UBIFS errors when file-system is full

[Richard Weinberger]

Am 14.07.2015 um 06:23 schrieb Bhuvanchandra DV:
> Hello Richard,
> 
> On 07/13/2015 07:39 PM, Richard Weinberger wrote:
> 
>> On Mon, Jul 13, 2015 at 3:55 PM, Bhuvanchandra DV
>> <bhuvanchandradv at gmail.com> wrote:
>>> Hello,
>>>
>>> Facing some issues with UBIFS mounting when NAND flash is almost near to
>>> full.
>>>
>>> Hardware:
>>> SOC: Freescale Vybrid
>>> COM: Colibri VF50
>>> Flash: Macronic NAND 128MB, 3.3V, 8-Bit
>>>
>>> By default the Linux Image flashed on the NAND flash almost occupies ~93% of
>>> the flash.
>>> Observed this issue when power resetting the module for about 500 times,
>>> with journalctl
>>> enabled, after around 500 power cycles the debug log fills up the nand
>>> flash. When the
>>> flash menory is about to full, the ubifs partation is un-mountable.
>>>
>>> The nand flash controller driver is under review and passed all MTD tests.
>>> https://lkml.org/lkml/2015/6/18/847
>>>
>>>
>>> Debug log during UBIFS mounting from U-Boot:
>>>
>>> U-Boot 2015.04 (May 18 2015 - 15:37:34)
>>>
>>> CPU: Freescale Vybrid VF500 at 396 MHz
>>> Reset cause: POWER ON RESET
>>> DRAM:  128 MiB
>>> NAND:  128 MiB
>>> MMC:   FSL_SDHC: 0
>>> *** Warning - bad CRC, using default environment
>>>
>>> In:    serial
>>> Out:   serial
>>> Err:   serial
>>> Model: Toradex Colibri VF50 128MB V1.1A, Serial# 04815679
>>> Net:   FEC
>>> Hit any key to stop autoboot:  0
>>> Booting from NAND...
>>> UBI: default fastmap pool size: 50
>>> UBI: default fastmap WL pool size: 25
>>> UBI: attaching mtd1 to ubi0
>>> UBI: attached by fastmap
>> Is this a recent kernel with all fastmap fixes applied?
> 
> I'm sorry! I didn't mention about the kernel version before.
> 
> Tested with 4.0.5 kernel[1]. I think your fastmap fixes[2] are not available with
> this kernel version.
> 
> Lately also tested on 4.1.1 kernel[3], observed the same issues.

I'm confused. Does this mean that you can _reproduce_ the issue from a clean
working UBI image also with 4.1.1?
Or does it mean that the already broken image does also not work with 4.1.1?

Thanks,
//richard