[Bluez-devel] Kernel vs Userland AVDTP

[Bluez-devel] Kernel vs Userland AVDTP

[Fabien Chevalier]

Hi all,

Browsing the archives i've seen there has been some interest around the 
end of 2004 for an avdtp kernel implementation.

However recent trends seem to try to solve the issue in userland 
(Frederic's a2dpd or bluez's btaudiod for instance).

Is there a reason for this change of trends (except for "it's easier to 
have it done in user space for now" ;-) ) ?

Wouldn't it make more sense to have it in kernel due to multiplexing 
issues ?

Cheers,

Fabien

-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
_______________________________________________
Bluez-devel mailing list
Bluez-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Re: [Bluez-devel] Kernel vs Userland AVDTP

[Brad Midgley]

[-- Attachment #1.1: Type: text/plain, Size: 473 bytes --]

Fabien

What are the multiplexing issues? Something like two apps trying to use the
same remote device?

Having avdtp sockets offered by the os would enforce better separation
between the layers and it would change the landscape for
bluetooth.orgvalidation of the stack. It does feel sort of funny to
write code to
assemble packets in userspace... maybe once we have something other than
a2dp running on top of avdtp it will be even more compelling to change
things.

Brad

[-- Attachment #1.2: Type: text/html, Size: 533 bytes --]

[-- Attachment #2: Type: text/plain, Size: 286 bytes --]

-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

[-- Attachment #3: Type: text/plain, Size: 164 bytes --]

_______________________________________________
Bluez-devel mailing list
Bluez-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Re: [Bluez-devel] Kernel vs Userland AVDTP

[Fabien Chevalier]
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Re: [Bluez-devel] Kernel vs Userland AVDTP

[Marcel Holtmann]

Hi Fabien,

> Browsing the archives i've seen there has been some interest around the 
> end of 2004 for an avdtp kernel implementation.
> 
> However recent trends seem to try to solve the issue in userland 
> (Frederic's a2dpd or bluez's btaudiod for instance).
> 
> Is there a reason for this change of trends (except for "it's easier to 
> have it done in user space for now" ;-) ) ?
> 
> Wouldn't it make more sense to have it in kernel due to multiplexing 
> issues ?

at some point, yes. Currently it makes no sense to start with a kernel
version of AVDTP. Also Bluetooth 2.1 and 3.0 are gonna have high impact
on L2CAP and I wanna have that settled first before considering a kernel
based AVDTP, because it is not that simple. It would be as complex as
RFCOMM and so keeping it in userspace is better for now.

Regards

Marcel



-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
_______________________________________________
Bluez-devel mailing list
Bluez-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Re: [Bluez-devel] Kernel vs Userland AVDTP

[Fabien Chevalier]

Hi Marcel,

> at some point, yes. Currently it makes no sense to start with a kernel
> version of AVDTP. Also Bluetooth 2.1 and 3.0 are gonna have high impact
> on L2CAP and I wanna have that settled first before considering a kernel
> based AVDTP, because it is not that simple. It would be as complex as
> RFCOMM and so keeping it in userspace is better for now.
> 
I would tend to agree with you, specially on the fact that it would be 
quite complex to implement.
But what about a user space library? Do you thing that would make sense ?

Cheers,

Fabien

-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
_______________________________________________
Bluez-devel mailing list
Bluez-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-devel