Date: Fri, 02 Jan 2009 17:17:43 +0100 From: Lokadamus <Lokadamus@gmx.de> To: Raphael Becker <rabe@uugrn.org> Cc: freebsd-hardware@freebsd.org Subject: Re: UDMA vs. VIA 82C686B Message-ID: <495E3E27.8020303@gmx.de> In-Reply-To: <20090102021129.GA7331@ma.sigsys.de> References: <20090102021129.GA7331@ma.sigsys.de>
next in thread | previous in thread | raw e-mail | index | archive | help
Raphael Becker wrote:
> Hi all,
>
> I have had some problems with a VIA 82C686B UDMA100 controller with an
> attached 2.5" Harddisk (Samsung-foo, 40GB) running 7.1-RELEASE (GENERIC):
>
> UDMA-access is not possible even tough the HDD is able to do (in my
> Laptop) and the controller claims to support UDMA100.
>
> I tried a CF-Card with the same hardware running 6.2 or 6.3-RELEASE with
> the same effect: some bad g_vfs-errors.
>
> I got around this problem by adding (at least one of) the lines to
> /boot/loader.conf:
>
> hw.ata.atapi_dma=0
> hw.ata.ata_dma=0
>
>
> Is this a known issue?
> Is there a workaround/patch for UDMA on this VIA-chipset?
>
> Do you need more details on it?
> It's kind of difficult to get the exact errors, because hw.ata.* seems
> to be a read-only ("boot-only") sysctl and I have limited (serial)
> access to the box. In case of hw.ata.ata_dma=1 I'm not able to unset
> this value in /boot/loader.conf because I cannot write to the disk
> because of the udma-errors then.
>
> Regards
> Raphael
>
>
Have you look for a bios update for the mainboard?
Chipset 82C686B was buggy, known as 686B-Bug.
Regards
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?495E3E27.8020303>