Bricked E8450 while installing

[grauerfuchs]

Ok. At this point, it looks like the flash chip is defective. Although there are still things that can be done for testing and probing, it's looking increasingly unlikely that you'll end up with a reliable device unless that chip is replaced. The flash chip has just shown it won't reliably read data. The failed upgrade procedure showed that it won't reliably write data. Even if you can get it to take an image after a hard wipe, there's a good chance that simply adding or changing user data will eventually result in failure all over again.

If you still want to go through with testing for academic purposes, you can try the command mtd dump.raw.oob spi-nand0 0x1c0000 and see if anything comes back. If it doesn't return anything at all, there's nothing to be done except wipe the chip flat and know that the factory data has been lost.

Completely resetting the chip

The flash chip can be completely wiped and reset with the command mtd erase.dontskipbad spi-nand0. No data will be retained. After issuing this command, you will need to manually reload all data partitions before the router will be able to boot.

Restoration

If you're willing to start with 23.05, the instructions to restore a router from a bad flash are in the main topic for this router under the label 'hard recovery instructions'. There is also a link to that post from the wiki. Said post also has a link to a surrogate file that can be used to replace the factory data and regain most if not all function. However, there may be unseen issues in doing so since the surrogate is not a complete backup of the factory data, but is instead a rebuild containing only the data that was listed within the device drivers. Please also note that the files are definitely version-specific. Files for 24.10 can't be used to load to 23.05, and files for 23.05 or older can't be used to load 24.10.

The process for restoring 24.10 from a bad flash is a bit more involved. Although all of the bits and pieces can be found within the same forum topic, I don't recall seeing a ready-made guide to do so.

[russell]

@grauerfuchs

Restoration

If you're willing to start with 23.05, the instructions to restore a router from a bad flash are in the main topic for this router under the label 'hard recovery instructions'. There is also a link to that post from the wiki.

Do you have a link? I'm not certain what you are referring to. Looks like this.

This is totally a backburner thing for me at this point, but it's behaving like it's not even talking to the NAND chip, like the SPI transceiver is misconfigured or isn't working. Weirdly, the stock firmware was working and the installer was working right up to the moment of the reboot in the console log.

[grauerfuchs]

The link to the specific post is here:

Belkin RT3200/Linksys E8450 WiFi AX discussion

Hard recovery of the Linksys e8450 / Belkin RT-3200 If the router's firmware has been damaged beyond the point of simple patching and operation, you can still restore the router to full function as long as you have previously taken a backup of your factory partition. The instructions for doing so are below. Be warned that if something goes wrong with the below, it could leave your router unable to boot and without data at all. Therefore, the steps below should only be followed if you know what y…

I've also just updated it with the initial version of the direct-to-24.10 restoration instructions.

This particular flash chip is not wired for SPI communication. Instead, it talks to the CPU via the on-SOC SNFI implementation. However, the bigger issue is the apparent low quality of the flash chip used. It seems excessively prone to bit-rot. If that weren't enough, the stock firmware wasn't properly using error correction on the flash chip, so the stock firmware was both more likely to be affected by issues and also less likely to recognize when issues were present. You say this router was acquired as a refurb? Well, this is probably the reason for it. It's quite possible the chip is/was so close to the edge of failure that although the original data write could still be mostly read, any further attempts could have been outside of tolerance.