Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 30 Sep 1998 23:12:10 +0100 (BST)
From:      Andrew Gordon <arg@arg1.demon.co.uk>
To:        Hallam Oaks P/L list account <mlnn4@oaks.com.au>
Cc:        "freebsd-isdn@FreeBSD.ORG" <freebsd-isdn@FreeBSD.ORG>
Subject:   Re: HSCX query
Message-ID:  <Pine.BSF.3.96.980930224713.20359A-100000@server.arg.sj.co.uk>
In-Reply-To: <199809301807.EAA25933@mail.aussie.org>

next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, 1 Oct 1998, Hallam Oaks P/L list account wrote:
> 
> I'm working on a new device support for i4b. This device (a PCI ISDN card) 
> does _not_ have a HSCX in the normal manner of speaking. Most of the HSCX 
> functions are handled by the PCI ISAC, the rest (HDLC decoding) has to be 
> done in software.

I don't think you've really described your card accurately enough for us
to help.

The 'classic' Siemens-chipset cards (Teles, AVM etc) contain:

 - HSCX (SAB8525).  This is a dual-channel HDLC controller, used to
   receive the B channels only.  It demultiplexes the two channels
   under the control of timing strobes from the ISAC.
   It also has a non-HDLC mode, where it simply acts as a FIFO for
   bytes of raw B channel data (byte timing being supplied by the
   ISAC again).  Raw mode is used for voice applications, HDLC mode
   is used for IP (if they were ever implemented in I4B, V.110 would
   use raw mode and V.120 would use HDLC mode).

 - ISAC (PEB2085).  This handles the clock recovery and mux/demux
   for the ISDN S-bus, and also has an HDLC controller (very similar
   register-wise to one half of the HSCX) for receiving the D channel.

So, if your card really has no hardware support for HDLC on the B
channels, it is going to be _very_ inefficient for handling IP: although
the HDLC 'protocol' is quite simple, and CRC-16 can be implemented quite
efficiently by table look-up, doing the zero-bit insertion/deletion in
software would use a lot of CPU.   If there is hardware HDLC for the B
channels but not for the D channel, that would require just the same
development effort but would have a much less serious impact on
performance since the average data rate on the D channel is extremely low.


To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-isdn" in the body of the message



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.3.96.980930224713.20359A-100000>