On Mon, Oct 29, 2012 at 06:11:58PM +0000, Luck, Tony wrote:

quoted

What I would recommend is adding a 

#define FS_CORRUPTED_FL		0x01000000 /* File is corrupted */

... and which could be accessed and cleared via the lsattr and chattr
programs.

Good - but we need some space to save the corrupted range information
too. These errors should be quite rare, so one range per file should be
enough.

New file systems should plan to add space in their on-disk format. The
corruption isn't going to go away across a reboot.

No, not at all. if you want to store something in the filesystem
permanently, then use xattrs. You cannot rely on the filesystem
being able to store random application specific data in their
on-disk format. That's the *exact purpose* that xattrs were
invented for - they are an extensible, user-defined, per-file
metadata storage mechanism that is not tied to the filesystem
on-disk format.

The kernel already makes extensive use of xattrs for such metadata -
just look at all the security and integrity code that uses xattrs to
store their application-specific metadata.  Hence *anything* that
the kernel wants to store on permanent storage should be using
xattrs because then the application has complete control of what is
stored without caring about what filesystem it is storing it on.

Cheers,

Dave.
-- 
Dave Chinner
david@fromorbit.com