Hi Dag,
While doing testing today (2.4.0-test7, multilink), and I've
identified the following problems...
1) Colliding discoveries
It does happen. I've made a fix for sligthly offset one (see
previous patch), but when they collide, they always collide (clock are
too well synched), and we are blind and never discover the box on this
link.
Maybe we should introduce some random jitter in the time
between discoveries...
---------------------------------
22:30:18.690174 xid:cmd b6f5b132 > ffffffff S=6 s=0 (14)
22:30:18.690565 xid:cmd ffffffff < 18113e58 S=6 s=0 (14)
22:30:18.780087 xid:cmd b6f5b132 > ffffffff S=6 s=1 (14)
22:30:18.780446 xid:cmd ffffffff < 18113e58 S=6 s=1 (14)
22:30:18.870128 xid:cmd b6f5b132 > ffffffff S=6 s=2 (14)
22:30:18.870479 xid:cmd ffffffff < 18113e58 S=6 s=2 (14)
22:30:18.960132 xid:cmd b6f5b132 > ffffffff S=6 s=3 (14)
22:30:18.960480 xid:cmd ffffffff < 18113e58 S=6 s=3 (14)
22:30:19.050140 xid:cmd b6f5b132 > ffffffff S=6 s=4 (14)
22:30:19.050509 xid:cmd ffffffff < 18113e58 S=6 s=4 (14)
22:30:19.140083 xid:cmd b6f5b132 > ffffffff S=6 s=5 (14)
22:30:19.140434 xid:cmd ffffffff < 18113e58 S=6 s=5 (14)
22:30:19.230159 xid:cmd b6f5b132 > ffffffff S=6 s=* tanguy hint=4400 [ Computer LAN
Access ] (22)
22:30:19.240002 xid:cmd ffffffff < 18113e58 S=6 s=* lagaffe hint=4400 [ Computer LAN
Access ] (23)
---------------------------------
2) Various warning messages
Probably harmless, but just for you to know...
------------------------
irlmp_state_dtr(), Unknown event LM_LAP_CONNECT_CONFIRM
irlmp_state_setup(), Unknown event LM_LAP_CONNECT_CONFIRM
irlmp_state_u_connect(), Unknown event LM_LAP_DISCOVERY_CONFIRM
------------------------
3) Usage count weirdness
Something is increasing the usage count of the irda.o module,
making it impossible to remove without a reboot. I did inspect the
code and did some test, I can't see why.
---------------------------
actisys 1692 2 (autoclean)
irtty 7588 4 (autoclean)
irda 145761 8 (autoclean) [actisys irtty]
---------------------------
4) ifconfig weirdness
Look at HWaddr for second dongle. Maybe ifconfig bug.
Also, look at Tx packet count ;-)
---------------------------------------------
irda0 Link encap:IrLAP HWaddr 1b:0f:88:cb
UP RUNNING NOARP MTU:2048 Metric:1
RX packets:4139 errors:0 dropped:0 overruns:0 frame:0
TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:8
irda1 Link encap:IrLAP HWaddr 1b:0f:88:cbf0:3a:77:db
UP RUNNING NOARP MTU:2048 Metric:1
RX packets:3640 errors:0 dropped:0 overruns:0 frame:0
TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:8
---------------------------------------------
5) Socket weirdness
The latest e-Squirt program on the page below fail to connect
to the incomming call the second time (but work fine first time).
http://www.hpl.hp.com/personal/Jean_Tourrilhes/IrDA/squirt.html
In sumary, nothing dramatic, but just keep that in your archive...
Jean
_______________________________________________
Linux-IrDA mailing list - [EMAIL PROTECTED]
http://www4.pasta.cs.UiT.No/mailman/listinfo/linux-irda
