Hi Marcel, On Sun, Sep 15, 2013, Marcel Holtmann wrote: > >>> The convention of the L2CAP code is to return EFAULT when a CID in the > >>> request packet is invalid. This patch fixes the l2cap_config_req to use > >>> that error code instead of ENOENT. > >>> > >>> Signed-off-by: Johan Hedberg <johan.hedberg@xxxxxxxxx> > >>> --- > >>> net/bluetooth/l2cap_core.c | 2 +- > >>> 1 file changed, 1 insertion(+), 1 deletion(-) > >>> > >>> diff --git a/net/bluetooth/l2cap_core.c b/net/bluetooth/l2cap_core.c > >>> index 9ec1561f..d134501 100644 > >>> --- a/net/bluetooth/l2cap_core.c > >>> +++ b/net/bluetooth/l2cap_core.c > >>> @@ -3978,7 +3978,7 @@ static inline int l2cap_config_req(struct l2cap_conn *conn, > >>> > >>> chan = l2cap_get_chan_by_scid(conn, dcid); > >>> if (!chan) > >>> - return -ENOENT; > >>> + return -EFAULT; > >> > >> do we really want to use EFAULT. That is normally only used if a > >> memory copy from userspace to kernel or vice versa fails. > > > > EFAULT seemed to be the most common error code in this case, but now > > that I did some research it seems it's mostly from newish AMP related > > code. Would you be ok if I just switch all places to use ENOENT then? > > EFAULT is really for memory address errors. It is a pretty bad idea to use CID errors. > > I am fine with keep using ENOENT if we used that before. If that is > what the original code used, then lets keep using it. Otherwise we > better find a more descriptive error. If we ignore new AMP code there's only one "legacy" reference for this kind of situation which is l2cap_config_req. Since it has always used ENOENT for a failed CID lookup I'll follow that lead and use the same error for all places that need it. Johan -- To unsubscribe from this list: send the line "unsubscribe linux-bluetooth" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
