Have not tried myself yet
but the addresses can be changed under Setup > Network ( at least the last device specific bytes) and are stored in the nvram
normally the addresses should be restored with a backup
Posted: Sun Mar 21, 2021 10:05 Post subject: WRT1900ACSV2
Due to COVID-19 outbreak, I try to show the result for DD-WRT improvement quickly.
The firmware was updated from version r36168 to r46069.
Updated Reason from r36168: 5.8G Wifi is down.
Updated to r46069: Wifi Ok.
Issue from r46069:
1. Active Client cannot refresh when the device is shut down until reboot.
2. Creat Bridge was set to the value "0". After reboot, no internet. However, PPTP have still continued work.
3. PPTP Server load the pictures with Instagram (Android App). The performance is very low.
This problem had happened more 2 years but It does not fix since 2019. It is a big problem.
only one regulatory domain is selectable in the GUI because it makes no sense that the radios use different regulatory domains
That makes sense, thanks! Any idea why changing the GUI selection from UNITED_STATES to CANADA only changed wlan0 (5G) to CANADA in nvram? wlan1 (2.4G) was still UNITED_STATES (before my nvram editing). Could it be that CANADA and UNITED_STATES are considered interchangeabe for 2.4G (e.g., NORTH_AMERICA or FCC)?
ho1Aetoo wrote:
"wlan$_txpwrdbm=" is your GUI Setting for TX Power
The allowed transmission power depends on the regulatory domain and the frequency range (and has nothing to do with interferences or similar).
the regdb is also not stored in the nvram but is located in the normal file system under
Code:
/lib/crda/regulatory.bin
Quote:
"The only difference in the version 2 is where the Wi-Fi driver looks for the wireless power tables - meaning the v1 and v2 have the power tables in 2 different locations."
marvel devices have another power table, because they have something like a region lock
but i am relatively sure that the dd-wrt driver does not consider EEPROM restrictions anyway
and the power table is also not stored in nvram
it can be loaded from the DTS for example or is stored directly in the WLAN EEPROM
Ah, thanks very much for the explanation! I use Wireless Channel: Auto due to the congested local environment (and perhaps some laziness), and had noticed the dbm changing between 23 and 30 at different times, but I'd failed to notice that the change corresponded to different channel-specific limits. Great to know, thanks!
EDIT:
ho1Aetoo wrote:
Wasn't it somehow that the mac addresses are also stored in the nvram and cloned with?
My v1 shows its correct MAC, and my v2 (restored from v1 backup) shows its own correct MAC. My various wlan MACs look correct too. I'm not using any MAC cloning, though. _________________ My DD-WRT Routers:
Linksys WRT3200ACM - Marvell
Linksys WRT1900ACS - Marvell
Netgear R9000 - Atheros
Netgear R7000 - Broadcom
PC x86-64 VM - Atheros
That makes sense, thanks! Any idea why changing the GUI selection from UNITED_STATES to CANADA only changed wlan0 (5G) to CANADA in nvram? wlan1 (2.4G) was still UNITED_STATES (before my nvram editing). Could it be that CANADA and UNITED_STATES are considered interchangeabe for 2.4G (e.g., NORTH_AMERICA or FCC)
I wrote that your manual change for the nvram variable "wlan1_regdomain=" is useless and will not be taken into account
strictly speaking, the variable "wlan1_regdomain=" is without function! and a relic from old builds!
Something serious goign on with he NAS on this one using SMB. It corrupting my NTFS HDD after a period of time and I need to physically remove the drive and scan it with a windows client which reports errors and repairs them. Plugging it back in it works..and then the same issue reappears a couple hours later.
Joined: 04 Aug 2018 Posts: 1447 Location: Appalachian mountains, USA
Posted: Sun Mar 28, 2021 20:56 Post subject:
I'll edit this into a proper report when I get done (days to weeks) rebuilding and testing my rather involved config for 46069 (first re-enter of everything for a couple of years), but meanwhile, here's an interesting heads-up:
On an WRT1900ACSv2 running 46069 (my first build since 44048) I just measured (dslreports.com/speedtest) 113 Mbps down through the router's OpenVPN client from an AirVPN (airvpn.org) server over 800mi / 1300km from here. (Accessed 200 Mbps ISP connection through a second router, so double NAT.) That's nearly twice my previous OpenVPN speed record for this little router. Edit: The OpenVPN speed gain went away once I got my firewall rules in. Apparently I have work to do in that area.
FWIW, the Air server was a 1 Gb/s server running OpenVPN 2.5.0. All their servers are on OpenVPN 2.5 now and have the CHACHA20-POLY1305 cipher. They have a couple of recently added 10 Gb/s servers in Europe, but the rest are 1 Gb/s setups for now. The AirVPN/OpenVPN speed record measured by a client is around 700 Mbps, and that was not actually on the new high-speed servers. I have a bash script that creates a complete dd-wrt OpenVPN config of nvram commands from an AirVPN .ovpn file provided by their online config generator. Willing to share with beta testers. See link in my sig for what else you'd need to know and PM me if you'd like to try the script.
Edit: iPhones (all WPA2-AES, AC/N-Mixed and NG-Mixed, SP. SGI, SUB, no protection, defaults elsewhere) hang many times per day, always correctable by cycling the phone's wifi switch to force a reconnect. Most of my experience is with 5G, but it hangs at least occasionally on 2.4G as well. I've patiently worked my way through trying a long list of "fixes" that have appeared in the forum. None were at all useful. No log clues spotted yet, but I need to search more thoroughly. The only noteworthy aspect is, as was noted by @geps in the 46259 new-build thread, seldom do busy processes hang. It's new activity that reveals the problem. Curiously, a music stream will usually keep playing even as entering its app to browse finds it hung.I had to add this because the 46329 new-build thread is beginning to fill with tales of 46069 being wifi salvation. It isn't. _________________ 2x Netgear XR500 and 3x Linksys WRT1900ACSv2 on 53544: VLANs, VAPs, NAS, station mode, OpenVPN client (AirVPN), wireguard server (AirVPN port forward) and clients (AzireVPN, AirVPN, private), 3 DNSCrypt providers via VPN.
Joined: 04 Aug 2018 Posts: 1447 Location: Appalachian mountains, USA
Posted: Mon Mar 29, 2021 15:19 Post subject:
How much NVRAM space is actually left?
My (WRT1900CSv2) Status page is now reporting nvram usage as "48 KiB / 128 KiB," and "nvram show >/dev/null" reports "size: 50072 bytes (81000 left)." But this router has 64K of nvram space, verified by my having run out of it once, necessitating a restore from backup in order to boot correctly. Has BS found a way to rearrange memory partitioning? I didn't think this was possible at the build level. Or is this a simple reporting error?
Edit: Let me add this: PLEASE FIND MORE NVRAM SPACE if it is actually possible! My routers are configured to within inches of their lives, and my main home router has <300 bytes of nvram left if it really tops out at 65536 bytes.
Last edited by SurprisedItWorks on Fri Apr 16, 2021 15:08; edited 1 time in total
I tried this release again after some people recently reported it worked with Apple clients. Last month when I tried this release, my Apple clients hung but I did not do a full reset at that time. See my note on page-1 of this thread.
Flashed my 1900ACSv1 via CLI from 44048. I did several 'nvram erase' and reboots before and after. Manually re-entered my config but I kept all the wireless defaults except for channel and 80mhz channel width on 5ghz. No issues with the upgrade. No crash dumps in dmesg like with the recent releases. Wifi performance is great when it is working.
Problems with Apple devices remain: I was hoping a St Patrick's Day release combined with a release number ending in 69 would be good luck, but alas. Wifi stays connected but most traffic stops flowing. Pings stop between the router and device and visa-versa. Speedtests on one or two Apple devices can reproduce the problem within (30) minutes. Recycling the wifi on the affected device fixes it right away.
iPhone SE 2020 - IOS 14.4.2
Macbook Pro 2018 - MACOS 11.2.3
Macbook Air 2010 - music streamer seems stable at HT40 5ghz
Others...
Nothing in syslog or dmesg.
I confirmed that the wifi firmware and drivers are the same with 44048 and 46069:
Wifi stays connected but most traffic stops flowing. Pings stop between the router and device and visa-versa. Speedtests on one or two Apple devices can reproduce the problem within (30) minutes. Recycling the wifi on the affected device fixes it right away.
iPhone SE 2020 - IOS 14.4.2
Macbook Pro 2018 - MACOS 11.2.3
Macbook Air 2010 - music streamer seems stable at HT40 5ghz
Others...
Nothing in syslog or dmesg.
I confirmed that the wifi firmware and drivers are the same with 44048 and 46069:
driver name: mwlwifi
chip type: 88W8864
hw version: 7
driver version: 10.3.8.0-20181120
firmware version: 0x0702091b
power table loaded from dts: yes
firmware region code: 0x0
mac address: 00:25:9c:13:90:d0
2g: disable
5g: enable
antenna: 4 4
irq number: 44
ap macid support: 0000ffff
sta macid support: 00010000
macid used: 00000001
radio: enable
iobase0: e0c80000
iobase1: e0e00000
tx limit: 768
rx limit: 64
qe trigger number: 2630
...
Linksys WRT1900ACSv1: router mode, hardwired ethernet lan to a WRT3200ACM. WIFI: 2.4ghz: NG-mixed, 20mhz channel width, channel 11, WPA2-CCMP-128. 5ghz: AC/N mixed, 80mhz channel width, channel 36+upper, WPA2-CCMP-128.
...
Hmmm... Thanks for checking! There must be something significantly different between our configs and/or client devices that's yielding divergent results. Just some observations:
Our Apple iOS clients/supplicants are working quite well on this r46069 build, but we haven't had any Apple MacOS devices lately. Are your MacOS devices both updated?
I'm unclear as to whether you have 2.4G disabled (your driver info seems to conflict with your summary). Ours roam between 2.4G and 5G, which I suppose could possibly be masking something.
Another possible difference is channel width, especially if you're using 80 MHz! Due to our overcrowded wifi environment (combined with fewer 5G allocations to begin with in the Canada regulatory domain), we're actually using 20 MHz.
We're also using FreeRadius and separate vlans for each encryption algorithm (WPA3 Enterprise on one, WPA2 Enterprise on another).
Other potentially determinative differences might be harder to spot, but I'll offer my 5G wifi settings as a screenshot.
EDIT: Screenshot attached as a pdf.
EDIT #2: As a follow-up, I've borrowed the same iPad again, which seemed to be running just fine with this build on our ACSv1 production unit, and I've been specifically testing it with this build on our ACSv2 test unit. So far, after over an hour of uptime and several roughly evenly spaced visits to speedtest.net, all's well. I can also confirm that it's stayed on the 5G radio so far (I think it only roams to 2.4G on its own when farther and/or more walls to AP). I'll update if anything changes. _________________ My DD-WRT Routers:
Linksys WRT3200ACM - Marvell
Linksys WRT1900ACS - Marvell
Netgear R9000 - Atheros
Netgear R7000 - Broadcom
PC x86-64 VM - Atheros
thanks for checking and helping debug this. In my past testing, it did seem to be more stable with 20ghz channel width, but I have no hard evidence. Also, I did test WPA3 on a release in February IIRC, and it was very stable but it only had two Apple devices connected - so I don't know if it was WPA3 that fixed it, or the fact that all the other WPA2-only devices in the environment were connected to another router during testing.
My two main MacoOS laptops are running the latest Big Sur release. I do have that old Macbook Air from 2010 that is a few releases behind and only connects at HT40 width, but these hanging problems existed before I pulled that MB-Air out of storage and repurposed a couple months ago. When my modern Apple devices start hanging, the old Macbook Air is still running fine - it's a music streamer and then music keeps playing.
The cat of the driver info above is only for the phy0 (5g) antenna - it's the same for the phy1 (2.4ghz) antenna accept it says 5g disable, 2g enable:
I did test WPA3 on a release in February IIRC, and it was very stable but it only had two Apple devices connected - so I don't know if it was WPA3 that fixed it, or the fact that all the other WPA2-only devices in the environment were connected to another router during testing.
My two main MacoOS laptops are running the latest Big Sur release. I do have that old Macbook Air from 2010 that is a few releases behind and only connects at HT40 width, but these hanging problems existed before I pulled that MB-Air out of storage and repurposed a couple months ago. When my modern Apple devices start hanging, the old Macbook Air is still running fine - it's a music streamer and then music keeps playing.
The cat of the driver info above is only for the phy0 (5g) antenna - it's the same for the phy1 (2.4ghz) antenna accept it says 5g disable, 2g enable:
Ah, that makes sense, thanks! I've got a hunch: In Wireless Security, have you tried setting 802.11w Management Frame Protection to "Auto"? If it is already, you might also try enabling WPA Strict Rekeying. _________________ My DD-WRT Routers:
Linksys WRT3200ACM - Marvell
Linksys WRT1900ACS - Marvell
Netgear R9000 - Atheros
Netgear R7000 - Broadcom
PC x86-64 VM - Atheros
I tried setting 802.11w Management Frame Protection to enable first - 3 Apple devices connected but all other devices were shut out because they didn't support it. I also turned on strict rekeying. With just those 3 devices connected to the 5ghz-80vht antenna, pings never stopped but different apps or sessions within an app on my macbook would hang. For example, on my macbook,I could get to google on the browser but not facebook, whereas the facebook app on my iphone was working great. I saw similar weirdness on the iphone.
I kept management frame protection on enable on the 5ghz antenna but set it to auto on the 2.4ghz antenna, and a lot of devices were able to connect to the 2.4ghz antenna now. All traffic and pings hung up fairly quickly on my apple devices connected to the 5ghz antenna.
I disabled all that stuff on the wifi security page and then set my wifi advanced settings to exactly what you had - except for canada domain and I selected a channel for each antenna. The 20mhz 5ghz channel width seemed to work ok but performance stinks. I ran out of time to play around but I will try more tests with the 5ghz channel set to 20mhz wide tomorrow. Maybe I'll try shutting off the 5ghz antenna and force everything to the 2.4ghz antenna, and then see if I can reproduce it.