Re: [PATCH v3 33/35] dmaengine: plx_dma: convert tasklets to use new tasklet_setup() API

[Logan Gunthorpe <logang@deltatee.com>]

On 2020-08-31 9:56 a.m., Allen wrote:
> Logan,
>>> In preparation for unconditionally passing the
>>> struct tasklet_struct pointer to all tasklet
>>> callbacks, switch to using the new tasklet_setup()
>>> and from_tasklet() to pass the tasklet pointer explicitly.
>>>
>>> Signed-off-by: Allen Pais <allen.lkml@gmail.com>
>>> ---
>>>  drivers/dma/plx_dma.c | 7 +++----
>>>  1 file changed, 3 insertions(+), 4 deletions(-)
>>>
>>> diff --git a/drivers/dma/plx_dma.c b/drivers/dma/plx_dma.c
>>> index db4c5fd453a9..f387c5bbc170 100644
>>> --- a/drivers/dma/plx_dma.c
>>> +++ b/drivers/dma/plx_dma.c
>>> @@ -241,9 +241,9 @@ static void plx_dma_stop(struct plx_dma_dev *plxdev)
>>>       rcu_read_unlock();
>>>  }
>>>
>>> -static void plx_dma_desc_task(unsigned long data)
>>> +static void plx_dma_desc_task(struct tasklet_struct *t)
>>>  {
>>> -     struct plx_dma_dev *plxdev = (void *)data;
>>> +     struct plx_dma_dev *plxdev = from_tasklet(plxdev, t, desc_task);
>>
>> The discussion I saw on another thread suggested the private macro
>> from_tasklet() would be replaced with something generic. So isn't this
>> patchset a bit premature?
> 
>  Yes, but efforts to replace it with something generic was not well accepted.
> We were left with either using a private macro from_tasklet() or use
> well known container_of().  I have left it to the maintainers to decide
> which one they prefer.

Well my vote would be for using container_of() directly. But I don't
really have much clout here.

Logan