[v4,11/11] staging: r8188eu: Remove unused macros

  Simple variants of macros PlatformEFIOWrite and PlatformEFIORead are
defined but never used. As they do not appear to be designed for anything
significant, we can remove them to avoid unexpected usage.

Suggested-by: Julia Lawall <julia.lawall@inria.fr>
Signed-off-by: Deepak R Varma <drv@mailo.com>
---

Changes in v4:
   1. Patch newly added to the patch set.


 drivers/staging/r8188eu/include/rtw_io.h | 14 --------------
 1 file changed, 14 deletions(-)

--
2.30.2
  

On Fri, Oct 21, 2022 at 07:56:05AM +0200, Greg KH wrote:
> On Fri, Oct 21, 2022 at 11:21:06AM +0530, Praveen Kumar wrote:
> > On 21-10-2022 03:02, Deepak R Varma wrote:
> > > Simple variants of macros PlatformEFIOWrite and PlatformEFIORead are
> > > defined but never used. As they do not appear to be designed for anything
> > > significant, we can remove them to avoid unexpected usage.
> > >
> > > Suggested-by: Julia Lawall <julia.lawall@inria.fr>
> > > Signed-off-by: Deepak R Varma <drv@mailo.com>
> > > ---
> > >
> > > Changes in v4:
> > >    1. Patch newly added to the patch set.
> > >
> > >
> > >  drivers/staging/r8188eu/include/rtw_io.h | 14 --------------
> > >  1 file changed, 14 deletions(-)
> > >
> > > diff --git a/drivers/staging/r8188eu/include/rtw_io.h b/drivers/staging/r8188eu/include/rtw_io.h
> > > index 87fcf6c94ff3..e9744694204b 100644
> > > --- a/drivers/staging/r8188eu/include/rtw_io.h
> > > +++ b/drivers/staging/r8188eu/include/rtw_io.h
> > > @@ -285,18 +285,4 @@ void bus_sync_io(struct io_queue *pio_q);
> > >  u32 _ioreq2rwmem(struct io_queue *pio_q);
> > >  void dev_power_down(struct adapter *Adapter, u8 bpwrup);
> > >
> > > -#define PlatformEFIOWrite1Byte(_a, _b, _c)		\
> > > -	rtw_write8(_a, _b, _c)
> > > -#define PlatformEFIOWrite2Byte(_a, _b, _c)		\
> > > -	rtw_write16(_a, _b, _c)
> > > -#define PlatformEFIOWrite4Byte(_a, _b, _c)		\
> > > -	rtw_write32(_a, _b, _c)
> > > -
> > > -#define PlatformEFIORead1Byte(_a, _b)		\
> >
> > Can the naming be reworked to make more Linux friendly ? something like PLATFORM_EFIO_READ1BYTE or better if there are other suggestions?
> > Rest others as applicable ?
>
> All of these should just be removed entirely and the normal calls here
> made instead.  There is no need for these #defines at all.

Hello Greg,
The current patch scope results in removal of these #defines. Can you please
clarify what you mean by "normal calls here made instead"? Do you mean the
calling function should directly call the wrapped functions OR you are
suggesting converting these #defines to inline functions?

Thank you,
./drv

>
> thanks,
>
> greg k-h
>
  

On Fri, Oct 21, 2022 at 10:33:57AM +0200, Greg KH wrote:
> On Fri, Oct 21, 2022 at 05:04:37AM +0530, Deepak R Varma wrote:
> > On Fri, Oct 21, 2022 at 07:56:05AM +0200, Greg KH wrote:
> > > On Fri, Oct 21, 2022 at 11:21:06AM +0530, Praveen Kumar wrote:
> > > > On 21-10-2022 03:02, Deepak R Varma wrote:
> > > > > Simple variants of macros PlatformEFIOWrite and PlatformEFIORead are
> > > > > defined but never used. As they do not appear to be designed for anything
> > > > > significant, we can remove them to avoid unexpected usage.
> > > > >
> > > > > Suggested-by: Julia Lawall <julia.lawall@inria.fr>
> > > > > Signed-off-by: Deepak R Varma <drv@mailo.com>
> > > > > ---
> > > > >
> > > > > Changes in v4:
> > > > >    1. Patch newly added to the patch set.
> > > > >
> > > > >
> > > > >  drivers/staging/r8188eu/include/rtw_io.h | 14 --------------
> > > > >  1 file changed, 14 deletions(-)
> > > > >
> > > > > diff --git a/drivers/staging/r8188eu/include/rtw_io.h b/drivers/staging/r8188eu/include/rtw_io.h
> > > > > index 87fcf6c94ff3..e9744694204b 100644
> > > > > --- a/drivers/staging/r8188eu/include/rtw_io.h
> > > > > +++ b/drivers/staging/r8188eu/include/rtw_io.h
> > > > > @@ -285,18 +285,4 @@ void bus_sync_io(struct io_queue *pio_q);
> > > > >  u32 _ioreq2rwmem(struct io_queue *pio_q);
> > > > >  void dev_power_down(struct adapter *Adapter, u8 bpwrup);
> > > > >
> > > > > -#define PlatformEFIOWrite1Byte(_a, _b, _c)		\
> > > > > -	rtw_write8(_a, _b, _c)
> > > > > -#define PlatformEFIOWrite2Byte(_a, _b, _c)		\
> > > > > -	rtw_write16(_a, _b, _c)
> > > > > -#define PlatformEFIOWrite4Byte(_a, _b, _c)		\
> > > > > -	rtw_write32(_a, _b, _c)
> > > > > -
> > > > > -#define PlatformEFIORead1Byte(_a, _b)		\
> > > >
> > > > Can the naming be reworked to make more Linux friendly ? something like PLATFORM_EFIO_READ1BYTE or better if there are other suggestions?
> > > > Rest others as applicable ?
> > >
> > > All of these should just be removed entirely and the normal calls here
> > > made instead.  There is no need for these #defines at all.
> >
> > Hello Greg,
> > The current patch scope results in removal of these #defines. Can you please
> > clarify what you mean by "normal calls here made instead"? Do you mean the
> > calling function should directly call the wrapped functions OR you are
> > suggesting converting these #defines to inline functions?
>
> The calling function should call the wrapped function instead, there is
> no need for another pointless level of indirection here.

Sounds good. Thank you for the clarification.

./drv

>
> thanks,
>
> greg k-h
  

On 21-10-2022 03:02, Deepak R Varma wrote:
> Simple variants of macros PlatformEFIOWrite and PlatformEFIORead are
> defined but never used. As they do not appear to be designed for anything
> significant, we can remove them to avoid unexpected usage.
> 
> Suggested-by: Julia Lawall <julia.lawall@inria.fr>
> Signed-off-by: Deepak R Varma <drv@mailo.com>
> ---
> 
> Changes in v4:
>    1. Patch newly added to the patch set.
> 
> 
>  drivers/staging/r8188eu/include/rtw_io.h | 14 --------------
>  1 file changed, 14 deletions(-)
> 
> diff --git a/drivers/staging/r8188eu/include/rtw_io.h b/drivers/staging/r8188eu/include/rtw_io.h
> index 87fcf6c94ff3..e9744694204b 100644
> --- a/drivers/staging/r8188eu/include/rtw_io.h
> +++ b/drivers/staging/r8188eu/include/rtw_io.h
> @@ -285,18 +285,4 @@ void bus_sync_io(struct io_queue *pio_q);
>  u32 _ioreq2rwmem(struct io_queue *pio_q);
>  void dev_power_down(struct adapter *Adapter, u8 bpwrup);
> 
> -#define PlatformEFIOWrite1Byte(_a, _b, _c)		\
> -	rtw_write8(_a, _b, _c)
> -#define PlatformEFIOWrite2Byte(_a, _b, _c)		\
> -	rtw_write16(_a, _b, _c)
> -#define PlatformEFIOWrite4Byte(_a, _b, _c)		\
> -	rtw_write32(_a, _b, _c)
> -
> -#define PlatformEFIORead1Byte(_a, _b)		\

Can the naming be reworked to make more Linux friendly ? something like PLATFORM_EFIO_READ1BYTE or better if there are other suggestions?
Rest others as applicable ?

> -		rtw_read8(_a, _b)
> -#define PlatformEFIORead2Byte(_a, _b)		\
> -		rtw_read16(_a, _b)
> -#define PlatformEFIORead4Byte(_a, _b)		\
> -		rtw_read32(_a, _b)
> -
>  #endif	/* _RTL8711_IO_H_ */
> --
> 2.30.2
> 
> 
> 
Regards,

~Praveen.
  

On Fri, Oct 21, 2022 at 11:21:06AM +0530, Praveen Kumar wrote:
> On 21-10-2022 03:02, Deepak R Varma wrote:
> > Simple variants of macros PlatformEFIOWrite and PlatformEFIORead are
> > defined but never used. As they do not appear to be designed for anything
> > significant, we can remove them to avoid unexpected usage.
> > 
> > Suggested-by: Julia Lawall <julia.lawall@inria.fr>
> > Signed-off-by: Deepak R Varma <drv@mailo.com>
> > ---
> > 
> > Changes in v4:
> >    1. Patch newly added to the patch set.
> > 
> > 
> >  drivers/staging/r8188eu/include/rtw_io.h | 14 --------------
> >  1 file changed, 14 deletions(-)
> > 
> > diff --git a/drivers/staging/r8188eu/include/rtw_io.h b/drivers/staging/r8188eu/include/rtw_io.h
> > index 87fcf6c94ff3..e9744694204b 100644
> > --- a/drivers/staging/r8188eu/include/rtw_io.h
> > +++ b/drivers/staging/r8188eu/include/rtw_io.h
> > @@ -285,18 +285,4 @@ void bus_sync_io(struct io_queue *pio_q);
> >  u32 _ioreq2rwmem(struct io_queue *pio_q);
> >  void dev_power_down(struct adapter *Adapter, u8 bpwrup);
> > 
> > -#define PlatformEFIOWrite1Byte(_a, _b, _c)		\
> > -	rtw_write8(_a, _b, _c)
> > -#define PlatformEFIOWrite2Byte(_a, _b, _c)		\
> > -	rtw_write16(_a, _b, _c)
> > -#define PlatformEFIOWrite4Byte(_a, _b, _c)		\
> > -	rtw_write32(_a, _b, _c)
> > -
> > -#define PlatformEFIORead1Byte(_a, _b)		\
> 
> Can the naming be reworked to make more Linux friendly ? something like PLATFORM_EFIO_READ1BYTE or better if there are other suggestions?
> Rest others as applicable ?

All of these should just be removed entirely and the normal calls here
made instead.  There is no need for these #defines at all.

thanks,

greg k-h
  

On Fri, Oct 21, 2022 at 05:04:37AM +0530, Deepak R Varma wrote:
> On Fri, Oct 21, 2022 at 07:56:05AM +0200, Greg KH wrote:
> > On Fri, Oct 21, 2022 at 11:21:06AM +0530, Praveen Kumar wrote:
> > > On 21-10-2022 03:02, Deepak R Varma wrote:
> > > > Simple variants of macros PlatformEFIOWrite and PlatformEFIORead are
> > > > defined but never used. As they do not appear to be designed for anything
> > > > significant, we can remove them to avoid unexpected usage.
> > > >
> > > > Suggested-by: Julia Lawall <julia.lawall@inria.fr>
> > > > Signed-off-by: Deepak R Varma <drv@mailo.com>
> > > > ---
> > > >
> > > > Changes in v4:
> > > >    1. Patch newly added to the patch set.
> > > >
> > > >
> > > >  drivers/staging/r8188eu/include/rtw_io.h | 14 --------------
> > > >  1 file changed, 14 deletions(-)
> > > >
> > > > diff --git a/drivers/staging/r8188eu/include/rtw_io.h b/drivers/staging/r8188eu/include/rtw_io.h
> > > > index 87fcf6c94ff3..e9744694204b 100644
> > > > --- a/drivers/staging/r8188eu/include/rtw_io.h
> > > > +++ b/drivers/staging/r8188eu/include/rtw_io.h
> > > > @@ -285,18 +285,4 @@ void bus_sync_io(struct io_queue *pio_q);
> > > >  u32 _ioreq2rwmem(struct io_queue *pio_q);
> > > >  void dev_power_down(struct adapter *Adapter, u8 bpwrup);
> > > >
> > > > -#define PlatformEFIOWrite1Byte(_a, _b, _c)		\
> > > > -	rtw_write8(_a, _b, _c)
> > > > -#define PlatformEFIOWrite2Byte(_a, _b, _c)		\
> > > > -	rtw_write16(_a, _b, _c)
> > > > -#define PlatformEFIOWrite4Byte(_a, _b, _c)		\
> > > > -	rtw_write32(_a, _b, _c)
> > > > -
> > > > -#define PlatformEFIORead1Byte(_a, _b)		\
> > >
> > > Can the naming be reworked to make more Linux friendly ? something like PLATFORM_EFIO_READ1BYTE or better if there are other suggestions?
> > > Rest others as applicable ?
> >
> > All of these should just be removed entirely and the normal calls here
> > made instead.  There is no need for these #defines at all.
> 
> Hello Greg,
> The current patch scope results in removal of these #defines. Can you please
> clarify what you mean by "normal calls here made instead"? Do you mean the
> calling function should directly call the wrapped functions OR you are
> suggesting converting these #defines to inline functions?

The calling function should call the wrapped function instead, there is
no need for another pointless level of indirection here.

thanks,

greg k-h