I discovered that the card is an 8875.
Well this is the weird part.
When I put this card in my 2.4.19 box it works fine.
When I put it on the server that needs it, because of a stupid dongle,
all of a sudden it doesn't work. Granted the server is running kernel
2.4.3 but I added the sio_ite_8872_probe function and everything
else. So I know the function is called. However it doesn't recognize
it as a 8875. So I decided to force it to be a 8875 but it won't
recognize EPP or TRISTATE like it does in my 2.4.19 box.
I'm willing to give up on this card. However is there another PCI-X
card that will work on 2.4.3? Anybody have good reccomendations?
Thanks,
David
On 02:04AM Thursday, 04/24/03, Duncan Haldane wrote:
>
> On 23-Apr-2003 David Euresti wrote:
> >
> > Hello,
> >
> > I have an ITE 8872 PCI-X card that I'm putting into a compaq DL580.
> > Unfortunately when I try to add the module I get this in dmesg.
> >
> > initing superIO
> > sio_ite_8872_probe()
> > parport_pc: unknown ITE887x 15
> > parport_pc: please mail 'lspci -nvv' output to Rich.Liu@ite.com.tw
>
> Hi
>
> I looked into the ITE887x support a few moths ago, when I was trying to get
> the ECP mode on an ITE8875 PCI parport card working.
>
> (1) I managed to find Rich Liu: he no longer works at ITE, and says he can
> no longer support these cards in any way.
>
> (2) There is in fact extensive documentation available for these chips at
> ite: see http://www.ite.com.tw/productInfo/Download.html
>
> (3): The Linux ITE support in parport_pc.c only activates the EPP mode: from
> the ITE docs I managed to easily find how to activate ECP mode, but any attempt
> to use it crashes the system (which explains why Rich Liu only activated EPP).
> This occurred when interrupt use got activated -I suspect that the linux
> ECP support for PIO (non DMA) ECP in parport_pc.c is broken, rather than this
> being a problem with the ITE chip. The card works fine in EPP mode.
>
>
> The error you mention is at the beginning of the ITE support, where the type of
> card is identified by reading the contents of some registers. All this is
> quite well documented in the available ITE specs.
>
> You probably want to just give up on the card, but you might like to check out
> the specs to see if its just some small problem you can fix.
>
>
>
> >
> > I modified it to print out the type that it's switching on.
> >
> > Here is the relevant part for lspci -nvv
> > 0a:02.0 Class 0700: 1283:8872 (rev 01) (prog-if 02)
> > Control: I/O Mem BusMaster SpecCycle- MemWINV- VGASnoop- ParErr Step
> > ping- SERR FastB2B-
> > Status: Cap- 66Mhz- UDF- FastB2B ParErr- DEVSEL+AD0-medium
> > +AD4-TAbort- +ADw-TAbort - +ADw-MAbort- +AD4-SERR- +ADw-PERR-
> > Latency: 0
> > Interrupt: pin A routed to IRQ 24
> > Region 0: I/O ports at a400 +AFs-size+AD0-256+AF0-
> > Region 1: I/O ports at a800 +AFs-size+AD0-256+AF0-
> > Region 2: I/O ports at ac00 +AFs-size+AD0-256+AF0-
> > Region 3: I/O ports at b000 +AFs-size+AD0-256+AF0-
> > Region 4: I/O ports at b400 +AFs-size+AD0-256+AF0-
> > Region 5: I/O ports at b800 +AFs-size+AD0-256+AF0-
> >
> > I tried emailing Rich.Liu@ite.com.tw but it didn't work. neither did
> > richliu@poorman.org
> >
> > Can anyone here help?
> >
> > Thanks,
> > David Euresti
> >
> >
> >
> >
> >
> >
> > -- To unsubscribe, send mail to: linux-parport-request@torque.net --
> > -- with the single word "unsubscribe" in the body of the message. --
>
> ----------------------------------
> E-Mail: Duncan Haldane <f.duncan.m.haldane@worldnet.att.net>
> Date: 24-Apr-2003
> Time: 02:06:58
>
> This message was sent by XFMail
> ----------------------------------
-- To unsubscribe, send mail to: linux-parport-request@torque.net --
-- with the single word "unsubscribe" in the body of the message. --
This archive was generated by hypermail 2b29 : Thu Apr 24 2003 - 16:52:41 EDT