Re: [PATCH] staging: rtl8192u: Convert __le16 to cpu before casting to u32

[Greg Kroah-Hartman]

On Sun, May 07, 2017 at 02:56:36PM +0200, Guillaume Brogi wrote:
> On Sat, Apr 08, 2017 at 08:32:36PM +0200, Guillaume Brogi wrote:
> > On Sat, Apr 08, 2017 at 12:31:25PM +0200, Greg Kroah-Hartman wrote:
> > > On Sun, Mar 26, 2017 at 12:24:14AM +0100, Guillaume Brogi wrote:
> > > > 
> > > > This patch fixes the following sparse warnings:
> > > > drivers/staging/rtl8192u/r8192U_dm.c:2307:49: warning: cast from restricted __le16
> > > > drivers/staging/rtl8192u/r8192U_dm.c:2308:44: warning: cast from restricted __le16
> > > > drivers/staging/rtl8192u/r8192U_dm.c:2309:44: warning: cast from restricted __le16
> > > > 
> > > > Those three members of qos_parameters are indeed __le16 so they should
> > > > be converted to the cpu's endianness before being cast to u32.
> > > > 
> > > > The lines are over the 80 character limit. They already were, and, for
> > > > the sake of readability, I don't think they should be split.
> > > 
> > > Please fix up your subject: line to match other patches for this driver,
> > > I almost missed it :(
> > > 
> > Sorry about that. I realised my mistake after sending the email.
> > 
> > > > 
> > > > Signed-off-by: Guillaume Brogi <gui-gui@xxxxxxxxxxxxxxx>
> > > > ---
> > > >  drivers/staging/rtl8192u/r8192U_dm.c | 6 +++---
> > > >  1 file changed, 3 insertions(+), 3 deletions(-)
> > > > 
> > > > diff --git a/drivers/staging/rtl8192u/r8192U_dm.c b/drivers/staging/rtl8192u/r8192U_dm.c
> > > > index 9209aad0515e..5946c1f8d37d 100644
> > > > --- a/drivers/staging/rtl8192u/r8192U_dm.c
> > > > +++ b/drivers/staging/rtl8192u/r8192U_dm.c
> > > > @@ -2304,9 +2304,9 @@ static void dm_check_edca_turbo(
> > > >  				/*  For Each time updating EDCA parameter, reset EDCA turbo mode status. */
> > > >  				dm_init_edca_turbo(dev);
> > > >  				u1bAIFS = qos_parameters->aifs[0] * ((mode&(IEEE_G|IEEE_N_24G)) ? 9 : 20) + aSifsTime;
> > > > -				u4bAcParam = (((u32)(qos_parameters->tx_op_limit[0])) << AC_PARAM_TXOP_LIMIT_OFFSET)|
> > > > -					(((u32)(qos_parameters->cw_max[0])) << AC_PARAM_ECW_MAX_OFFSET)|
> > > > -					(((u32)(qos_parameters->cw_min[0])) << AC_PARAM_ECW_MIN_OFFSET)|
> > > > +				u4bAcParam = (((u32)le16_to_cpu(qos_parameters->tx_op_limit[0])) << AC_PARAM_TXOP_LIMIT_OFFSET) |
> > > > +					(((u32)le16_to_cpu(qos_parameters->cw_max[0])) << AC_PARAM_ECW_MAX_OFFSET) |
> > > > +					(((u32)le16_to_cpu(qos_parameters->cw_min[0])) << AC_PARAM_ECW_MIN_OFFSET) |
> > > >  					((u32)u1bAIFS << AC_PARAM_AIFS_OFFSET);
> > > >  				/*write_nic_dword(dev, WDCAPARA_ADD[i], u4bAcParam);*/
> > > >  				write_nic_dword(dev, EDCAPARA_BE,  u4bAcParam);
> > > 
> > > How are you sure that this change is correct?  How did you verify it?
> > > 
> > Sadly, I don't have the hardware to test that. I only checked the code
> > and compiled it.
> > 
> > > thanks,
> > > 
> > Cheers,
> > 
> Hello again,
> 
> Since there was no news for a while, I figured I'd ask what's the status
> of this patch. I hope that's OK. Has it simply been forgotten or is
> there something blocking/unacceptable?

I don't see it in my queue, sorry.  If you can't test this, or verify
that it is correct some other way, I don't want to take endian fixes,
sorry.

thanks,

greg k-h