PCMCIA issues ---> cs: unable to map card memory 0x1000!
Mathieu Bertrand
Bertrand at tcl.ite.mee.com
Mon Feb 14 10:44:26 EST 2005
Hi,
So the "unable to map card memory 0x1000" error is ok.
Now, I try to test both an IDE compact flash card in socket 0 and a 3com LAN
10/100 (3c574_cs) card in socket 1.
Calling cardmgr, I have the following :
# cardmgr -v
cardmgr[3063]: cannot access /lib/modules/2.6.10-rc2: No such file or
directory
cardmgr[3063]: watching 2 sockets
# Dec 31 17:04:24 cardmgr[3068]: starting, version is 3.2.7
Dcs: memory probe 0xc0000000-0xc1ffffff: excluding 0xc0000000-0xc1ffffff
cs: memory probe 0xc2000000-0xc3ffffff: excluding 0xc2000000-0xc3ffffff
cs: memory probe 0xc4000000-0xc4ffffff: excluding 0xc4000000-0xc4ffffff
cs: memory probe 0xc5000000-0xc5ffffff: excluding 0xc5000000-0xc5ffffff
cs: memory probe 0xc6000000-0xc9ffffff:<7>exca_readb: 8d2da000 0006 00
eDc 31 17:04:24 cardmgr[3068]: excluding 0xc6000000-0xc83fffff<7>exca_readb:
8d2da000 0006 01
initializing socket 0
0xc9000000-0xc9ffffff
Dec 31 17:04:24 cardmgr[3068]: socket 0: ATA/IDE Fixed Disk
Dec 31 17:04:24 cardmgr[3068]: product info: "TOSHIBA
Tkmalloc(sizeof(*stk), GFP_KERNEL); ok
HTHNCF256
ide-cs: GetNextTuple: No more items
MBA ", "", "", ""
Dec 31 17:04:24 cardmgr[3068]: manfid: 0x0098, 0x0000 function: 4 (fixed
disk)
Dec 31 17:04:24 cardmgr[3068]: get dev info on socket 0 failed: No such
device
Dec 31 17:04:24 cardmgr[3068]: initializing socket 1
Dec 31 17:04:24 cardmgr[3068]: socket 1: 3Com 572/574 Fast Ethernet
Dec 31 17:04:24 cardmgr[3068]: product info: "3Com", "3c574_attach()
Me3c574_cs: RequestIO: Resource in use
gahertz 574B", "B", "001"
Dec 31 17:04:24 cardmgr[3068]: manfid: 0x0101, 0x0574 function: 6
(network)
Dec 31 17:04:25 cardmgr[3068]: get dev info on socket 1 failed: Resource
temporarily unavailable
So with the ide-cs driver, I have a "ide-cs: GetNextTuple: No more items"
and "get dev info on socket 0 failed: No such device"
With the 3c574_cs driver, I have a "RequestIO: Resource in use" and a "get
dev info on socket 1 failed: Resource temporarily unavailable"
An idea ?
I'm not sure concerning my IO space configuration :
# lspci -v
01:0f.0 Class 0604: 104c:ac28 (rev 02)
Flags: bus master, medium devsel, latency 64
Bus: primary=01, secondary=02, subordinate=0a, sec-latency=0
I/O behind bridge: 00000000-00003fff
Memory behind bridge: c4000000-c8ffffff
Prefetchable memory behind bridge: 00000000c0000000-00000000c3f00000
Capabilities: [dc] Power Management version 2
Capabilities: [e4] #06 [0000]
02:00.0 Class 0607: 104c:ac55 (rev 01)
Flags: bus master, medium devsel, latency 168, IRQ 361
Memory at c8000000 (32-bit, non-prefetchable) [size=4K]
Bus: primary=02, secondary=03, subordinate=06, sec-latency=176
Memory window 0: c0000000-c1fff000 (prefetchable)
Memory window 1: c4000000-c5fff000
I/O window 0: 00000000-00000fff
I/O window 1: 00001000-00001fff
16-bit legacy interface ports at 0001
02:00.1 Class 0607: 104c:ac55 (rev 01)
Flags: bus master, medium devsel, latency 168, IRQ 361
Memory at c8001000 (32-bit, non-prefetchable) [size=4K]
Bus: primary=02, secondary=07, subordinate=0a, sec-latency=176
Memory window 0: c2000000-c3fff000 (prefetchable)
Memory window 1: c6000000-c7fff000
I/O window 0: 00002000-00002fff
I/O window 1: 00003000-00003fff
16-bit legacy interface ports at 0001
and
# cat /proc/ioports
e0000000-e7ffffff : MV64340 PCI1 IO MEM
e0000000-e0003fff : PCI Bus #02
e0000000-e0000fff : PCI CardBus #03
e0001000-e0001fff : PCI CardBus #03
e0002000-e0002fff : PCI CardBus #07
e0003000-e0003fff : PCI CardBus #07
e8000000-efffffff : MV64340 PCI0 IO MEM
Here is my config in config.opts :
include port 0x0000-0x0fff, port 0x1000-0x1fff
include port 0x2000-0x2fff, port 0x3000-0x3fff
include port 0x4000-0x4fff, port 0x5000-0x5fff
include port 0x6000-0x7fff, port 0x8000-0x9fff
include port 0xa000-0xcfff, port 0xd000-0xffff
So from the MIPS address space, the IO space is located at 0xE0000000 and is
then remapped physically to the PCI address 0.
Not sure it works...
In fact, can I use an another range than 0-0xffff for the IO space ? It
seems that we can't use IO addresses above 0xffff in config.opts.
Thanks for your help.
Regards.
Mathieu BERTRAND
MITSUBISHI ITE TCL, France
-----Original Message-----
From: Dominik Brodowski [mailto:linux at dominikbrodowski.net]
Sent: mercredi 9 février 2005 21:13
To: Mathieu Bertrand
Cc: 'linux-pcmcia at lists.infradead.org'
Subject: Re: PCMCIA issues ---> cs: unable to map card memory 0x1000!
Hi,
On Tue, Feb 08, 2005 at 05:40:51PM +0100, Mathieu Bertrand wrote:
> Memory behind bridge: c4000000-c8ffffff
> Prefetchable memory behind bridge:
00000000c0000000-00000000c3f00000
> Memory window 0: c0000000-c1fff000 (prefetchable)
> Memory window 1: c4000000-c5fff000
> Memory window 0: c2000000-c3fff000 (prefetchable)
> Memory window 1: c6000000-c7fff000
> include port 0x00000000-0x00000fff, port 0x00001000-0x00001fff
> include port 0x00002000-0x00002fff, port 0x00003000-0x00003fff
> include memory 0xc0000000-0xc1ffffff, memory 0xc4000000-0xc5ffffff
> include memory 0xc2000000-0xc3ffffff, memory 0xc6000000-0xc7ffffff
>
> I don't really know what to do.
You need to make more iomem and ioport resources available. Especially some
which are _not_ used as Memory windows or I/O windows on the socket by
default. E.g. on my notebook's yenta sockets there is
Memory window 0: 30400000-307ff000 (prefetchable)
Memory window 1: 30800000-30bff000
I/O window 0: 00004000-000040ff
I/O window 1: 00004400-000044ff
but my /etc/pcmcia/config.opts does not even contain these resources:
include port 0x100-0x4ff, port 0x800-0x8ff, port 0xc00-0xcff
include memory 0xc0000-0xfffff
include memory 0xa0000000-0xa0ffffff, memory 0x60000000-0x60ffffff
Hope that helps,
Dominik
_______________________________________________
Linux PCMCIA reimplementation list
http://lists.infradead.org/mailman/listinfo/linux-pcmcia
More information about the linux-pcmcia
mailing list