Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 5 Jan 2001 20:13:44 +0100 (CET)
From:      =?ISO-8859-1?Q?G=E9rard_Roudier?= <groudier@club-internet.fr>
To:        Matthew Jacob <mjacob@feral.com>
Cc:        Yoriaki FUJIMORI <fujimori@grafin.fujimori.cache.waseda.ac.jp>, freebsd-alpha@FreeBSD.ORG
Subject:   Re: bus_dmamap_load 
Message-ID:  <Pine.LNX.4.10.10101052002060.2363-100000@linux.local>
In-Reply-To: <Pine.LNX.4.21.0101041642370.16662-100000@zeppo.feral.com>

next in thread | previous in thread | raw e-mail | index | archive | help


On Thu, 4 Jan 2001, Matthew Jacob wrote:

> On Thu, 4 Jan 2001, [ISO-8859-1] G=E9rard Roudier wrote:
>=20
> >=20
> >=20
> > On Thu, 4 Jan 2001, Matthew Jacob wrote:
> >=20
> > > The trick now is to figure out who is passing a bogus value
> > > to bus_dmamap_load. The value of buf_len (0xfaae8000) looks
> > > suspicious.
> > >=20
> > > Since I hve essentially the same h/w (except for the symbios card) in=
 a PC164
> > > that runs 4.2 just fine, I would suspect the sym driver. But needs to=
 happen
> > > is some better debugging for this. I take it that you aren't really r=
unning on
> > > this h/w yet (i.e, you haven't been able to install)?
> >=20
> > Btw, you missed the `de' driver that also performs bus_dmamap_load() on
> > mbuf's as a possible candidate.
>=20
> That's why I said 'possibly'.... And because pretty much the sole differe=
nce
> between his PC164 and mine is the Symbios card.....

Hmmm... May-be `sym' (or the Symbios card) is related to the problem, but
suspecting it in a few seconds looked to me as relevant as suspecting the
moon to be the cause of night.;-)
There are probably much other differences between the _actual_ hardware
config + software config + use pattern of the 2 systems.

> > Would be interesting to shut the `de' board and to give a try with a
> > different network board long enough to see if it makes difference.
>=20
>=20
> No, no, no, since he's actually running with this, the smart thing to do =
is to
> put a panic to DDB to get a traceback, i.e.:
>=20
> options DDB
>=20
> and change the printf in alpha/busdma_machdep.c to a panic.

Ok. This will give some infos.

  G=E9rard.



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




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.LNX.4.10.10101052002060.2363-100000>