Jump to content
XPEnology Community

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


Recommended Posts

30 minutes ago, wer said:

I might add it myself to extra.lzma the same way? I'm just not sure how to compress/pack it again.

i documented thesteps here, for just adding a file and repacking you will not need chroot, just any (live-) linux will do that

https://xpenology.com/forum/topic/7187-how-to-build-and-inject-missing-drivers-in-jun-loader-102a/

 

 

Link to comment
Share on other sites

10 hours ago, IG-88 said:

i documented thesteps here, for just adding a file and repacking you will not need chroot, just any (live-) linux will do that


Thank you, including driver worked that way.

However just bringing up the second NIC crashed the settings completely!

It's no longer possible to use the network configuration at the web login, I get operation fails. Connection by SSH fails too, works again after removing second NIC. Settings aren't possible anymore. /var/log/messages is floated by synoscgi_SYNO.Core,.Network error messages.

Link to comment
Share on other sites

On 11/29/2021 at 8:55 PM, IG-88 said:

try with two vm's own nic's first before using the pass through nic

also add 2nd nic with its mac into grub.cfg and check /var/log/messages if the nic's are found as eth0 and eth1


Second NIC even works without MAC in grub.cfg with fresh install. I don't know what went wrong, what crashed whole config. There was no other way than fresh install.
I tried switching to 918 but no matter what (vid & pid in grub.cfg), I always got error 13 while installing pat.
After complete cleaning 3615 is now running again and backup is restoring.

Edited by wer
Link to comment
Share on other sites

8 hours ago, maqros said:

hi guys. my question  is to IG-88. First of all thanks a lot for your great job. Is there any chance to get support for mobile 10th gen video core i3- 10110U (8086:9B41)?

we use UHD610 and UHD630 GPU, yours cpu seems to be having UHD620

for more precise info https://pci-ids.ucw.cz/read/PC/8086

3e92 CometLake-S GT2 [UHD Graphics 630]

3e93 CoffeeLake-S GT1 [UHD Graphics 610]

yours is

9b41 CometLake-U GT2 [UHD Graphics]

so we loog for a similar typ in the 3eXX area that is UHD820 GT2

the entry in the list that comes close (and is supported in the driver - we see that entry in the table we are patching in the binary driver)

3ea0 WhiskeyLake-U GT2 [UHD Graphics 620]

 

d2724 h - A0 3E changed to 41 9B

i patched one of the files from 6.2.3u3 mod2 to add it to the other two (now "9BA8 9B41 9BC8 - UHD 610 620 630 10th low end")

if that does not work we can try to patch 3E92 to 9B41 (position d2644 h)

 

918plus_dsm623u3_i915_mod3 - 3EA0_to_9B41-UHD620GT2.zip

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

12 часа назад, IG-88 сказал:

we use UHD610 and UHD630 GPU, yours cpu seems to be having UHD620

for more precise info https://pci-ids.ucw.cz/read/PC/8086

3e92 CometLake-S GT2 [UHD Graphics 630]

3e93 CoffeeLake-S GT1 [UHD Graphics 610]

yours is

9b41 CometLake-U GT2 [UHD Graphics]

so we loog for a similar typ in the 3eXX area that is UHD820 GT2

the entry in the list that comes close (and is supported in the driver - we see that entry in the table we are patching in the binary driver)

3ea0 WhiskeyLake-U GT2 [UHD Graphics 620]

 

d2724 h - A0 3E changed to 41 9B

i patched one of the files from 6.2.3u3 mod2 to add it to the other two (now "9BA8 9B41 9BC8 - UHD 610 620 630 10th low end")

if that does not work we can try to patch 3E92 to 9B41 (position d2644 h)

 

918plus_dsm623u3_i915_mod3 - 3EA0_to_9B41-UHD620GT2.zip 588 \u041a\u0431 · 0 downloads

thanks. replaced /usr/lib/modules/i915.ko with this one, unfortunately hangs on reboot with no even DHCP lease(. ready to go further

Link to comment
Share on other sites

The r8125 works for me at DSM 6.2.3, however I can't get the hard disks to spin down, no matter what I try (using tmpfs for several mounts, disabling logs and all possible services).
I was happy with DSM 6.1-15101 Update 4 for several years and it was the last version that I had no problems with hard disk spinning down after  selected time of no activity.

 

I was able to compile r8125.ko for DSM 6.1-15101, however it doesn't load (error: "region #1 not an MMIO resource", reported for other realtek adapters and different systems too).

@IG-88
What driver source did you use for r8125?

Link to comment
Share on other sites

Hello,

 

I have an MSI motherboard (B460I gaming edge wifi) with an integrated NIC RTL8125B.

 

I'm trying to install the DSM 6.2.3 V25426 for 918+


I have followed the tutorial for the loader 1.04b with the migration from DSM 5.2 to DSM 6.1.7 (I'm doing a fresh installation).


Reading about the problem with realtek, I have downloaded the extra/extra2 files from the first page  and i have extract the RD and zlmage files from the dsm corresponding image. These files have replaced the ones that were in the folder opened via osfmount.

I have mount an usb stick and it seems to be ok until I use find.synology.com, it isn't discovered on the network.

Have I missed something ?

Thanks in advance for your help !

Link to comment
Share on other sites

Thanks for your answer !

 

I have tried with an other usb stick and followed the tutorial on reddit but still the same not found on the network.

 

Looking back at the first page, it seems that something is wrong as the screen stops at the line "screen will stop quickly..." I don't see the line : "early console..."and the next one

  • Confused 1
Link to comment
Share on other sites

7 hours ago, vince104 said:

Thanks for your answer !

 

I have tried with an other usb stick and followed the tutorial on reddit but still the same not found on the network.

 

Looking back at the first page, it seems that something is wrong as the screen stops at the line "screen will stop quickly..." I don't see the line : "early console..."and the next one

 

I see that you are confused.

I wrote that because on a video I saw these lines in addition to the previous ones.

 

I also see this video (https://www.youtube.com/watch?v=iFUCpebVYAc) as you can see at the very beginning it shows a number of lines and now I don't know if in case of correct installation I should see this...

In fact I have the impression that the installation is too quick. It goes from the screen with choice of the installation (reinstallation and the third choice) to the next screen and that's it.

 

Maybe I have made everything correctly and there is a "problem" with the driver for NIC ? There ismaybe a difference between the rtl8125 and the rtl8125b ?

 

Edited by vince104
Link to comment
Share on other sites

On 12/12/2021 at 3:22 PM, vince104 said:

"screen will stop quickly..." I don't see the line : "early console..."and the next one

there is no message like this, its just about stop updating shortly and to open http://find.synology.com to continue

(i still prefer the synolgy assistant)

21 hours ago, vince104 said:

There ismaybe a difference between the rtl8125 and the rtl8125b

the driver is not recent (r8125-9.003.05 from 6/2020) so there will be versions that might not work

the old source used in extra.lzma already references to 8125b so its not just that

old driver from 6/2020
r.8125_n.c
static const struct {
        const char *name;
        u8 mcfg;
        u32 RCR_Cfg;
        u32 RxConfigMask;   /* Clears the bits supported by this chip */
        u32 jumbo_frame_sz;
} rtl_chip_info[] = {
        _R("RTL8125A",
        CFG_METHOD_2,
        BIT_30 | BIT_22 | BIT_23 | (RX_DMA_BURST << RxCfgDMAShift),
        0xff7e5880,
        Jumbo_Frame_9k),

        _R("RTL8125A",
        CFG_METHOD_3,
        BIT_30 | BIT_22 | BIT_23 | (RX_DMA_BURST << RxCfgDMAShift),
        0xff7e5880,
        Jumbo_Frame_9k),

        _R("RTL8125B",
        CFG_METHOD_4,
        BIT_30 | BIT_22 | BIT_23 | (RX_DMA_BURST << RxCfgDMAShift),
        0xff7e5880,
        Jumbo_Frame_9k),

        _R("RTL8125B",
        CFG_METHOD_5,
        BIT_30 | BIT_22 | BIT_23 | (RX_DMA_BURST << RxCfgDMAShift),
        0xff7e5880,
        Jumbo_Frame_9k),

        _R("Unknown",
        CFG_METHOD_DEFAULT,
        (RX_DMA_BURST << RxCfgDMAShift),
        0xff7e5880,
        Jumbo_Frame_1k)
};


new driver fom 11/2021
r.8125_n.c
} rtl_chip_fw_infos[] = {
        /* PCI-E devices. */
        [CFG_METHOD_2] = {"RTL8125A"				},
        [CFG_METHOD_3] = {"RTL8125A",		FIRMWARE_8125A_3},
        [CFG_METHOD_4] = {"RTL8125B",                       },
        [CFG_METHOD_5] = {"RTL8125B",		FIRMWARE_8125B_2},
        [CFG_METHOD_DEFAULT] = {"Unknown",                  },
};

 

both refrence to a/b and up to cfg_method5, but there might be other differences, ive seen a difference about a dac detected in the source but there might be more like diffrent phy chips too

 

if you want o know if the driver is the problem you would need to use the serial console (but you board does not have serial port, not even a connector on board) or you use a 2nd nic that is supported in the pcie slot

but it seems likely that the r8125 is not supported in the old driver

afair the r8125 driver was still building with 918+ and a did a new extra in april but never tested it

maybe i should just do a new version with just the r8125 driver a it seem the most wanted atm

 

Link to comment
Share on other sites

26 minutes ago, Florian83260 said:

do you think it will work for my I3-8100 ?

if its just "plain dsm" thw 1.03b 3615/17 and 1.04/918+ will work but 1.03b need CSM mode in bios and the ledacy usb boot device to work

https://xpenology.com/forum/topic/13333-tutorialreference-6x-loaders-and-platforms/

https://xpenology.com/forum/topic/52709-from-dsm_ds3615xs_15284-617-juns-mod-v102b-to-dsm_ds3615xs_25426-623/?do=findComment&comment=255924

 

 

but i assume you talk about 1.04b/918+ an intel quicksync as its the one cpu specific thing (beside 918+ needing min 4th gen intel cpu)

https://ark.intel.com/content/www/us/en/ark/products/126688/intel-core-i38100-processor-6m-cache-3-60-ghz.html

-> iGPU 0x3E91/x92

 

in 1st post there is a list of supported iGPU devices from the driver synology has in 6.2.3

"  iGPU device ID's supported by synology's i915 driver (Reveal hidden contents)"

3E91/92 is among them

 

and also if you read careful

"...

edit: i made a modded i195 driver were the pci device id of the 9th gen UHD 630 iGPU (3E92/3E93) is replaces with the device id's of the newer/different UHD 610/630 iGPU's that are unsupported

..."

so if 3E92 is the base that will be patched for the "newer CPU''s then your iGPU is supported ootb with the default driver in 6.2.3

 

  • Like 1
Link to comment
Share on other sites

I am having issues with the patched i915.ko  (downloaded from https://gofile.io/d/4fFJA5)

CPU i5 10400, MB Gigabyte H470M DS3H). Installed as usual and then used insmod to test that the i915 works (sadly its crashing).

 

Loader 1.04b with extra/extra2.lzma ver 13.3

DS 918+ 6.2.3 25426  (both this and U3 give a hard crash)

I used the 9BC8 version of the patched file (havent tried the others). I checked with lspci -nn to ensure that its the correct iGPU version for the processor on hand.

 

I have also tried the following DSM versions

a) Used the same patched i915.ko from the thread with DSM 6.2.0 - couldnt load modules (kernel version mismatch - symbols error)

b) DSM version 6.2.2 24922 (using juns i915.ko - which i manually patched 3E92 => 9BC8) - had a partial success. i do get /dev/dri and videostation does transcode, but i get soft lockups and NMI watchdog errors.

 

Any help will be really appreciated for DSM 6.2.3. dmesg output as attached

 

 

dmesg.txt

Link to comment
Share on other sites

Sorry to disturb you. I use cpu 10400 with a UHD 630 8086:9BC5 GPU and motherboard biostar z490gtn.
I install ds918 6.23,then upgrade to u3 successfully.
Then i replaced /usr/lib/modules/i915.ko with [9BC5 UHD 630 10th High End/i915.ko],unfortunately dsm hangs on reboot and there is nothing written in dmesg.
How to solve this problem?hope your reply.

Link to comment
Share on other sites

On 12/17/2021 at 8:10 PM, IG-88 said:

 

i added a new version 13.5 for 918+ that has the latest 8125 driver so in theory it should work with your onboard nic

 

Appreciate your time and effort in loading Relatek 8125 driver in the new version 13.5. I just loaded it on my PC running MSI B560M Pro MB (Relatek 8125) and i3 10105 cpu. It works now. I was using DS3617xs on AMD laptop until now and my cpu fan started misbehaving after a year.

 

Would it be possible to load 8125 drivers on extra.lzma for 3617 in future? Alternatively, is there a guide that I can follow and try to load them?

Link to comment
Share on other sites

1 hour ago, Sapan said:

Would it be possible to load 8125 drivers on extra.lzma for 3617 in future?

 

mentioned earlier in this thread, newer versions do not compile against 3615/17 dsm 6.2 source (i can't fix this)

 

CC:  cc
CCVERSION:  4.9.2
KERNEL_GCC_VERSION:
KVER:  3.10.105
KMAJ:  3
KMIN:  10
KREV:  105
BASEDIR:  /lib/modules/3.10.105
DRIVERDIR:  /lib/modules/3.10.105/kernel/drivers/net/ethernet/realtek
PWD:  /test3/r8125-9.007.01/src
RTKDIR:  kernel/drivers/net/ethernet/realtek

make -C /lib/modules/3.10.105/build M=/test3/r8125-9.007.01/src clean
make[1]: Entering directory '/test3/linux-3.10.x'
  CLEAN   /test3/r8125-9.007.01/src/.tmp_versions
make[1]: Leaving directory '/test3/linux-3.10.x'
make -C /lib/modules/3.10.105/build M=/test3/r8125-9.007.01/src modules
make[1]: Entering directory '/test3/linux-3.10.x'
/test3/linux-3.10.x/arch/x86/Makefile:162: CONFIG_RETPOLINE=y, but not supported by the compiler. Toolchain update recommended.
  CC [M]  /test3/r8125-9.007.01/src/r8125_n.o
In file included from /test3/r8125-9.007.01/src/r8125.h:44:0,
                 from /test3/r8125-9.007.01/src/r8125_n.c:83:
/test3/r8125-9.007.01/src/r8125_ptp.h:39:31: fatal error: linux/timecounter.h: No such file or directory
 #include <linux/timecounter.h>
                               ^
compilation terminated.
scripts/Makefile.build:308: recipe for target '/test3/r8125-9.007.01/src/r8125_n.o' failed
make[2]: *** [/test3/r8125-9.007.01/src/r8125_n.o] Error 1
Makefile:1229: recipe for target '_module_/test3/r8125-9.007.01/src' failed
make[1]: *** [_module_/test3/r8125-9.007.01/src] Error 2
make[1]: Leaving directory '/test3/linux-3.10.x'
Makefile:176: recipe for target 'modules' failed
make: *** [modules] Error 2

 

 

1 hour ago, Sapan said:

Alternatively, is there a guide that I can follow and try to load them?

you wanted to say "make" them ;-)

https://xpenology.com/forum/topic/7187-how-to-build-and-inject-missing-drivers-in-jun-loader-102a/

https://xpenology.com/forum/topic/7341-tutorial-compile-xpenology-drivers-in-windows-10/

 

Link to comment
Share on other sites

5 hours ago, IG-88 said:

Oh! I thought one would have to "unpack" original drivers from say Realtek or Intel and tweak and load it on the extra.lzma. Nevermind in that case, if you couldn't do it, I am not going to attempt to make it. I have no clue how to go about writing the code for the driver. :D

I will be happy with DS918+ for now.. Also, I thought 918+ is a 4 bay NAS so I would only be able to attach 4 drives - but I was so wrong. When I booted it, it showed me I have  16 slots available to "load" my drives. I have currently connected 5 drives and it shows I have room for 11 more!!

Link to comment
Share on other sites

6 hours ago, Sapan said:

When I booted it, it showed me I have  16 slots available to "load" my drives. I have currently connected 5 drives and it shows I have room for 11 more!!

you can have up to 24 if you manually tweak the synoinfo.conf and thats for all three types, so even 3615/17 can do more the the 12 defaut (as jun made the loader he did not touch the drive count in 3615/17 as it seemed high enough at this time for most cases and as 916+/918+ came out with its 4 drive default it was naturally needed to tweak that)

also to mention is that 918+ has still a 2 nic limit from its default configuration thats need to be tweaked manually if there are more then 2 nic ports

 

3617 is the choice when it comes to more then 8 cores (its often pretty ok to just disable HT in bios as "HT core" just give 25% performance of a real core) OR it you are using lsi sas controller

you can change dsm type just be exchanging the loader and dsm will offer a migration to the new system (its a official process you can read about on synology's website)

 

there are other options for >1Gbit, either have more then one 1G nic and manually active smb multicahnnel in smb.conf (synology might offer this in the gui in dsm 7.1) or have a 802.3bz/NBASE-T usb bases adapter (pcie like asus XG-C100C or even based usb realtek 8152/8156 have drivers working for all three)

intel 2.5G is no option at all, intel doues not deliver driver source outside from kernel 5.x

 

 

also i had a idea, @pocopico offers precompiled modules not just for dsm 7 but also 6.2 and his r8125.ko is newer then mine so i could just add his module and you test if its working, i was thinking about using his modules to make a new driver extension for 7.0 but why not also using it for 6.2.3's driver extension too, his 8152 driver i 9.006 and thats from 2021 source, very like any 8125 nic around will work with that driver

https://github.com/pocopico/3.10.105-modules

(and i should ask him about anything special ob building that module like modding the source)

  • Like 1
Link to comment
Share on other sites

4 minutes ago, IG-88 said:

also i had a idea, @pocopico offers precompiled modules not just for dsm 7 but also 6.2 and his r8125.ko is newer then mine so i could just add his module and you test if its working, i was thinking about using his modules to make a new driver extension for 7.0 but why not also using it for 6.2.3's driver extension too, his 8152 driver i 9.006 and thats from 2021 source, very like any 8125 nic around will work with that driver

https://github.com/pocopico/3.10.105-modules

(and i should ask him about anything special ob building that module like modding the source)

 

Some modules required patching, but most of them complile fine using the original linux in-tree or vendor provided sources.

 

Most Realtek modules, if i recall correctly, compile fine OOTB. On the other hand, most Intel modules needed some tweeking. 

Edited by pocopico
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.   Restore formatting

  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...