| Message ID | c2fe73b04dac0827adf135e669ed441f9ed1d13c.1518126991.git.lorenzo.bianconi@redhat.com (mailing list archive) |
|---|---|
| State | Accepted |
| Commit | 032a552e8dc9d2147e34157bb3b0b23e8f418e43 |
| Delegated to: | Kalle Valo |
| Headers | show |
On Thu, 8 Feb 2018 23:08:09 +0100, Lorenzo Bianconi wrote: > If mac80211 adds a vif with a different mac address respect to > the eeprom one, the device will not be able to connect to the ap > since the hw address has not been updated. > Fix the issue updating hw mac address in mt7601u_add_interface routine > > BugLink: https://bugzilla.redhat.com/show_bug.cgi?id=1516935 > Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi@redhat.com> > --- > drivers/net/wireless/mediatek/mt7601u/main.c | 3 +++ > 1 file changed, 3 insertions(+) > > diff --git a/drivers/net/wireless/mediatek/mt7601u/main.c b/drivers/net/wireless/mediatek/mt7601u/main.c > index 43ebd460ba86..3c9ea40d9584 100644 > --- a/drivers/net/wireless/mediatek/mt7601u/main.c > +++ b/drivers/net/wireless/mediatek/mt7601u/main.c > @@ -64,6 +64,9 @@ static int mt7601u_add_interface(struct ieee80211_hw *hw, > */ > mvif->idx = idx; > > + if (!ether_addr_equal(dev->macaddr, vif->addr)) > + mt7601u_set_macaddr(dev, vif->addr); > + > if (dev->wcid_mask[wcid / BITS_PER_LONG] & BIT(wcid % BITS_PER_LONG)) > return -ENOSPC; > dev->wcid_mask[wcid / BITS_PER_LONG] |= BIT(wcid % BITS_PER_LONG); Sorry, my recollection of mac80211 code is waning, but can't we have more than one vif as long as they are on the same channel?
On Feb 08, Jakub Kicinski wrote: > On Thu, 8 Feb 2018 23:08:09 +0100, Lorenzo Bianconi wrote: > > If mac80211 adds a vif with a different mac address respect to > > the eeprom one, the device will not be able to connect to the ap > > since the hw address has not been updated. > > Fix the issue updating hw mac address in mt7601u_add_interface routine > > > > BugLink: https://bugzilla.redhat.com/show_bug.cgi?id=1516935 > > Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi@redhat.com> > > --- > > drivers/net/wireless/mediatek/mt7601u/main.c | 3 +++ > > 1 file changed, 3 insertions(+) > > > > diff --git a/drivers/net/wireless/mediatek/mt7601u/main.c b/drivers/net/wireless/mediatek/mt7601u/main.c > > index 43ebd460ba86..3c9ea40d9584 100644 > > --- a/drivers/net/wireless/mediatek/mt7601u/main.c > > +++ b/drivers/net/wireless/mediatek/mt7601u/main.c > > @@ -64,6 +64,9 @@ static int mt7601u_add_interface(struct ieee80211_hw *hw, > > */ > > mvif->idx = idx; > > > > + if (!ether_addr_equal(dev->macaddr, vif->addr)) > > + mt7601u_set_macaddr(dev, vif->addr); > > + > > if (dev->wcid_mask[wcid / BITS_PER_LONG] & BIT(wcid % BITS_PER_LONG)) > > return -ENOSPC; > > dev->wcid_mask[wcid / BITS_PER_LONG] |= BIT(wcid % BITS_PER_LONG); > > Sorry, my recollection of mac80211 code is waning, but can't we have > more than one vif as long as they are on the same channel? Hi Jakub, yep, you are right, but according to my understanding (please correct me if it is wrong) current implementation supports just one interface in sta mode (i.e. mvif->idx is always 0) so my patchset just fixes the issue highlighted in the bugzilla since I do not know if the hw supports multiple concurrent vifs in client mode. If so, I can extend the support to multiple client vifs if it is a way to properly configure the rx filters to allow reception from multiple mac addresses. Regards, Lorenzo
On Fri, 9 Feb 2018 10:49:21 +0100, Lorenzo Bianconi wrote: > On Feb 08, Jakub Kicinski wrote: > > On Thu, 8 Feb 2018 23:08:09 +0100, Lorenzo Bianconi wrote: > > > If mac80211 adds a vif with a different mac address respect to > > > the eeprom one, the device will not be able to connect to the ap > > > since the hw address has not been updated. > > > Fix the issue updating hw mac address in mt7601u_add_interface routine > > > > > > BugLink: https://bugzilla.redhat.com/show_bug.cgi?id=1516935 > > > Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi@redhat.com> > > > --- > > > drivers/net/wireless/mediatek/mt7601u/main.c | 3 +++ > > > 1 file changed, 3 insertions(+) > > > > > > diff --git a/drivers/net/wireless/mediatek/mt7601u/main.c b/drivers/net/wireless/mediatek/mt7601u/main.c > > > index 43ebd460ba86..3c9ea40d9584 100644 > > > --- a/drivers/net/wireless/mediatek/mt7601u/main.c > > > +++ b/drivers/net/wireless/mediatek/mt7601u/main.c > > > @@ -64,6 +64,9 @@ static int mt7601u_add_interface(struct ieee80211_hw *hw, > > > */ > > > mvif->idx = idx; > > > > > > + if (!ether_addr_equal(dev->macaddr, vif->addr)) > > > + mt7601u_set_macaddr(dev, vif->addr); > > > + > > > if (dev->wcid_mask[wcid / BITS_PER_LONG] & BIT(wcid % BITS_PER_LONG)) > > > return -ENOSPC; > > > dev->wcid_mask[wcid / BITS_PER_LONG] |= BIT(wcid % BITS_PER_LONG); > > > > Sorry, my recollection of mac80211 code is waning, but can't we have > > more than one vif as long as they are on the same channel? > > Hi Jakub, > > yep, you are right, but according to my understanding (please correct me if > it is wrong) current implementation supports just one interface in sta mode > (i.e. mvif->idx is always 0) so my patchset just fixes the issue highlighted > in the bugzilla since I do not know if the hw supports multiple concurrent vifs > in client mode. If so, I can extend the support to multiple client vifs if it is > a way to properly configure the rx filters to allow reception from multiple mac > addresses. Ah, you're probably right, perhaps we would have one vif but multiple stas.
diff --git a/drivers/net/wireless/mediatek/mt7601u/main.c b/drivers/net/wireless/mediatek/mt7601u/main.c index 43ebd460ba86..3c9ea40d9584 100644 --- a/drivers/net/wireless/mediatek/mt7601u/main.c +++ b/drivers/net/wireless/mediatek/mt7601u/main.c @@ -64,6 +64,9 @@ static int mt7601u_add_interface(struct ieee80211_hw *hw, */ mvif->idx = idx; + if (!ether_addr_equal(dev->macaddr, vif->addr)) + mt7601u_set_macaddr(dev, vif->addr); + if (dev->wcid_mask[wcid / BITS_PER_LONG] & BIT(wcid % BITS_PER_LONG)) return -ENOSPC; dev->wcid_mask[wcid / BITS_PER_LONG] |= BIT(wcid % BITS_PER_LONG);
If mac80211 adds a vif with a different mac address respect to the eeprom one, the device will not be able to connect to the ap since the hw address has not been updated. Fix the issue updating hw mac address in mt7601u_add_interface routine BugLink: https://bugzilla.redhat.com/show_bug.cgi?id=1516935 Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi@redhat.com> --- drivers/net/wireless/mediatek/mt7601u/main.c | 3 +++ 1 file changed, 3 insertions(+)