Subject: Re: [Bluez-users] MTU in l2test
From: Marcel Holtmann <marcel@holtmann.org>
To: BlueZ Mailing List <bluez-users@lists.sourceforge.net>
In-Reply-To: <2de4657605010303287a6eb6db@mail.gmail.com>
References: <2de4657604123007015c2c9461@mail.gmail.com>
	 <1104674721.8894.10.camel@pegasus>
	 <2de4657605010302474a44519e@mail.gmail.com>
	 <1104750031.8894.48.camel@pegasus>
	 <2de4657605010303287a6eb6db@mail.gmail.com>
Content-Type: text/plain
Message-Id: <1104752606.8894.51.camel@pegasus>
Mime-Version: 1.0
Sender: bluez-users-admin@lists.sourceforge.net
Errors-To: bluez-users-admin@lists.sourceforge.net
Reply-To: bluez-users@lists.sourceforge.net
Date: Mon, 03 Jan 2005 12:43:26 +0100

Hi Manjunath,

> > > I am using anycom-120 dongles and using the l2test utility.  I am also
> > > setting the incoming and outgoing mtu values (imtu and omtu variables)
> > > to the same value, say 65000 bytes. My 'datasize' is about 45000
> > > bytes. I get the error as mentioned above.
> > 
> > what is the output of "hciconfig -a" of the Anycom dongle. Check when
> > this error occurs. Do you see it only when calling the send()/write()
> > function? Maybe there is a buffer problem. What do you see with hcidump?
> > Is the MTU correctly negotiated?
> 
> hi marcel,
> 
> the "hciconfig -a" shows
> 
> hci0:   Type: USB
>         BD Address: 00:0B:0D:30:5D:EA ACL MTU: 120:20  SCO MTU: 64:0
>         UP RUNNING PSCAN ISCAN
>         RX bytes:114526 acl:28 sco:0 events:15460 errors:0
>         TX bytes:4857089 acl:39552 sco:0 commands:1400 errors:0
>         Features: 0xff 0xff 0x05 0x00 0x00 0x00 0x00 0x00
>         Packet type: DH5
>         Link policy:
>         Link mode: SLAVE ACCEPT
>         Name: ''
>         Class: 0x000000
>         Service Classes: Unspecified
>         Device Class: Miscellaneous,
>         HCI Ver: 1.1 (0x1) HCI Rev: 0x0 LMP Ver: 1.1 (0x1) LMP Subver: 0x757
>         Manufacturer: Silicon Wave (11)

leave the packet types to the default (all on).

> I get this error when i use the send() function to pump the data out.
> May be there is a problem with the HCI buffer. Is there some way to
> add a delay. The L2CAP connection is correctly established and the MTU
> is also correctly negotiated to the set value (65000) as seen in the
> hcidump on both sides.

The problem is maybe inside the l2cap_do_send() kernel function. I never
played with MTUs such big.

You should think about using RFCOMM, because this is a stream and so MTU
less.

Regards

Marcel


-------------------------------------------------------
The SF.Net email is sponsored by: Beat the post-holiday blues
Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek.
It's fun and FREE -- well, almost....http://www.thinkgeek.com/sfshirt
_______________________________________________
Bluez-users mailing list
Bluez-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-users