On Tue, Jun 2, 2009 at 1:35 PM, Johannes Berg <johannes@xxxxxxxxxxxxxxxx> wrote: > On Tue, 2009-06-02 at 16:30 -0400, Luis R. Rodriguez wrote: >> Lets just check for the STA before its addition through >> sta_info_insert() and make this mandatory. It simpifies >> code. >> >> In our mac80211 cfg80211 callback for device addition we >> also can simplify the code by first checking for the STA >> before trying to add it and then checking for -EEXIST which >> we were not doing. If that actualy would happen we could >> end up potentially with a stale sta and the rate info was >> never updated. We should now be updating accordingly. >> >> Since we are making part of the API to check for the STA >> prior to addition lets warn in sta_info_insert() if the >> sta does already exist instead of passing an -EEXIST. >> >> Signed-off-by: Luis R. Rodriguez <lrodriguez@xxxxxxxxxxx> >> --- >> >> johill, this is what I meant. Still not the right approach? > > This may sound flippy, but you do need to understand RCU first. This is > simply not correct, even with rcu_read_lock() the sta info might be > unlinked from the list, your pointer will just point to something no > longer on the list. Thanks, yeah I see where I was making an incorrect assumption here. There still is a case here where userpsace can try to add a sta twice and the second parameters passed would be ignored. Luis -- To unsubscribe from this list: send the line "unsubscribe linux-wireless" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
