Voxel Custom firmware build for R7800 v. 1.0.2.93SF

[Voxel]

Continuation of

. . .
https://www.snbforums.com/threads/custom-firmware-build-for-r7800-v-1-0-2-91sf-v-1-0-2-91-1sf.76803/
https://www.snbforums.com/threads/custom-firmware-build-for-r7800-v-1-0-2-92sf.77498/

New version of my custom firmware build: 1.0.2.93SF.

Changes (vs 1.0.2.92SF):

1. Toolchain: Go is upgraded 1.17.7->1.18.
2. OpenVPN is upgraded 2.5.5->2.5.6 (fixing CVE-2022-0547, score 9.8, Critical).

https://nvd.nist.gov/vuln/detail/CVE-2022-0547​

3. OpenSSL v. 1.1.1 package is upgraded 1.1.1m->1.1.1n (fixing CVE-2022-0778, score 7.5, High).

https://nvd.nist.gov/vuln/detail/CVE-2022-0778​

4. minidlna package is upgraded 1.3.0->1.3.1 (fixing CVE-2022-26505, score 7.4, High).

https://nvd.nist.gov/vuln/detail/CVE-2022-26505​

5. libxml2 package is upgraded 2.9.12->2.9.13 (fixing CVE-2022-23308, score 7.5, High).

https://nvd.nist.gov/vuln/detail/CVE-2022-23308​

6. libpcap package is upgraded 1.1.1->1.10.1 many CVE fixes (v1.1.1 is kept as well, it is used exclusevily by QCA QoS).

CVE - Search Results

7. expat package is upgraded 2.4.4->2.4.8 Relax fix to CVE-2022-25236 (introduced with release 2.4.5).
8. dropbear package is upgraded 2020.81->2022.82.
9. zlib package is upgraded 1.2.11->1.2.12.
10. libgpg-error package is upgraded 1.42->1.43.
11. curl package is upgraded 7.81.0->7.82.0.
12. wget package is upgraded 1.21.2->1.21.3.
13. ethtool package is upgraded 5.16->5.17.
14. util-linux package is upgraded 2.37.4->2.38.
15. ubus package is upgraded 2021-08-90->2022-02-28.
16. dbus package is upgraded 1.13.20->1.13.22.
17. libflac package is upgraded 1.3.3->1.3.4.
18. libogg package is upgraded 1.3.4->1.3.5.
19. sysstat package is upgraded 12.4.3->12.4.5.
20. usbmode package is updated 2017-12-19->2022-02-24.
21. libjpeg package is upgraded 9d->9e.
22. Fix NG/DNI 'igmpproxy' modified codes (to provide compatibility with 'libpcap').
23. Make an order in various Makefiles.
24. Host tools: upgrade mkimage/u-boot to 2022.01.

The link is:

https://www.voxel-firmware.com (thanks to vladlenas for his help with hosting).

Thanks to HELLO_wORLD for his help (making order in Makefiles).

Voxel.

 

[IAMP373R]

Welcome back

 

[bartj12]

Good to see you online again, hope you are doing well!

 

[fossil]

Thank you. Hope you are doing well.

Here are few findings:

• Host tools mklibs can be upgraded to 0.1.44
• detcable package can be upgraded/synced with NG ver .90 [It has detcable binary as part of package that NG has upgraded in last few FW releases]
• detwan package can be upgraded/synced with NG ver .90

I am not sure if detcable & detwan packages you have not synced due to some NG bug. I have compiled Voxel FW with upgraded versions of above 3 and FW builds fine.

 

[fossil]

Thank you @HELLO_wORLD and @Voxel for fixing busybox & ipset Makefiles. I had to resolve them manually whenever I was syncing the code.

 

[Voxel]

• Host tools mklibs can be upgraded to 0.1.44
• detcable package can be upgraded/synced with NG ver .90 [It has detcable binary as part of package that NG has upgraded in last few FW releases]
• detwan package can be upgraded/synced with NG ver .90

I do not see any necessity to use anything from the stock .90 or so. Additional not known bugs. E.g. detcable mentioned by you has the same bug in init script (.90, /etc/init.d/detcable):

#!/bin/sh /etc/rc.common
START=20

start() {
    # Start detcable
    /usr/bin/detcable 2
}

stop() {
    killall /usr/bin/detcable
}

restart() {
    stop
    start
}

"killall /usr/bin/detcable" is a bug. Should be "killall detcable".

Voxel.

 

[fossil]

Its your call you have the final say.

In case of detcable my emphasis was on pre-built src/detcable NG binary. init script can be the one you are already using.

 

[Voxel]

In case of detcable my emphasis was on pre-built src/detcable NG binary. init script can be the one you are already using.

We should be spectic a bit trusting GPL from NG released few weeks after actual release of the resulting binary firmware.

Voxel.

 

[Luizk]

Thank you kindly

 

[Remy561]

Thanks @Voxel! Another smooth update for both my R7800s!

 

[StarCaller]

thank you Voxel, your efforts here are much appreciated!

 

[LeKeiser]

Again and Again, A Big Thanx to you Mr Voxel and to you Mr HELLO_wORLD for all your work and dedication
Crossing fingers you'll be able to continue for many years

 

[h1ms3lf]

A very, very big Thank You, for all your work!! I really like the R7800 and how flawless it works due to your constant updates. Wish you all the best!

 

[LeKeiser]

Funny, first time I have this :

V1.0.2.93SF (2022-04-05) (Voxel). Kamoj Add-on:220305-194131-5.5b20. New Voxel Release:Not found/slow Internet Latest Netgear:V1.0.2.90 (2022-01-11)

I'm downloading a big file on the side, but I didn't think it would slow my R7800 :O

 

[Bati]

I might be overlooking things but isn't there a version of Kamoj add-on available on the Voxel Firmware site anymore?
Installed 93SF about an hour ago.
@Voxel Thanks a lot for this great stuff!

 

[kokishin]

Hello @Voxel

Just installed 93SF a couple of minutes ago.
No issues.
Mobile devices connected to wifi quickly after router reboot.
Smooth update as usual.

Thanks!

 

[3rdOne]

Continuation of

. . .
https://www.snbforums.com/threads/custom-firmware-build-for-r7800-v-1-0-2-91sf-v-1-0-2-91-1sf.76803/
https://www.snbforums.com/threads/custom-firmware-build-for-r7800-v-1-0-2-92sf.77498/

New version of my custom firmware build: 1.0.2.93SF.

Changes (vs 1.0.2.92SF):

1. Toolchain: Go is upgraded 1.17.7->1.18.
2. OpenVPN is upgraded 2.5.5->2.5.6 (fixing CVE-2022-0547, score 9.8, Critical).

https://nvd.nist.gov/vuln/detail/CVE-2022-0547​

3. OpenSSL v. 1.1.1 package is upgraded 1.1.1m->1.1.1n (fixing CVE-2022-0778, score 7.5, High).

https://nvd.nist.gov/vuln/detail/CVE-2022-0778​

4. minidlna package is upgraded 1.3.0->1.3.1 (fixing CVE-2022-26505, score 7.4, High).

https://nvd.nist.gov/vuln/detail/CVE-2022-26505​

5. libxml2 package is upgraded 2.9.12->2.9.13 (fixing CVE-2022-23308, score 7.5, High).

https://nvd.nist.gov/vuln/detail/CVE-2022-23308​

6. libpcap package is upgraded 1.1.1->1.10.1 many CVE fixes (v1.1.1 is kept as well, it is used exclusevily by QCA QoS).

CVE - Search Results

7. expat package is upgraded 2.4.4->2.4.8 Relax fix to CVE-2022-25236 (introduced with release 2.4.5).
8. dropbear package is upgraded 2020.81->2022.82.
9. zlib package is upgraded 1.2.11->1.2.12.
10. libgpg-error package is upgraded 1.42->1.43.
11. curl package is upgraded 7.81.0->7.82.0.
12. wget package is upgraded 1.21.2->1.21.3.
13. ethtool package is upgraded 5.16->5.17.
14. util-linux package is upgraded 2.37.4->2.38.
15. ubus package is upgraded 2021-08-90->2022-02-28.
16. dbus package is upgraded 1.13.20->1.13.22.
17. libflac package is upgraded 1.3.3->1.3.4.
18. libogg package is upgraded 1.3.4->1.3.5.
19. sysstat package is upgraded 12.4.3->12.4.5.
20. usbmode package is updated 2017-12-19->2022-02-24.
21. libjpeg package is upgraded 9d->9e.
22. Fix NG/DNI 'igmpproxy' modified codes (to provide compatibility with 'libpcap').
23. Make an order in various Makefiles.
24. Host tools: upgrade mkimage/u-boot to 2022.01.

The link is:

https://www.voxel-firmware.com (thanks to vladlenas for his help with hosting).

Thanks to HELLO_wORLD for his help (making order in Makefiles).

Voxel.

Just upgraded to version 1.0.2.93SF - awesome as always, thank you Voxel.

 

[sppmaster]

Unfortunately the Netgear stock firmware contains a bug (@Voxel firmware is affected too) that causes WAN/LAN performance degradation.
Full details are here https://www.snbforums.com/threads/r...-when-a-client-is-connected-at-100mbps.78375/
I summarized the issue and the steps to reproduce it in this post https://www.snbforums.com/threads/r...ent-is-connected-at-100mbps.78375/post-758348
Probably many people will never notice this issue because they don't have the required setup to observe the WAN/LAN performance degradation. But it's simply there and the paradox is that a single 100Mbps device connected by cable to any LAN port in certain conditions can badly destroy the WAN/LAN performance of your network.

 

[TheLostSwede]

Unfortunately the Netgear stock firmware contains a bug (@Voxel firmware is affected too) that causes WAN/LAN performance degradation.
Full details are here https://www.snbforums.com/threads/r...-when-a-client-is-connected-at-100mbps.78375/
I summarized the issue and the steps to reproduce it in this post https://www.snbforums.com/threads/r...ent-is-connected-at-100mbps.78375/post-758348
Probably many people will never notice this issue because they don't have the required setup to observe the WAN/LAN performance degradation. But it's simply there and the paradox is that a single 100Mbps device connected by cable to any LAN port in certain conditions can badly destroy the WAN/LAN performance of your network.

Never noticed that and our Samsung TV is connected via 10/100Mbps Ethernet directly to the R7800. Everything else is connected via a switch one floor up though, so maybe I'm not seeing any issues due to only using two ports in total.
I'm curious though, why would you only use a four wire cable? I still run normal Cat 5e cabling to the TV.

 

[sppmaster]

Never noticed that and our Samsung TV is connected via 10/100Mbps Ethernet directly to the R7800. Everything else is connected via a switch one floor up though, so maybe I'm not seeing any issues due to only using two ports in total.
I'm curious though, why would you only use a four wire cable? I still run normal Cat 5e cabling to the TV.

You can see the details in my posts. This issue is difficult to be detected with just one 100Mbps device but it remains hidden and spoils the WAN/LAN performance differently depending on the current network traffic. And it is compulsory to have simultaneous WAN and LAN traffic in order to trigger the issue.
I needed one week to find the reason and to find the right setup in order to replicate the issue with 100% certainty.
I only use four fire cable just to be able to force the connection to 100Mbps for a Laptop that has 1Gbps NIC in order to be able to replicate the issue. The issue is not triggered when a device is connected at 1Gbps.
In my network all cables are Cat 5e to all devices but I have four 100Mbps devices in my LAN (TVs, Android TV boxes).
You can read all my posts. Here is the first one when I discovered the issue
https://forum.openwrt.org/t/ipq806x...k-c2600-linksys-ea8500/82525/1877?u=sppmaster
There are several other posts explaining my tests and later findings.
In general I have initially noticed that when someone watches a movie, streamed over LAN from PC file server to a TV set, on a second TV HBOMax cannot be watched normally because the WAN speed to the second TV is low. HBOMax needs 20Mbps average bitrate but the second TV device can achieve maximum 5Mbps or even lower.
It's really tricky when there is an additional switch (even a gigabit one) connected to the router because I can see the link speed is 1000Mbps, but some clients have only 100Mbps NICs and when there is LAN traffic between the clients this becomes a nightmare because all the clients connected to the second switch cannot download from Internet. The WAN speed sometimes is just 1-2Mbps with huge ping 200-500 ms.
For Voxel and Netgear stock firmware in order to replicate the issue with 100% certainty I need at least three 100Mbps devices, because with this two firmwares when I have 2 or less 100Mbps devices, LAN traffic is cut first and I thouht there was a problem with my cabling. But with 3 or more 100Mbps clients the WAN traffic is affected too. I have a Gigabit WAN connection and wondered how it cannot keep up with just a single HBOMax movie stream.