From: Izumi Tsutsui <tsutsui@ceres.dti.ne.jp>
Date: Mon, 12 Jan 2004 02:42:03 +0900

> >>Galileo Technology GT-64011 System Controller (miscellaneous memory, \
> >>revision 0x11) at ? dev 0 function 0 (intrswiz 0, intrpin 0x1, \
> >>i/o off, mem on, no quirks): Galileo GT-64111 System Controller, rev 1
> 
> BTW, could you try PCI siop (or some other devices which use mem space)
> on your Qube2? On my RaQ2, siop0 works with mem base==0x00102000
> with patch I posted:
> http://mail-index.netbsd.org/port-cobalt/2003/11/24/0002.html
> but I don't know how it works...

I don't have siop(4) card. I tried some cards...

    1. fxp(?) & tlp is hangup in the firmware. 
    2. vr is an error repeatedly.   X-<
       remove '#define VR_USEIOSPACE' at if_vr.c

PCI: bus 0, device 10, function 0: id=30431106: Vendor=1106, Product=3043
PCI: bus 0, device 10, function 0: Found dev 0x1106 0x3043 -- really probing.
PCI: bus 0, device 10, function 0: Register 0x10, I/O size 128
PCI: bus 0, device 10, function 0: MEM32 BAR 0x14 has size 80
PCI: bus 0, device 10, function 0: Register 0x14, memory size 128
PCI: bus 0, device 10, function 0: MEM32 BAR 0x18 is void
PCI: bus 0, device 10, function 0: MEM32 BAR 0x1c is void
PCI: bus 0, device 10, function 0: MEM32 BAR 0x20 is void
PCI: bus 0, device 10, function 0: MEM32 BAR 0x24 is void

PCI: bus 0, device 10, function 0: Putting 128 I/O bytes @ 0x10000080 (reg 10)
PCI: bus 0, device 10, function 0: Putting 128 MEM bytes @ 0x12000800 (reg 14)

PCI: bus 0, device 10, function 0: Configuring device.

vr0: interrupting at irq 9
vr0: reset never completed!
vr0: Ethernet address: 8c:8c:8c:8c:8c:8c


# ifconfig vr0 192.168.1.XXX
vr0: reset never completed!
vr0: transmit aborted
vr0: restarting
     .
     .
     .




    3. pdcide is OK. (Although not recognized last night...)

PCI: bus 0, device 10, function 0: id=4d38105a: Vendor=105a, Product=4d38
PCI: bus 0, device 10, function 0: Found dev 0x105a 0x4d38 -- really probing.
PCI: bus 0, device 10, function 0: Register 0x10, I/O size 8
PCI: bus 0, device 10, function 0: Register 0x14, I/O size 4
PCI: bus 0, device 10, function 0: Register 0x18, I/O size 8
PCI: bus 0, device 10, function 0: Register 0x1c, I/O size 4
PCI: bus 0, device 10, function 0: Register 0x20, I/O size 64
PCI: bus 0, device 10, function 0: MEM32 BAR 0x24 has size 20000
PCI: bus 0, device 10, function 0: Register 0x24, memory size 131072

PCI: bus 0, device 10, function 0: Putting 64 I/O bytes @ 0x10000100 (reg 20)
PCI: bus 0, device 10, function 0: Putting 8 I/O bytes @ 0x10000150 (reg 10)
PCI: bus 0, device 10, function 0: Putting 8 I/O bytes @ 0x10000158 (reg 18)
PCI: bus 0, device 10, function 0: Putting 4 I/O bytes @ 0x10000160 (reg 14)
PCI: bus 0, device 10, function 0: Putting 4 I/O bytes @ 0x10000164 (reg 1c)
PCI: bus 0, device 10, function 0: Putting 131072 MEM bytes @ 0x12000000 (reg 24)

PCI: bus 0, device 10, function 0: Configuring device.

pdcide0 at pci0 dev 10 function 0
pdcide0: Promise Ultra66/ATA Bus Master IDE Accelerator (rev. 0x01)
pdcide0: bus-master DMA support present
pdcide0: primary channel configured to native-PCI mode
pdcide0: using irq 9 for native-PCI interrupt




aa...
Ethernet Address cannot reading (8c:8c:8c:8c:8c:8c). '8c' is happened
while writing this patch, my misstake at range. The value was reverse
at mem & io.

fwohci0: OHCI 1.0, 8c:8c:8c:8c:8c:8c:8c:8c, 400Mb/s, 2048 max_rec, 8 ir_ctx, 8 it_ctx



I don't understand. X-<
--
kiyohara