TinyCore RedPill Loader (TCRP)

[logze]

Hi @pocopico

I'm using the TCRP stable version, and everything works fine except the SATA sequence.

My motherboard(B250 chipset) has 1 sata controller that support 6 sata ports.

But there are only 4 physical ports sata3~sata6 so it shows DISK3-DISK6 in the DSM.

SATA1&SATA2 was not show even in the BIOS settings.

I have try add  sata_remap  into grub.cfg boot argument but it seems not work.

Does TCRP support this boot argument  or is there any other way to remap the SATA sequence?

Edited September 29, 2022 by logze

[pocopico]

7 minutes ago, logze said:

Hi @pocopico

I'm using the TCRP stable version, and everything works fine except the SATA sequence.

My motherboard(B250 chipset) has 1 sata controller that support 6 sata ports.

But there are only 4 physical ports sata3~sata6 so it shows DISK3-DISK6 in the DSM.

SATA1&SATA2 was not show even in the BIOS settings.

I have try add  sata_remap  into grub.cfg boot argument but it seems not work.

Does TCRP support this boot argument  or is there any other way to remap the SATA sequence?

 

Hi,

If you have a hotplug option in your BIOS setup then you can disable it.

 

There is also a linux command line option named HddHotplug=0 and should be 0 i assume that since this is the default you have not changed that.

 

sata_remap should work but it does not work for all platforms. You can also use a dtb based platform like DS920 that uses the dtb file to assign the disks to ports automatically 

[Peter Suh]

24 minutes ago, logze said:

Hi @pocopico

I'm using the TCRP stable version, and everything works fine except the SATA sequence.

My motherboard(B250 chipset) has 1 sata controller that support 6 sata ports.

But there are only 4 physical ports sata3~sata6 so it shows DISK3-DISK6 in the DSM.

SATA1&SATA2 was not show even in the BIOS settings.

I have try add  sata_remap  into grub.cfg boot argument but it seems not work.

Does TCRP support this boot argument  or is there any other way to remap the SATA sequence?

 

A muxed Sataport is expected, but as a last resort, try setting SataportMap="",DiskIdxMap="" .
If this setting does not work as desired, try switching to DS920+ as the next best work.

DS920+ is a Device Tree-based model and does not involve SataportMap settings.

[DidaJI]

Hi,

I have a problem with the bootloader starting correctly after the DS3622xs+ build. I use "RedPill DS3622xs+ v7.1.1-42962 Beta (USB, Verbose) booting properly". 

It ends up listing:
Loading Linux...

Loading initramfs...

Starting kernel with USB boot

 

That should be fine, but the device won't show up on the network. I connected to the serial console and the log ends with this one:

[    3.057422] Trying to unpack rootfs image as initramfs...
[    3.076107] Freeing initrd memory: 25912K
[    3.081179] redpill: loading out-of-tree module taints kernel.
[    3.087073] redpill: module verification failed: signature and/or required key missing - tainting kernel
[    3.097874] Kernel panic - not syncing: Fatal exception
[    3.103114] CPU: 5 PID: 769 Comm: insmod Tainted: G           OE   4.4.180+ #42962
[    3.110714] Hardware name: Gigabyte Technology Co., Ltd. B560M DS3H V2/B560M DS3H V2, BIOS F7 03/25/2022
[    3.120171]  0000000000000000 ffff880893207c60 ffffffff812dce6b ffffffffa000c0a2
[    3.127810]  ffff8808958290c0 ffff880893207cd8 ffffffff811217f7 0000000000000008
[    3.135437]  ffff880893207ce8 ffff880893207c88 ffff880893207ce8 ffffffffa000c3a0
[    3.143111] Call Trace:
[    3.145611]  [<ffffffff812dce6b>] dump_stack+0x4d/0x72
[    3.150744]  [<ffffffff811217f7>] panic+0xc6/0x21e
[    3.155544]  [<ffffffffa0013000>] ? 0xffffffffa0013000
[    3.160688]  [<ffffffffa0013019>] ? init_+0x19/0xde [redpill]
[    3.166441]  [<ffffffff81000340>] ? do_one_initcall+0x80/0x130
[    3.172321]  [<ffffffff810bf26b>] ? do_init_module+0x5b/0x1d0
[    3.178116]  [<ffffffff810c1044>] ? load_module+0x1be4/0x2080
[    3.183896]  [<ffffffff810bdef0>] ? __symbol_put+0x40/0x40
[    3.189382]  [<ffffffff8118a4ac>] ? kernel_read+0x3c/0x50
[    3.194817]  [<ffffffff810c169d>] ? SYSC_finit_module+0x7d/0xa0
[    3.200720]  [<ffffffff810c16d9>] ? SyS_finit_module+0x9/0x10
[    3.206470]  [<ffffffff81566d8a>] ? entry_SYSCALL_64_fastpath+0x1e/0x8e
[    3.213294] ---[ end Kernel panic - not syncing: Fatal exception

 

HW:
Gigabyte B560M DS3H V2
CPU: Intel i5 10400
RAM: 32GB

 

Thanks for the answer, Zdenek.

[pocopico]

36 minutes ago, DidaJI said:

[    3.081179] redpill: loading out-of-tree module taints kernel.
[    3.087073] redpill: module verification failed: signature and/or required key missing - tainting kernel

 

Your problem is there. The kernel is not patched correctly. I'm looking at it right now ....

 

EDIT: I've reproduced and the patching issue is still there. Can you please start with 42661 withfriend and update using DSM in the meantime ? 

Edited September 29, 2022 by pocopico

[DidaJI]

59 minutes ago, pocopico said:

 

Your problem is there. The kernel is not patched correctly. I'm looking at it right now ....

 

EDIT: I've reproduced and the patching issue is still there. Can you please start with 42661 withfriend and update using DSM in the meantime ? 

 

Thank you very much, I will try and let you know. Zdenek.

[DidaJI]

On 9/29/2022 at 3:15 PM, DidaJI said:

 

Thank you very much, I will try and let you know. Zdenek.

 

On 9/29/2022 at 2:15 PM, pocopico said:

 

Your problem is there. The kernel is not patched correctly. I'm looking at it right now ....

 

EDIT: I've reproduced and the patching issue is still there. Can you please start with 42661 withfriend and update using DSM in the meantime ? 

 

So I tried version 7.1.0-42661 and unfortunately it ends the same.

I use this command to build: "./rploader.sh build broadwellnk-7.1.0-42661"

 

[    3.154385] redpill: loading out-of-tree module taints kernel.
[    3.160292] redpill: module verification failed: signature and/or required key missing - tainting kernel
[    3.171050] Kernel panic - not syncing: Fatal exception
[    3.176286] CPU: 3 PID: 1010 Comm: insmod Tainted: G           OE   4.4.180+ #42661
[    3.183929] Hardware name: Gigabyte Technology Co., Ltd. B560M DS3H V2/B560M DS3H V2, BIOS F7 03/25/2022
[    3.193442]  0000000000000000 ffff88089339fc60 ffffffff812dca8b ffffffffa000c0a2
[    3.201162]  ffff880895a43350 ffff88089339fcd8 ffffffff811216d7 0000000000000008
[    3.208841]  ffff88089339fce8 ffff88089339fc88 ffff88089339fce8 ffffffffa000c3a0
[    3.216559] Call Trace:
[    3.219058]  [<ffffffff812dca8b>] dump_stack+0x4d/0x72
[    3.224184]  [<ffffffff811216d7>] panic+0xc6/0x21e
[    3.228973]  [<ffffffffa0013000>] ? 0xffffffffa0013000
[    3.234109]  [<ffffffffa0013019>] ? init_+0x19/0xde [redpill]
[    3.239857]  [<ffffffff81000340>] ? do_one_initcall+0x80/0x130
[    3.245725]  [<ffffffff810bf16b>] ? do_init_module+0x5b/0x1d0
[    3.251515]  [<ffffffff810c0f44>] ? load_module+0x1be4/0x2080
[    3.257298]  [<ffffffff810bde40>] ? __symbol_put+0x40/0x40
[    3.262783]  [<ffffffff8118a3ac>] ? kernel_read+0x3c/0x50
[    3.268216]  [<ffffffff810c159d>] ? SYSC_finit_module+0x7d/0xa0
[    3.274125]  [<ffffffff810c15d9>] ? SyS_finit_module+0x9/0x10
[    3.279872]  [<ffffffff8156380a>] ? entry_SYSCALL_64_fastpath+0x1e/0x8e
[    3.383583] ---[ end Kernel panic - not syncing: Fatal exception

[pocopico]

3 minutes ago, DidaJI said:

So I tried version 7.1.0-42661 and unfortunately it ends the same:

 

Well,i only recently encountered this issue with other broadwellnk platform and we have to discuss this with other devs. 

Edited September 29, 2022 by pocopico

[DidaJI]

2 minutes ago, pocopico said:

 

Well,i only recently encountered this issue with other broadwellnk platform and we have to discuss this with other devs. 

 

Thanks, I'll look forward to any solution.

[pocopico]

Just now, DidaJI said:

 

Thanks, I'll look forward to any solution.

 

Do you mind trying with ARPL also ?  

[DidaJI]

4 minutes ago, pocopico said:

 

Do you mind trying with ARPL also ?  

 

5 minutes ago, pocopico said:

 

Do you mind trying with ARPL also ?  

 

It didn't matter.

 

Do you have a link with the procedure?

[pocopico]

Just now, DidaJI said:

 

 

It didn't matter.

 

Do you have a link with the procedure?

 

[DidaJI]

Just now, pocopico said:

 

 

Ok, I'll try and let you know.

[DidaJI]

1 hour ago, DidaJI said:

 

Ok, I'll try and let you know.

 

So ARPL works normally with version 7.1.1-42962

 

Zdenek.

[pocopico]

3 hours ago, DidaJI said:

 

So ARPL works normally with version 7.1.1-42962

 

Zdenek.

 

OK ! After two and a half hours, i've found the issue .... There was a typo in the config files. 

 

the netif_num": kernel command didnt match the macX values.  It should now be OK.

 

@DidaJI Thanks for the help and testing ! Could you please go back to TCRP and test again ? Please remember to run

 

./rploader.sh clean 

 

before running the build 

 

EDIT: This time you can also use withfriend

 

Edited September 29, 2022 by pocopico

[Peter Suh]

@pocopico

Today I tried some more NVMe cache with my TCRP friend on baremetal on my Intel MOBO.

 

Both MOBO's NVMe cache operation was successful as capture.

 

MOBOs are ASUS ROG STRIX Z490-I GAMING / ASROCK H310M ITX/AC.

 

ASUS ROG STRIX Z490-I GAMING has two M.2 internal slots,

 

ASROCK H310M ITX/AC has 1 M.2 built-in slot.

 

 

 

 

 

We don't have to worry about NVMe cache not working in TCRP friend loader.

 

As soon as the TCRP friend loader was released, I immediately applied it to my main XPE.

 

It seems like a special feature of my industrial motherboard (ASUS IOT H310Mi-IM-A R2.0).

 

NMVe is not common and has a rather unusual shape controller.

 

The Ryzen MOBO is still awaiting testing, and I will report the results as soon as the serial block arrives.

[Peter Suh]

1 hour ago, pocopico said:

 

OK ! After two and a half hours, i've found the issue .... There was a typo in the config files. 

 

the netif_num": kernel command didnt match the macX values.  It should not be OK.

 

@DidaJI Thanks for the help and testing ! Could you please go back to TCRP and test again ? Please remember to run

 

./rploader.sh clean 

 

before running the build 

 

EDIT: This time you can also use withfriend

 

 

I've always wondered if this number, which is pre-recorded in netif_num in config.json, is greater than 1.

 

It was only recorded in advance, but I thought there would be no problem if it was updated again during the loader setting or build process.

 

https://github.com/pocopico/redpill-load/commit/3e80bd51bcf04f06ed9b3eb768f3fa84747493a5

[pocopico]

On 9/29/2022 at 7:52 PM, Peter Suh said:

 

I've always wondered if this number, which is pre-recorded in netif_num in config.json, is greater than 1.

 

It was only recorded in advance, but I thought there would be no problem if it was updated again during the loader setting or build process.

 

https://github.com/pocopico/redpill-load/commit/3e80bd51bcf04f06ed9b3eb768f3fa84747493a5

 

Yes, if the user has set this in user_config.json according to what he has, there should be no issue, it will be overwritten by the user defined variable.

 

If the netif_num does not match the number of macX variables, then redpill will warn you and it stop at boot and kernel panic DSM. If you have production redpill, then you will loose the debug info and instead you will only see the kernel panic. It took me three hours to figure this out cause i had the prod version of redpill. 

 

A check has been added to TCRP Friend for that, it will warn you and will exit if netif_num and number of macs do not match.

[DidaJI]

2 hours ago, pocopico said:

 

OK ! After two and a half hours, i've found the issue .... There was a typo in the config files. 

 

the netif_num": kernel command didnt match the macX values.  It should now be OK.

 

@DidaJI Thanks for the help and testing ! Could you please go back to TCRP and test again ? Please remember to run

 

./rploader.sh clean 

 

before running the build 

 

EDIT: This time you can also use withfriend

 

 

TCRP works now, great job!
Thank you very much for the very quick resolution.
With regards, Zdenek.

 

[josi20]

Hello,

 

I want to use Xpenology 6 instead of 7.

 

I have installed Xpenology 6.2.4 with TCRP 0.4.6.

How can I update to the current version Xpenology 6.2.4 update 6?

 

Best regards, josi20

[DidaJI]

Use NVME SSD as storage..
So I'm here again with another question. Is it possible to somehow remap an M.2 nvme ssd to standard storage? No Cache. 

It would be more useful for me to have nvme for data rather than cache.

 

Something like this here: https://www.reddit.com/r/synology/comments/a7o44l/guide_use_nvme_ssd_as_storage_volume_instead_of/?ref=share&ref_source=embed&utm_content=title&utm_medium=post_embed&utm_name=c32094bf0a304e83831a4926260d5d1f&utm_source=embedly&utm_term=a7o44l

[logze]

18 hours ago, pocopico said:

 

Hi,

If you have a hotplug option in your BIOS setup then you can disable it.

 

There is also a linux command line option named HddHotplug=0 and should be 0 i assume that since this is the default you have not changed that.

 

sata_remap should work but it does not work for all platforms. You can also use a dtb based platform like DS920 that uses the dtb file to assign the disks to ports automatically 

 

18 hours ago, Peter Suh said:

 

A muxed Sataport is expected, but as a last resort, try setting SataportMap="",DiskIdxMap="" .
If this setting does not work as desired, try switching to DS920+ as the next best work.

DS920+ is a Device Tree-based model and does not involve SataportMap settings.

 

Hi @pocopico @Peter Suh

 

I have switch the platform to the DS920+ and the disk sequence shows as my expected.(disk1~disk4).

But there is another problem, the nvme SSD does not recognized by the DSM.

With DS918+ platform, it can be done by modifying the libsynonvme.so.1.

[DidaJI]

4 hours ago, logze said:

 

 

Hi @pocopico @Peter Suh

 

I have switch the platform to the DS920+ and the disk sequence shows as my expected.(disk1~disk4).

But there is another problem, the nvme SSD does not recognized by the DSM.

With DS918+ platform, it can be done by modifying the libsynonvme.so.1.

 

For visibility in DSM, this helped me:

 

[logze]

On 9/30/2022 at 1:16 PM, DidaJI said:

 

For visibility in DSM, this helped me:

 

 

Not work for me. 

DS920+ does not have the file /etc.defaults/extensionPorts, try to add manual and reboot the DSM but still not working.

[Peter Suh]

3 minutes ago, logze said:

 

Not work for me. 

DS920+ does not have the file /etc.defaults/extensionPorts, try to add manual and reboot the DSM but still not working.

 

 

 

Since DS920+ is a Device Tree-based model, the dtbpatch ext driver automatically detects and adds it.
This process should have been done automatically, but something seems to have gone wrong.
DS918+, DS3622xs+'s NVMe cache activation method should be viewed as another method.

 

When building the DS920+, what command did you use with ./rploader.sh and what options?

Edited September 30, 2022 by Peter Suh