Re: "No such device" with PCI card

[Jan-Benedict Glaw <jbglaw@lug-owl.de>]

[-- Attachment #1: Type: text/plain, Size: 5457 bytes --]

On Fri, 2004-06-18 10:41:55 +0800, jospehchan <jospehchan@yahoo.com.tw>
wrote in message <20040618024155.35970.qmail@web16605.mail.tpe.yahoo.com>:
> Hi Jan-Benedict,
>  Thanks. Please refer to the follownig replies. 

(By the way, think about changing you email client's configuration...)

> - What kind of MIPS system do you use *exactly*? What
> board? Which
>   kernel version? From where did you get your sources.
> 
> >>>The MIPS system is R3000 and uses an ADI Media
> Adapter MB.
> The kernel is 2.4.16 from the vendor and plus an USB
> patch which backported from kernel 2.4.26.

First off, I didn't easily find information for that board...

Then, porting direction was wrong. You want to diff out vendor's changes
ontop vanilla 2.4.16 (probably they've started off the mvista or the
linux-mips.org kernel) and port *those* to current 2.4.x (of same
vendor, preferring linux-mips.org ...). If they only added drivers and
bootup-code for that board, just port that over to 2.6.x.

Sounds like the vendor did make linux waddle on that board and never
cared for it again :(

> - A USB2.0 card is IMHO driven by the ehci driver, but
> I may be wrong.
>   I'm not exactly a regular USB user...
> 
> >>>Yes, you're right. But the USB2.0 card also can be
> driven by usb-uhci if ehci-hcd is not loaded.
> In the my problem, I can load the usbcore, but both of
> usb-uhci and ehci-hcd can not be loaded. 

I'd guess usb-core has nothing else to do than to accept loaded host and
client drivers, so it should just load and do nothing. I guess they just
broke the whole PCI interface in some way or another (weren't there
general MIPS bugs at that time in 2.4.x? Even with endianess? It's so
long ago...)

> - Do you have output of "lspci", "lspci -v", "lspci
> -n", "lspci -vn" and
>   "lspci -nxxx" at your hand, once from your i386 test
> machine, once
>   from the MIPS board? Right, those commands mostly
> give the same
>   output, but each style eases reading for specific
> values:)
> 
> >>> MIPS
> # lspci
> 00:00.0 Class 0c03: 1106:3038 (rev 61)

Only *one* PCI device? I'm shocked...

	1106:	VIA
	3038:	USB

> # lspci -v
> 00:00.0 Class 0c03: 1106:3038 (rev 61)
>         Subsystem: 1106:3038
>         Flags: bus master, medium devsel, latency 22, IRQ 4
>         I/O ports at <ignored>
>         Capabilities: [80] Power Management version 2

Seems the device didn't get any I/O assigned. Of course, that won't
fly. Depending on what's expected, either Linux' PCI core should assign
I/O for devices, or the board's firmware.

> # lspci -n

> >>> i386 (RH7.2, kernel 2.4.16 plus USB patch from kernel 2.4.26)

Dito, quite outdated:)

> #lspci 
> 00:14.0 USB Controller: VIA Technologies, Inc. UHCI USB (rev 61)
> 00:14.1 USB Controller: VIA Technologies, Inc. UHCI USB (rev 61)
> 00:14.2 USB Controller: VIA Technologies, Inc.: Unknown device 3104 (rev 62)

So on the MIPS board, there's only *one* single device, but on your
i386 machine, this board registers as three separate subdevices? Sounds
there's something seriously broken in the MIPS PCI code as of 2.4.16...

But *that* doesn't make me wonder:)


> #lspci -v
> 00:14.0 USB Controller: VIA Technologies, Inc. UHCI
> USB (rev 61) (prog-if 00 [UHCI])
> 	Subsystem: VIA Technologies, Inc. UHCI USB
> 	Flags: bus master, medium devsel, latency 32, IRQ 11
> 	I/O ports at e400 [size=32]
> 	Capabilities: [80] Power Management version 2
> 
> 00:14.1 USB Controller: VIA Technologies, Inc. UHCI
> USB (rev 61) (prog-if 00 [UHCI])
> 	Subsystem: VIA Technologies, Inc. UHCI USB
> 	Flags: bus master, medium devsel, latency 32, IRQ 5
> 	I/O ports at e800 [size=32]
> 	Capabilities: [80] Power Management version 2
> 
> 00:14.2 USB Controller: VIA Technologies, Inc.:
> Unknown device 3104 (rev 62) (prog-if 20)
> 	Subsystem: VIA Technologies, Inc.: Unknown device 3104
> 	Flags: bus master, medium devsel, latency 32, IRQ 5
> 	Memory at ee003000 (32-bit, non-prefetchable) [size=256]
> 	Capabilities: [80] Power Management version 2

Also, if you're asked for output, please don't cut it down to what you
think is useful or related. For sure, your i386 PC as well as your MIPS
box *does* indeed have more PCI devices than only the USB card.

> #lspci -vn
> 00:14.0 Class 0c03: 1106:3038 (rev 61)
> 	Subsystem: 1106:3038
> 	Flags: bus master, medium devsel, latency 32, IRQ 11
> 	I/O ports at e400 [size=32]
> 	Capabilities: [80] Power Management version 2

See? On your PeeCee, it got I/O resources assigned.

> 
> - Does your MOPS board have working on-board PCI
> devices? These don't
>   neccessarily have a PCI plug as you know them from
> add-on cards,
>   because they're directly built into the chipset. For
> instance, does
>   your board have onboard IDE interfaces?
>   
>   >>>No, the PCI device can not work, such as (Realtek
> 8139 LAN card, Philips and VIA USB 2.0 card)
>   But there is a mini-PCI device seems workable,
> because it's driver can be loaded.

Does it have any PCI attached devices (modulo the USB card)?

MfG, JBG

-- 
   Jan-Benedict Glaw       jbglaw@lug-owl.de    . +49-172-7608481
   "Eine Freie Meinung in  einem Freien Kopf    | Gegen Zensur | Gegen Krieg
    fuer einen Freien Staat voll Freier Bürger" | im Internet! |   im Irak!
   ret = do_actions((curr | FREE_SPEECH) & ~(NEW_COPYRIGHT_LAW | DRM | TCPA));

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]