patch-2.4.11-dontuse linux/net/irda/irlap_event.c
Next file: linux/net/irda/irlmp.c
Previous file: linux/net/irda/irlan/irlan_common.c
Back to the patch index
Back to the overall index
- Lines: 23
- Date:
Thu Oct 4 18:41:09 2001
- Orig file:
v2.4.10/linux/net/irda/irlap_event.c
- Orig date:
Sun Sep 23 11:41:02 2001
diff -u --recursive --new-file v2.4.10/linux/net/irda/irlap_event.c linux/net/irda/irlap_event.c
@@ -701,7 +701,21 @@
* We are allowed to send two frames, but this may increase
* the connect latency, so lets not do it for now.
*/
- /* What the hell is this ? - Jean II */
+ /* This is full of good intentions, but doesn't work in
+ * practice.
+ * After sending the first UA response, we switch the
+ * dongle to the negociated speed, which is usually
+ * different than 9600 kb/s.
+ * From there, there is two solutions :
+ * 1) The other end has received the first UA response :
+ * it will set up the connection, move to state LAP_NRM_P,
+ * and will ignore and drop the second UA response.
+ * Actually, it's even worse : the other side will almost
+ * immediately send a RR that will likely collide with the
+ * UA response (depending on negociated turnaround).
+ * 2) The other end has not received the first UA response,
+ * will stay at 9600 and will never see the second UA response.
+ * Jean II */
irlap_send_ua_response_frame(self, &self->qos_rx);
#endif
FUNET's LINUX-ADM group, linux-adm@nic.funet.fi
TCL-scripts by Sam Shen (who was at: slshen@lbl.gov)