Belkin RT3200/Linksys E8450 WiFi AX discussion

Not sure if this is related to Openwrt or Philips Hue but I updated to snapshot r18803-2f024b7933 this morning and Alexa control and connectivity to my hue bridge is now intermittent. I have to tell Alexa multiple times to control lights and in my Hue account I have to refresh the bridge status multiple times before it connects. I noticed this snapshot updates firewall4 to 2022-02-12-53caa1a7-1 and wondering if anyone else is having issues with their Hue bridge with or without this latest update?

UPDATE: Reverted back to firewalll 3 by flashing release 0.6.1 then updating to the current snapshot and Alexa is now working reliably with the Philips Hue bridge.

1 Like

Hi PeterPan, you mentioned running your mesh on 5G 80MHz and getting about 250Mbps, could you please detail your OpenWrt version and overall settings?
With a single RT3200, 5G at 160Mhz, hardware and software offloading enabled, and with my laptops sitting right by the RT3200, I'm getting about 140Mbps via wifi. :frowning:

The only place there is 160MHz of contiguous bandwidth is on DFS channels. IIRC, this chip doesn't support 80+80MHz (split), just 160MHz contiguous bandwidth.

Unsure if/why it worked previously on a non-DFS channel, but I don't think it should work there.

Also, please only use the country code for the country you are in. You don't want to be broadcasting on frequencies that are not allowed in your country!

Just use 80Mhz. 160Mhz doesn't work well on this device.

1 Like

Woke this morning, 2.4G not working.

MacBook associates, no DHCP takes place, and then the Mac book disassociates.

Config unchanged from previously

config wifi-device 'radio0'
	option type 'mac80211'
	option path 'platform/18000000.wmac'
	option band '2g'
	option cell_density '0'
	option txpower '20'
	option distance '20'
	option country 'GB'
	option htmode 'HT20'
	option channel '2'
	option wps_pushbutton '1'

config wifi-iface 'wifinet4'
	option device 'radio0'
	option mode 'ap'
	option ssid 'lounge'
	option key 'XXXXXXX'
	option network 'lan'
	option encryption 'psk2'
	option skip_inactivity_poll '1'
	option disassoc_low_ack '0'
	option wps_pushbutton '1'

Could this be Firewall4 related, and if so what would be the best way to monitor and restart networks that fail.

Please post your system and kernel log to see if it contains any hints to what's wrong.

Will do it next time it freezes, but there nothing other than lan ports being enabled and disabled by devices waking

So I went from 160MHz to 80MHz and seeing an improvement from 140Mbps via iperf3 to 170Mbps, using 2 laptops connected via wifi, sitting about 5 inches away from my RT3200. I should be at 433Mbps in one direction, 867Mbps in the other...

(editing because as a new user I can only post 3 times in a topic apparently...)

Darksky, my RT3200 is a simple AP and the laptops are connected directly to it via 5G. Spectrum is good. (as a new user I'm limited to one image in each post...)

Thanks for the great explanation Daniel!!! Indeed with no other wifi client around, I get about 520Mbps on my 802.11ax laptop (with the other laptop connected via ethernet) and about 200Mbps on my 802.11ac laptop (with the other laptop connected via ethernet), with both laptops inches away from the RT3200.

(editing again)

Great news! I live in a condo with metal studs, and where I used to get 140Mbps at 20 yards with my 802.11ac-equipped laptop, I'm now getting 250+Mbps, thanks to replacing the stock antennas. :slight_smile: Clearly I'll have to set those up properly, but it's great to see!

I'm at 70Mbps with my current router (TP-Link AC1900) in the same circumstances.

6 Likes

What is the bug you guys are experiencing? Laptop1 to Laptop 2 is too slow over 5 GHz wifi in mesh mode?

1 Like

The physical RX and TX rate are not what you should expect to see as speed for payload. The physical rate is used for transfer, but keep in mind that there is a preamble, a guard interval and overhead of 802.11 headers, Ethernet headers and IP headers. So TCP throughput will be roundabout half of the physical rate or a bit better, but not much more.

As LAPTOP-QC0G... is only 802.11ac and apparently 1T2R you should manage to get around 500 MBit/s downstream and 250 MBit/s upstream speed from/to the AP if it's the only wifi device around.

Tje pop-os Laptop is 802.11ax 2T2R, so with that you should manage to get around 700-800 MBit/s downstream and probably a bit less than that upstream, but also that of course only if it's the only wifi device around.

Between the two wifi devices on the same network 170MBit/s is not so bad, given that 802.11ax is only performing well if all clients are 802.11ax (but the other Laptop is only 802.11ac).

3 Likes

I was having the same issue as you with the max upload being capped at 10mpbs (on a 1gbit symmetrical line). It turns out that a powered off computer I had connected to one of the ports was being detected as a 10Mbps/Full connection.

[ 2965.810819] mt7530 mdio-bus:00 lan4: Link is Up - 10Mbps/Full - flow control rx/tx

I physically unplugged the network cable of the powered off server from the router and I've got my full gig upload speed back.

That's interesting, that's also the case here. Why would one Ethernet port link speed determine the upload speed of another one (but not the download speed). Mega weird issue. Should help narrow down the mistake in the patch a lot, one would think

I am seeing similar to @justinkb.

Two RT3200 devices, connected via 1Gbps backhaul.

On each RT3200 there are a number of devices with 100mbps ethernet connections, and one backhaul link at 1gbps.

On the satellite AP, all ports LAN1 through LAN4, and WAN are connected to BR-LAN.
On the main router LAN1-LAN4 are bridged, and WAN is separate.

All ports connected to the same bridge are running at the lowest speed device in the bridge. with the current snapshot. To prove this I disconnected the two 100mbps devices leaving only the backhaul connected, and obtained the full 1gpbs link speed on the back haul,

1 Like

Reverting to @Daniel installer 0.6.2 version, the problem is gone.

@Mushoz @justinkb
Please check if this patch fixes the issue.

patch
--- a/drivers/net/dsa/mt7530.c
+++ b/drivers/net/dsa/mt7530.c
@@ -1582,6 +1582,21 @@ mt7530_hw_vlan_update(struct mt7530_priv *priv, u16 vid,
 	mt7530_vlan_cmd(priv, MT7530_VTCR_WR_VID, vid);
 }
 
+static int
+mt7530_setup_vlan0(struct mt7530_priv *priv)
+{
+	u32 val;
+
+	/* Validate the entry with independent learning, keep the original
+	 * ingress tag attribute.
+	 */
+	val = IVL_MAC | EG_CON | PORT_MEM(MT7530_ALL_MEMBERS) | FID(FID_BRIDGED) |
+	      VLAN_VALID;
+	mt7530_write(priv, MT7530_VAWD1, val);
+
+	return mt7530_vlan_cmd(priv, MT7530_VTCR_WR_VID, 0);
+}
+
 static void
 mt7530_port_vlan_add(struct dsa_switch *ds, int port,
 		     const struct switchdev_obj_port_vlan *vlan)
@@ -1595,6 +1610,9 @@ mt7530_port_vlan_add(struct dsa_switch *ds, int port,
 	mutex_lock(&priv->reg_mutex);
 
 	for (vid = vlan->vid_begin; vid <= vlan->vid_end; ++vid) {
+		if (!vid)
+			continue;
+
 		mt7530_hw_vlan_entry_init(&new_entry, port, untagged);
 		mt7530_hw_vlan_update(priv, vid, &new_entry,
 				      mt7530_hw_vlan_add);
@@ -1624,6 +1642,9 @@ mt7530_port_vlan_del(struct dsa_switch *ds, int port,
 	mutex_lock(&priv->reg_mutex);
 
 	for (vid = vlan->vid_begin; vid <= vlan->vid_end; ++vid) {
+		if (!vid)
+			continue;
+
 		mt7530_hw_vlan_entry_init(&target_entry, port, 0);
 		mt7530_hw_vlan_update(priv, vid, &target_entry,
 				      mt7530_hw_vlan_del);
@@ -2148,6 +2169,11 @@ mt7530_setup(struct dsa_switch *ds)
 			   PVC_EG_TAG(MT7530_VLAN_EG_CONSISTENT));
 	}
 
+	/* Setup VLAN ID 0 for VLAN-unaware bridges */
+	ret = mt7530_setup_vlan0(priv);
+	if (ret)
+		return ret;
+
 	/* Setup port 5 */
 	priv->p5_intf_sel = P5_DISABLED;
 	interface = PHY_INTERFACE_MODE_NA;
@@ -2318,6 +2344,11 @@ mt7531_setup(struct dsa_switch *ds)
 			   PVC_EG_TAG(MT7530_VLAN_EG_CONSISTENT));
 	}
 
+	/* Setup VLAN ID 0 for VLAN-unaware bridges */
+	ret = mt7530_setup_vlan0(priv);
+	if (ret)
+		return ret;
+
 	ds->configure_vlan_while_not_filtering = true;
 	ds->assisted_learning_on_cpu_port = true;
 	ds->mtu_enforcement_ingress = true;
--- a/drivers/net/dsa/mt7530.h
+++ b/drivers/net/dsa/mt7530.h
@@ -144,6 +144,8 @@ enum mt7530_vlan_cmd {
 #define  PORT_STAG			BIT(31)
 /* Independent VLAN Learning */
 #define  IVL_MAC			BIT(30)
+/* Egress Tag Consistent */
+#define  EG_CON				BIT(29)
 /* Per VLAN Egress Tag Control */
 #define  VTAG_EN			BIT(28)
 /* VLAN Member Control */

Save it as 999.patch and put it into target/linux/generic/backport-5.10 directory.

Reflects what I see I think:

root@OpenWrt:~# iperf3 -c 192.168.1.2
Connecting to host 192.168.1.2, port 5201
[  5] local 192.168.1.1 port 59086 connected to 192.168.1.2 port 5201
[ ID] Interval           Transfer     Bitrate         Retr  Cwnd
[  5]   0.00-1.00   sec   105 MBytes   877 Mbits/sec    0   1.89 MBytes
[  5]   1.00-2.00   sec  96.1 MBytes   807 Mbits/sec    0   1.89 MBytes
[  5]   2.00-3.00   sec   101 MBytes   849 Mbits/sec    0   1.89 MBytes
[  5]   3.00-4.00   sec  94.5 MBytes   793 Mbits/sec    0   1.89 MBytes
[  5]   4.00-5.00   sec  95.9 MBytes   804 Mbits/sec    0   1.89 MBytes
[  5]   5.00-6.00   sec   101 MBytes   845 Mbits/sec    0   1.89 MBytes
[  5]   6.00-7.00   sec   104 MBytes   872 Mbits/sec    0   1.89 MBytes
[  5]   7.00-8.00   sec   102 MBytes   854 Mbits/sec    0   1.89 MBytes
[  5]   8.00-9.00   sec  84.8 MBytes   712 Mbits/sec    0   1.89 MBytes
[  5]   9.00-10.00  sec  85.8 MBytes   720 Mbits/sec    0   1.89 MBytes
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bitrate         Retr
[  5]   0.00-10.00  sec   970 MBytes   813 Mbits/sec    0             sender
[  5]   0.00-10.01  sec   967 MBytes   811 Mbits/sec                  receiver

This is between two RT3200's on separate floors. The results are fairly similar between two RT3200's on the same floor at opposite ends of the house. If it's relevant I have all of my three RT3200's facing the same way and aligned in a parallel direction to one another. Covers whole house and surrounding garden and some of the path from the garden uphill towards the woods. So no complaints here about WiFi performance, albeit I wonder if I could have got by with just one of the Ubiquiti U6-LR UFOs.

@zdzT9G9P8tp6EEyHmltR it seems your environment is not entirely WiFi friendly!

1 Like

Could you share hardware details (part number of antennas you used, part number of the lead wires) and any relevant notes on the hardware modification (how you physically disconnected the factory stuff and connect the replaces) you made? Might be best to do in a separate thread dedicated to this mod.

I realize it's been 10 months since you posted, but I am wondering if you ever attempted to modify the RT3200 with external antennas?

If this patch is supposed to resolve the slow bridge speeds when mixing up 100mbps and 1Gbps ports, it does not work for me.

@Lynx Correct, my environment may be the worst possible scenario: large condo, concrete (of course) and metal studs for all walls. Worse than that, I have at least 22 Verizon FIOS routers around me broadcasting - out of a total of 36...

I could be wrong, but some time about 2+ years ago, it seems said FIOS routers started broadcasting much stronger than they used to, that's when I upgraded to a TP-Link AC1900. I originally got a repeater to get a better signal from my office (where the ISP entry point is) to my living room, but tweaking the channels on the TP-Link made that obsolete. I recently upgraded from 100Mbps Internet to 200 Mbps, so when I read the RT3200 would allow me to get both OpenWrt (for VLANs) and a stronger signal, I figured that would be the best, simplest, most cost effective solution. In my case though, adding external antennas to the RT3200 seems needed.