Yeah, this looks good to me.
Regards,
 Felix
On 2018-01-10 04:58 PM, Gustavo A. R. Silva wrote:
> Hi Felix,
>
> Quoting Felix Kuehling <felix.kuehling at amd.com>:
>
>> Hi Gustavo,
>>
>> Thanks for catching that. When returning a fault, I think you also need
>> to srcu_read_unlock(&kfd_processes_srcu, idx).
>>
>> However, instead of returning an error, I think I'd prefer to skip PDDs
>> that can't be found with continue statements. That way others would
>> still suspend and resume successfully. Maybe just print a WARN_ON for
>> PDDs that aren't found, because that's an unexpected situation,
>> currently. Maybe in the future it could be normal thing if we ever
>> support GPU hotplug.
>>
>
> I got it. In that case, what do you think about the following patch
> instead?
>
> index a22fb071..4ff5f0f 100644
> --- a/drivers/gpu/drm/amd/amdkfd/kfd_process.c
> +++ b/drivers/gpu/drm/amd/amdkfd/kfd_process.c
> @@ -461,7 +461,8 @@ int kfd_bind_processes_to_device(struct kfd_dev *dev)
> Â Â Â Â Â Â Â hash_for_each_rcu(kfd_processes_table, temp, p, kfd_processes) {
> Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â mutex_lock(&p->mutex);
> Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â pdd = kfd_get_process_device_data(dev, p);
> -Â Â Â Â Â Â Â Â Â Â Â Â Â Â if (pdd->bound != PDD_BOUND_SUSPENDED) {
> +
> +Â Â Â Â Â Â Â Â Â Â Â Â Â Â if (WARN_ON(!pdd) || pdd->bound != PDD_BOUND_SUSPENDED) {
> Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â mutex_unlock(&p->mutex);
> Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â continue;
> Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â }
> @@ -501,6 +502,11 @@ void kfd_unbind_processes_from_device(struct
> kfd_dev *dev)
> Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â mutex_lock(&p->mutex);
> Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â pdd = kfd_get_process_device_data(dev, p);
>
> +Â Â Â Â Â Â Â Â Â Â Â Â Â Â if (WARN_ON(!pdd)) {
> +Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â mutex_unlock(&p->mutex);
> +Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â continue;
> +Â Â Â Â Â Â Â Â Â Â Â Â Â Â }
> +
> Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â if (pdd->bound == PDD_BOUND)
> Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â pdd->bound = PDD_BOUND_SUSPENDED;
> Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â mutex_unlock(&p->mutex);
>
>
> Thank you for the feedback.
> --
> Gustavo
>
>> Regards,
>> Â Felix
>>
>>
>> On 2018-01-10 11:50 AM, Gustavo A. R. Silva wrote:
>>> In case kfd_get_process_device_data returns null, there are some
>>> null pointer dereferences in functions kfd_bind_processes_to_device
>>> and kfd_unbind_processes_from_device.
>>>
>>> Fix this by null checking pdd before dereferencing it.
>>>
>>> Addresses-Coverity-ID: 1463794 ("Dereference null return value")
>>> Addresses-Coverity-ID: 1463772 ("Dereference null return value")
>>> Signed-off-by: Gustavo A. R. Silva <garsilva at embeddedor.com>
>>> ---
>>> Â drivers/gpu/drm/amd/amdkfd/kfd_process.c | 12 ++++++++++++
>>> Â 1 file changed, 12 insertions(+)
>>>
>>> diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process.c
>>> b/drivers/gpu/drm/amd/amdkfd/kfd_process.c
>>> index a22fb071..29d51d5 100644
>>> --- a/drivers/gpu/drm/amd/amdkfd/kfd_process.c
>>> +++ b/drivers/gpu/drm/amd/amdkfd/kfd_process.c
>>> @@ -461,6 +461,13 @@ int kfd_bind_processes_to_device(struct kfd_dev
>>> *dev)
>>> Â Â Â Â hash_for_each_rcu(kfd_processes_table, temp, p, kfd_processes) {
>>> Â Â Â Â Â Â Â Â mutex_lock(&p->mutex);
>>> Â Â Â Â Â Â Â Â pdd = kfd_get_process_device_data(dev, p);
>>> +
>>> +Â Â Â Â Â Â Â if (!pdd) {
>>> +Â Â Â Â Â Â Â Â Â Â Â pr_err("Process device data doesn't exist\n");
>>> +Â Â Â Â Â Â Â Â Â Â Â mutex_unlock(&p->mutex);
>>> +Â Â Â Â Â Â Â Â Â Â Â return -EFAULT;
>>> +Â Â Â Â Â Â Â }
>>> +
>>> Â Â Â Â Â Â Â Â if (pdd->bound != PDD_BOUND_SUSPENDED) {
>>> Â Â Â Â Â Â Â Â Â Â Â Â mutex_unlock(&p->mutex);
>>> Â Â Â Â Â Â Â Â Â Â Â Â continue;
>>> @@ -501,6 +508,11 @@ void kfd_unbind_processes_from_device(struct
>>> kfd_dev *dev)
>>> Â Â Â Â Â Â Â Â mutex_lock(&p->mutex);
>>> Â Â Â Â Â Â Â Â pdd = kfd_get_process_device_data(dev, p);
>>>
>>> +Â Â Â Â Â Â Â if (!pdd) {
>>> +Â Â Â Â Â Â Â Â Â Â Â mutex_unlock(&p->mutex);
>>> +Â Â Â Â Â Â Â Â Â Â Â return;
>>> +Â Â Â Â Â Â Â }
>>> +
>>> Â Â Â Â Â Â Â Â if (pdd->bound == PDD_BOUND)
>>> Â Â Â Â Â Â Â Â Â Â Â Â pdd->bound = PDD_BOUND_SUSPENDED;
>>> Â Â Â Â Â Â Â Â mutex_unlock(&p->mutex);
>
>
>
>
>
>
