On Monday 05 January 2009, Bob Copeland wrote: > On Sun, Jan 4, 2009 at 9:18 PM, Geoffrey McRae <ge...@rabidhost.com> wrote: > > Isn't that patch slightly wrong anyway? > > > > If the length of the ssid is 0 why check its first byte? Without looking > > at the surrounding code, if that buffer is not allocated and we check > > the bytes value, it could cause a seg fault. > > As it happens, ssid is a statically allocated array, part of struct > ieee80211_bss. But I agree it should be written differently if it > is to go in.
The patch for an empty hidden SSID is already in wireless-testing (commit bae888b3ea89097dcb8b28573d8bd6fa1147bc31) but it's slightly different from the one I had proposed in the thread recalled by Bob. Basically a SSID is considered hidden when its string is composed only of '\0'. The hidden SSID is then accepted only when its length is 0 or has the same length of the set SSID for the association. Regards, Fabio _______________________________________________ ath5k-devel mailing list ath5k-devel@lists.ath5k.org https://lists.ath5k.org/mailman/listinfo/ath5k-devel