• State
  • Locked Locked
  • Replies 2 replies
  • Subscribers 98 subscribers
  • Views 1838 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

How to enable ECC and change caching policy on kernel?

Guilherme Costa
Intellectual 635 points

Hello,

I'm working with an application that must guarantee that data is saved (even if partially) in case of a power failure. We've opted to use JFFS2, and are currently facing some issues with ECC. Looking through the boot messages, this is one of the first ones shown:

[    0.000000] Memory policy: ECC disabled, Data cache writeback

So it appears I have ECC disabled. I've poked around the internet, but did not find out how to turn it on. Also, I've stumbled upon a forum thread stating that changing the data caching policy should make my system more robust (I only found some information regarding changing some parameters on fstab, but all the examples I've seen of this were for ext filesystems). As these two issues seem to be deeply related, I would like to know how to enable these features on the kernel.


Also, we are using U-Boot, and kernel version 3.2.61, on an AM335x.

Regards,

Guilherme

Source:

ez.analog.com/.../10730

  • 0
    TI__Guru** 116100 points
    This message is not related to ecc for flash based devices, it's for core.

    You no need to worry about this line.

    You can tell us how and when you are getting into ecc failures?
  • 0 in reply to Titusrathinaraj Stalin
    Intellectual 635 points

    Hi,

    This is the message that pops up more often: 

    [10939.320388] mtd->read(0x1044 bytes from 0x38bd430) returned ECC error
[10939.327780] Data CRC failed on REF_PRISTINE data node at 0x038bd430: Read 0x6520845b, calculated 0x71a211ba

    Although after some code inspection, I am inclined to believe that this is more of a informational message, rather than a indication of something wrong with the ECC. However, I am not 100% sure about it, so your opinion would be appreciated.

    I am currently running a test on the board, and this message shows up every now and then, but the test is not failing. The test itself consists of writing a file to disk several times, and the message appears sometimes, without stopping the test.

    By the way, we are using a custom board with AM335x, and a modified 3.2.61-arch-psp04.06.00.11 kernel.

    Regards,
    Guilherme