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.

JFFS2 slow to mount, should I change to YAFFS/YAFFS2?

My DM6467 EVM derived custom board with NAND flash and slightly customized linux 2.6.32 kernel takes over TWENTY FIVE SECONDS just to mount JFFS2.

Console output:
...
[Sun Mar 18 07:46:34.277 2012] [ 3.340000] async_continuing @ 1 after 2 usec

[Sun Mar 18 07:47:04.913 2012] [ 28.400000] VFS: Mounted root (jffs2 filesystem) on device 31:3.
...

I'm wondering about changing to YAFFS/YAFFS2, but I just read at http://processors.wiki.ti.com/index.php/Create_a_YAFFS_Target_Image that "YAFFS/YAFFS2 scans the entire file system during mount to check for valid data and recreate the file system tree. This can result in longer boot times, particularly for large flash devices".  My flash is 128MBtyes and "df" reports that it's 125568 1K-blocks. (1GBit flash NAND01GW3B2AZA6E.)

Or perhaps there's some other problem.  Just earlier today, I shaved 51 seconds off my 2:17 boot time (argh!) simply by replacing TraceUtil_start() with Engine_start().  So I'm concerned that the whole 25 second thing with jffs2 might be another dumb snafu, where yaffs/yaffs2 won't fix it but some other trivial change will.

Note also my file system that I'm flashing to NAND is about 106MB.  I don't currently have the knowledge to pare that down smaller.  Is it simply that this is unreasonably large? However, even if it is too large, the device is liable to create large files anyway, which would slow things back down even if reducing this size sped them up temporarily.

Thanks in advance for your help.