Re: [PATCH v3 1/3] staging: r8188eu: remove _io_ops structure

[Greg Kroah-Hartman]

On Mon, Sep 06, 2021 at 08:19:05PM +0300, Pavel Skripkin wrote:
> On 9/6/21 16:56, Greg Kroah-Hartman wrote:
> > On Sun, Sep 05, 2021 at 12:00:46AM +0200, Fabio M. De Francesco wrote:
> > > -void _rtw_read_mem(struct adapter *adapter, u32 addr, u32 cnt, u8 *pmem)
> > > -{
> > > -	void (*_read_mem)(struct intf_hdl *pintfhdl, u32 addr, u32 cnt, u8 *pmem);
> > > -	struct io_priv *pio_priv = &adapter->iopriv;
> > > -	struct	intf_hdl		*pintfhdl = &pio_priv->intf;
> > > -
> > > -
> > > -	if (adapter->bDriverStopped || adapter->bSurpriseRemoved)
> > > -	     return;
> > > -	_read_mem = pintfhdl->io_ops._read_mem;
> > > -	_read_mem(pintfhdl, addr, cnt, pmem);
> > > -
> > > -}
> > 
> > This is odd, in that it resolves down to usb_read_mem which does
> > nothing at all.
> > 
> > And then no one calls this at all either?
> > 
> > How about removing the io ops that are not used at all first, one at a
> > time, making it obvious what is happening, and then convert the ones
> > that are used one at a time, and when all is done, then removing the
> > structure?
> > 
> 
> Just have started to cut one big patch to smaller ones and does it make
> sense to group changes like: one for usb_read*, one for usb_write* and one
> for usb_port*? I think, it would be cleaner and series won't be too big.
> 
> 
> What do you think?

I will not know until I see the patches, so no need to ask :)