JFFS3 & performance
Artem B. Bityuckiy
dedekind at infradead.org
Wed Dec 22 13:20:44 EST 2004
I didn't investigate this. But may be I will do. Anyway, it should be
But not all flashes are protected by ECC and I believe we want to have the
same nodes format for any flash. From the another hand, those who not
protected are robust.
Now I may only say about CRC32 - with properly selected polnome it
protects 1, 2 bit errors and burst of consequitive bit errors. At least papers
I've read stand this.
On Wed, 22 Dec 2004, xemc wrote:
> > adler32 is strong agains 1bit errors:
> > adler32 is relatively strong against 2bit erros:
> > For higher level errors, adler32 should have the usual a-priori
> > chances of catching them.
> > Overall, it is not as good as crc32, but pretty close, esp. for long
> > data. I would entrust my personal data to it, provided that the
> > device in general is good enough and the first adler32 checksum
> > failure (ignoring power-failure during writes) results in a message to
> > discard the broken flash chips or something similar.
> Please forgive me if this is a naive question, but isn't this data
> also protected by ECC in the first place? (or is that just for NAND?)
> How strong is this ECC compared to CRC32 or Adler32?
> If the ECC can handle a few bit errors, then wouldn't a simple
> checksum handle the case where the file was completely corrupt or
> partially written?
> Please correct any wrong assumptions.
> Linux MTD discussion mailing list
Artem B. Bityuckiy,
More information about the linux-mtd