Re: [EXT] [PATCH v2 3/3] mtd: rawnand: micron: Address the shallow erase issue

[Richard Weinberger <richard@xxxxxx>]

----- Ursprüngliche Mail -----
> Von: "Bean Huo, beanhuo" <beanhuo@xxxxxxxxxx>
> An: "richard" <richard@xxxxxx>, "Boris Brezillon" <boris.brezillon@xxxxxxxxxxxxx>
> CC: "Miquel Raynal" <miquel.raynal@xxxxxxxxxxx>, "Vignesh Raghavendra" <vigneshr@xxxxxx>, "Tudor Ambarus"
> <Tudor.Ambarus@xxxxxxxxxxxxx>, "linux-mtd" <linux-mtd@xxxxxxxxxxxxxxxxxxx>, "Steve deRosier" <derosier@xxxxxxxxx>,
> "Thomas Petazzoni" <thomas.petazzoni@xxxxxxxxxxx>, "tglx" <tglx@xxxxxxxxxxxxx>, "Zoltan Szubbocsev, zszubbocsev"
> <zszubbocsev@xxxxxxxxxx>, "Piotr Wojtaszczyk" <WojtaszczykP@xxxxxxxxxxxxxxxxxx>
> Gesendet: Donnerstag, 7. Mai 2020 11:28:59
> Betreff: RE: [EXT] [PATCH v2 3/3] mtd: rawnand: micron: Address the shallow erase issue

> Hi Richard
> Thanks.
> 
> How about this special  situation:
> 
> Page 0(EC) is good;
> Page 1(VID) is bad;
> Page 2 (data) is good;
> Page 3( data) is bad, or contain filling pattern.
> Page 4 (data) is good, or empty;
> Page 5( data) is bad, or contain filling pattern.
> Page 6 (data) is good, or empty;

"bad" means ECC errors upon read?

UBI will be unhappy in scanning mode if VID header is bad but payload
does not show ECC errors nor bitflips and is not 0xFF.

See check_corruption() in drivers/mtd/ubi/attach.c
I'm not super eager to soften these checks but as last resort
we can modify them.

Fastmap is more forgiving and just checks whether the VID header is corrupted.
While reading this code I think we can actually use check_corruption() there too, hmmm.

Thanks,
//richard

______________________________________________________
Linux MTD discussion mailing list
http://lists.infradead.org/mailman/listinfo/linux-mtd/