Re: [PATCH 58/65] media: zoran: Remove access to __fh

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On 8/7/25 14:31, Laurent Pinchart wrote:
> On Thu, Aug 07, 2025 at 08:58:01AM +0200, Hans Verkuil wrote:
>> On 02/08/2025 11:23, Jacopo Mondi wrote:
>>> The __fh parameter is assigned to an unsued variable. Remove it
>>> and remove the unused struct zoran_fh type.
>>>
>>> Signed-off-by: Jacopo Mondi <jacopo.mondi@xxxxxxxxxxxxxxxx>
>>> ---
>>>  drivers/media/pci/zoran/zoran.h        | 6 ------
>>>  drivers/media/pci/zoran/zoran_driver.c | 3 +--
>>>  2 files changed, 1 insertion(+), 8 deletions(-)
>>>
>>> diff --git a/drivers/media/pci/zoran/zoran.h b/drivers/media/pci/zoran/zoran.h
>>> index 1cd990468d3de9db8b14b72483972041c57bfee2..d05e222b392156bf1b3b4c83c6591db642c3c377 100644
>>> --- a/drivers/media/pci/zoran/zoran.h
>>> +++ b/drivers/media/pci/zoran/zoran.h
>>> @@ -154,12 +154,6 @@ struct zoran_jpg_settings {
>>>  
>>>  struct zoran;
>>>  
>>> -/* zoran_fh contains per-open() settings */
>>> -struct zoran_fh {
>>> -	struct v4l2_fh fh;
>>> -	struct zoran *zr;
>>> -};
>>> -
>>>  struct card_info {
>>>  	enum card_type type;
>>>  	char name[32];
>>> diff --git a/drivers/media/pci/zoran/zoran_driver.c b/drivers/media/pci/zoran/zoran_driver.c
>>> index f42f596d3e6295e31e3b33cd83c5f7243911bd30..ec7fc1da4cc02f5a344cb49bb9a783c41c758195 100644
>>> --- a/drivers/media/pci/zoran/zoran_driver.c
>>> +++ b/drivers/media/pci/zoran/zoran_driver.c
>>> @@ -511,12 +511,11 @@ static int zoran_s_fmt_vid_cap(struct file *file, void *__fh,
>>
>> This driver uses __fh as the name for the second argument of the ioctl
>> callbacks. Can you take this opportunity to rename it to either 'fh' or 'priv'?
> 
> It's a bit of yak shaving, but not too difficult so it will be addressed
> in the next version (Jacopo already wrote the patches, thanks).
> 
>> Generally it's not a good idea to prefix variables with __ for no good reason.
>>
>> Grepping for __fh also shows two other drivers:
>>
>> drivers/media/platform/chips-media/coda/coda-common.c:#define fh_to_ctx(__fh)   container_of(__fh, struct coda_ctx, fh)
>> drivers/media/platform/samsung/s5p-mfc/s5p_mfc_common.h:#define fh_to_ctx(__fh) container_of(__fh, struct s5p_mfc_ctx, fh)
>>
>> I think it is a good idea to rename __fh to fh there as well.
> 
> That can't be done, otherwise things like
> 
> #define fh_to_ctx(fh)	container_of(fh, struct coda_ctx, fh)
> 
> static inline struct coda_ctx *file_to_ctx(struct file *filp)
> {
> 	return fh_to_ctx(file_to_v4l2_fh(filp));
> }
> 
> will expand in the file_to_ctx() function to
> 
> 	return container_of(file_to_v4l2_fh(filp), struct coda_ctx, file_to_v4l2_fh(filp))

If the fh_to_ctx #define is changed to a static inline, then there is no problem.

That said, it is overkill.

Regards,

	Hans

> 
> We could rename the __fh argument to vfh or v4l2fh, but I think __fh is
> equally good there. It's just the macro argument name.
> 
>>>  			       struct v4l2_format *fmt)
>>>  {
>>>  	struct zoran *zr = video_drvdata(file);
>>> -	struct zoran_fh *fh = __fh;
>>>  	int i;
>>>  	int res = 0;
>>>  
>>>  	if (fmt->fmt.pix.pixelformat == V4L2_PIX_FMT_MJPEG)
>>> -		return zoran_s_fmt_vid_out(file, fh, fmt);
>>> +		return zoran_s_fmt_vid_out(file, __fh, fmt);
>>>  
>>>  	for (i = 0; i < NUM_FORMATS; i++)
>>>  		if (fmt->fmt.pix.pixelformat == zoran_formats[i].fourcc)
> 





[Index of Archives]     [Linux SoC Development]     [Linux Rockchip Development]     [Linux for Synopsys ARC Processors]    
  • [Linux on Unisoc (RDA Micro) SoCs]     [Linux Actions SoC]     [Linux USB Development]     [Video for Linux]     [Linux Audio Users]     [Linux SCSI]     [Yosemite News]

  •   Powered by Linux