e2fsck -c to check for bad blocks

[Haines Brown <brownh@hartford-hwp.com>]

I gather one can use the e2fsck with -c option to have it call
/sbin/badblocks to report bad blocks on an unmounted partition. 

1. Although the -c option causes fsck to use badblocks to identify any
bad blocks present, does e2fsck then proceed to use this information to
fix corruption as usual? That is, how does 

          # e2fsck -cy /dev/sda1

differ from simply:

          # e2fsck -y /dev/sda1
           
2. If badblocks is non-destructive, why does the targeted filesystem
have to be unmounted? 

3. While e2fsck is run on an unmounted file system, the man page says,
"If this [-c] option is specified twice, then the bad block scan will be
done using a non-destructive read-write test." Does this "specified
twice" simply mean "-cc"? If the test is non-destructive, can it be run
on a mounted filesystem? I assume not, but wanted to be sure.

4. Both badblocks and e2fsck -c can identify bad blocks as part of a
check of hard disk viability. Is the difference only that while
badblocks just reports bad blocks, e2fsck -c actually goes ahead and
tries to fix them? 
 
-- 
 
       Haines Brown, KB1GRM