Can't format JFFS partition =(

[zidler]

Hello,
I am trying to enable the jffs partition on my RT-AC56U using Administration menu. No luck =(

I think the most interesting part of my log is this

start jffs2: 4, 65798144
/ # nand_erase_nand: attempt to erase a bad block at page 0x000092c0
brcmnand: Input/output error

What options do I have? Thx in advance!

 

[wocram]

Hello,
I am trying to enable the jffs partition on my RT-AC56U using Administration menu. No luck =(

I think the most interesting part of my log is this

start jffs2: 4, 65798144
/ # nand_erase_nand: attempt to erase a bad block at page 0x000092c0
brcmnand: Input/output error

What options do I have? Thx in advance!

Following thread might be interesting for you:
http://www.snbforums.com/threads/jffs-not-mounting-read-only-error.8376/

 

[john9527]

Since you are using an AC56U, as a last resort, you might try loading my 374 LTS fork. It's an older code base and handles the jffs partition slightly different from the current code. Load via the ASUS Firmware Recovery Tool then factory reset before trying to format jffs. Reboot twice after selecting to do the format.

 

[Jozek]

Same problem here with ac68u

 

[sfx2000]

Since you are using an AC56U, as a last resort, you might try loading my 374 LTS fork. It's an older code base and handles the jffs partition slightly different from the current code. Load via the ASUS Firmware Recovery Tool then factory reset before trying to format jffs. Reboot twice after selecting to do the format.

At some point - the vendors should be seriously consider eMMC vs. direct NAND, then JFFS would never be a problem

 

[Jozek]

Which version is the last one that allows to mount the partition manually?

 

[L&LD]

At some point - the vendors should be seriously consider eMMC vs. direct NAND, then JFFS would never be a problem

eMMC would be alright, but an actual ssd would be better.

 

[sfx2000]

eMMC would be alright, but an actual ssd would be better.

eMMC for the core OS and settings, and an SATA M2 or mini-PCI slot perhaps?

 

[RMerlin]

eMMC, SSD... Aren't high-end routers expensive enough already?

 

[L&LD]

eMMC, SSD... Aren't high-end routers expensive enough already?

Exactly. That's why they need to use better hardware.

 

[sfx2000]

eMMC, SSD... Aren't high-end routers expensive enough already?

eMMC adds about 3 bucks in volume for a 4GB chip compared to regular 256MB NAND...

 

[L&LD]

eMMC adds about 3 bucks in volume for a 4GB chip compared to regular 256MB NAND...

View attachment 7799

At those prices, the latest routers (from all manufacturers) should have the 16GB MLC chips inside.

 

[Jozek]

We need real solutions

 

[L&LD]

We need real solutions

Uhm, to what exactly?

 

[Jozek]

Uhm, to what exactly?

Mount and format partition... There are serveral posts with this problem without solutions long time ago... Except The people that a mage fix it

 

[RMerlin]

eMMC adds about 3 bucks in volume for a 4GB chip compared to regular 256MB NAND...

View attachment 7799

Does that include the memory controller for it, and all the other external circuit?

 

[sfx2000]

Does that include the memory controller for it, and all the other external circuit?

Remove the discrete NAND, and put an eMMC, and like I said, it's about 3 bucks, perhaps less... as eMMC is on the right side of the pricing/supply curve, and the old school NAND is on the downside, esp as FAB's are moving in that direction... moving from SLC NAND to MLC/TLC and 3D-VNAND technology - at some point, producing old-school NAND will not be profitable...

It's not that much of a cost adder, as MMC is similar to SDIO - the eMMC is self-hosting with it's own controller, so it's a matter of running the lines out to the single eMMC chip, and enabling the driver in the OS...

In any event, discrete NAND chips, both for expansion via JFFS, and real-time (where that address block must be mirrored into RAM) is 2005 technology in applications built around the Broadcom SDK, which I'm reasonably certain that we can both agree is getting long in tooth - being based largely on the Linksys WRT54G GPL drops from even before (and OPKG being based on Unslung which goes back to... NSLU2...)

Think about the advantages of eMMS and a real file system - one can do OverlayFS (or AUFS if one wants to be really bold) - this offers the opportunity to do targeted fixes without having to rebuild the entire flash image - and this is huge considering the security risks that we're facing these days - just fix the lib or application in place... Entware/Optware is kinda there already, but this is an add-on, can you imagine the benefits here for both off-the-shelf and carrier provided equipment?

I can...

 

[Jozek]

Still Broken... Nobody knows how to fix it or a miracle?


Enviado desde mi iPhone utilizando Tapatalk

 

[sfx2000]

Since you are using an AC56U, as a last resort, you might try loading my 374 LTS fork. It's an older code base and handles the jffs partition slightly different from the current code. Load via the ASUS Firmware Recovery Tool then factory reset before trying to format jffs. Reboot twice after selecting to do the format.

Take a look at this patch to mdt.c from OpenWRT-Devel - it's been around for about 3 years, so good chance that might have been picked up already by either Broadcom directly, or perhaps Asus has looked at it...

https://lists.openwrt.org/pipermail/openwrt-devel/2013-December/022744.html

 

[RMerlin]

Take a look at this patch to mdt.c from OpenWRT-Devel - it's been around for about 3 years, so good chance that might have been picked up already by either Broadcom directly, or perhaps Asus has looked at it...

https://lists.openwrt.org/pipermail/openwrt-devel/2013-December/022744.html

I did implement bad block support a few years ago, but I can't remember if it was only in the mtd-erase code or in the other operations as well.