mung
Posts: 506
Joined: Fri Nov 18, 2011 10:49 am

raspbian dpkg problems on new 20160513 jessie-lite

Sat May 14, 2016 12:50 pm

I posted about problems in the '13th May 2016 Changes to the Raspbian Jessie Desktop' thread, but I suppose the jessie-lite image is not really the desktop as X is not part of the image.

So I start new thread.....

I assume this shows there is problems in the repo dependencies?

I try to add X as shown in my post viewtopic.php?p=974103#p974103
This give serious problems, which I will not go into yet.

So I try a minimal package install on fresh jessie-lite install:

Code: Select all

sudo apt-get install cmake
This give the following error:

Code: Select all

Reading package lists...
Building dependency tree...
Reading state information...
The following extra packages will be installed:
  cmake-data libarchive13 liblzo2-2
Suggested packages:
  codeblocks eclipse ninja-build lrzip
The following NEW packages will be installed:
  cmake cmake-data libarchive13 liblzo2-2
0 upgraded, 4 newly installed, 0 to remove and 0 not upgraded.
Need to get 0 B/3,076 kB of archives.
After this operation, 15.1 MB of additional disk space will be used.
Do you want to continue? [Y/n] database /var/lib/apt/listchanges.db failed to load.
Exception bsddb.db.DBRunRecoveryError: DBRunRecoveryError(-30973, 'BDB0087 DB_RUNRECOVERY: Fatal error, run database recovery -- BDB0060 PANIC: fatal region error detected; run recovery') in  ignored
Selecting previously unselected package cmake-data.
(Reading database ... 
(Reading database ... 5%
(Reading database ... 10%
(Reading database ... 15%
(Reading database ... 20%
(Reading database ... 25%
(Reading database ... 30%
(Reading database ... 35%
(Reading database ... 40%
(Reading database ... 45%
(Reading database ... 50%
(Reading database ... 55%
(Reading database ... 60%
dpkg: unrecoverable fatal error, aborting:
 files list file for package `colord-data' contains empty filename
E: Sub-process /usr/bin/dpkg returned an error code (2)
Also there seems to be problems with the default .bashrc showing some commands unavailable?

I tested the SD card with h2testw.exe and no errors shown, its a brand new kingston 16Gb class 10 'IU1' sdhc.

I did notice that the mirror redirector shows some servers unavailable.

Anyone else seen this problems?

DirkS
Posts: 10017
Joined: Tue Jun 19, 2012 9:46 pm
Location: Essex, UK

Re: raspbian dpkg problems on new 20160513 jessie-lite

Sat May 14, 2016 2:23 pm

mung wrote:I tested the SD card with h2testw.exe and no errors shown, its a brand new kingston 16Gb class 10 'IU1' sdhc.
There have been several reports from users having similar problems with some Kingston cards.
@PhilE mentioned he was researching this and there seems to be a bugfix in the latest kernel / firmware (https://github.com/raspberrypi/linux/co ... 1d19c1d8d3)

You could try to write Jessie Lite again and then as the first action use

Code: Select all

sudo rpi-update
to see if this fixes the problem.

mung
Posts: 506
Joined: Fri Nov 18, 2011 10:49 am

Re: raspbian dpkg problems on new 20160513 jessie-lite

Sat May 14, 2016 5:22 pm

DirkS wrote:
mung wrote:I tested the SD card with h2testw.exe and no errors shown, its a brand new kingston 16Gb class 10 'IU1' sdhc.
There have been several reports from users having similar problems with some Kingston cards.
@PhilE mentioned he was researching this and there seems to be a bugfix in the latest kernel / firmware (https://github.com/raspberrypi/linux/co ... 1d19c1d8d3)

You could try to write Jessie Lite again and then as the first action use

Code: Select all

sudo rpi-update
to see if this fixes the problem.
How can I check what the latest raspbian repo kernel version is and check it against the git kernel update to see if patch is applied? I found this: http://unix.stackexchange.com/questions ... ontains-it but I think I need to install git to try that and using apt-get seems to corrupt the packages required :(

I already tried reflashing the image and trying the update again and I get similar problems, I think you maybe correct about corruption as after three attempts I get different errors (bashrc errors, network config, and dpkg config, probably others too that I have not noticed)

DirkS
Posts: 10017
Joined: Tue Jun 19, 2012 9:46 pm
Location: Essex, UK

Re: raspbian dpkg problems on new 20160513 jessie-lite

Sat May 14, 2016 10:26 pm

How can I check what the latest raspbian repo kernel version is and check it against the git kernel update to see if patch is applied?
It's possible to check that, but it's not relevant here ATM
The Jessie (Lite) has kernel 4.4.9 #884 (from 6/5/16) installed.
That's from before the bugfix I mentioned.
So I would still suggest you test with a fresh image and then run 'sudo rpi-update' to load that fix and take it from there.

mung
Posts: 506
Joined: Fri Nov 18, 2011 10:49 am

Re: raspbian dpkg problems on new 20160513 jessie-lite

Sun May 15, 2016 11:45 pm

DirkS wrote:
How can I check what the latest raspbian repo kernel version is and check it against the git kernel update to see if patch is applied?
It's possible to check that, but it's not relevant here ATM
The Jessie (Lite) has kernel 4.4.9 #884 (from 6/5/16) installed.
That's from before the bugfix I mentioned.
So I would still suggest you test with a fresh image and then run 'sudo rpi-update' to load that fix and take it from there.
So I am guessing the rpi-update does not fix the problem or maybe when I then apt-get a newer kernel is overwritten the rpi-update version?

The fresh install and eventual install of 'apt-get install rpi-update' followed by 'rpi-update' looked as though it worked on reboot but after doing some more apt-get install the system seemed to crash again with config file read errors.

So I am not sure what should be done next, I would rather not make bug reports on github, and wonder if this is something I could get an amazon refund for, is this a faulty SDcard issue or are the kingston cards within spec but the linux drivers are buggy?

Is there an easy to understand test guide for SDcard issues?

mung
Posts: 506
Joined: Fri Nov 18, 2011 10:49 am

Re: raspbian dpkg problems on new 20160513 jessie-lite

Mon May 16, 2016 2:58 pm

Okay I have tried and double checked on a few different rpi (3/B+/zero) and same errors occur, I think the patch does not fix the errors only reduces them.
The card details as shown on the card as as below:
Kingston ®
soc1-g2/16gb
no579-003 a00lf
taiwan FⒸo
The output from udevadm is as follows:

Code: Select all

Udevadm info starts with the device specified by the devpath and then
walks up the chain of parent devices. It prints for every device
found, all possible attributes in the udev rules key format.
A rule to match, can be composed by the attributes of the device
and the attributes from one single parent device.

  looking at device '/devices/platform/soc/20202000.sdhost/mmc_host/mmc0/mmc0:0003/block/mmcblk0':
    KERNEL=="mmcblk0"
    SUBSYSTEM=="block"
    DRIVER==""
    ATTR{ro}=="0"
    ATTR{size}=="30253056"
    ATTR{stat}=="    2629     1599   199226    22140      225      198    55241   186660        0    24130   208800"
    ATTR{range}=="32"
    ATTR{discard_alignment}=="0"
    ATTR{force_ro}=="0"
    ATTR{ext_range}=="32"
    ATTR{alignment_offset}=="0"
    ATTR{inflight}=="       0        0"
    ATTR{removable}=="0"
    ATTR{capability}=="10"

  looking at parent device '/devices/platform/soc/20202000.sdhost/mmc_host/mmc0/mmc0:0003':
    KERNELS=="mmc0:0003"
    SUBSYSTEMS=="mmc"
    DRIVERS=="mmcblk"
    ATTRS{cid}=="4134325344313647307b7807970102d7"
    ATTRS{csd}=="400e005a5b59000073677f800a400027"
    ATTRS{scr}=="0235800300000000"
    ATTRS{date}=="02/2016"
    ATTRS{name}=="SD16G"
    ATTRS{type}=="SD"
    ATTRS{preferred_erase_size}=="12582912"
    ATTRS{fwrev}=="0x0"
    ATTRS{hwrev}=="0x3"
    ATTRS{oemid}=="0x3432"
    ATTRS{manfid}=="0x000041"
    ATTRS{serial}=="0x7b780797"
    ATTRS{erase_size}=="512"

  looking at parent device '/devices/platform/soc/20202000.sdhost/mmc_host/mmc0':
    KERNELS=="mmc0"
    SUBSYSTEMS=="mmc_host"
    DRIVERS==""

  looking at parent device '/devices/platform/soc/20202000.sdhost':
    KERNELS=="20202000.sdhost"
    SUBSYSTEMS=="platform"
    DRIVERS=="sdhost-bcm2835"
    ATTRS{driver_override}=="(null)"

  looking at parent device '/devices/platform/soc':
    KERNELS=="soc"
    SUBSYSTEMS=="platform"
    DRIVERS==""
    ATTRS{driver_override}=="(null)"

  looking at parent device '/devices/platform':
    KERNELS=="platform"
    SUBSYSTEMS==""
    DRIVERS==""


Udevadm info starts with the device specified by the devpath and then
walks up the chain of parent devices. It prints for every device
found, all possible attributes in the udev rules key format.
A rule to match, can be composed by the attributes of the device
and the attributes from one single parent device.

  looking at device '/devices/platform/soc/20202000.sdhost/mmc_host/mmc0/mmc0:0003/block/mmcblk0/mmcblk0p1':
    KERNEL=="mmcblk0p1"
    SUBSYSTEM=="block"
    DRIVER==""
    ATTR{ro}=="0"
    ATTR{size}=="129024"
    ATTR{stat}=="      77       10     1624      160        1        0        1       10        0      160      170"
    ATTR{partition}=="1"
    ATTR{start}=="8192"
    ATTR{discard_alignment}=="8388608"
    ATTR{alignment_offset}=="0"
    ATTR{inflight}=="       0        0"

  looking at parent device '/devices/platform/soc/20202000.sdhost/mmc_host/mmc0/mmc0:0003/block/mmcblk0':
    KERNELS=="mmcblk0"
    SUBSYSTEMS=="block"
    DRIVERS==""
    ATTRS{ro}=="0"
    ATTRS{size}=="30253056"
    ATTRS{stat}=="    2629     1599   199226    22140      228      203    55305   186680        0    24150   208820"
    ATTRS{range}=="32"
    ATTRS{discard_alignment}=="0"
    ATTRS{force_ro}=="0"
    ATTRS{ext_range}=="32"
    ATTRS{alignment_offset}=="0"
    ATTRS{inflight}=="       0        0"
    ATTRS{removable}=="0"
    ATTRS{capability}=="10"

  looking at parent device '/devices/platform/soc/20202000.sdhost/mmc_host/mmc0/mmc0:0003':
    KERNELS=="mmc0:0003"
    SUBSYSTEMS=="mmc"
    DRIVERS=="mmcblk"
    ATTRS{cid}=="4134325344313647307b7807970102d7"
    ATTRS{csd}=="400e005a5b59000073677f800a400027"
    ATTRS{scr}=="0235800300000000"
    ATTRS{date}=="02/2016"
    ATTRS{name}=="SD16G"
    ATTRS{type}=="SD"
    ATTRS{preferred_erase_size}=="12582912"
    ATTRS{fwrev}=="0x0"
    ATTRS{hwrev}=="0x3"
    ATTRS{oemid}=="0x3432"
    ATTRS{manfid}=="0x000041"
    ATTRS{serial}=="0x7b780797"
    ATTRS{erase_size}=="512"

  looking at parent device '/devices/platform/soc/20202000.sdhost/mmc_host/mmc0':
    KERNELS=="mmc0"
    SUBSYSTEMS=="mmc_host"
    DRIVERS==""

  looking at parent device '/devices/platform/soc/20202000.sdhost':
    KERNELS=="20202000.sdhost"
    SUBSYSTEMS=="platform"
    DRIVERS=="sdhost-bcm2835"
    ATTRS{driver_override}=="(null)"

  looking at parent device '/devices/platform/soc':
    KERNELS=="soc"
    SUBSYSTEMS=="platform"
    DRIVERS==""
    ATTRS{driver_override}=="(null)"

  looking at parent device '/devices/platform':
    KERNELS=="platform"
    SUBSYSTEMS==""
    DRIVERS==""


Udevadm info starts with the device specified by the devpath and then
walks up the chain of parent devices. It prints for every device
found, all possible attributes in the udev rules key format.
A rule to match, can be composed by the attributes of the device
and the attributes from one single parent device.

  looking at device '/devices/platform/soc/20202000.sdhost/mmc_host/mmc0/mmc0:0003/block/mmcblk0/mmcblk0p2':
    KERNEL=="mmcblk0p2"
    SUBSYSTEM=="block"
    DRIVER==""
    ATTR{ro}=="0"
    ATTR{size}=="30115840"
    ATTR{stat}=="    2506     1589   196802    21860      235      206    55392   188410        0    24350   210270"
    ATTR{partition}=="2"
    ATTR{start}=="137216"
    ATTR{discard_alignment}=="5242880"
    ATTR{alignment_offset}=="0"
    ATTR{inflight}=="       0        0"

  looking at parent device '/devices/platform/soc/20202000.sdhost/mmc_host/mmc0/mmc0:0003/block/mmcblk0':
    KERNELS=="mmcblk0"
    SUBSYSTEMS=="block"
    DRIVERS==""
    ATTRS{ro}=="0"
    ATTRS{size}=="30253056"
    ATTRS{stat}=="    2629     1599   199226    22140      236      206    55393   188420        0    24470   210560"
    ATTRS{range}=="32"
    ATTRS{discard_alignment}=="0"
    ATTRS{force_ro}=="0"
    ATTRS{ext_range}=="32"
    ATTRS{alignment_offset}=="0"
    ATTRS{inflight}=="       0        0"
    ATTRS{removable}=="0"
    ATTRS{capability}=="10"

  looking at parent device '/devices/platform/soc/20202000.sdhost/mmc_host/mmc0/mmc0:0003':
    KERNELS=="mmc0:0003"
    SUBSYSTEMS=="mmc"
    DRIVERS=="mmcblk"
    ATTRS{cid}=="4134325344313647307b7807970102d7"
    ATTRS{csd}=="400e005a5b59000073677f800a400027"
    ATTRS{scr}=="0235800300000000"
    ATTRS{date}=="02/2016"
    ATTRS{name}=="SD16G"
    ATTRS{type}=="SD"
    ATTRS{preferred_erase_size}=="12582912"
    ATTRS{fwrev}=="0x0"
    ATTRS{hwrev}=="0x3"
    ATTRS{oemid}=="0x3432"
    ATTRS{manfid}=="0x000041"
    ATTRS{serial}=="0x7b780797"
    ATTRS{erase_size}=="512"

  looking at parent device '/devices/platform/soc/20202000.sdhost/mmc_host/mmc0':
    KERNELS=="mmc0"
    SUBSYSTEMS=="mmc_host"
    DRIVERS==""

  looking at parent device '/devices/platform/soc/20202000.sdhost':
    KERNELS=="20202000.sdhost"
    SUBSYSTEMS=="platform"
    DRIVERS=="sdhost-bcm2835"
    ATTRS{driver_override}=="(null)"

  looking at parent device '/devices/platform/soc':
    KERNELS=="soc"
    SUBSYSTEMS=="platform"
    DRIVERS==""
    ATTRS{driver_override}=="(null)"

  looking at parent device '/devices/platform':
    KERNELS=="platform"
    SUBSYSTEMS==""
    DRIVERS==""

0x3432

and the info in /sys/devices/platform/soc/20202000.sdhost/mmc_host/mmc0/mmc0\:0003/oemid is 0x3432

I may try and post similar bug report comment on github raspberrypi/linux if I don't need to setup an account.

mung
Posts: 506
Joined: Fri Nov 18, 2011 10:49 am

Re: raspbian dpkg problems on new 20160513 jessie-lite

Wed May 18, 2016 2:20 pm

kinstonsd2_e.jpg
kinstonsd2_e.jpg (63.36 KiB) Viewed 4180 times
kinst3.jpg
kinst3.jpg (4.79 KiB) Viewed 4180 times
kinstonsd1_4.jpg
kinstonsd1_4.jpg (37.21 KiB) Viewed 4180 times
I post here photos of the cards, I bought all at same time from amazon.co.uk, 4 16Gb cards and 1 32Gb about three weeks ago, I have not opened the 32gb and 2 other 16gb cards.

Seems very strange that cards from same batch should be different (code numbers and packaging differ and no laser etch on reverse of one or the cards.

I get the feeling they could be dodgy cards but amazon listed them as coming direct from kingston.

I did notice that the cards seemed warmer than other cards.

BerryPicker
Posts: 177
Joined: Tue Oct 16, 2012 3:03 pm
Location: The East of England

Re: raspbian dpkg problems on new 20160513 jessie-lite

Wed May 18, 2016 4:20 pm

mung wrote:I tested the SD card with h2testw.exe and no errors shown, its a brand new kingston 16Gb class 10 'IU1' sdhc. [snip] Anyone else seen this problems?
My same Kingston 16GB class 10 card by from MyMemory has similar behaviour. It will accept the Jessie lite image, but then throws errors when trying to install pikrellcam. Errors start to show after processing triggers for man-db. I was going to return it, but no card errors show when the card is tested fsck, and it works fine in a sony camera. I've had no such problem with Jessie light on a San Disk class 10 16GB.

DirkS
Posts: 10017
Joined: Tue Jun 19, 2012 9:46 pm
Location: Essex, UK

Re: raspbian dpkg problems on new 20160513 jessie-lite

Wed May 18, 2016 4:41 pm

BerryPicker wrote:
mung wrote:I tested the SD card with h2testw.exe and no errors shown, its a brand new kingston 16Gb class 10 'IU1' sdhc. [snip] Anyone else seen this problems?
My same Kingston 16GB class 10 card by from MyMemory has similar behaviour. It will accept the Jessie lite image, but then throws errors when trying to install pikrellcam. Errors start to show after processing triggers for man-db. I was going to return it, but no card errors show when the card is tested fsck, and it works fine in a sony camera. I've had no such problem with Jessie light on a San Disk class 10 16GB.
fsck is not great for checking out faulty cards.
A full write / read cycle is more effective (on Windows use e.g .h2testw).
Even that may not show up any errors. AIUI some of the Kingston cards have problems with eerasing small blocks (https://github.com/raspberrypi/linux/co ... 1d19c1d8d3)

mung
Posts: 506
Joined: Fri Nov 18, 2011 10:49 am

Re: raspbian dpkg problems on new 20160513 jessie-lite

Wed May 18, 2016 11:25 pm

I just thought I would round off the issue with another link on github.

https://github.com/raspberrypi/linux/issues/1416

Seems a number of kingston SDHC cards have problems with the 'erase' command, the patch is not applied to the kernel available through 'rpi-update' so grab the images linked on the first github link in this thread, copy the kernels to your boot partition and all seems to work okay.

I am not sure though if this patch should be used, maybe all those with effected cardss should complain to kingston and get a refund?

I may search kingstons web site and see if I can find a spec datasheet for the cards(though I am probably too lazy to bother), I will have a play over the weekend and try to decide if I should return to amazon for a refund.

DirkS
Posts: 10017
Joined: Tue Jun 19, 2012 9:46 pm
Location: Essex, UK

Re: raspbian dpkg problems on new 20160513 jessie-lite

Wed May 18, 2016 11:43 pm

mung wrote: the patch is not applied to the kernel available through 'rpi-update'
According to Phil's comment in https://github.com/raspberrypi/linux/is ... -219081163 it should be...

mung
Posts: 506
Joined: Fri Nov 18, 2011 10:49 am

Re: raspbian dpkg problems on new 20160513 jessie-lite

Thu May 19, 2016 9:24 am

DirkS wrote:
mung wrote: the patch is not applied to the kernel available through 'rpi-update'
According to Phil's comment in https://github.com/raspberrypi/linux/is ... -219081163 it should be...
In the commit comments https://github.com/raspberrypi/linux/co ... 1d19c1d8d3 phil says he thinks the rpi-update build stopped short of his commit.

And unfortunately I left a build with jessie and new kernel running overnight last night and checked this morning to find filesystem corruption in the build script.

So still problems still occuring, but of the two cards shown in the photos on this thread the overnight build failure was the card without the laser etched serial number on the back, the other card seemed to run the build script okay last night (about 2 hour build time). I seems strange that the build script is corrupted as it obviously was okay last night as I read the text file, I am guessing there maybe some 'wear leveling' and the script file got moved to a faulty area of the memory card?

I think this has totally fininished any chance of having confidence in the kingston cards, I shall have to return the whole batch of them to amazon unless someone can offer some help on this.

texy
Forum Moderator
Forum Moderator
Posts: 5160
Joined: Sat Mar 03, 2012 10:59 am
Location: Berkshire, England

Re: raspbian dpkg problems on new 20160513 jessie-lite

Thu May 19, 2016 9:47 am

Perhaps this thread should move to a more appropriate forum in order to gain attention from the right people?
Texy
Various male/female 40- and 26-way GPIO header for sale here ( IDEAL FOR YOUR PiZero ):
https://www.raspberrypi.org/forums/viewtopic.php?f=93&t=147682#p971555

martinhj
Posts: 23
Joined: Mon Jan 09, 2012 1:05 pm

Re: raspbian dpkg problems on new 20160513 jessie-lite

Tue Feb 14, 2017 12:33 pm

Took these old cards back out of the drawer to try again and they do indeed work now...that said I did have a Class 4 Kingston 16GB card that I think was starting to fail....thankfully I had an image of my card but couldnt find another Kingston card they would fit on (batches can vary in size and all my others had slightly less sectors...bought a SanDisk to try before spending time faffing to reduce the image size to fit and it worked. I will prob just bu SanDisk in future!

Return to “General discussion”