6244

Just to prevent confusion here: 0x20000 (hexadecimal) == 131072 (decimal) == 128kiB erase block size.

The flash does have a 2kiB page layout when it comes to reading, and 2k+64 means 2048 bytes data + 64 bytes out-of-band area for error correction and detection (on this device: used by MediaTek's in-SoC BCH engine).

1 Like
6245

This proves that my issue is different. I had no errors.

Anyway, many thanks for the write-up; it removed most of the unknowns from my understanding of bad blocks and will surely help others.

6246 
root@OpenWrt:~# grep "(release)" /dev/mtd0ro
v2.10.0 (release):OpenWrt v2024.01.17~bacca82a-3 (mt7622-snand-ubi-1ddr)
v2.10.0 (release):OpenWrt v2024.01.17~bacca82a-3 (mt7622-snand-ubi-1ddr)
v2.10.0 (release):OpenWrt v2024.01.17~bacca82a-3 (mt7622-snand-ubi-1ddr)
v2.10.0 (release):OpenWrt v2024.01.17~bacca82a-3 (mt7622-snand-ubi-1ddr)
root@OpenWrt:~#

I'm currently on snapshots, want to move to
24.10.0-rc2, am i good to go with sysupgrade?

6247

Yes, you can go ahead, sysupgrade should work as expected.

1 Like
6248

@daniel

Checked beamforming and BSS color on snapshot latest as i thought it's ON built-in and there's no need to input in wireless Config manually

Here are outcomes

beamforming FINE

root@OpenWrt:~# grep -i beamform /tmp/run/hostapd*.conf
/tmp/run/hostapd-wl1.conf:vht_capab=[RXLDPC][SHORT-GI-80][TX-STBC-2BY1][SU-BEAMFORMER][SU-BEAMFORMEE][MU-BEAMFORMER][MU-BEAMFORMEE][RX-ANTENNA-PATTERN][TX-ANTENNA-PATTERN][RX-STBC-1][SOUNDING-DIMENSION-4][BF-ANTENNA-4][MAX-MPDU-7991][MAX-A-MPDU-LEN-EXP7]
/tmp/run/hostapd-wl1.conf:he_su_beamformer=1
/tmp/run/hostapd-wl1.conf:he_su_beamformee=1
/tmp/run/hostapd-wl1.conf:he_mu_beamformer=1
root@OpenWrt:~#

BSS_Color says 128 which is NOT a valid value between 1 to 63

root@OpenWrt:~# grep he_bss_color /tmp/run/hostapd*.conf
/tmp/run/hostapd-wl1.conf:he_bss_color=128                                  root@OpenWrt:~#
6249

Hi There and thank you for the step. In Step 5, what is this recovery environment? Will I need a serial port and a terminal client?

Also, where do I store the 24.10,0-rc2 that is referenced in step 5?

Sorry for the basic questions... I have bricked one e8450, I down't want to brick a 2nd.

6250

I don't know the correct term, but it's like a temporal firmware that runs on RAM and it's only meant to flash the actual sysupgrade. You don't need anything. The router should be accessible at 192.168.1.1 on the browser like the regular OpenWRT image.

You store the 24.10.0-rc2 sysupgrade file on your PC and you install it the same way like you have flashed the 1.1.3 installer by now: in Luci > System > Backup/Flash Firmware. Don't keep any setttings and after flashing the sysupgrade image, the router will reboot at 192.168.1.1 and you can set it up as usual or restore the backed up config with the modifications outlined.

1 Like
6251

If BSS color is greater than 63, hostapd chooses a random value between 1 and 63, which is probably what most people want. Using a default of 128 makes it obvious that it was done intentionally (maybe not that obvious, but it definitely sticks out as being the wrong value if you're reading the docs).

4 Likes
6252

Thank you for clarifying this and sharing blob link, thank you :slight_smile:

6253

I just upgraded from 23.05x to 24.10 (compiled build). I followed the “In case the configuration wants to be preserved:” instructions from the Wiki and within minutes I was back up and running. Everything went perfectly smooth.

Thank you for keeping the Wiki up to date!! I was really dreading this update but this just made my weekend!

Edit:

I did deviate from the instructions by removing /etc/config/ubootenv and /etc/fw_env.config from a running 23.05 system. I then created a backup and proceeded to flash the ubi-installer. It was an unintentional mistake but harmless and prevents you from having to edit the backup. YMMV.

2 Likes
6254

Sorry if adding to this thread is not appropriate. Just let me know and and I will start a separate topic. I am trying to recover a LinkSys E8450 also. U-Boot not correct but I can boot via mtk_uartboot. But when I try to reload anything via tftp I just keep getting this.

Press ENTER to return to menu

    ( ( ( OpenWrt ) ) )       U-Boot 2023.07.02-OpenWrt-r24106-10cc5fcd00 (Sep 23 2
  1. Run default boot command.
  2. Boot system via TFTP.
  3. Boot production system from flash.
  4. Boot recovery system from flash.
  5. Load production system via TFTP then write to flash.
  **6. Load recovery system via TFTP then write to flash.**
  7. Load BL31+U-Boot FIP via TFTP then write to flash.
  8. Load BL2 preloader via TFTP then write to flash.
  9. Reboot.
  a. Reset all settings to factory defaults.
  0. U-Boot console

Press UP/DOWN to move, ENTER to select, ESC to quit

Error: ethernet@1b100000 address not set.
Error: ethernet@1b100000 address not set.
Error: ethernet@1b100000 address not set.
Error: ethernet@1b100000 address not set.
No ethernet found.

MT7622> setenv ipaddr 192.168.1.1
MT7622> setenv serverip 192.168.1.254
MT7622> tftpboot $loadaddr factory.bin

Error: ethernet@1b100000 address not set.
Error: ethernet@1b100000 address not set.
Error: ethernet@1b100000 address not set.
Error: ethernet@1b100000 address not set.
Error: ethernet@1b100000 address not set.
Error: ethernet@1b100000 address not set.
Error: ethernet@1b100000 address not set.
Error: ethernet@1b100000 address not set.

6255

I was up very late last night. Helps if I read the troubleshooting section. Below was the fix.

Q: Getting “Error: ethernet@e6800000 address not set.”?

A: Use commands inside console:

setenv bootcmd
setenv bootargs
setenv ethaddr xx:xx:xx:xx:xx:xx #(MAC from the sticker)
saveenv

1 Like
6256

The MAC address not being set in U-Boot is a symptom of missing or broken factory partition (in v1.0.x) or UBI volume (in v1.1.x).
So the procedure to manually set it in the U-Boot console makes sense mostly as a work-around so you are able to transfer a backup of the factory volume/partition onto the device and write it.

6258

Is there anything special about the ethernet ports on the Belkin RT3200 compared to the ZyXEL T-56 (Odido branded EX5601-T1)?

I’m trying to setup the ZyXEL as dumb AP in the same way as I have my Belkin. I connect to my main router via the WAN port and use the remaining ports for different VLANs. Firewall and DHCP is disabled. This works on the Belkin, but on the ZyXEL my config is rolled back when I connect to my main router via the WAN port. The config will stick when I use the first LAN port to connect to my main router.

Any comment regarding the difference between these two routers for the dumb AP scenario is appreciated.

6259

They are all connected to the same switch and that is different from the Zyxel. Best to continue the discussion in that thread and not over here.

1 Like
6260

I just noticed a change in Post Install Instructions In Openwrt E8450 Page, previously i remember it used to say, change governor to Ondemand and now it has been modified to User Space which sets it to Mediatek defaults, anyone can tell me by what values it sets CpU to Mediatek defaults? How it can be compared to performance governor which sets it at max? Any opinions ?

1000097761
10000977611220×1204 158 KB

6261

I did a demo this evening and, for the first time, out of the sealed box comes a refurbished device with v1.0.0 vendor firmware on it. I use the v1.1.3 unsigned installer firmware, it reboots, I try to sysupgrade from the recovery and things are haywire: reboots into the recovery as default. From a u-boot console, I can reboot to the production firmware, but if I sysupgrade to openwrt-24.10.0-rc4-mediatek-mt7622-linksys_e8450-squashfs-sysupgrade.bin, it kernel panics during boot. If the production firmware is a snapshot from circa August 23, 2024, I can select it from the u-boot menu, and it boots to something that looks normal, with the rootfs_data partition mounted. Another oddity, is that warm reboots fail, hanging at:

root@OpenWrt:~# reboot
root@OpenWrt:~# [   81.631862] mt7530-mdio mdio-bus:1f lan1: left allmulticast mode
[   81.637898] mt7530-mdio mdio-bus:1f lan1: left promiscuous mode
[   81.644093] br-lan: port 1(lan1) entered disabled state
[   81.708414] mt7530-mdio mdio-bus:1f lan2: left allmulticast mode
[   81.714552] mt7530-mdio mdio-bus:1f lan2: left promiscuous mode
[   81.720685] br-lan: port 2(lan2) entered disabled state
[   81.768128] mt7530-mdio mdio-bus:1f lan3: left allmulticast mode
[   81.774252] mt7530-mdio mdio-bus:1f lan3: left promiscuous mode
[   81.780393] br-lan: port 3(lan3) entered disabled state
[   81.837203] mt7530-mdio mdio-bus:1f lan4: left allmulticast mode
[   81.843321] mtk_soc_eth 1b100000.ethernet eth0: left allmulticast mode
[   81.849888] mt7530-mdio mdio-bus:1f lan4: left promiscuous mode
[   81.855911] br-lan: port 4(lan4) entered disabled state
[   82.366265] mt7530-mdio mdio-bus:1f wan: Link is Down
[   82.373904] mtk_soc_eth 1b100000.ethernet eth0: Link is Down
[   86.439990] reboot: Restarting system

F0: 102B 0000
F6: 0000 0000
V0: 0000 0000 [0001]
00: 0000 0000
BP: 0400 0041 [0000]
G0: 1190 0000
T0: 0000 02F1 [000F]
Jump to BL

NOTICE:  BL2: v2.10.0   (release):OpenWrt v2024.01.17~bacca82a-3 (mt7622-snand-ubi-1ddr)
NOTICE:  BL2: Built : 22:48:06, Aug 12 2024
NOTICE:  WDT: [40000000] Software reset (reboot)
NOTICE:  CPU: MT7622

Power cycle brings me back to a u-boot menu.

Looking closer at the serial console logs, the 24.10.0-rc4 is hanging at not finding a bunch of /dev/mtdblockN's, so it looks like it's a non-UBI version (which I thought didn't exist anymore). So, that starts to make sense.

EDIT: Oh, lol, I was blind. Squinting harder, there IS a ubi version. I somehow got the impression that there weren't any non-UBI versions anymore. Can't read and talk at the same time.

EDIT2: /dev/mtdblock0 (bl2), /dev/ubi0_0 (fip), /dev/ubi0_3 (ubootenv2) and /dev/ubi0_4 (recovery) sha256sums are the same as on a working device. Now reboot is rebooting into the 24.10.0-rc4 rather than recovery, so that's good. But I'm still seeing the hang, quoted above, during a warm reboot.

A different device running the same firmware build does not hang at that spot, but continues on, thusly:

NOTICE:  BL2: v2.10.0 (release):OpenWrt v2024.01.17~bacca82a-3 (mt7622-snand-ubi-1ddr)
NOTICE:  BL2: Built : 22:48:06, Aug 12 2024
NOTICE:  WDT: [40000000] Software reset (reboot)
NOTICE:  CPU: MT7622
NOTICE:  SPI-NAND: adjusting SPI-NAND pin drive strength to 12mA
NOTICE:  SPI-NAND: W25N01GV (128MB)
NOTICE:  UBI: scanning [0x80000 - 0x8000000] ...
NOTICE:  UBI: scanning is finished
NOTICE:  UBI: PEB size: 131072 bytes (128 KiB), LEB size: 126976 bytes
NOTICE:  UBI: VID header offset: 2048 (aligned 2048), data offset: 4096
NOTICE:  UBI: Volume fip (Id #0) size is 1015884 bytes
NOTICE:  BL2: Booting BL31
NOTICE:  BL31: v2.10.0  (release):OpenWrt v2024.01.17~bacca82a-3 (mt7622-snand-ubi-1ddr)
NOTICE:  BL31: Built : 22:48:06, Aug 12 2024


U-Boot 2024.07-OpenWrt-r27137-f51cb74473 (Aug 12 2024 - 22:48:06 +0000)

CPU:   MediaTek MT7622
Model: mt7622-linksys-e8450-ubi
DRAM:  512 MiB
Core:  49 devices, 21 uclasses, devicetree: separate
MMC:   
Loading Environment from UBI... SPI-NAND: W25N01GV (128MB)
Read 126976 bytes from volume ubootenv to 000000005f7bf0c0
Read 126976 bytes from volume ubootenv2 to 000000005f7de100
[...]

What might cause that hang?

Here's a diff of the cold-boot boot logs (with timestamps trimmed) between the problematic device and the happily warm booting device:

--- /tmp/screenlog.2.txt        2025-01-03 04:11:14.347589979 -0800
+++ /tmp/screenlog.9.txt        2025-01-03 04:11:25.488154784 -0800
@@ -1,11 +1,11 @@
 F0: 102B 0000
 F6: 0000 0000
 V0: 0000 0000 [0001]
 00: 0000 0000
 BP: 0400 0041 [0000]
 G0: 1190 0000
-T0: 0000 02EE [000F]
+T0: 0000 02EF [000F]
 Jump to BL

 NOTICE:  BL2: v2.10.0   (release):OpenWrt v2024.01.17~bacca82a-3 (mt7622-snand-ubi-1ddr)
@@ -14,7 +14,7 @@
 NOTICE:  CPU: MT7622
 NOTICE:  WDT: disabled
 NOTICE:  SPI-NAND: adjusting SPI-NAND pin drive strength to 12mA
-NOTICE:  SPI-NAND: FM35Q1GA (128MB)
+NOTICE:  SPI-NAND: W25N01GV (128MB)
 NOTICE:  UBI: scanning [0x80000 - 0x8000000] ...
 NOTICE:  UBI: scanning is finished
 NOTICE:  UBI: PEB size: 131072 bytes (128 KiB), LEB size: 126976 bytes
@@ -32,7 +32,7 @@
 DRAM:  512 MiB
 Core:  49 devices, 21 uclasses, devicetree: separate
 MMC:
-Loading Environment from UBI... SPI-NAND: FM35Q1GA (128MB)
+Loading Environment from UBI... SPI-NAND: W25N01GV (128MB)
 Read 126976 bytes from volume ubootenv to 000000005f7bf0c0
 Read 126976 bytes from volume ubootenv2 to 000000005f7de100
 OK
@@ -265,7 +265,7 @@
  mtk_rng 1020f000.rng: registered RNG driver
  loop: module loaded
  mtk-ecc 1100e000.ecc: probed
- spi-nand spi2.0: Fidelix SPI NAND was found.
+ spi-nand spi2.0: Winbond SPI NAND was found.
  spi-nand spi2.0: 128 MiB, block size: 128 KiB, page size: 2048, OOB size: 64
  mtk-snand 1100d000.spi: ECC strength: 4 bits per 512 bytes
  2 fixed-partitions partitions found on MTD device spi2.0
@@ -284,13 +284,13 @@
  ubi0: VID header offset: 2048 (aligned 2048), data offset: 4096
  ubi0: good PEBs: 1020, bad PEBs: 0, corrupted PEBs: 0
  ubi0: user volume: 8, internal volumes: 1, max. volumes count: 128
- ubi0: max/mean erase counter: 2/0, WL threshold: 4096, image sequence number: 147074388
+ ubi0: max/mean erase counter: 7/4, WL threshold: 4096, image sequence number: 405897712
  ubi0: available PEBs: 0, total reserved PEBs: 1020, PEBs reserved for bad PEB handling: 20
  ubi0: background thread "ubi_bgt0d" started, PID 228
  OF: Bad cell count for /spi@1100d000/flash@0/partitions
  OF: Bad cell count for /spi@1100d000/flash@0/partitions
  block ubiblock0_5: created from ubi0:5(fit)
- mtk_soc_eth 1b100000.ethernet eth0: mediatek frame engine at 0xffffffc081300000, irq 126
+ mtk_soc_eth 1b100000.ethernet eth0: mediatek frame engine at 0xffffffc081440000, irq 126
  i2c_dev: i2c /dev entries driver
  mtk-wdt 10212000.watchdog: Watchdog enabled (timeout=31 sec, nowayout=0)
  NET: Registered PF_INET6 protocol family
@@ -416,14 +416,14 @@
 Press the [f] key and hit [enter] to enter failsafe mode
 Press the [1], [2], [3] or [4] key and hit [enter] to select the debug level
  UBIFS (ubi0:7): Mounting in unauthenticated mode
- UBIFS (ubi0:7): background thread "ubifs_bgt0_7" started, PID 732
+ UBIFS (ubi0:7): background thread "ubifs_bgt0_7" started, PID 729
  UBIFS (ubi0:7): recovery needed
  UBIFS (ubi0:7): recovery completed
  UBIFS (ubi0:7): UBIFS: mounted UBI device 0, volume 7, name "rootfs_data"
  UBIFS (ubi0:7): LEB size: 126976 bytes (124 KiB), min./max. I/O unit sizes: 2048 bytes/2048 bytes
  UBIFS (ubi0:7): FS size: 94724096 bytes (90 MiB, 746 LEBs), max 757 LEBs, journal size 4698112 bytes (4 MiB, 37 LEBs)
  UBIFS (ubi0:7): reserved for root: 4474052 bytes (4369 KiB)
- UBIFS (ubi0:7): media format: w5/r0 (latest is w5/r0), UUID B0E57F54-C5CC-4B4E-BA98-185D6690CF85, small LPT model
+ UBIFS (ubi0:7): media format: w5/r0 (latest is w5/r0), UUID 9C96635F-40FE-471D-8AB2-EB103C904C44, small LPT model
  mount_root: switching to ubifs overlay
  overlayfs: null uuid detected in lower fs '/', falling back to xino=off,index=off,nfs_export=off.
  urandom-seed: Seeding with /etc/urandom.seed
@@ -440,8 +440,8 @@
  ieee80211 phy0: Selected rate control algorithm 'minstrel_ht'
  mt7915e 0000:01:00.0: assign IRQ: got 130
  mt7915e 0000:01:00.0: enabling device (0000 -> 0002)
- urngd: v1.0.2 started.
  mt7915e 0000:01:00.0: enabling bus mastering
+ urngd: v1.0.2 started.
  mt7622-wmac 18000000.wmac: HW/SW Version: 0x8a108a10, Build Time: 20190801210006a

  mtk-pcie 1a143000.pcie: msi#0 address_hi 0x0 address_lo 0x44fb50c0

I'm not seeing anything that looks significant, other than (maybe?) the flash vendor, but I have other devices with Fidelix without apparent problems.

6262

Could try changing the CPU governor or raising the min speed to 600 MHz. Was noted in the thread about a similar problem.

6263

I found this: Belkin RT3200/Linksys E8450 WiFi AX discussion - #3089 by daniel

Adding the two lines:

echo 600000 > /sys/devices/system/cpu/cpufreq/policy0/scaling_min_freq
echo ondemand > /sys/devices/system/cpu/cpufreq/policy0/scaling_governor

to /etc/rc.local seems to fix the warm boot hang.

1 Like
6264

That's great! One note is that ondemand is the current default in OpenWrt for this device so no need to set that.