Jump to content
XPEnology Community

Driver extension jun 1.03b/1.04b for DSM6.2.3 for 918+ / 3615xs / 3617xs


IG-88

Recommended Posts

3 minutes ago, -iliya- said:

which controller is better for syno for 16 drives ?

"native" should do fine and thats lsi for 3617

there are 16 port lsi controller (16i instead of 8i at the end of the name like 9201-16i) but as they are not that common i guess it would be cheaper to use two  8 port controller, there seem to be some 9201-16i on ebay for 200-230 bugs

 

Link to comment
Share on other sites

Thanks IG-88,
i made my lspci:
# lspci
0000: 00: 00.0 Class 0600: Device 8086: 0bf1 (rev 03)
0000: 00: 02.0 Class 0300: Device 8086: 0be1 (rev 09)
0000: 00: 1b.0 Class 0403: Device 8086: 27d8 (rev 02)
0000: 00: 1c.0 Class 0604: Device 8086: 27d0 (rev 02)
0000: 00: 1c.1 Class 0604: Device 8086: 27d2 (rev 02)
0000: 00: 1c.2 Class 0604: Device 8086: 27d4 (rev 02)
0000: 00: 1d.0 Class 0c03: Device 8086: 27c8 (rev 02)
0000: 00: 1d.1 Class 0c03: Device 8086: 27c9 (rev 02)
0000: 00: 1d.2 Class 0c03: Device 8086: 27ca (rev 02)
0000: 00: 1d.3 Class 0c03: Device 8086: 27cb (rev 02)
0000: 00: 1d.7 Class 0c03: Device 8086: 27cc (rev 02)
0000: 00: 1e.0 Class 0604: Device 8086: 2448 (rev e2)
0000: 00: 1f.0 Class 0601: Device 8086: 27bc (rev 02)
0000: 00: 1f.2 Class 0106: Device 8086: 27c1 (rev 02)
0000: 00: 1f.3 Class 0c05: Device 8086: 27da (rev 02)
0000: 01: 00.0 Class 0200: Device 10ec: 8136 (rev 05)
0000: 03: 00.0 Class ff00: Device 10ec: 5209 (rev 01)
0001: 07: 00.0 Class 0000: Device 1b4b: 9235 (rev ff)
0001: 08: 00.0 Class 0000: Device 1b4b: 9235 (rev ff)
0001: 09: 00.0 Class 0000: Device 1b4b: 9235 (rev ff)
0001: 0a: 00.0 Class 0000: Device 1b4b: 9235 (rev ff)

unfortunately no other NIC declared :(
I will try with a live ubuntu, we'll see ...

Link to comment
Share on other sites

Maybe a strange question, but is it possible to remove everything from the driver package except the drivers for hardware transcoding?

My RAID controller runs without the additional drivers without problems, with the driver extensions I have no more SMART values and no TRIM support for the SSDs.
But without the driver extensions i have no possibility for hardware transcoding so /dev/dri is missing.

I tried to combine the original extra.lzma and the extra.lzma from IG-88 but I think I don't know enough about linux drivers because DSM couldn't boot after that.

 

Link to comment
Share on other sites

3 hours ago, Leroy said:

Maybe a strange question, but is it possible to remove everything from the driver package except the drivers for hardware transcoding?

My RAID controller runs without the additional drivers without problems, with the driver extensions I have no more SMART values and no TRIM support for the SSDs.

 

i was thinking about making a additional version with the older scsi/sas driver for people taking the risk or disabling hibernation

 

the one you ask is more like jun's original without his i915 driver, also possible, it just needs some minor changes and the changed patch to get rid of already installed i915 drivers, its not that much to do, just removing some entry's in rc.modules and the files and copying the patch from my recent extra/extra2

Link to comment
Share on other sites

В 02.07.2020 в 22:15, NiGGaZ сказал:

Hi.

I've prepared extra.lzma based on Juns patch and Custom loader for DS1019+ for DSM 6.2.3-25426, link below:

https://mega.nz/file/pao0CQrb#_pLGNjTl-f7eneoWQgDkPklDoRCyqxJ_vhx_d79kIKg (only extra.lzma ~ 5.7Mb)

You can find loader for DS1019+ on this forum or another place, you can remove both extra files from partition and use re-compiled.

Then you need to download PAT files from https://archive.synology.com/download/DSM/release/6.2.3/25426/DSM_DS1019+_25426.pat, and just replace zImage and rd.gz

 

Tested on ASUS N3150M-E (Lan: Realtek 8111H) and it works better then on loader for DS918+ (including extra)

 

Comparing Read/Write speed between DS1019+ (chinese, repacked) and DS918+ (from this forum loaders).

Hardware: ASUS N3150M-E, Lan: Realtek 8111H, SSD: Samsung 128Gb (Sata 3Gbps).

File: approx. 2,52 GB img

Strange, but DS1019+ (W: ~63-65 Mb/s, R: ~66-67 Mb/s) copying faster then DS918+ (W: ~50-58 Mb/s, R: ~60-62 Mb/s).

Checked on another HDD WD 1Tb: DS1019+ (W: ~39-45 Mb/s, R: ~40-47 Mb/s) copying faster then DS918+ (W: ~30-36 Mb/s, R: ~36-38 Mb/s).

 

Maybe someone can explain why it so!?

DS918+ Wrinting (SSD Samsung 128Gb)

918_write.PNG

DS918+ Reading (SSD Samsung 128Gb)

918_read.PNG

DS1019+ Wrinting (SSD Samsung 128Gb)

1019_write.PNG

DS1019+ Reading (SSD Samsung 128Gb)

1019_read.PNG

Edited by NiGGaZ
Link to comment
Share on other sites

On 8/13/2020 at 10:27 AM, NiGGaZ said:

Maybe someone can explain why it so!?

 

when trying to estimate network speed its best to use a tool that only does exactly this, keeping out buffers cache and read/write speed problems of disks, even ssd's have cache and can have write speed decreased when needing to empty a block (4k) before writing it again, to cover that up and emptying early there is trim but that usually kicks in when there is nothing to do

the kernel and settings for 918+ and 1019+ will be the same as its the same hardware platform (apollo lake), when used on the same hardware i'd expect the same performance, so my explanation would be systematic problems to measure reliably under constant conditions and one method to cope with this is to exclude as much factors as possible (like using a tool that just meaures the network speed and does not write anywhere else then /dev/nul)

 

Link to comment
Share on other sites

20 часов назад, IG-88 сказал:

 

when trying to estimate network speed its best to use a tool that only does exactly this, keeping out buffers cache and read/write speed problems of disks, even ssd's have cache and can have write speed decreased when needing to empty a block (4k) before writing it again, to cover that up and emptying early there is trim but that usually kicks in when there is nothing to do

the kernel and settings for 918+ and 1019+ will be the same as its the same hardware platform (apollo lake), when used on the same hardware i'd expect the same performance, so my explanation would be systematic problems to measure reliably under constant conditions and one method to cope with this is to exclude as much factors as possible (like using a tool that just meaures the network speed and does not write anywhere else then /dev/nul)

 

Curious! I've tested network speed with iperf3 (on synology with docker). I'm shocked! Then why copying through ordinary network not so good!?

iperf3_test.PNG

Edited by NiGGaZ
Loader 918+
Link to comment
Share on other sites

On 7/14/2020 at 10:02 PM, Stan_G said:

hi  i`v bought rtl8125.also   i·v updated the extra.and i found the "lan" in the control panel.but i cant get a dhcp.the ip is still 169.254.58.x the network port light flashes.and i also set a static ip,but still can`t find the nas.why

Hi, It seems that I have met the same problem with post id:396

 

I bought TP-LINK NG421 2.5G nic with chip RTL8125B. After updated the extra v0.11_test(http://s000.tinyupload.com/?file_id=04147311964729253115), PCIE Lan3 successfully recognized in the control panel. but dhcp failed. the ip is still 169.254.182.x, Changed to a static ip but problem still.

newly install DSM 6.2.3 25426U2 with loader 1.03b for DS3615xs

Mainboard:SuperMicro x10slv-q with 2 nic I217-V and I210-AT onboard

PCIE: TP-LINK NG421 2.5G(RTL8125B)

CPU: G3460T

please help, thanks.

Edited by ztielong
Link to comment
Share on other sites

Anybody has realtek 8168 network card?? I have faced with some issue on the last extra.lzma(13.3), the maximum speed of copying one big file is 17MB/s via the samba.

Also I have tried USB 3 network adapter, with the same file and cable and receive 70-80 MB/s  It looks very strange for me. The system shows connection speed as 1000Mbps,Full duplex, MTU 2000

 

 

Here is some details about hardware:

0000:03:00.0 Class 0200: Device 10ec:8168 (rev 15)
        Subsystem: Device 1b0a:01a5
        Flags: bus master, fast devsel, latency 0, IRQ 90
        I/O ports at d000 [size=256]
        Memory at d0604000 (64-bit, non-prefetchable) [size=4K]
        Memory at d0600000 (64-bit, non-prefetchable) [size=16K]
        Capabilities: [40] Power Management version 3
        Capabilities: [50] MSI: Enable+ Count=1/1 Maskable- 64bit+
        Capabilities: [70] Express Endpoint, MSI 01
        Capabilities: [b0] MSI-X: Enable- Count=4 Masked-
        Capabilities: [100] Advanced Error Reporting
        Capabilities: [140] Virtual Channel
        Capabilities: [160] Device Serial Number 01-00-00-00-68-4c-e0-00
        Capabilities: [170] Latency Tolerance Reporting
        Capabilities: [178] L1 PM Substates
        Kernel driver in use: r8168

 

Link to comment
Share on other sites

Morning all

 

I'm going to give updating to 6.2.3 a go (from 6.2.2).

Update reports pages suggest it should work fine  - it's a Asrock J3455 board. DS918+

(I have a couple of others to do, but this one first)

 

However, I currently have a custom extra.lzma. It's been recommended this is removed before upgrade?

Do I need to replace it with anything in particular - or just remove it? Can't recall if there was one beforehand!

 

I also need to disable the cache and I'll probably break any NIC teams I have in place (just in case).

 

So I guess I would remove/replace the custom.lzma

Upgrade to 6.2.3

Replace with newest custom lzma (I guess after a couple of reboots - and especially if it doesn't come back up).

That should be it.

 

Sound about right?

 

Cheers,

Link to comment
Share on other sites

On 8/26/2020 at 12:00 PM, ztielong said:

Hi, It seems that I have met the same problem with post id:396

 

I bought TP-LINK NG421 2.5G nic with chip RTL8125B. After updated the extra v0.11_test(http://s000.tinyupload.com/?file_id=04147311964729253115), PCIE Lan3 successfully recognized in the control panel. but dhcp failed. the ip is still 169.254.182.x, Changed to a static ip but problem still.

newly install DSM 6.2.3 25426U2 with loader 1.03b for DS3615xs

Mainboard:SuperMicro x10slv-q with 2 nic I217-V and I210-AT onboard

PCIE: TP-LINK NG421 2.5G(RTL8125B)

CPU: G3460T

please help, thanks.

I encountered the same problem as you. The RLT8125B chip can be used normally under DSM 918+, but DHCP fails under DSM3615x. It's a pity that I need to use LSI SAS to link SAS hard drives, and DSM 918+ system is not suitable for me.

Link to comment
Share on other sites

3 hours ago, bzfzis said:

I encountered the same problem as you. The RLT8125B chip can be used normally under DSM 918+, but DHCP fails under DSM3615x. It's a pity that I need to use LSI SAS to link SAS hard drives, and DSM 918+ system is not suitable for me.

I received a new extra from Stan_G, if you need, please leave your email address, I'll forward to you.

Link to comment
Share on other sites

Hello,

 

maybe a stupid question but how can I update the driver package on an already installed stick. I mean the boot loader is copied on the stick and the system is running. How can I update to the latest driver version?

I use bootloader 1.04b and if I plug the stick to a Windows-PC or MacBook the stick will not be recognized.

 

Thanks in advance

Link to comment
Share on other sites

Hi, and again thanks to IG-88 for the amazing work ...

 

As regularly registered tester for the DSM7.0 preview, I got link to download/test the pre-release on an official Synology 920+  ... it works fine (not a real revolution on GUI, certainly more on the core features especially on rebuild optimisations etc., but not tested obviously).

 

By chance the overall Synology versions are available for the registered preview testers (I will not share, please don't ask), and I tried the 3617XS on an Xpenology baremetal under DSM 6.2.3-25426 Update 2.

 

The update is well accepted, and installation worked fine (btw June Loader 1.03b is still compatible 👌), with final automated reboots etc. but then it block exactly as for the install from 6.2.2 to 6.2.3 = the new DSM7.0 install reset the additional drivers and a new extra.lzma aligned with DSM7.0 is needed to restore NICs support (and I assume the other requested drivers).

 

Hope our driver extension master will be kind to provide in the future with new release compatible of DSM7.0 😃

 

  • Like 2
Link to comment
Share on other sites

10 hours ago, satdream said:

(I will not share, please don't ask),

 

to make any driver i would at least need to see a *.ko file from the new 7.0 or you would have to look up whats after then string "vermagic=" in such a *.ko file

with 918+ its "4.4.59+" (920+ would be the same as 918+)

with 3615/17 its "3.10.105"

whats in the new *.ko files?

 

Link to comment
Share on other sites

13 hours ago, Huberer said:

update to the latest driver version?

I use bootloader 1.04b and if I plug the stick to a Windows-PC or MacBook the stick will not be recognized.

 

https://xpenology.com/forum/topic/29872-tutorial-mount-boot-stick-partitions-in-windows-edit-grubcfg-add-extralzma/

oder auch in deutsch

https://xpenology.com/forum/topic/29670-howto-1-partition-des-bootsticks-in-windows-mountenbearbeiten-grubcfg/

 

  • Like 1
Link to comment
Share on other sites

Il y a 2 heures, IG-88 a dit :

 

to make any driver i would at least need to see a *.ko file from the new 7.0 or you would have to look up whats after then string "vermagic=" in such a *.ko file

with 918+ its "4.4.59+" (920+ would be the same as 918+)

with 3615/17 its "3.10.105"

whats in the new *.ko files?

 

 

Both shows "vermagic=4.4.180+"  if I am correct ... 

 

[Edit] both= checked only 918+ and 3617 :)

 

Edited by satdream
Link to comment
Share on other sites

On 9/10/2020 at 9:11 PM, IG-88 said:

whats in the new *.ko files?

 

the build number of that dsm beta version it 40850, a lot of changes compared to the 25426 of dsm 6.2 (if there is no new start at a higher number like 40000 like when starting to use  new base kernel)

 

918+  -> 4.4.180+

3615 -> 3.10.108

(i guess 3617 will be 3.10.108 too)

edit: and i was wrong with 3617, its 4.4.180+

i guess @satdream's "both" might be 918+ and 3617?

its a little surprise to see that 3617 got a really "new" kernel,

 

so all three versions use a new base kernel and old extra drivers will not work for that reason

as there is no new kernel source from synology we could start with a vanilla kernel from kernel.org, i guess at least the network drivers should work that way because synologys patches in there custom kernel are usually not in that area

Edited by IG-88
  • Thanks 1
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...