Skip to contentSkip to conversation

wireless: Add ifname to STOP_AP event.

PATCH In review [email protected] 2 participants 2 messages Jun 10, 2015 – Jun 17, 2015 802.11 (including cfg80211/NL80211)

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 wireless: Add ifname to STOP_AP event.

From: Ben Greear <[email protected]>

In case user-space wants to know the name of the device, this will save a name lookup by index, and in case where kernel manages to fully delete the interface before user-space can determine the name, this will ensure user-space has proper name.

---
 net/wireless/nl80211.c | 1 +
 1 file changed, 1 insertion(+)
net/wireless/nl80211.c +1 -0
index 736c474..ba73def 100644
@@ -12992,6 +12992,7 @@ void nl80211_send_ap_stopped(struct wireless_dev *wdev)
 		goto out;
 
 	if (nla_put_u32(msg, NL80211_ATTR_WIPHY, rdev->wiphy_idx) ||
+	    nla_put_string(msg, NL80211_ATTR_IFNAME, wdev->netdev->name) ||
 	    nla_put_u32(msg, NL80211_ATTR_IFINDEX, wdev->netdev->ifindex) ||
 	    nla_put_u64(msg, NL80211_ATTR_WDEV, wdev_id(wdev)))
 		goto out;
Signature · 1 line
1.7.11.7
maintainer
#1 lore raw
Quoted 7 lines · On Wed, 2015-06-10 at 15:17 -0700, [email protected] wrote:

From: Ben Greear <[email protected]>

In case user-space wants to know the name of the device, this will save a name lookup by index, and in case where kernel manages to fully delete the interface before user-space can determine the name, this will ensure user-space has proper name.

None of this reasoning is specific to the STOP_AP event, and I'd hate us adding it one by one...

I still think there isn't all that much reason to have it (if userspace really needs the names so badly it could also subscribe to the rtnetlink and keep a table of index/names!), but I could get behind a refactoring of all these places (I count at least 10 events that include the wdev id) to combine the four lines in the patch below into a single helper.

johannes