[feisty] regression: ti mmc card reader not working (worked flawlessly in edgy)

I can also confirm that tifm7xx1 stopped working on my system also after an upgrade to Feisty.

I have the same issue, except I don't remember it working entirely correctly in Edgy, either -- I'd get duplicate /media/SD /media/SD_ /media/SD__ ...

My chip is the PCI7412. All the tutorials I can find refer to some function
    xx.xx.3 Texas Instruments PCIxx12 SDA Standard Compliant SD Host Controller
but my system does not have this. It's a Gateway M685.

04:09.0 CardBus bridge: Texas Instruments PCIxx12 Cardbus Controller
04:09.1 FireWire (IEEE 1394): Texas Instruments PCIxx12 OHCI Compliant IEEE 1394 Host Controller
04:09.2 Mass storage controller: Texas Instruments 5-in-1 Multimedia Card Reader (SD/MMC/MS/MS PRO/xD)

Also, in dmesg I see
tifm_7xx1: sd card detected in socket 1
but I don't get a /dev/mmcblk0 .

I have seen this working after the update to kernel version 2.6.20-8, but the directories inside the mmc card where shown as empty files - however a reboot made the mmc card disappear again.

Can you attach the output of 'cat /proc/devices' ?

Actually, there appears to be no difference between the two.

Vincenzo & DanaGoyette: you seem to be describing two separate and different issues here.

DanaGoyette: Your /proc/devices doesn't show " 254 mmc".

This is probably because you have not got "mmc_block" and/or "mmc_core" loaded.

Try "lsmod | grep mmc"

if it doesn't show both mmc_core and mmc_block try

modprobe mmc_core
modprobe mmc_block

This may fix your problem. It if does then this needs further investigation as to why udev/modprobe didn't load it automatically.

Vincenzo: your problem is different. It is likely your machine isn't loading the correct low level mmc driver. Please attach a full dmesg and the full output from lsmod. Thanks.

Viencenzo,

Can you attach the output of both "lspci" and "lspci -nv".

Thanks.

Now I inserted every module that I had in edgy but not in feisty, but nothing changed, I even tried to restart hal but the device in /dev is still missing and creating it will not allow me to mount it (not a valid block device).

If I remove and insert tifm_7xx1 I get in my logs:

remove:
[ 1457.008000] ACPI: PCI interrupt for device 0000:03:0b.2 disabled

insert
[ 1466.336000] ACPI: PCI Interrupt 0000:03:0b.2[D] -> GSI 23 (level, low) -> IRQ 20

so I guess this is the right module and something strange is happening...

Even if I modprobe mmc_block (note that tifm_sd already pulls in mmc_core), the /dev/mmcblk0 device still isn't created, and I see nothing new in dmesg.

Here's lspci -nnv (nn makes it show both names and numbers.)

DanaGoyette:

I wanted to know if mmc_block was present in lsmod output or not. From your response, the implication is it wasn't. Correct? You had to modprobe it yourself?

After modprobe mmc_block I expected to see a "254 mmc" line in /proc/devices? Is it there?

If so, then after modprobing, you're in the same situation as Vincenzo.

That is correct. I do now see a "254 mmc" line after manually modprobing mmc_block, but there still is no /dev/mmcblk0 file created.

hi, Im having the same problem like DanaGoyette.
my Laptop is a HP nx6325 and according to several tutorials it should have worked with edgy.

> lspci:
02:04.2 Mass storage controller: Texas Instruments 5-in-1 Multimedia Card Reader (SD/MMC/MS/MS PRO/xD)
02:04.3 Generic system peripheral [0805]: Texas Instruments PCIxx12 SDA Standard Compliant SD Host Controller

> dmesg on card insertion:
[ 278.356000] tifm_7xx1: xd card detected in socket 0
[ 278.848000] tifm_7xx1: demand removing card from socket 0
[ 279.144000] tifm_7xx1: sd card detected in socket 1

but no device is being created.

like DanaGoyette I dont get mmc_block loaded and only get that "254 mmc" line if I load it manually.

a bit of digging pulled this out:
https://lists.berlios.de/pipermail/tifm21-devel/2007-February/000238.html

From what i can tell from that thread, especially this message:

https://lists.berlios.de/pipermail/tifm21-devel/2007-February/000276.html

just disabling CONFIG_SYSFS_DEPRECATED in kernel should resolve the problem in feisty, unfortunately this is not the case (I recompiled my kernel) and so we know that we have a different problem.

the message you cited, is about linux 2.6.20 with older hal & udev versions, which only applies to ubuntu 6.06.

I think the patch in attached to the message I initially referred to is what we need is what we need:
https://lists.berlios.de/pipermail/tifm21-devel/2007-February/000239.html

but I was too lazy to recompile the kernel as yet...

everything works here after running manually the following command from Konsole:

sudo setpci -s nn:nn.n 4c.b=0x02 (of course I had to replace the "n"
with the Mass storage controller device number visible from "lspci", in my case it is: 02:06.3 )

with the previous kubuntu stable release everything worked out of the box, is it a kernel bug?

Most of the headache seems to come from CONFIG_SYSFS_DEPRECATED

https://lists.berlios.de/pipermail/tifm21-devel/2007-February/000276.html

I tried and forgot to post here the result: disabling CONFIG_SYSFS_DEPRECATED in feisty did not change anything i.e. the bug is still there. Since the thread you linked was about edgy, and the problem was solved installing feisty packages and disabling CONFIG_SYSFS_DEPRECATED, I suppose these are two different bugs.

There are a lot of conflicting reports against this bug.

1. The CONFIG_SYSFS_DEPRECATED thread is not applicable to this bug.

2. The sudo setpci -s nn:nn.n 4c.b=0x02 fix doesn't fix the underlying problem. The TI MMC card readers support two interfaces: the proprietary TI interface, and the generic "Secure Digital Host Controller Interface". Only one of these interfaces are connected to the physical pins of the MMC reader at any one time. Normally the physical pins are connected to the proprietary interface, which is supported by the broken tifm_sd driver.

The setpci command switches the physical pins to the "Secure Digital Host Controller Interface" supported by the sdhci driver. This command in effect moves the responsibility of reading/writing the MMC card to the working sdhci driver.

3. The tifm_sd driver is obviously broken. Hopefully, the patch referred to in a previous comment will fix the driver, although this is far from being definate.

I will evaluate the patch, and consider applying it to the Feisty kernel. Obviously without the necessary hardware, I am relying on one of the bug reporters to test the patch, perhaps independently, or once (if) the patch appears in a released Feisty kernel.

To confirm this analysis, it would be useful for someone to unload the tifm_sd driver (modprobe -r tifm_sd), and then to issue the setpci command. If the sdhci driver is in use, the MMC card should still be accessible.

I did what you asked for, the result in my dmesg follows, I inserted sdhci and mmc_block but I got no /dev/mmcwhatever, and I don't know how to mount the card. Restarting hald using "/etc/dbus-1/event.d/20hal restart" does not help. Don't know if the logs confirm that my card is working, and there are other things to do to complete the fix, or if it is just "still broken".

[ 135.304000] ACPI: PCI interrupt for device 0000:03:0b.2 disabled
[ 176.300000] ACPI: PCI interrupt for device 0000:03:0b.3 disabled
[ 200.256000] sdhci: Secure Digital Host Controller Interface driver, 0.12
[ 200.256000] sdhci: Copyright(c) Pierre Ossman
[ 200.256000] sdhci: SDHCI controller found at 0000:03:0b.3 [104c:803c] (rev 0)
[ 200.256000] ACPI: PCI Interrupt 0000:03:0b.3[D] -> GSI 23 (level, low) -> IRQ 19
[ 200.256000] sdhci:slot0: Controller reports > 25 MHz base clock, but no high speed support.
[ 200.256000] mmc0: SDHCI at 0x58006800 irq 19 DMA

As you said, "The sudo setpci -s nn:nn.n 4c.b=0x02 fix doesn't fix the underlying problem."
That is true for me -- I don't have the SDHCI controller at all, so that fix doesn't apply.

My card reader is the TI PCI7412 -- 1 Cardbus, 2 firewire (though Gateway only gives me 1), and one media card reader. Perhaps Gateway disabled the SDHCI controller.

Vincenzo:

My comment was probably unclear about what I wanted you to do. Following the steps below should work. Please execute these commands in sequence:

sudo modprobe sdhci
sudo modprobe mmc_block
sudo modprobe -r tifm_sd
sudo setpci -s 03:0b.2 4c.b=0x02

*then* insert a SD card into the MMC reader. On insertion of the SD card you should get a /dev/mmcblk0. You should also see some information in the dmesg log after insertion of the card.

Yes, I did all these steps (I now re-did those in the exact sequence you wrote, however I already did more or less the same thing yesterday), but the log is what I posted above.

Me too, TI card readers don't work. Can you help me?

dmesg |grep Failed
[ 67.932557] PCI: Failed to allocate mem resource #10:4000000@e4000000 for 0000:02:04.0
[ 67.932572] PCI: Failed to allocate mem resource #10:4000000@e4000000 for 0000:02:04.1

lspci -v|grep Texas
02:00.0 FireWire (IEEE 1394): Texas Instruments TSB43AB21 IEEE-1394a-2000 Controller (PHY/Link) (prog-if 10 [OHCI])
02:04.0 CardBus bridge: Texas Instruments PCI1620 PC Card Controller (rev 01)
02:04.1 CardBus bridge: Texas Instruments PCI1620 PC Card Controller (rev 01)
02:04.2 System peripheral: Texas Instruments PCI1620 Firmware Loading Function (rev 01)

Here's the deal:

Some newer variants of the TIFM 7xx1 will NOT revert to sdhci-compatible operation with the setpci hack.

The tifm_7xx1 driver in our Feisty kernel is somehow broken. Replacing it with this one : http://prdownload.berlios.de/tifmxx/tifm-0.8b.tar.bz2 (newer upstream) fixes it.

This is a confirmed bug.

Here's what output from tifm7xx1 version 0.8 on Feisty shows:
[48722.364000] tifm_core: MMC/SD card detected in socket 0:1
[48722.484000] mmc0: Problem switching card into high-speed mode!
[48722.484000] mmcblk0: mmc0:9ffc SD01G 992000KiB
[48722.484000] mmcblk0: p1

And indeed, GNOME pops up telling me I inserted a photo card.

I walked two IRC'ers through compiling the newer tifm, and both reported back success stories.

If we can get this new tifm release in Feisty's kernel, this bug will be closed. The tifm7xx1 controller is highly popular on laptops these days that come with n-in-1 card readers.

I installed the module and after rebooting twice (???) I have the mmc card icon on my desktop, and it works. Thanks for finding a solution.

I rebooted again, and it does not work again. I didn't upgrade my kernel.

Vincenzo:

This is with the new version of the driver reported as working by "John Dong"? If so (and if you can get the module working again) please post the dmesg from when it does work, and when it doesn't work. Thanks

search thru your /lib/modules/`uname -r` and make sure you only have one
copy of the TIFM modules.... It works just fine for me still. When this is
in the Ubuntu kernel it'll be more reliable to test.

On 3/6/07, Vincenzo Ciancia <email address hidden> wrote:
>
> I rebooted again, and it does not work again. I didn't upgrade my
> kernel.
>
> --
> [feisty] regression: ti mmc card reader not working (worked flawlessly in
> edgy)
> https://launchpad.net/bugs/82680
>

I discovered that the Makefile in the linked download puts the modules in the wrong place when you make install. It puts them in misc/ instead of kernel/drivers/misc.
Once I corrected it, I found that it does work for me -- now I can read SD cards. I believe MMC cards may still not be supported -- I tried one and got nothing in dmesg.

You where right, I had two copies of them. I removed both, and reinstalled the module. Unfortunately it is not working anymore - do not really know why, logs are uninformative.

It don't work at home.... Can it be a problem of allocation ?
I repeat messages of 'dmesg' :

[ 67.932557] PCI: Failed to allocate mem resource #10:4000000@e4000000 for 0000:02:04.0
[ 67.932572] PCI: Failed to allocate mem resource #10:4000000@e4000000 for 0000:02:04.1

It is not a problem of module ? Are there options of kernel on boot ?

Tifm driver updated to 0.8 release (20070222). The new release should fix the problems raised here. When the kernel is released, please retest. Thanks.

Confirming new kernel fixes it.

works but had to manually load tifm_sd

Yeah, Pavel, it's always been that way with this driver :-/

On 3/13/07, Pavel Rojtberg <email address hidden> wrote:
>
> works but had to manually load tifm_sd
>
> --
> [feisty] regression: ti mmc card reader not working (worked flawlessly in
> edgy)
> https://launchpad.net/bugs/82680
>

it works now thank you all!!

Unfortunately it does not work for me. Can you tell me something more to debug the problem? My kernel version is 2.6.20-10.

Ok, it works now, I had to reboot twice again... really don't know why twice, but it works :) Thank you

After manually loading load "tifm_sd" this hardware works well on my lenovo-thinkpad z61m
and kubuntu edgy .

After inserting a sd-card, i get the popup which helps mounting the device.

My hardware:

$ lspci -v|less -p Mass
-----
15:00.2 Mass storage controller: Texas Instruments Unknown device 803b
        Subsystem: Lenovo Unknown device 207c
        Flags: bus master, medium devsel, latency 64, IRQ 169
        Memory at e4301000 (32-bit, non-prefetchable) [size=4K]
        Capabilities: [44] Power Management version 2
-----

Adding it to "/etc/modules" should solve this problem for me......

No, I was wrong, I rebooted in edgy by mistake. In feisty it is not working. I will try a full reinstall as soon as possible.

Perhaps we need a modprobe script to load tifm_sd whenever tifm_7xx1 is loaded. Something like,

/etc/modprobe.d/tifm_7xx1
install tifm_7xx1 /sbin/modprobe --ignore-install tifm_7xx1 $CMDLINE_OPTS && { /sbin/modprobe tifm_sd ; }

I agree with DanaGoyette -- it doesn't make sense to probe the tifm controller without turning on SD support. Perhaps file a new bug for it?

udev ought to handle this - the driver sends out events on card
insertion.

Yes, please file a new udev bug.

Nope, I made a full reinstall of feisty herd 5, and upgraded the system, now I have kernel 2.6.20-11-generic but inserting tifm_sd or sdhci or mmc_block changes nothing. Please let's have a look at this together, I am really scared about getting to release without mmc reader working :) I am leaving next friday, so if you need interaction it's better to do this in the next days - or else after the 2nd of April.

Hmm, I've talked to many people who have this cardreader and it all works for them now, just by adding tifm_sd to /etc/modules

Are you sure you have this cardreader?

0a:09.2 Mass storage controller: Texas Instruments 5-in-1 Multimedia Card Reader (SD/MMC/MS/MS PRO/xD)
        Subsystem: Acer Incorporated [ALI] Unknown device 0094
        Flags: bus master, medium devsel, latency 57, IRQ 20
        Memory at c8406000 (32-bit, non-prefetchable) [size=4K]
        Capabilities: <access denied>

Numeric (lspci -nv):
0a:09.2 0180: 104c:803b
        Subsystem: 1025:0094
        Flags: bus master, medium devsel, latency 57, IRQ 20
        Memory at c8406000 (32-bit, non-prefetchable) [size=4K]
        Capabilities: <access denied>

This is strange:

lspci -v gives:

03:0b.2 Mass storage controller: Texas Instruments 5-in-1 Multimedia Card Reader (SD/MMC/MS/MS PRO/xD)
        Subsystem: Toshiba America Info Systems Unknown device 0001
        Flags: medium devsel, IRQ 19
        Memory at 58005000 (32-bit, non-prefetchable) [size=4K]
        Capabilities: [44] Power Management version 2

03:0b.3 Generic system peripheral [0805]: Texas Instruments PCIxx12 SDA Standard Compliant SD Host Controller (prog-if 01)
        Subsystem: Toshiba America Info Systems Unknown device 0001
        Flags: bus master, medium devsel, latency 64, IRQ 19
        Memory at 58006800 (32-bit, non-prefetchable) [size=256]
        Capabilities: [80] Power Management version 2

which is different than yours, but lspci -nv gives

03:0b.2 0180: 104c:803b
        Subsystem: 1179:0001
        Flags: medium devsel, IRQ 19
        Memory at 58005000 (32-bit, non-prefetchable) [size=4K]
        Capabilities: [44] Power Management version 2

03:0b.3 0805: 104c:803c (prog-if 01)
        Subsystem: 1179:0001
        Flags: bus master, medium devsel, latency 64, IRQ 19
        Memory at 58006800 (32-bit, non-prefetchable) [size=256]
        Capabilities: [80] Power Management version 2

Notice that the PCI id of the first device is the same as yours.

By digging the web deeper, I found that removing all modules, then giving the following command:

 setpci -s 03:0b.2 4c=0x22

then inserting sdhci only (and not the tifm_* drivers) makes my sd cards work, and that's enough as a workaround. Notice that in the above comments it was

 4c=0x02

which didn't work for me.

I would surely like this to be supported out of the box, on one hand, and the tifm drivers to support my reader on the other hand, because this should make cards different than SD work. I wait further instructions.

Very peculiar. My cardreader does not respond properly to any of the setpci
hacks.

udev bug found and updated: bug #65042

the update is already available to be downloaded/installed via adept?
because I have just updated my kubuntu, and I still have to manually load the tifm_sd module to get the card reader to work..

bye!

No, it isn't I just found the relevant bug in launchpad, and amended it.

I added a udev rule to the bug that fixes the problem, but the udev developers need to incorporate it.

I've got a HP Compaq nx6325 EY349EA and my card reader still does not work :/
lspci -v|less -p Mass
02:04.2 Mass storage controller: Texas Instruments 5-in-1 Multimedia Card Reader (SD/MMC/MS/MS PRO/xD)
        Subsystem: Hewlett-Packard Company Unknown device 30b0
        Flags: bus master, medium devsel, latency 64, IRQ 16
        Memory at d4018000 (32-bit, non-prefetchable) [size=4K]
        Capabilities: <access denied>

02:04.3 Generic system peripheral [0805]: Texas Instruments PCIxx12 SDA Standard Compliant SD Host Controller
        Subsystem: Hewlett-Packard Company Unknown device 30b0
        Flags: bus master, medium devsel, latency 64, IRQ 16
        Memory at d4019000 (32-bit, non-prefetchable) [size=256]
        Capabilities: <access denied>
<<azrael@laptop:~>> uname -a
Linux laptop 2.6.20-12-generic #2 SMP Sun Mar 18 03:07:14 UTC 2007 i686 GNU/Linux

None of the setpci hacks worked. I used one in edgy (setpci -s 02:04.2 4c=0x22).

dmesg shows just this upon inserting and removing an SD card:
Mar 19 11:59:17 laptop kernel: [ 2350.600000] tifm_core: SmartMedia/xD card detected in socket 0:0
Mar 19 11:59:18 laptop kernel: [ 2351.720000] tifm0 : demand removing card from socket 0:0
Mar 19 12:03:13 laptop kernel: [ 2586.392000] tifm_core: SmartMedia/xD card detected in socket 0:0
Mar 19 12:03:13 laptop kernel: [ 2586.840000] tifm0 : demand removing card from socket 0:0

Przemyslaw: It looks like you are using an xD card. These aren't supprted yet by the tifm driver

http://openfacts.berlios.de/index-en.phtml?title=TI_FlashMedia_xx12/xx21_driver

I have a SD card, not XD. I can assure you that I don't mistake these. Besides, it worked in edgy.

Given that SDHCI and tifm both detect a xD card being inserted, I also have
to strongly suspect that it is indeed an xD card, or a hardware malfunction.

On 3/19/07, Przemysław Kulczycki <email address hidden> wrote:
>
> I have a SD card, not XD. I can assure you that I don't mistake these.
> Besides, it worked in edgy.
>
> --
> [feisty] regression: ti mmc card reader not working (worked flawlessly in
> edgy)
> https://launchpad.net/bugs/82680
>

> Given that SDHCI and tifm both detect a xD card being inserted, > I also have to strongly suspect that it is indeed an xD card, or a > hardware malfunction.

I guess it's a driver malfunction.
I've attached a photo of my card for the unbelievers ;)

I reported the bug and it's definitely not fixed :)

Just to clarify: in my case, in edgy, I didn't have to insert any module. With my great surprise, my mmc card reader worked out-of-the-box in the livecd.

Vincenzo, on Edgy I needed to load the module. I have it in /etc/modules now so when 2.6.20-12 came out it worked fine.

Wait, did you ever test it with *installed* Edgy? It's always required loading the modules in the installed version.

I don't remember if I had to add the module to /etc/modules in installed edgy, but I don't think so, it worked in the live cd and the kernel autodetected the module, I said this to discern between what seem to be two different TI card readers, mine which was supported in edgy out of the box and does not work in feisty, and another one that works in feisty, but always required the module to be loaded.

adding tifm_sd in /etc/modules worked for me, thanks..

i atach output of lspci -nv

-Jose

i forgot to menction im using Feisty 2.6.20-12-generic #2 SMP Wed Mar 21 20:55:46 UTC 2007 i686 GNU/Linux

Got mine to work?!
ok i have not really posted anything to this bug but.. i was in the same place as everyone else here. I had a problem with my hda_intel sound card bug #94373. but i also had a problem with my this TI card reader... Me comming from 5 years of being a gentoo guy i desided to build my own kernel. I am at work now and then flying from Miami to Uzhgorod, Ukraine and i really don't have the time to figure out what i exactly did with my kernel to get it working. but here are my .config "i totally just made it for what my computer needs" and a dmesg as well as a lsmod

hope it helps figuring out what is up with all of this.. wish i could be more help but i am just taking a 15 minute break to get this on here. and in Ukriane i won't have an internet connection

Mine tifm worked very well on 2.6.20-12 (with tifm_sd on /etc/modules), but on 2.6.20-13 it stopped to work now.

On my dmesg, the detection works ok:

[ 1632.312000] tifm_7xx1: sd card detected in socket 1
[ 1632.404000] mmcblk0: mmc1:a95c SD01G 992000KiB
[ 1632.404000] mmcblk0: p1

But gnome-volume-manager does not mount on /media/SD.

Someone had the same problem?

pereira > this is the whole purpose of #53923

The driver still reports my SD card as xD - WTF???
[66332.376000] tifm_7xx1: xd card detected in socket 0
[66332.556000] tifm_7xx1: demand removing card from socket 0

Built-in card reader does not work here either.
Im on a Dell Inspiron 6000.
Kubuntu Feisty Fawn *Beta* with latest updates (via apt), kernel 2.6.20-13-generic

The lsmod above was with the card inserted.

The card is a 256mb SD Card "Lexar".

When the card is inserted, this is added to my /var/log/syslog:
Apr 7 15:05:25 localhost kernel: [18715.472000] mmcblk0: mmc0:b368 D1018 249856KiB
Apr 7 15:05:25 localhost kernel: [18715.472000] mmcblk0: p1

I hope it helps.

when i insert my sony memory stick duo pro (i have no idea if it is supported - i use kernel 2.6.20.15), in dmesg i can see this line:
[ 8551.388000] tifm_7xx1: ms card detected in socket 2
When i remove the stick another entry appears:
[ 8555.924000] tifm_7xx1: demand removing card from socket 2

in my /var/log/syslog i can see these lines and a lot more similar (with differences in sector numbers):
Apr 21 18:06:08 FerrariUbuntu kernel: [14696.256000] mmcblk0: error 1 sending read/write command
Apr 21 18:06:08 FerrariUbuntu kernel: [14696.256000] end_request: I/O error, dev mmcblk0, sector 105
Apr 21 18:06:08 FerrariUbuntu kernel: [14696.256000] mmcblk0: error 1 sending read/write command
Apr 21 18:06:08 FerrariUbuntu kernel: [14696.256000] end_request: I/O error, dev mmcblk0, sector 106
Apr 21 18:06:08 FerrariUbuntu kernel: [14696.256000] mmcblk0: error 1 sending read/write command
Apr 21 18:06:08 FerrariUbuntu kernel: [14696.256000] end_request: I/O error, dev mmcblk0, sector 16
Apr 21 18:06:08 FerrariUbuntu kernel: [14696.256000] mmcblk0: error 1 sending read/write command
Apr 21 18:06:08 FerrariUbuntu kernel: [14696.256000] end_request: I/O error, dev mmcblk0, sector 24
Apr 21 18:06:08 FerrariUbuntu kernel: [14696.256000] mmcblk0: error 1 sending read/write command
Apr 21 18:06:08 FerrariUbuntu kernel: [14696.256000] end_request: I/O error, dev mmcblk0, sector 0
Apr 21 18:06:08 FerrariUbuntu NetworkManager: <debug info>^I[1177167968.825884] nm_hal_device_removed (): Device removed (hal udi is '/org/freedesktop/Hal/devices/pci_104c_8033_mmc_host_mmc_card_rca40401').
Apr 21 18:06:08 FerrariUbuntu NetworkManager: <debug info>^I[1177167968.834043] nm_hal_device_removed (): Device removed (hal udi is '/org/freedesktop/Hal/devices/pci_104c_8033_mmc_host').
Apr 21 18:06:08 FerrariUbuntu NetworkManager: <debug info>^I[1177167968.834947] nm_hal_device_removed (): Device removed (hal udi is '/org/freedesktop/Hal/devices/pci_104c_8033_mmc_host_mmc_card_rca40401_storage').

I have added these lines to /etc/modules:
#card reader
tifm_core
tifm_7xx1
tifm_sd
mmc_core
mmc_block
sdhci

(sudo lspci -nnv):
06:09.3 Mass storage controller [0180]: Texas Instruments PCIxx21 Integrated FlashMedia Controller [104c:8033]
        Subsystem: Acer Incorporated [ALI] Unknown device [1025:007e]
        Flags: bus master, medium devsel, latency 64, IRQ 16
        Memory at c0306000 (32-bit, non-prefetchable) [size=8K]
        Capabilities: [44] Power Management version 2

and in my laptop this makes the insert/remove events not to show on dmesg
sudo setpci -s 06:09.3 4c.b=0x22

while sudo setpci -s 06:09.3 4c.b=0x02 does nothing in particular...

What can i do now? I upgraded to feisty two days ago. In Edgy my card reader did not work out of the box but i did not try to fix it.
Now i am only a few steps away from completely abandoning Win XP and i want so much to just fix everything and get rid of them...

having similar problem.

Compaq nx6325

uname -a
Linux 2.6.20-15-generic #2 SMP Sun Apr 15 06:17:24 UTC 2007 x86_64 GNU/Linux

I have added these lines to /etc/modules:
#card reader
tifm_core
tifm_7xx1
tifm_sd
mmc_core
mmc_block
sdhci

root@mgm-laptop:/dev# lsmod | grep tifm
tifm_sd 13960 0
mmc_core 31624 3 mmc_block,tifm_sd,sdhci
tifm_7xx1 11520 0
tifm_core 11520 2 tifm_sd,tifm_7xx1

-------
lspci -nnv

02:04.2 Mass storage controller [0180]: Texas Instruments 5-in-1 Multimedia Card Reader (SD/MMC/MS/MS PRO/xD) [104c:803b]
        Subsystem: Hewlett-Packard Company Unknown device [103c:30b0]
        Flags: bus master, medium devsel, latency 64, IRQ 20
        Memory at d4018000 (32-bit, non-prefetchable) [size=4K]
        Capabilities: [44] Power Management version 2

02:04.3 Generic system peripheral [0805]: Texas Instruments PCIxx12 SDA Standard Compliant SD Host Controller [104c:803c]
        Subsystem: Hewlett-Packard Company Unknown device [103c:30b0]
        Flags: bus master, medium devsel, latency 64, IRQ 20
        Memory at d4019000 (32-bit, non-prefetchable) [size=256]
        Capabilities: [80] Power Management version 2

output from tail /var/log/messages

Apr 23 14:33:50 mgm-laptop kernel: [ 349.021022] tifm_7xx1: sd card detected in socket 1

no devices in dev are created, everything else looks like it is working ok... am i missing something?

I have HP NC8230 with

02:06.3 Mass storage controller: Texas Instruments PCIxx21 Integrated FlashMedia Controller
        Subsystem: Hewlett-Packard Company Unknown device 0934
        Flags: bus master, medium devsel, latency 64, IRQ 21
        Memory at c8408000 (32-bit, non-prefetchable) [size=8K]
        Capabilities: <access denied>

02:06.4 Generic system peripheral [0805]: Texas Instruments PCI6411/6421/6611/6621/7411/7421
/7611/7621 Secure Digital Controller
        Subsystem: Hewlett-Packard Company Unknown device 0934
        Flags: bus master, medium devsel, latency 64, IRQ 19
        Memory at c840a000 (32-bit, non-prefetchable) [size=256]
        Memory at c840b000 (32-bit, non-prefetchable) [size=256]
        Memory at c840c000 (32-bit, non-prefetchable) [size=256]
        Capabilities: <access denied>

02:06.5 Communication controller: Texas Instruments PCI6411/6421/6611/6621/7411/7421/7611/76
21 Smart Card Controller
        Subsystem: Hewlett-Packard Company Unknown device 0934
        Flags: medium devsel, IRQ 11
        Memory at c840d000 (32-bit, non-prefetchable) [size=4K]
        Memory at c840e000 (32-bit, non-prefetchable) [size=4K]
        Memory at c840f000 (32-bit, non-prefetchable) [size=4K]
        Memory at c8410000 (32-bit, non-prefetchable) [size=4K]
        Capabilities: <access denied>

When I insert SD card:
[54417.900000] mmcblk0: mmc3:0001 SD 988160KiB
[54417.900000] mmcblk0: p1

it does not mount, but soon dmesg gets flooded with:
[54447.928000] mmcblk0: error 1 sending read/write command
[54447.928000] end_request: I/O error, dev mmcblk0, sector 254
[54447.928000] mmcblk0: error 1 sending read/write command
[54447.928000] end_request: I/O error, dev mmcblk0, sector 255
[54447.928000] mmcblk0: error 1 sending read/write command
[54447.928000] end_request: I/O error, dev mmcblk0, sector 256
[54447.928000] mmcblk0: error 1 sending read/write command
[54448.020000] end_request: I/O error, dev mmcblk0, sector 1976319
[54448.020000] mmcblk0: error 1 sending read/write command
[54448.020000] end_request: I/O error, dev mmcblk0, sector 1976315
[54448.020000] mmcblk0: error 1 sending read/write command
[54448.020000] end_request: I/O error, dev mmcblk0, sector 1976316
and so on...

This card mounts and reads fine in an external SD reader (YX-RWc01-B).

I "installed" the new upstream tifm drivers http://prdownload.berlios.de/tifmxx/tifm-0.8b.tar.bz2 (added the proper entries to /etc/modules and modprobed in the new modules...) but my card reader still was not working properly. dmesg showed the card insert is being properly detected by the driver but the card did not get mounted after insert. So i gave up for a while on the problem.

Then ones i was booting my laptop i noticed that at one point the boot log says (i have the "quiet" kernel option removed so i see the boot messages):
Mounting local file systems ..... FAIL

Further research showed that it attempts to mount /dev/mmcblk0 to /media/mmcblk0 on every boot, which isn't right since the card isn't always present. So i commented out the following line in /etc/fstab file:
 /dev/mmcblk0 /media/mmcblk0 vfat defaults,utf8,umask=007,gid=46 0 1
on next laptop boot the boot log error is gone (of course) and my card reader now properly mounts my MMC card if i insert it and i even get the "icon on the desktop" :D ... I only tested MMC card since i don't have any SD ones :P

I used a script for installing the new tifm driver posted in this howto thread:
http://ubuntuforums.org/showpost.php?p=2520594&postcount=1

Inserting card while booted and logged in now works properly for me.
Cold start with card inserted also mounts the card properly.
Reboot with card inserted works as well.

And i am running 32bit Ubuntu 7.04 Feisty Fawn with a 2.6.20-16 kernel on an HP nx6125 laptop with a 5in1 built in card reader.

Is this bug fixed for everybody else than me, I mean bug #82680, not its "master" bug. If not, this should be not a duplicate of the other one. Please reply soon, I would like to clarify the situation ASAP.