Re: [PATCH] wifi: mt76: mt7996: prevent uninit return in mt7996_mac_sta_add_links

[Qasim Ijaz <qasdev00@gmail.com>]

On Sat, Mar 22, 2025 at 12:51:57PM +0100, Jonas Gorski wrote:
> Hi,
> 
> On Thu, Mar 20, 2025 at 9:19 PM Qasim Ijaz <qasdev00@gmail.com> wrote:
> >
> > If link_conf_dereference_protected() or mt7996_vif_link()
> > or link_sta_dereference_protected() fail the code jumps to
> > the error_unlink label and returns ret which is uninitialised.
> >
> > Fix this by setting err before jumping to error_unlink.
> >
> > Fixes: c7e4fc362443 ("wifi: mt76: mt7996: Update mt7996_mcu_add_sta to MLO support")
> > Fixes: dd82a9e02c05 ("wifi: mt76: mt7996: Rely on mt7996_sta_link in sta_add/sta_remove callbacks")
> > Signed-off-by: Qasim Ijaz <qasdev00@gmail.com>
> > ---
> >  drivers/net/wireless/mediatek/mt76/mt7996/main.c | 12 +++++++++---
> >  1 file changed, 9 insertions(+), 3 deletions(-)
> >
> > diff --git a/drivers/net/wireless/mediatek/mt76/mt7996/main.c b/drivers/net/wireless/mediatek/mt76/mt7996/main.c
> > index 91c64e3a0860..78f7f1fc867e 100644
> > --- a/drivers/net/wireless/mediatek/mt76/mt7996/main.c
> > +++ b/drivers/net/wireless/mediatek/mt76/mt7996/main.c
> > @@ -998,16 +998,22 @@ mt7996_mac_sta_add_links(struct mt7996_dev *dev, struct ieee80211_vif *vif,
> >                         continue;
> >
> >                 link_conf = link_conf_dereference_protected(vif, link_id);
> > -               if (!link_conf)
> > +               if (!link_conf) {
> > +                       err = -EINVAL;
> >                         goto error_unlink;
> > +               }
> >
> >                 link = mt7996_vif_link(dev, vif, link_id);
> > -               if (!link)
> > +               if (!link) {
> > +                       err = -EINVAL;
> >                         goto error_unlink;
> > +               }
> >
> >                 link_sta = link_sta_dereference_protected(sta, link_id);
> > -               if (!link_sta)
> > +               if (!link_sta) {
> > +                       err = -EINVAL
> 
> You are missing a semicolon at the end of the line.
> 

Good spot! not sure how that happened but I will resend a v2 patch.

> To also do some bike shedding, you could initialize err (with 0 or
> -EINVAL), and then change the err return to return err ? : -EINVAL;.
> But your way has an explicit error code assigned to each failure, even
> if it is always the same. So I won't claim my suggestion is better.
> 

Thanks for the suggestion, I think this could work however I think my 
current approach is more a bit more readable, so I will leave it as is.

Thanks,
Qasim

> Best regards,
> Jonas