Dag wrote :
> Hi,
>
> You should try Linux-2.4.2-ac20. I'm using the ACTiSYS dongle
> without any problems.
>
> -- Dag
2.4.2-ac20 contains an older version of the driver that
doesn't handle Tx timeout as good as mine. Let first try to understand
what's happening before asking the poor guy to try all version of the
driver in existence...
By the way, there are blatant errors in the 2.4.2-ac20 driver :
------------------------------------------
+/*
+ * Maximum number of URB on the Rx and Tx path, a number larger than 1
+ * is required for handling back-to-back (brickwalled) frames
+ */
+#define IU_MAX_RX_URBS 1 /* Most devices can only handle one */
------------------------------------------
This is wrong, it's possible to support back-to-back
(brickwalled) frames in the driver with only one receive URB. The
dongle has its own frame buffer so will wait for us.
The second comment is also wrong. The number of Rx URB has
nothing to do with the USB device, it is only the way we want to
interact with the USB controller and control latency.
Remember : USB is a serial bus, not a memory mapped interface,
so the device has no visibility on what's happening on the host.
Back to the topic...
> On Thu, 15 Mar 2001 17:32:07 -0600, Syed, Hassan (ETD) wrote:
> > Hi,
> >
> > I can't get the Actisys IR-2000U working and I was hoping that
> > someone will be able to help me out.
> >
> > I got the irda-usb.c and .h files which Jean Tourrilhes posted a
> > few days ago. I built the driver. I am running Red Hat 7.1 (Kernel 2.4.2)
> >
> > modprobe usb-uhci
> > insmod irda-usb.o
Should be "modprobe irda-usb".
> > irattach irda0 -s 1
> >
> > /var/log/message contains the following info:
> >
> > Mar 15 16:05:02 etd-e500 kernel: usb.c: registered new driver irda-usb
> > Mar 15 16:05:02 etd-e500 kernel: IRDA-USB found at address 3
> > Mar 15 16:05:02 etd-e500 kernel: usb-uhci.c: interrupt, status 3, frame# 187
> > Mar 15 16:05:02 etd-e500 kernel: usb-uhci.c: interrupt, status 2, frame# 193
> > Mar 15 16:05:02 etd-e500 kernel: usb-uhci.c: interrupt, status 3, frame# 194
> > Mar 15 16:05:02 etd-e500 kernel: IrDA: Registered device irda0
> > Mar 15 16:05:02 etd-e500 kernel: USB IrDA support registered
> > Mar 15 16:05:26 etd-e500 irattach: executing: '/sbin/modprobe irda0'
> > Mar 15 16:05:27 etd-e500 irattach: executing: 'echo 1 >
> > /proc/sys/net/irda/discovery'
> > Mar 15 16:05:27 etd-e500 irattach: Starting device irda0
> > Mar 15 16:05:27 etd-e500 irattach: executing: 'echo etd-e500 >
> > /proc/sys/net/irda/devname'
> > Mar 15 16:05:30 etd-e500 kernel: NETDEV WATCHDOG: irda0: transmit timed out
> > Mar 15 16:06:01 etd-e500 last message repeated 279 times
This is a bit short. Obviously, your distribution doesn't save
all the messages in the log. Please use "dmesg" to get also debug
messages that will tell an interesting story.
Note : I have a few broken dongle that produce exactly this
behaviour (that people returned to me because not working OK under
Windows). I looks like a hardware problem to me. But let's check
first...
> > Side note:
> > I had to make a change in the irda-usb.c code to make it compile:
> > irlap_open does not take 3 params; it only takes 2
Yes. This is related to the hw_name patch, which is not
required for you (it's just a nice addition). You did the right thing.
> > Thanks for your help,
> > Hassan
Good luck...
Jean
_______________________________________________
Linux-IrDA mailing list - [EMAIL PROTECTED]
http://www.pasta.cs.UiT.No/mailman/listinfo/linux-irda
