struct amdgpu_task_info task_info;
>>> +    struct amdgpu_task_info *task_info;
>>>   -    memset(&task_info, 0, sizeof(struct amdgpu_task_info));
>>> -    amdgpu_vm_get_task_info(dev->adev, pasid, &task_info);
>>>       /* Report VM faults from user applications, not retry from 
>>> kernel */
>>> -    if (!task_info.pid)
>>> -        return;
>>> +    task_info = amdgpu_vm_get_task_info(dev->adev, pasid);
>>> +    if (!task_info || !task_info->pid)
>>> +        goto unref;
>>>         kfd_smi_event_add(0, dev, KFD_SMI_EVENT_VMFAULT, "%x:%s\n",
>>> -              task_info.pid, task_info.task_name);
>>> +              task_info->pid, task_info->task_name);
>>> +
>>> +unref:
>>> +    amdgpu_vm_put_task_info(dev->adev, pasid);
>>>   }
>>>     void kfd_smi_event_page_fault_start(struct kfd_node *node, pid_t 
>>> pid,
>


More information about the amd-gfx mailing list