Gentoo Forums
Gentoo Forums
Gentoo Forums
Quick Search: in
Error segfault when cloning sata between x86 boards
View unanswered posts
View posts from last 24 hours

Goto page Previous  1, 2, 3  Next  
Reply to topic    Gentoo Forums Forum Index Installing Gentoo
View previous topic :: View next topic  
Author Message
NeddySeagoon
Administrator
Administrator


Joined: 05 Jul 2003
Posts: 43194
Location: 56N 3W

PostPosted: Sat Jul 14, 2018 12:28 pm    Post subject: Reply with quote

ranran,

Use a live CD or live CD image on a USB stick.
System Rescue CD is highly regarded as its Gentoo.

Any live CD will do.
_________________
Regards,

NeddySeagoon

Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.
Back to top
View user's profile Send private message
krinn
Watchman
Watchman


Joined: 02 May 2003
Posts: 7071

PostPosted: Sat Jul 14, 2018 12:30 pm    Post subject: Reply with quote

http://www.system-rescue-cd.org/ provide a "kick ass" gentoo base generic livecd, and better than anything, it could be use with a simple usb key.
Back to top
View user's profile Send private message
Chiitoo
Administrator
Administrator


Joined: 28 Feb 2010
Posts: 1771
Location: Here and Away Again

PostPosted: Sat Jul 14, 2018 1:29 pm    Post subject: Reply with quote

ranran wrote:
Probably becuase my 2 previous seperated questions were merged to one it is makes it now less understood to anyone who read the above messages.

While I agree that the other topics merged in may be somewhat confusing, which is often unavoidable in cases like this, I felt that the information in them was important enough to have here for the full context, to still go for it.

Depending on how strongly you feel that they should be separate topics, we can always consider de-merging them.

ranran wrote:
We clone using clonezilla. So as I understand there souldn't be any issue of open files when using this tool. So I am not sure yet why there is segfault on open files.

Could you confirm if the source for the clone was online/mounted, and in use during the operation, or unmounted/offline?

A quick look at Clonezilla tells me this:

https://clonezilla.org/ wrote:
Limitations:
  • [...]
  • Online imaging/cloning is not implemented yet. The partition to be imaged or cloned has to be unmounted.
  • [...]

_________________
Kind Regards,
~ The Noob Unlimited ~

Sore wa sore, kore wa kore.
Back to top
View user's profile Send private message
ranran
n00b
n00b


Joined: 11 Jul 2018
Posts: 69

PostPosted: Sat Jul 14, 2018 2:10 pm    Post subject: Reply with quote

I would please like to ask additional question on the same topic. It is more related to the second stage, after we (hopefully) shall manage to boot with sata.
In the next step we need to move to mmc, becuase the internall mcc on the board is the actual boot medium. This is actualkly what we need to achieve (booting of board from mmc).

So, my question is:
Currently we don't use any initramfs in image.
Is it required to use initramfs for mmc boot , or can we use mmc partitions for all boot sequence, without adding initramfs:
BIOS->GRUB (mmc)->ROOTFS (mmc)

Thanks,
ranran
Back to top
View user's profile Send private message
NeddySeagoon
Administrator
Administrator


Joined: 05 Jul 2003
Posts: 43194
Location: 56N 3W

PostPosted: Sat Jul 14, 2018 2:50 pm    Post subject: Reply with quote

ranran,

1. The BIOS must support booting from mmc.
2. The kernel must have everything needed to boot built in.
3. It just works.

There may be a complication with boot loaders. Boot loaders have to make their own arrangements for reading the files needed to boot.
I don't know if grub supports mmc. It may just make BIOS calls, so it won't even know.
Boot loaders cannot rely on the kernel, since their purpose is to load the kernel.

In short, an initrd is optional, depending on your kernel configuration.
_________________
Regards,

NeddySeagoon

Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.
Back to top
View user's profile Send private message
ranran
n00b
n00b


Joined: 11 Jul 2018
Posts: 69

PostPosted: Sat Jul 14, 2018 2:52 pm    Post subject: Reply with quote

Chiitoo wrote:

A quick look at Clonezilla tells me this:

https://clonezilla.org/ wrote:
Limitations:
  • [...]
  • Online imaging/cloning is not implemented yet. The partition to be imaged or cloned has to be unmounted.
  • [...]


Hi,

I boot from clonezilla live CD, so as far as I understand, the partitions are not mounted. (It is not that I boot into gentoo and then do the clone, but I boot directly into clonezilla live).
Yet, I will verify next time I do the clonezilla image, if I miss something in setting.
I also want to find if I can add sata plug into the 1st board, (which is the source of clone) , so that I will verify first that my image on sata works without any issues on the 1st board (original , source), from which I cloned the image.
This will give a verification for the cloned image.

Thanks,
ranran
Back to top
View user's profile Send private message
ranran
n00b
n00b


Joined: 11 Jul 2018
Posts: 69

PostPosted: Sun Jul 15, 2018 9:14 am    Post subject: Reply with quote

Hello,

I have put:

portwell_cpuinfo
congatec_cpuinfo
portwell_lspci
congatec_lspci
+
emerge_info

portwell is the source (1st board) , congatec is the 2nd (new board)
https://drive.google.com/drive/folders/1DekaQEfUmyJo_HeaaDfL2ynwvtFZFlok?usp=sharing

Thank for any advise,
ranran
Back to top
View user's profile Send private message
NeddySeagoon
Administrator
Administrator


Joined: 05 Jul 2003
Posts: 43194
Location: 56N 3W

PostPosted: Sun Jul 15, 2018 9:47 am    Post subject: Reply with quote

ranran,

Sorry but I'm not about to download random things trom the internet.
You need to put the files onte a pastebin site so that the downloads are sandboxed in my browser.

Thats why I asked you to use wgetpaste or put things onte a pastebin site some other way.

Code:
$ wgetpaste -S
Services supported: (case sensitive):
   Name:        | Url:
   =============|=================
    bpaste      | https://bpaste.net/
    codepad     | http://codepad.org/
    dpaste      | http://dpaste.com/
    gists       | https://api.github.com/gists
   *poundpython | https://paste.pound-python.org/
Try one of those sites.

As you boot with Clonezillia, the drive you want to image will not be mounted.
_________________
Regards,

NeddySeagoon

Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.
Back to top
View user's profile Send private message
ranran
n00b
n00b


Joined: 11 Jul 2018
Posts: 69

PostPosted: Sun Jul 15, 2018 10:05 am    Post subject: Reply with quote

sorry , here it is:

congatec info
https://paste.pound-python.org/show/EeKJvZgvw3sAhNGMFhoq/

portwell info
https://paste.pound-python.org/show/3RfEJWfgjg3G6XCRNB8j/

gentoo info
https://paste.pound-python.org/show/aXshnckCaz3bPSOOxjJ2/

Thanks,
ranran
Back to top
View user's profile Send private message
NeddySeagoon
Administrator
Administrator


Joined: 05 Jul 2003
Posts: 43194
Location: 56N 3W

PostPosted: Sun Jul 15, 2018 10:33 am    Post subject: Reply with quote

ranran,

Your destination board has
Code:
00:0f.0 Communication controller: Intel Corporation Celeron N3350/Pentium N4200/Atom E3900 Series Trusted Execution Engine (rev 0b)
   Subsystem: Intel Corporation Celeron N3350/Pentium N4200/Atom E3900 Series Trusted Execution Engine
   Kernel driver in use: mei_me
   Kernel modules: mei_me

which is not fitted to your source board.

That Communication controller mediates lots of things, so if your kernel is missing mei_me, I would expect your clone to fail in its new home.

Both boards have the silicon for an ahci SATA interface. Thats not to say that they have the tracking and connectors.
As the South Bridges are BGAs, if the tracking is missing, you can't add the SATA connectors.

The flags differ between the two CPUs. I don't know enough about the flags to say if it matters. However, if the source supports instructions that the destination does not and those instructions are used in the code, the code won't run in its new home.
You need to understand what the differences in the flags entries in both /proc/cpuinfo files mean. You can ignore common ones.

From your emerge --info
Code:
Timestamp of repository gentoo: Sun, 20 Mar 2016 12:00:01 +0000
Thats a very old Gentoo ... over two years.
Is there a reason that you need to use such an old system?

Something appears to be updated somewhere as /proc/cpuinfo contains
Code:
bugs      : monitor cpu_meltdown spectre_v1 spectre_v2


Do you need the real time kernel patch set with what appears to be a KDE desktop system?
_________________
Regards,

NeddySeagoon

Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.
Back to top
View user's profile Send private message
ranran
n00b
n00b


Joined: 11 Jul 2018
Posts: 69

PostPosted: Sun Jul 15, 2018 11:30 am    Post subject: Reply with quote

NeddySeagoon wrote:
ranran,

Your destination board has
Code:
00:0f.0 Communication controller: Intel Corporation Celeron N3350/Pentium N4200/Atom E3900 Series Trusted Execution Engine (rev 0b)
   Subsystem: Intel Corporation Celeron N3350/Pentium N4200/Atom E3900 Series Trusted Execution Engine
   Kernel driver in use: mei_me
   Kernel modules: mei_me

which is not fitted to your source board.

That Communication controller mediates lots of things, so if your kernel is missing mei_me, I would expect your clone to fail in its new home.


I will check this !

NeddySeagoon wrote:


The flags differ between the two CPUs. I don't know enough about the flags to say if it matters. However, if the source supports instructions that the destination does not and those instructions are used in the code, the code won't run in its new home.
You need to understand what the differences in the flags entries in both /proc/cpuinfo files mean. You can ignore common ones.


Do you think that when building kernel, it automatically build with this flags ?

Do I need to build kernel only or other tools & libraries too / gentoo from scratch ?

NeddySeagoon wrote:

From your emerge --info
Code:
Timestamp of repository gentoo: Sun, 20 Mar 2016 12:00:01 +0000
Thats a very old Gentoo ... over two years.
Is there a reason that you need to use such an old system?

Something appears to be updated somewhere as /proc/cpuinfo contains
Code:
bugs      : monitor cpu_meltdown spectre_v1 spectre_v2


Do you need the real time kernel patch set with what appears to be a KDE desktop system?


Yes, we use RT kernel with desktop.

Thanks for all comments above,
ranran
Back to top
View user's profile Send private message
NeddySeagoon
Administrator
Administrator


Joined: 05 Jul 2003
Posts: 43194
Location: 56N 3W

PostPosted: Sun Jul 15, 2018 12:12 pm    Post subject: Reply with quote

ranran,

ranran wrote:
Do you think that when building kernel, it automatically build with this flags ?
Do I need to build kernel only or other tools & libraries too / gentoo from scratch ?


The kernel build system is quite conservative. It does not optimise for the host processor by default.
It can be made to do so though.

About the other tools & libraries, I need to understand what the flags do.
From your emerge --info, you have
Code:
CFLAGS="-O2 -pipe"
That means that CPU optimisations are not set, so the code should run on any amd64 CPU.
If your install was build that way, it should just work ... other than bits of kernel missing.

You didn't say why you need the real time kernel, nor why you need such an old install.
It will be full of security issues. Does that matter?
_________________
Regards,

NeddySeagoon

Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.
Back to top
View user's profile Send private message
ranran
n00b
n00b


Joined: 11 Jul 2018
Posts: 69

PostPosted: Sun Jul 15, 2018 2:56 pm    Post subject: Reply with quote

Hi,

I made more checks:
1. mei_me modules exist in kernel (as kernel module)
2. trying to recompile and install kernel on target (from live cd) did not help. I get the same results.
3. The old gentoo image is required because we have our installations in this gentoo and we don't want to restore what changes we need to do now, if we install gentoo from scratch.
4. We use RT with application, it is a requirement for a specific application running on gentoo.

Seems that I am a bit lost here with this issue, I don't actually know what changes are required on kernel/gentoo to make this starting image work on the new hardware.

Anyway, Thanks a lot for this great help !!
ranran
Back to top
View user's profile Send private message
NeddySeagoon
Administrator
Administrator


Joined: 05 Jul 2003
Posts: 43194
Location: 56N 3W

PostPosted: Sun Jul 15, 2018 3:39 pm    Post subject: Reply with quote

ranran,

Post the output of
Code:
ls -l /boot
from the donor system.
You may need to mount /boot first.

Pastebin the kernel .config file from the donor system too.

With the lspci from your earlier posts, I can look your .config over and see if it should boot on both systems or not.

I understand your problem and why you want to use this solution.
Its a very bad thing to use old insecure installs this way. You really need to upgrade.
_________________
Regards,

NeddySeagoon

Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.
Back to top
View user's profile Send private message
ranran
n00b
n00b


Joined: 11 Jul 2018
Posts: 69

PostPosted: Sun Jul 15, 2018 4:13 pm    Post subject: Reply with quote

Hi NeddySeagoon,

Thanks a lot for the suggestions !!!

ls -l /boot
total 14220
-rw-r--r-- 1 root root 2587342 May 31 2018 System.map-3.18.11-rt7
-rw-r--r-- 1 root root 2586525 May 28 2017 System.map-3.18.11-rt7.old
-rw-r--r-- 1 root root 99418 May 31 2018 config-3.18.11-rt7
-rw-r--r-- 1 root root 99429 May 28 2017 config-3.18.11-rt7.old
-rw-r--r-- 1 root root 4588256 May 31 2018 vmlinuz-3.18.11-rt7
-rw-r--r-- 1 root root 4586624 May 28 2017 vmlinuz-3.18.11-rt7.old


.config
https://paste.pound-python.org/show/wLHuGedczFd5HBlGPaJw/


Yes, I understand I might have to do it from scratch if there's no other way...

Thanks,
ranran
Back to top
View user's profile Send private message
NeddySeagoon
Administrator
Administrator


Joined: 05 Jul 2003
Posts: 43194
Location: 56N 3W

PostPosted: Sun Jul 15, 2018 5:01 pm    Post subject: Reply with quote

ranran,

Thats good. There is no initrd.
That means that everything needed to boot must be built into your kernel binary. Modules cannot be loaded until after the root filesystem is mounted because they are in /lib/modules/`uname -r`/

We know it works on portwell, so I only need to look at congatec and see what is missing.

Code:
00:0e.0 Audio device: Intel Corporation Celeron N3350/Pentium N4200/Atom E3900 Series Audio Cluster (rev 0b)
   Subsystem: Intel Corporation Celeron N3350/Pentium N4200/Atom E3900 Series Audio Cluster
   Kernel driver in use: snd_hda_intel
   Kernel modules: snd_hda_intel, snd_soc_skl

snd_soc_skl is not in your kernel. Its SND_SOC_INTEL_SKYLAKE which appeared in the 4.3 kernel. You probably won't get sound on the new board without a kernel update.

Code:
00:1b.0 SD Host controller: Intel Corporation Celeron N3350/Pentium N4200/Atom E3900 Series SDXC/MMC Host Controller (rev 0b)
   Subsystem: Intel Corporation Celeron N3350/Pentium N4200/Atom E3900 Series SDXC/MMC Host Controller
   Kernel driver in use: sdhci-pci
   Kernel modules: sdhci_pci

Needs # CONFIG_MMC is not set to be turned on. Thats a menu. Inside that menu, you need at least
Code:
  │ │     <*>   MMC block device driver                                              │ │ 
  │ │     (8)     Number of minors per block device

  │ │     <*>   Secure Digital Host Controller Interface support                     │ │ 
  │ │     <*>     SDHCI support on PCI bus

The Number of minors per block device sets the makimun number of partitions on an MMC card. Here its 7 as the whole device uses a minor device number too.
If you eventually want to boot from this device, everything must be built in.

The on board
Code:
 00:1c.0 SD Host controller: Intel Corporation Celeron N3350/Pentium N4200/Atom E3900 Series eMMC Controller (rev 0b)
needs the same drivers.

Code:
00:1f.1 SMBus: Intel Corporation Celeron N3350/Pentium N4200/Atom E3900 Series SMBus Controller (rev 0b)
   Subsystem: Intel Corporation Celeron N3350/Pentium N4200/Atom E3900 Series SMBus Controller
   Kernel modules: i2c_i801


Thats off in your kernel
Code:
# CONFIG_I2C_I801 is not set


Rebuild your kernel, using make menuconfig to add in the kernel options above.
Test it on the portwell system. The additions should nave no ill effects.

If that works, clone it again. On the new system, it should boot and you should be able to see the on board eMMC and the MMC slot.

If both boards support booting from USB, I would be tempted to try booting from USB as a half way house.
Mounting root from USB requires rootdelay=10 on the kernel command line.
_________________
Regards,

NeddySeagoon

Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.
Back to top
View user's profile Send private message
ranran
n00b
n00b


Joined: 11 Jul 2018
Posts: 69

PostPosted: Sun Jul 15, 2018 6:40 pm    Post subject: Reply with quote

Hi NeddySeagoon,

That's interesting, I will try that.
Yet, I have doubt if it will also solve the segfault,
I got the segfault on boot, even when trying to boot from sata (not mmc yet) in the 2nd board (congatec).
So, I assume there is no relation between these kernel options to the segfault:
"/etc/init.d/net.lo line 36: wait_for no record of process 0 "
, Right ?


Thanks,
ranran
Back to top
View user's profile Send private message
NeddySeagoon
Administrator
Administrator


Joined: 05 Jul 2003
Posts: 43194
Location: 56N 3W

PostPosted: Sun Jul 15, 2018 6:47 pm    Post subject: Reply with quote

ranran,

I'm not sure. I don't know what that error means.
net.lo is the loopback interface.

Missing i2c_i801 can cause lots of problems as the system management won't work.

Try it and see.
_________________
Regards,

NeddySeagoon

Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.
Back to top
View user's profile Send private message
krinn
Watchman
Watchman


Joined: 02 May 2003
Posts: 7071

PostPosted: Sun Jul 15, 2018 7:20 pm    Post subject: Reply with quote

nobody could really help you with the error, because even the line36 may help (seeing where the error is triggered), nobody use an openrc that is this old ; the network handling has change a lot.
you should use a more recent openrc version, but this may mean, update other components too (which may endup with the "update everything")
Back to top
View user's profile Send private message
NeddySeagoon
Administrator
Administrator


Joined: 05 Jul 2003
Posts: 43194
Location: 56N 3W

PostPosted: Sun Jul 15, 2018 7:25 pm    Post subject: Reply with quote

krinn,

We have a few diehards on sys-apps/openrc-0.13 but not many. I'm on 0.17 but I'm a stick-in-the-mud.
Like you say, I can see updating/reinstalling being the only way ... but we are not there yet.
_________________
Regards,

NeddySeagoon

Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.
Back to top
View user's profile Send private message
NeddySeagoon
Administrator
Administrator


Joined: 05 Jul 2003
Posts: 43194
Location: 56N 3W

PostPosted: Sun Jul 15, 2018 7:27 pm    Post subject: Reply with quote

ranran,

I missed one.
Code:
# CONFIG_LPC_ICH is not set
is needed to be on too.
_________________
Regards,

NeddySeagoon

Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.
Back to top
View user's profile Send private message
krinn
Watchman
Watchman


Joined: 02 May 2003
Posts: 7071

PostPosted: Sun Jul 15, 2018 7:28 pm    Post subject: Reply with quote

I could get reason why people keep an old openrc version, the problem is support, if you keep an old openrc version and you're ready to fix and manage that yourself, it's fine.
That user is not able to do so, leaving only two possibilities:
* an openrc user using that 0.13 coming to help him
* user upgrade openrc to a version any users could help him with, meaning latest
Back to top
View user's profile Send private message
NeddySeagoon
Administrator
Administrator


Joined: 05 Jul 2003
Posts: 43194
Location: 56N 3W

PostPosted: Sun Jul 15, 2018 7:34 pm    Post subject: Reply with quote

krinn,

I'm not convinced that there is a real problem with OpenRC.

It works on one host and fails on another with different hardware, which we know the kernel does not support properly.
The OpenRC error may just be a manifestation of the broken kernel.

So, fix the kernel and it all works again except sound because the 3.18 kernel does not support that sound hardware.
_________________
Regards,

NeddySeagoon

Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.
Back to top
View user's profile Send private message
ranran
n00b
n00b


Joined: 11 Jul 2018
Posts: 69

PostPosted: Mon Jul 16, 2018 9:04 pm    Post subject: Reply with quote

Hi,

I eventually decided to install it from scratch.
Seems that everything works now.

Thanks,
ranran
Back to top
View user's profile Send private message
NeddySeagoon
Administrator
Administrator


Joined: 05 Jul 2003
Posts: 43194
Location: 56N 3W

PostPosted: Mon Jul 16, 2018 9:19 pm    Post subject: Reply with quote

ranran,

What does
Code:
emerge --info
say now ?
_________________
Regards,

NeddySeagoon

Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.
Back to top
View user's profile Send private message
Display posts from previous:   
Reply to topic    Gentoo Forums Forum Index Installing Gentoo All times are GMT
Goto page Previous  1, 2, 3  Next
Page 2 of 3

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum