On Thu, Oct 23, 2008 at 4:08 AM, Stephane Lapie <lapie@aozora-is.co.jp> wrote:
quoted text
> On Oct 22, 2008, at 4:49 PM, Alexey Suslikov wrote:
>
>> On Wed, Oct 22, 2008 at 9:15 AM, Stephane Lapie <lapie@aozora-is.co.jp>
>> wrote:
>>>
>>> On Oct 21, 2008, at 8:01 PM, Alexey Suslikov wrote:
>>>
>>> Stephane Lapie wrote:
>>>
>>> Hello,
>>> I am currently working with a Portwell NAR-5530 (network appliance
>>> running off Intel hardware, can use regular HDs or CF cards as boot
>>> device).
>>> We want to use this at work for network appliances, but end up
>>> bumping in the following problem : the kernel detects any device
>>> plugged to the controller (SATA or CF) as UltraDMA-5, even though the
>>> BIOS specifies otherwise clearly.
>>>
>>> You definitely should try to install recent snapshot on a CF
>>> card and boot GENERIC.MP with APM disabled since that
>>> box may mis-behave on interrupts without APCI.
>>> Issue "bsd.mp -c" in UKC than "disable apm" and "exit".
>>>
>>> Since the Intel controller got two modes (Enhanced, which successfully
>>> activates, but provokes issues with wd0 being recognized as UDMA5 ;
>>> Compatible, which can't allocate properly an IRQ, probably due to an ACPI
>>> issue), I tried booting with both modes :
>>> Here is the dmesg trace for a bsd.mp kernel (I'll build a full fledged MP
>>> based ramdisk once I confirm this works) in Enhanced mode :
>>
>> You need to sendbug about this (including your researches)
>> if you want this to be fixed in tree.
>
> FYI, I filed the bug under kernel/5961.
>
> On another note, simply detecting proper mode for the disk doesn't fix all
> of the issues ; it still seems there is a problem with the pciide driver, as
> I have confirmed the following error :
>
> wd0(pciide0:0:0): timeout
>        type: ata
>        c_bcount: 16384
>        c_skip: 0
> pciide0:0:0: bus-master DMA error: status=0x22
> pciide0:0:0: bus-master DMA error: missing interrupt, status=0x22

Check if that still true for GENERIC.MP from -current.

Alexey