Bug 5678 - endless loop/kernel crash after memory card insert into 6-in-1 pcmcia memory card reader reader
Summary: endless loop/kernel crash after memory card insert into 6-in-1 pcmcia memory ...
Status: REJECTED WILL_NOT_FIX
Alias: None
Product: IO/Storage
Classification: Unclassified
Component: IDE (show other bugs)
Hardware: i386 Linux
: P2 normal
Assignee: Bartlomiej Zolnierkiewicz
URL:
Keywords:
: 5680 (view as bug list)
Depends on:
Blocks:
 
Reported: 2005-11-29 15:31 UTC by Khashayar Naderehvandi
Modified: 2007-06-18 08:03 UTC (History)
0 users

See Also:
Kernel Version: 2.6.13*, 2.6.14*
Subsystem:
Regression: ---
Bisected commit-id:


Attachments
logs+config; 2.6.14.3-vanilla (37.30 KB, application/x-gzip)
2005-11-30 15:00 UTC, Khashayar Naderehvandi
Details

Description Khashayar Naderehvandi 2005-11-29 15:31:18 UTC
Most recent kernel where this bug did not occur: Not sure; it might have worked
with 2.6.12 :-/

Distribution: ArchLinux

Hardware Environment: 

00:1f.1 IDE interface: Intel Corp.: Unknown device 24ca (rev 03) (prog-if 8a
[Master SecP PriP])
	Subsystem: Citicorp TTI: Unknown device 11ab
	Flags: bus master, medium devsel, latency 0, IRQ 11
	I/O ports at <unassigned>
	I/O ports at <unassigned>
	I/O ports at <unassigned>
	I/O ports at <unassigned>
	I/O ports at 2420 [size=16]
	Memory at 34000000 (32-bit, non-prefetchable) [size=1K]
</snip>
1:0a.0 CardBus bridge: Ricoh Co Ltd RL5c476 II (rev ab)
	Subsystem: Citicorp TTI: Unknown device 1254
	Flags: bus master, medium devsel, latency 168, IRQ 11
	Memory at d0203000 (32-bit, non-prefetchable) [size=4K]
	Bus: primary=01, secondary=02, subordinate=05, sec-latency=176
	Memory window 0: 30000000-31fff000 (prefetchable)
	Memory window 1: 36000000-37fff000
	I/O window 0: 00003400-000034ff
	I/O window 1: 00003800-000038ff
	16-bit legacy interface ports at 0001

01:0a.1 CardBus bridge: Ricoh Co Ltd RL5c476 II (rev ab)
	Subsystem: Citicorp TTI: Unknown device 1254
	Flags: bus master, medium devsel, latency 168, IRQ 11
	Memory at d0204000 (32-bit, non-prefetchable) [size=4K]
	Bus: primary=01, secondary=06, subordinate=09, sec-latency=176
	Memory window 0: 32000000-33fff000 (prefetchable)
	Memory window 1: 38000000-39fff000
	I/O window 0: 00003c00-00003cff
	I/O window 1: 00001000-000010ff
	16-bit legacy interface ports at 0001

Software Environment:
GNU bash, version 3.00.16(1)-release (i686-pc-linux-gnu)
KDE/Gnome/E17 - any software environment really

Problem Description:

I have been trying this with various 2.6.13 and 2.6.14
kernels including (archck/ck/kanotix/vanilla etc). 

I start the pcmcia services with /etc/rc.d/pcmcia start, after
which dmesg gives this:

------------------------------------------
Yenta: CardBus bridge found at 0000:01:0a.0 [10cf:1254]
Yenta: ISA IRQ mask 0x04b8, PCI irq 11
Socket status: 30000006
pcmcia: parent PCI bridge I/O window: 0x3000 - 0x3fff
pcmcia: parent PCI bridge Memory window: 0xd0200000 -
0xd02fffff
pcmcia: parent PCI bridge Memory window: 0x30000000 -
0x33ffffff
ACPI: PCI Interrupt 0000:01:0a.1[B] -> Link [LNKB] ->
GSI 11 (level, low) -> IRQ 11
Yenta: CardBus bridge found at 0000:01:0a.1 [10cf:1254]
Yenta: ISA IRQ mask 0x04b8, PCI irq 11
Socket status: 30000006
pcmcia: parent PCI bridge I/O window: 0x3000 - 0x3fff
pcmcia: parent PCI bridge Memory window: 0xd0200000 -
0xd02fffff
pcmcia: parent PCI bridge Memory window: 0x30000000 -
0x33ffffff
---------------------------------------------

Now, if I would insert a memory card (I have only tried this
with SD-cards, as those are the only ones I have at the
moment), I get this:

---------------------------------------------
Probing IDE interface ide2...
hde: Memory Card Adapter, CFA DISK drive
ide2 at 0x3100-0x3107,0x310e on irq 11
hde: max request size: 128KiB
hde: 29120 sectors (14 MB) w/1KiB Cache, CHS=455/2/32
hde: cache flushes not supported
hde:hde: status error: status=0x7f { DriveReady
DeviceFault SeekComplete DataRequest CorrectedError
Index Error }
hde: status error: error=0x00 { }
ide: failed opcode was: unknown
hde: drive not ready for command
ide2: reset: master: ECC circuitry error

ide-cs: hde: Vcc = 3.3, Vpp = 0.0
hde:
hde:
hde:
hde:
hde:
hde:
hde:
hde:
-------------------------------------------------

and 'hde:' continues endlessly, sometimes causing a
very noticeable overall slowdown of the entire system.
Rarely, it even causes a kernel crash and a forced cold
reboot.

Steps to reproduce:
1) start pcmcia service
2) insert SD-memory card

(Note: This bugreport was written a while back but I never posted it. Just to
make sure, I tried it with 2.6.15-rc2-mm1 right now and the results were even
worse. After services are started I get a message saying "no pcmcia driver in
/proc/devices", and after card insertion the kernel will oops badly and then crash)
Comment 1 Bartlomiej Zolnierkiewicz 2005-11-29 23:50:14 UTC
Please post:
- OOPS from 2.6.15-rc2-mm1
- kernel configs for -mm and vanilla 2.6.14
- full dmesgs for -mm and vanilla 2.6.14

Could you also try if 2.6.12 (or some earlier kernel) works?
Comment 2 Khashayar Naderehvandi 2005-11-30 02:00:50 UTC
This is gonna take some time for me; I will most probably need to make a clean
installation of some distro in order to try the vanilla kernels since my root
file system runs on reiser4.

However, I've tried using the card reader under a session of Damn Small Linux,
since it runs a 2.4 kernel. That gave no problems with card reader whatsoever.
Do you want me to try it with an earlier kernel in the 2.6 branch as well?

Last but no least; how do do I capture the OOPS? The kernel does crash
seriously, so there's really not much I can do after the crash. (Hmm.. just
realized I could boot with a live disk and grab the log-file, that way it won't
be overwritten).

I'll get to these things, but unortunately, it's gonna take some time..
Comment 3 Khashayar Naderehvandi 2005-11-30 15:00:41 UTC
Created attachment 6725 [details]
logs+config; 2.6.14.3-vanilla

I hope this helps. These are all the logs from 2.6.14.3. I tried first with a
2.6.13.4-ARCH, and it led to oops+kernel crash. But there were no logs
containing those messages. These aren't in /var/log/kernel.log?

Anyway, I'm a bit in a confusion here. The pcmcia-cs packages isn't supposed to
be used? Cause that's what I've been trying with. If I don't start the pcmcia
services (with /etc/rc.d/pcmcia; part of pcmcia-cs), the kernel gives
absolutely no reaction on my insertion of a memory card.

Am I missing something here?
Comment 4 Khashayar Naderehvandi 2005-12-04 08:50:09 UTC
OK, upon removing pcmcia-cs the result really is that nothing happens
whatsoever. I insert a card and there's no response from anything anywhere.
Also, I don't have the /sys/bus/pcmcia directory.

But :
----------------------------------
[khashayar@nightfrost ~]$ dmesg | grep -i pcmc
pcmcia: parent PCI bridge I/O window: 0x3000 - 0x3fff
pcmcia: parent PCI bridge Memory window: 0xd0200000 - 0xd02fffff
pcmcia: parent PCI bridge Memory window: 0x30000000 - 0x33ffffff
pcmcia: parent PCI bridge I/O window: 0x3000 - 0x3fff
pcmcia: parent PCI bridge Memory window: 0xd0200000 - 0xd02fffff
pcmcia: parent PCI bridge Memory window: 0x30000000 - 0x33ffffff
----------------------------------

Anything?
Comment 5 Bartlomiej Zolnierkiewicz 2005-12-05 00:05:09 UTC
The easiest way to capture OOPS is to take photo of the screen with digital
camera :), the more time consuming method is to use serial console (or netcosole).

-mm contains bugfix for registering removable devices but it seems it is
something different (unsupported command causes device to hang? - we should
try to find which command causes ECC error), anyway kernel shouldn't OOPS
in such situation.
Comment 6 Adrian Bunk 2005-12-05 18:43:18 UTC
*** Bug 5680 has been marked as a duplicate of this bug. ***
Comment 7 Dominik Brodowski 2005-12-08 01:36:02 UTC
To get PCMCIA working now, you need to issue these two commands:
modprobe pcmcia
modprobe yenta_socket

Also, see http://kernel.org/pub/linux/utils/kernel/pcmcia/pcmcia.html for more
information on the pcmcia-cs -> pcmicautils change
Comment 8 Khashayar Naderehvandi 2005-12-08 02:20:18 UTC
It seems I have had two problems. Firstly, the 'pcmcia' module was not being
loaded. Secondly, udev 068. With udev 068 (the version I used when originally
reporting the bug) the kernel would oops. But with udev 076 there is no such
problem. I'm still getting this in the logs though: 

-------------------------
cs: memory probe 0x0c0000-0x0fffff: excluding 0xc0000-0xcffff 0xdc000-0xfffff
cs: memory probe 0x30000000-0x33ffffff: excluding 0x30000000-0x33ffffff
cs: memory probe 0x60000000-0x60ffffff: clean.
cs: memory probe 0xa0000000-0xa0ffffff: clean.
cs: memory probe 0xd0200000-0xd02fffff: excluding 0xd0200000-0xd020ffff
Probing IDE interface ide2...
hde: Memory Card Adapter, CFA DISK drive
ide2 at 0x3100-0x3107,0x310e on irq 11
hde: max request size: 128KiB
hde: 29120 sectors (14 MB) w/1KiB Cache, CHS=455/2/32
hde: cache flushes not supported
 hde:hde: status error: status=0x7f { DriveReady DeviceFault SeekComplete
DataRequest CorrectedError Index Error }
hde: status error: error=0x00 { }
ide: failed opcode was: unknown
hde: drive not ready for command
ide2: reset: master: ECC circuitry error
 hde1
ide-cs: hde: Vcc = 3.3, Vpp = 0.0
 hde: hde1
-----------------------------------------

But the last output of "hde: hde1" only shows once (sometimes twice or thrice,
but no more), and I can mount the card no problem.

So, the question is: is this a bug with udev 068, and do you guys still want me
to post the oops (still looking for a digi-cam)?
Comment 9 Dominik Brodowski 2006-01-10 09:20:40 UTC
It seems to me that this bug is solved?
Comment 10 Bartlomiej Zolnierkiewicz 2006-01-12 06:20:45 UTC
Original IDE bug - yes but it seems that there are two other problems:
- udev-0.68 shouldn't OOPS the kernel
- IDE shouldn't send the unsupported command to the device

Khashayar, could you close this bug and open two new ones
(one for udev OOPS-ing kernel and one for IDE error messages).

Thanks!
Comment 11 Alan 2007-06-18 08:03:35 UTC
Closing as pushed to other bugs (Master ECC is the flash card hardware crashing usually)

Note You need to log in before you can comment on or make changes to this bug.