PCMCIA Problems with a Panasonic Toughbook CF-29 MK5 and Antix Linux

Hi Guys,

I'm new to the forum, although have been reading it for some time. I have a problem with a Panasonic Toughbook CF-29 MK5 and Antix Linux. Only 1 of the two PCMCIA slots work. I know it's not a hardware issue, as both slots work perfectly in Windows XP. I have also tried Mint Linux with the same results, the top PCMCIA slot does not work at all. Funny thing is that I also have a CF-29 MK2, and both slots DO work in Antix Linux, so something changed between the MK2 and the MK5 that is preventing Linux from seeing that second slot. Are there any Toughbook experts out there that may have experienced this? I've been using an Adaptec USB2connect (2 USB Port) card to test out the slots.

One interesting thing to note is that when I run lspci -k I get:

06:01.0 CardBus bridge: Ricoh Co Ltd RL5c476 II (rev 8b)
Subsystem: Matsushita Electric Industrial Co., Ltd. Panasonic CF-Y5 laptop
Kernel driver in use: yenta_cardbus
Kernel modules: yenta_socket

One thing I noted in the above is it is saying that "Subsystem" is a CF-Y5. Well a Panasonic CF-Y5 only has one PCMCIA type II slot. The CF-29 has 2 of them. Don't know if that's a clue or not. I'll apologize up front in that although I have been using Linux on an off for about 10 years, I have never really gotten deep into the nuts and bolts of it. Hopefully somebody else here has. Any help is greatly appreciated! Thanks in advance!

Trevor

 

Here is part of the results from lspci -k ran on my CF-52GUN (TS model ) using Linux Mint LMDE2:

Code:

0c:00.0 CardBus bridge: Ricoh Co Ltd RL5c476 II (rev ba)
    Subsystem: Matsushita Electric Industrial Co., Ltd. Panasonic CF-Y5 laptop
    Kernel driver in use: yenta_cardbus
0c:00.1 FireWire (IEEE 1394): Ricoh Co Ltd R5C832 IEEE 1394 Controller (rev 04)
    Subsystem: Matsushita Electric Industrial Co., Ltd. Device 8338
    Kernel driver in use: firewire_ohci
0c:00.2 SD Host controller: Ricoh Co Ltd R5C822 SD/SDIO/MMC/MS/MSPro Host Adapter (rev 21)
    Subsystem: Matsushita Electric Industrial Co., Ltd. Panasonic CF-Y5 laptop
    Kernel driver in use: sdhci-pci

I have used the pcmcia slot (only one, other is express card) with a USB3 adapter without a problem. I noticed that there is no "yenta_socket" entry in my read out.

I will fire up my CF-29 and see what shows.

I have to say that I have never had a problem with the pcmcia slots under Linux.

 

Thank you, that would be awesome. I have never had a problem with PCMCIA out of Linux either, but this is the first time I've installed Linux on a CF-29 MK5. Is your CF-29 a MK5 also? Thanks for the help!

Trevor

 

My CF-29 is a Mk4, CF-29LTQGZBM.

All I could find for a pcmcia card is a Belkin wireless N card. I disabled the wifi card in the BIOS and booted to the desktop, Linux Mint 17.3 XFCE, stuck in the wifi card, clicked on the network icon in the system bar and selected my wireless network. After entering the pass code I was connected. I did updates that were available and shut down. I removed the wireless N card from the bottom pcmcia port and booted the unit. Once back on the desk top, I stuck the N card into the top pcmcia slot and ...... nothing. I left the card in the pcmcia and shut the unit down. Rebooted with the N card still in the slot and upon getting to the desktop the card still does not work.

Here is the print out from lspci -k from the 29Mk4:

Code:

06:01.0 CardBus bridge: Ricoh Co Ltd RL5c476 II (rev 8b)
    Subsystem: Matsushita Electric Industrial Co., Ltd. Panasonic CF-Y5 laptop
    Kernel driver in use: yenta_cardbus
06:01.1 SD Host controller: Ricoh Co Ltd R5C822 SD/SDIO/MMC/MS/MSPro Host Adapter (rev 11)
    Subsystem: Matsushita Electric Industrial Co., Ltd. Panasonic CF-Y5 laptop
    Kernel driver in use: sdhci-pci
07:00.0 Network controller: Ralink corp. RT2800 802.11n PCI
    Subsystem: Belkin Device 8013
    Kernel driver in use: rt2800pci

It seems that I was always using the bottom slot, never realized there was a problem with the top slot.

There should be some other members with more experience with Linux on Toughbooks than I come along. Wait for them and their knowledge.

 

Hey thank you so much for testing that! I really do appreciate it. Now we know it's an issue with MK4 and MK5 CF-29s.

Trevor

 

Sheepman will know I bet, I haven't had a CF-29 in several years to be honest and never ran Linux on one.

I am very curious if the CF-30 has the same issue, I do run Linux on them but don't have one here at the moment.

 

Ya, I figured he would possibly know what the problem is, or figure out how to fix it.

Good question about the 30's. I don't have any atm either. Last one I had went to a brother to upgrade from his 29.

 

Thanks guys. Yep I know this CF-29 MK5 is an older model (March 2006 - this one) but I needed the LPT and RS232 ports. Mainly for an older EPROM burner and to run Motorola's RSS for programming older Maxtrac mobile radios. I have a dual boot on this one of MS-DOS 6.22 and Antix Linux. BTW Antix Linux runs fantastic on these older machines! I'm using a PATA to mSATA drive adapter that I got off of ebay with a little 32 GB Samsung mSATA SSD and running hdparm -Tt /dev/sda shows that I'm getting about 90 MB/s throughput. Not too bad for an older machine. If I can just get the second PCMCIA port working I think this'll work out nicely.

The drive adapter:
https://www.ebay.com/itm/mSATA-to-2...r-Converter-Card-44PIN-with-Case/263855160333

BTW, the mSATA Drives and adapters I tried in the MK2 CF-29 worked, but the speed wasn't as fast as on the MK5 CF-29. It seems there was an issue with BIOS (even updated the BIOS) or the hardware interface talking to the PATA to mSATA cards. I was only getting about 30 MB/s throughput. - Anyway I'm getting off topic here. Hopefully I can get the PCMCIA slots both working.

 

The only thing I have to test this is a Atheros wireless card and it works in lower pcmcia port. Upper shows no activity.

I messed with settings in bios but will revert back to default.

PCI Bridge is 82801 rev d4

Cardbus bridge RL5c476 II rev 8b if that means anything.

MXLinux 18.3 386. LiveUSB

antiX IS the weapon of choice for faster load time.

dmesg doesn't seem to see the upper but I will check again.

I might have antiX on a hdd in one of these rascals.

 

Try dmidecode and see if it looks like this........

Dmidecode sees it but non-functional.

Code:

m@m-CF-29NTQGZBM:~$ sudo dmidecode --type 9
# dmidecode 3.1
Getting SMBIOS data from sysfs.
SMBIOS 2.3 present.

Handle 0x0017, DMI type 9, 13 bytes
System Slot Information
    Designation: Lower PC Card
    Type: 32-bit PC Card (PCMCIA)
    Current Usage: Unknown
    Length: Other
    ID: Adapter 0, Socket 0
    Characteristics:
        5.0 V is provided
        3.3 V is provided
        PC Card-16 is supported
        Cardbus is supported
        PME signal is supported
        Hot-plug devices are supported

Handle 0x0018, DMI type 9, 13 bytes
System Slot Information
    Designation: Upper PC Card
    Type: 32-bit PC Card (PCMCIA)
    Current Usage: Unknown
    Length: Other
    ID: Adapter 0, Socket 1
    Characteristics:
        5.0 V is provided
        3.3 V is provided
        PC Card-16 is supported
        Cardbus is supported
        PME signal is supported
        Hot-plug devices are supported

 

Hey Sheepman, thanks for the help! OK, so I did that, and also included Handle 0x0019 as it looks related, and here's what I got with no card in either slot:

Handle 0x0017, DMI type 9, 13 bytes
System Slot Information
Designation: Lower PC Card
Type: 32-bit PC Card (PCMCIA)
Current Usage: Unknown
Length: Other
ID: Adapter 0, Socket 0
Characteristics:
5.0 V is provided
3.3 V is provided
PC Card-16 is supported
Cardbus is supported
PME signal is supported
Hot-plug devices are supported

Handle 0x0018, DMI type 9, 13 bytes
System Slot Information
Designation: Upper PC Card
Type: 32-bit PC Card (PCMCIA)
Current Usage: Unknown
Length: Other
ID: Adapter 0, Socket 1
Characteristics:
5.0 V is provided
3.3 V is provided
PC Card-16 is supported
Cardbus is supported
PME signal is supported
Hot-plug devices are supported

Handle 0x0019, DMI type 9, 13 bytes
System Slot Information
Designation: mini PCI Main
Type: 32-bit PCI
Current Usage: In Use
Length: Other
ID: 0
Characteristics:
5.0 V is provided
3.3 V is provided
PME signal is supported


Here's what I get with the card in the BOTTOM slot:

Handle 0x0017, DMI type 9, 13 bytes
System Slot Information
Designation: Lower PC Card
Type: 32-bit PC Card (PCMCIA)
Current Usage: Unknown
Length: Other
ID: Adapter 0, Socket 0
Characteristics:
5.0 V is provided
3.3 V is provided
PC Card-16 is supported
Cardbus is supported
PME signal is supported
Hot-plug devices are supported

Handle 0x0018, DMI type 9, 13 bytes
System Slot Information
Designation: Upper PC Card
Type: 32-bit PC Card (PCMCIA)
Current Usage: Unknown
Length: Other
ID: Adapter 0, Socket 1
Characteristics:
5.0 V is provided
3.3 V is provided
PC Card-16 is supported
Cardbus is supported
PME signal is supported
Hot-plug devices are supported

Handle 0x0019, DMI type 9, 13 bytes
System Slot Information
Designation: mini PCI Main
Type: 32-bit PCI
Current Usage: In Use
Length: Other
ID: 0
Characteristics:
5.0 V is provided
3.3 V is provided
PME signal is supported

Here's what I got with the card in the TOP slot:

Handle 0x0017, DMI type 9, 13 bytes
System Slot Information
Designation: Lower PC Card
Type: 32-bit PC Card (PCMCIA)
Current Usage: Unknown
Length: Other
ID: Adapter 0, Socket 0
Characteristics:
5.0 V is provided
3.3 V is provided
PC Card-16 is supported
Cardbus is supported
PME signal is supported
Hot-plug devices are supported

Handle 0x0018, DMI type 9, 13 bytes
System Slot Information
Designation: Upper PC Card
Type: 32-bit PC Card (PCMCIA)
Current Usage: Unknown
Length: Other
ID: Adapter 0, Socket 1
Characteristics:
5.0 V is provided
3.3 V is provided
PC Card-16 is supported
Cardbus is supported
PME signal is supported
Hot-plug devices are supported

Handle 0x0019, DMI type 9, 13 bytes
System Slot Information
Designation: mini PCI Main
Type: 32-bit PCI
Current Usage: In Use
Length: Other
ID: 0
Characteristics:
5.0 V is provided
3.3 V is provided
PME signal is supported

BTW I used my Adaptech USB card and also has a dongle for a logitech M130 wireless mouse plugged into one of the USB slots. The mouse works great on the bottom slot.

Looks like these was no change in the output of dmidecode no matter which slot the card was loaded into and whether it worked or not.

Now, there are a few major hardware differences between the MK2 CF-29 that I have and the MK5. They are as I know, DDR2 memory instead of DDR memory, a second USB slot was added on the back and PS2 port removed, and an SD Card slot was added under the PCMCIA card slots. I'm sure the chipset is different too along with other changes. I know from testing drives lately, that the IDE port behaves differently and takes advantage of some of the higher speed data transfer protocols in the MK5 - it doesn't in the MK2.

Any other ideas? Hopefully we can figure this out...

 

There is a group of members, led by SHEEPMAN!, that are working on trying to figuring out a solution. Hopefully we will figure out a solution, soon.

 

That's Awesome! Thanks Guys!

 

Both PC slots run and work fine on MK5 using XP.

So, not hardware.

Do you have any LiveUSB from earlier kernels? I may resort to my DVD stack if I can find it.

Jeff

 

Hey Jeff,

You know I grabbed a quite old copy of Puppy Linux - 4.12-K2.6.25.16-SEAMONKEY - and both slots worked! I also tried a live CD of Ubuntu 10.04.03 LTS and that worked with both slots as well. Tried the latest Tiny Core and that did not work, so my guess at the moment might be that one of the newer kernels has an issue. I'm going to download the latest Slackware linux and see if it works. Patrick is pretty good about checking stuff. Also maybe that PCMCIA support is now limited.

Let me know what you find and I'll let you know if the latest Slackware linux works after it downloads.

Thanks!,
Trev

 

Somewhere I saw where pc (pcmcia) was no longer important or something.

So no serial port.
No parallel port.
No pc (pcmia) port.

This is progress.

I can still tune radios with CF-27 (DOS) and 29-31 (W7)

Tablets are worthless for this stuff.

Small vent there.

Jeff

 

I feel the same way!

Trev

 

Just FYI, Slackware 14 did not work either, just the lower slot. Slackware 14.2 came out in 2016, so this may have been an issue for some time.

Trev

 

Also downloaded and tried Devaun linux as it's based on Debian Jessie which uses a version 3.16 kernel - same thing - top slot does not work, so if it is a kernel issue, it existed at least that far back, and it positively worked with kernel 2.6 and Ubuntu 10.04 LTS.

It's a strange issue, because if they were not supporting PCMCIA any longer, I would expect that neither slot would work, and to make matters even stranger, both slots work in the older MK2 CF-29. Anyway, I'm hitting the sack for tonight. It's my Friday, so I'll hit it again when I get up.

Trev

 

Just for info, the PCMCIA on the card might be a issue here or a bent pin in the slot though maybe not

Just pulled out a CF-29 MK4 and had Linux Puppy on it, and just saw on both top and bottom one have no damage (pins bent no pins just not there). So maybe check that? I’m 77.12% sure it will work. The other maybe not. However Swan might be able to check. Sold a MK2 to him and maybe a MK2 is laying around? But do check from him. @sswan28.

 

We already ruled that out. Both slots work great in Windows XP and any Linux distro that uses a kernel before 2.6.33. Several others have been able to recreate the same issue on their own machines. Also as stated earlier, this is NOT an issue for MK1 through MK3 CF-29's - They work fine. It's the MK4 and MK5 CF-29s with the later hardware that are having the issue. The top PCMCIA/Cardbus - yes 32 bits - slot does not work in a toughbook CF-29 MK4 or MK5 with any current Linux distribution. It appears Linux does not support the PCI bridge in these machines.

 

Should prob have read more than 5 posts

 

An update for those following this thread.

FX-Tech is now working with a group of us behind the scene to come up with a solution to the problem.
The thread will be updated when we achieve a solution.

 

It seems that the top PCMCIA support was dropped when the kernel went from 2.6.32 to 2.6.33.
There has been numerous attempts to compile a hybrid(?) kernel using using Debian 4.19 and the config file from the 2.6.32 kernel. FAIL on all attempts.
An email has been sent to the kernel maintainer laying out what has been attempted and asking if a fix can be implemented.

A BIG THANK YOU to Michael AU for all the time, effort and knowledge he has spent working on this problem with the group. Very much appreciated!

The group is hopeful that a solution can/will be found. Either way, this thread will get updated when there is something to report.

 

Email hasn't been sent yet due to being a little under the weather and not getting online for the last week. I have a little bit more work to do getting some information from various distros and will then send the email. Progress reports will be forthcoming.

 

Email was sent a short time ago directly to Greg Kroah-Hartman. Hopefully we might be able to get both slots working again on CF-29 Mark 4 and 5.

 

A patch has been developed by a kernel dev (not Greg Kroah-Hartmman) that we need to merged into the kernel. It has been developed for kernel 5.4-rc1 but should be able to be merged into 4.19 by individual users. I will get to it sometime this week and get some testing done, if both slots show up we will do some testing and then report back to the kernel devs to let them know how it went. If we get the all clear I will release a 4.19 and a 5.* kernel for Debian and Debian derivatives, anyone who uses a non-Debian based distro will be able to merge the patch and build their own kernel as I will post the approved patch here. If everything works as it is hoped it will it will possibly be part of the 5.4 kernel release.

 

I have been given the honour of letting everyone know that after alot of compiling, testing, emailing, patching, compiling, and testing a small band of Toughbook Warriors and kernel developers have worked together to successfully re-enable both PCMCIA slots on the CF-29 Mk4 and Mk5 Toughbooks.

It has been a huge learning experience but also a pleasure to have the opportunity to work with a great group of people to get something working that has been broken for at least 8 years.

I'm going to add "Never give up, never surrender" to my signature. If you have never seen Galaxy Quest you won't know what I mean.

 

Michael, thank you once again. You all are dedicated & determined, and going the extra mile certainly has paid off. Cheers!

 

Just a quick update after a long absence, earlier this year kernel devs provided some information that a setting regarding the Ricoh kernel module needed to be changed. Today I compiled the latest Debian Stable kernel with 1 change and posted it for a few users to test and report back on. If the modified kernel setting works we will have narrowed down the issue and then decide, with the guidance of the kernel devs, what the best solution to the issue is.

 

Have had confirmation that the specific kernel switch works as intended, now to let the kernel devs know.