sd card not recognized

[feydreva]

Yesterday, I used was with the ASU software.
I formatted my SD card, with a 6Bg partition in ext3. That was working fine.

Today I use the FSO software, and it is not recognize.
I did a dmesg to check what was going on :

EXT3-fs error (device mmcblk0): ext3_check_descriptors: Block bitmap for group 1 not in group (block 0)!
EXT3-fs: group descriptors corrupted!

[hedora]

Did you suspend/resume (is the power button menu set to "dim then lock")? A known bug causes suspend/resume to corrupt SD cards.

[feydreva]

I was definitely set up in "dim then lock" while I was running ASU.
Unfortunately, I haven't paid enough attention to sys it s because of that.

If we need more data about it, I can do the test....

Feydreva

[zecke]

Where is "dim then lock" in ASU? Could it be that you use the Qtopia image? Which kernel version do you use?

[hedora]

It's under the power button menu on 2007.2. It makes the linux kernel go into suspend mode instead of just turning off the screen. I don't know where the power management settings live in the other images.

It's a known issue in the current kernel. OpenMoko? is working on it:

(early report)
​http://kerneltrap.org/mailarchive/openmoko-community/2008/7/23/2652514

(explanation of problem from andy)
​http://kerneltrap.org/mailarchive/openmoko-community/2008/7/23/2653864

[rhk]

2007.2: My 8gb (one large ext3 partition) SD drops the partition every now and then: sometimes I feel like it has something to do with GPS (like using tangogps, loading maps from the card, then when I turn the GPS off, tango can't find the maps anymore -> the /media/card is empty) and sometimes.. it just drops.

Sometimes I can remount the partition (usually only by rebooting), sometimes I have to create the partition again with fdisk (but the data is still there and I can use it).

This is what my dmesg says:

(a huge list of rows like: 
evbug.c: Event. Dev: <NULL>, Type: 2, Code: 0, Value: 360
---clip clip--
evbug.c: Event. Dev: <NULL>, Type: 0, Code: 0, Value: 0
PM: Finishing wakeup.
Restarting tasks ... <6>soc-audio soc-audio: starting resume work
soc-audio soc-audio: resume work completed
dma2: channel has nothing loaded
glamo-mci glamo-mci.0: powered (vdd = 15) clk: 195kHz div=255 (req: 195kHz). Bus width=0
glamo-mci glamo-mci.0: powered (vdd = 15) clk: 195kHz div=255 (req: 195kHz). Bus width=0
glamo-mci glamo-mci.0: powered (vdd = 15) clk: 195kHz div=255 (req: 195kHz). Bus width=0
glamo-mci glamo-mci.0: powered (vdd = 15) clk: 195kHz div=255 (req: 195kHz). Bus width=0
glamo-mci glamo-mci.0: powered (vdd = 15) clk: 195kHz div=255 (req: 195kHz). Bus width=0
glamo-mci glamo-mci.0: powered (vdd = 15) clk: 16666kHz div=2 (req: 16666kHz). Bus width=0
glamo-mci glamo-mci.0: powered (vdd = 15) clk: 16666kHz div=2 (req: 16666kHz). Bus width=2
mmc0: new high speed SDHC card at address d555
evbug.c: Event. Dev: FIXME, Type: 1, Code: 116, Value: 0
evbug.c: Event. Dev: FIXME, Type: 0, Code: 0, Value: 0
mmcblk1: mmc0:d555 SU08G 7977472KiB
 mmcblk1: p1
Buffer I/O error on device mmcblk0, logical block 0
lost page write due to I/O error on mmcblk0
EXT2-fs warning: mounting unchecked fs, running e2fsck is recommended
done.
Alignment trap: phone-kit (1441) PC=0x0001214c Instr=0xe5970008 Address=0xbed1e663 FSR 0x013
--another huge list of these rows:
evbug.c: Event. Dev: <NULL>, Type: 3, Code: 0, Value: 153
evbug.c: Event. Dev: <NULL>, Type: 3, Code: 1, Value: 548

so as you see, it might have also something to do with suspend. Though it has happened also when the power save features have been turned off.

[andy]

What you pasted from dmesg looks exactly like your card takes too long to respond to the first bulk packet, something that was seen before in #1743.

But, your other syslog traffic is funny. What version is this kernel?

[rhk]

Replying to andy:

But, your other syslog traffic is funny. What version is this kernel?

Linux om-gta02 2.6.24 #1 PREEMPT Tue Aug 5 04:27:23 CEST 2008 armv4tl unknown

[rhk]

I'm not sure if this is the same bug as the one reported.. If not, I can file it as new:

The SD card seems to stop working every now and then: this time again it might have something to do with suspend: The phone has been connected with the USB on the computer and was suspended once before I switched to power management: disabled.

Here's some dmesg: ​http://pastebin.com/f55f4cc68

But the funny thing is that you never know what's answering when you type ls /media/card..

root@om-gta02:/media/card# ls /dev/|grep mmc
mmcblk1
mmcblk1p1
root@om-gta02:/media/card# sudo mount /dev/mmcblk1 /media/card/
-sh: sudo: not found
root@om-gta02:/media/card# mount /dev/mmcblk1 /media/card/
root@om-gta02:/media/card# ls
maps
root@om-gta02:/media/card# ls
maps
root@om-gta02:/media/card# cd maps/
root@om-gta02:/media/card/maps# ls
4free      g_road     g_terrain  osm
root@om-gta02:/media/card/maps# cd ..
root@om-gta02:/media/card# ls
20080805_074751.log  20080806_091247.log  20080807_081622.log  music
20080805_152715.log  20080806_165845.log  hp                   orrery
20080805_191632.log  20080806_170753.log  lost+found
20080805_202044.log  20080807_080610.log  maps
root@om-gta02:/media/card#

on the card I have the log files, hp, maps, music, orrery and so on, on /media/card (on the internal memory of the phone) I have only the maps folder (tangogps wrote stuff there when the card was not mounted).

So I mounted the card as /media/card but right after the mount command it didn't work. Then I did something else and suddenly it works again. So either it's very random or mounting is slow or something..

[andy]

Funny... but the logs you gave at pastebin don't show anything about it, that event debugging stuff is unrelated. It shows only clean mounts.

Is this micro SD card large and cheap, or is it the shipping 512MB one, something else?

[rhk]

The card is SDSDQR-8192-E12M, Sandisk 8GB microSDHC, class4

[andy]

Just based on the size, I could wildly guess this is somehow related to the timeout issue, but where are the noisy buffer IO errors that should be in the log then? I never saw any card here return from the mount action successfully with the old mountpoint showing.

Is it possible it was already mounted somewhere else in the filesystem before these actions? Mount's behaviour changes radically then.

[rhk]

Replying to andy:

Is it possible it was already mounted somewhere else in the filesystem before these actions? Mount's behaviour changes radically then.

Actually yes, for some reason it was automounted:
/dev/mmcblk1 on /media/mmcblk1 type ext2 (rw,sync)

This is what fstab looks like:

# Root and Pseudo
/dev/mtdblock4  /               jffs2   rw,noatime              1  1
proc            /proc           proc    defaults                0  0

# devpts?
# usb?

# Temporary
tmpfs           /var/volatile   tmpfs   mode=0755               0  0
tmpfs           /dev/shm                tmpfs   mode=0777                       0  0

# microSD slot
/dev/mmcblk0    /media/card     auto    defaults,async,noauto   0  0

# USB Storage
#/dev/sda1 /media/hdd         vfat    noauto,umask=000,noatime,iocharset=utf8,codepage=932    0       0

I'll restart and report what happens with powersave turned on.

[andy]

Huh that's new then, maybe it is some kind of automount that happens upstream of udev now in the rootfs. It's not coming out of fstab obviously. Anyway if that is the issue then it's not a real problem I think.

[rhk]

So I left the phone alone, powersave activated (so it was able to suspend) for some hours.

I come back and see only terminal (no X), messages shown below. The phone is stuck, no ssh in. Power button is useless as well, had to remove the battery.

{{
mmcblk1: mmc0:d555 SU08G 7977472KiB

mmcblk1: p1

dma2: channel has nothing loaded
glamo-mci glamo-mci.0: powered (vdd = 15) clk: 195kHz div=255 (req: 195kHz). Bus width=0
glamo-mci glamo-mci.0: powered (vdd = 15) clk: 195kHz div=255 (req: 195kHz). Bus width=0
glamo-mci glamo-mci.0: powered (vdd = 15) clk: 195kHz div=255 (req: 195kHz). Bus width=0
glamo-mci glamo-mci.0: powered (vdd = 15) clk: 195kHz div=255 (req: 195kHz). Bus width=0
glamo-mci glamo-mci.0: powered (vdd = 15) clk: 195kHz div=255 (req: 195kHz). Bus width=0
glamo-mci glamo-mci.0: powered (vdd = 15) clk: 16666kHz div=2 (req: 16666kHz). Bus width=0
glamo-mci glamo-mci.0: powered (vdd = 15) clk: 16666kHz div=2 (req: 16666kHz). Bus width=0
mmc0: new high speed SDHC card at address d555
mmcblk1: mmc0:d555 SU08G 7977472KiB

mmcblk1: p1

dma2: channel has nothing loaded
glamo-mci glamo-mci.0: powered (vdd = 15) clk: 195kHz div=255 (req: 195kHz). Bus width=0
glamo-mci glamo-mci.0: powered (vdd = 15) clk: 195kHz div=255 (req: 195kHz). Bus width=0
glamo-mci glamo-mci.0: powered (vdd = 15) clk: 195kHz div=255 (req: 195kHz). Bus width=0
glamo-mci glamo-mci.0: powered (vdd = 15) clk: 195kHz div=255 (req: 195kHz). Bus width=0
glamo-mci glamo-mci.0: powered (vdd = 15) clk: 195kHz div=255 (req: 195kHz). Bus width=0
glamo-mci glamo-mci.0: powered (vdd = 15) clk: 16666kHz div=2 (req: 16666kHz). Bus width=0
glamo-mci glamo-mci.0: powered (vdd = 15) clk: 16666kHz div=2 (req: 16666kHz). Bus width=0
mmc0: new high speed SDHC card at address d555
mmcblk1: mmc0:d555 SU08G 7977472KiB

mmcblk1: p1

EXT2-fs warning: mounting unchecked fs, running e2fsk is recommended
dma2: channel has nothing loaded
glamo-mci glamo-mci.0: powered (vdd = 15) clk: 195kHz div=255 (req: 195kHz). Bus width=0
glamo-mci glamo-mci.0: powered (vdd = 15) clk: 195kHz div=255 (req: 195kHz). Bus width=0
glamo-mci glamo-mci.0: powered (vdd = 15) clk: 195kHz div=255 (req: 195kHz). Bus width=0
glamo-mci glamo-mci.0: powered (vdd = 15) clk: 195kHz div=255 (req: 195kHz). Bus width=0
glamo-mci glamo-mci.0: powered (vdd = 15) clk: 195kHz div=255 (req: 195kHz). Bus width=0
glamo-mci glamo-mci.0: powered (vdd = 15) clk: 16666kHz div=2 (req: 16666kHz). Bus width=0
glamo-mci glamo-mci.0: powered (vdd = 15) clk: 16666kHz div=2 (req: 16666kHz). Bus width=0
mmc0: new high speed SDHC card at address d555
mmcblk1: mmc0:d555 SU08G 7977472KiB

mmcblk1: p1

}}