Am 04.11.2016 um 10:22 schrieb Flora Cui:
> On Fri, Nov 04, 2016 at 09:52:16AM +0100, Christian König wrote:
>> Well clearly a NAK on this, why do you think this is just dummy code?
>>
>> Am 04.11.2016 um 08:32 schrieb Flora Cui:
>>> Change-Id: I3238a2520c9161b1c2f9cf176158bdeb9cb21cbb
>>> Signed-off-by: Flora Cui <Flora.Cui at amd.com>
>>> ---
>>> drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c | 15 +--------------
>>> drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 4 ----
>>> 2 files changed, 1 insertion(+), 18 deletions(-)
>>>
>>> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
>>> index 4068504..be2ad79 100644
>>> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
>>> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
>>> @@ -917,13 +917,6 @@ static int amdgpu_cs_ib_fill(struct amdgpu_device *adev,
>>> memcpy(ib->ptr, kptr, chunk_ib->ib_bytes);
>>> amdgpu_bo_kunmap(aobj);
>>> - } else {
>>> - r = amdgpu_ib_get(adev, vm, 0, ib);
>>> - if (r) {
>>> - DRM_ERROR("Failed to get ib !\n");
>>> - return r;
>>> - }
>>> -
>> This is the IB allocation and initialization for the VM case and vital to
>> command submission.
> flora: amdgpu_ib_get do nothing with param size=0
That's clearly a bug. The function should at least clear the structure
members.
Christian.
>>> }
>>> ib->gpu_addr = chunk_ib->va_start;
>>> @@ -1008,21 +1001,15 @@ static int amdgpu_cs_submit(struct amdgpu_cs_parser *p,
>>> job = p->job;
>>> p->job = NULL;
>>> - r = amd_sched_job_init(&job->base, &ring->sched, entity, p->filp);
>>> + r = amdgpu_job_submit(job, ring, entity, p->filp, &p->fence);
>>> if (r) {
>>> amdgpu_job_free(job);
>>> return r;
>>> }
>>> - job->owner = p->filp;
>>> - job->fence_ctx = entity->fence_context;
>>> - p->fence = fence_get(&job->base.s_fence->finished);
>>> cs->out.handle = amdgpu_ctx_add_fence(p->ctx, ring, p->fence);
>>> job->uf_sequence = cs->out.handle;
>>> - amdgpu_job_free_resources(job);
>>> -
>>> trace_amdgpu_cs_ioctl(job);
>>> - amd_sched_entity_push_job(&job->base);
>> This results in a race condition because the job might already be freed
>> after amdgpu_job_submit() succeeded.
> flora: OK. I'll drop this.
>
>>> return 0;
>>> }
>>> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
>>> index d6c2839..c6448e1 100644
>>> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
>>> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
>>> @@ -960,10 +960,6 @@ static int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev,
>>> ring = container_of(vm->entity.sched, struct amdgpu_ring, sched);
>>> - memset(¶ms, 0, sizeof(params));
>>> - params.adev = adev;
>>> - params.src = src;
>>> -
>> This indeed looks like a duplicate, probably a leftover merge issue.
>>
>> Regards,
>> Christian.
>>
>>> /* sync to everything on unmapping */
>>> if (!(flags & AMDGPU_PTE_VALID))
>>> owner = AMDGPU_FENCE_OWNER_UNDEFINED;
>>
