UBIFS: file data corruption during the power cut-off test

Sergei Poselenov sposelenov at emcraft.com
Thu Jun 6 11:08:03 PDT 2019

Hello Richard,

On Thu, 6 Jun 2019 16:55:27 +0200 Richard Weinberger <richard.weinberger at gmail.com> wrote:

> On Thu, Jun 6, 2019 at 11:10 AM Sergei Poselenov <sposelenov at emcraft.com> wrote:
> >
> > Hello,
> >
> > We have an embedded system based on i.MX 6ULL, with the root filesystem deployed on a UBI volume built on top of the SLC NAND device. The volume is mounted as "rw,sync,relatime".
> >
> > We are doing the following test:
> >
> >    - Copy the same file (/bin/busybox) to a different file names, in a sequence.
> >
> >    - Power is cut off during the copy sequence.
> >
> >    - After reboot, UBI mounts OK, and we see a number of valid copies of /bin/busybox (md5sum matches), and the last file with the short length (partially written), as expected.
> >
> > The question is: should we expect the last partially-written file to have the valid data of the original? We observe that quite often, the last 512 or so bytes in the short file doesn't match the original.  
> Hm, I don't fully understand your question.
> If you copy files and this is interrupted, the copy is not complete
> and therefore contents are missing.

This is understood. However, on the file length that is written to the partition, I'd expect that the file content will be the same as in the original file. This is not so.
Is it expected, or is it a deficiency of UBI?



More information about the linux-mtd mailing list