Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 17 Aug 2000 01:38:50 +0100 (BST)
From:      Gavan Fantom <gavan@coolfactor.org>
To:        Gary Jennejohn <garyj@muc.de>
Cc:        freebsd-isdn@FreeBSD.ORG
Subject:   Re: established sppp connection confused by incoming call 
Message-ID:  <Pine.NEB.4.10.10008170131400.453-100000@paper.durnsford.net>
In-Reply-To: <200008161822.UAA38424@peedub.muc.de>

next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, 16 Aug 2000, Gary Jennejohn wrote:

> Gavan Fantom writes:
> > On Tue, 1 Aug 2000, Gary Jennejohn wrote:
> > 
> > > Turn on ALL kernel trace using isdndebug and show us the results.
> > 
> >
> [trace mostly deleted]
> 
> > Aug 12 14:15:51 frog /netbsd: i4b-L1-avma1pp_hscx_intr: RAWHDLC rx 
> buffer overflow in RPF, in_len=2048
> > Aug 12 14:15:51 frog /netbsd: i4b-L4-i4b_idle_check: 966086151: 
> incoming-call, activity, last_active=966086151, max_idle=25
> > Aug 12 14:15:51 frog /netbsd: i4b-L4-i4btelread: i4btel0, mbuf (2048 
> bytes), uiomove 0!
> > Aug 12 14:15:51 frog /netbsd: i4b-L1-avma1pp_hscx_intr: RAWHDLC rx 
> buffer overflow in RPF, in_len=2048
> 
> This fragment of trace is very weird. It indicates that the receive buffer
> for the sppp connection (RAWHDLC) contains > 2048 bytes ==> no RME
> interrupt was ever received ! It almost looks like the interrupt status
> is being screwed up by the simultaneously running telephone call. I
> can't explain it :(

Weird.

> It might be good to reset the receiver in this case, rather than just
> resetting the count and data pointer. Try this patch and see if it
> helps.

I will, but prolly not for a few days; I'm a bit overwhelmed by work this
week.

> The patch looks OK, except the above line should read
> 		chan->state = HSCX_IDLE; (NO `&')
> but this probably doesn't really affect the behavior when _both_ channels
> are active (chan->state is set equal to HSCX_IDLE a few lines later).
> It will, however, totally screw up the deactivate logic later on when
> both channels _do_ go inactive !

Oh, the irony!

Last night I had to reboot poor little frog after it couldn't assign a
free channel when dialling out. Incoming voice calls were met with a
continuous tone from the exchange, indicating that something had gone
badly wrong.

Might it have been that which caused frog's misery last night?

[this has never happened before while running this version of i4b]

-- 
Gillette - the best a man can forget



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.NEB.4.10.10008170131400.453-100000>