Re: [GSoC 2013] Defragmentation for FFS

[Manuel Wiesinger <manuel%lungenarzt-wien.at@localhost>]

On 07/25/13 02:23, Mouse wrote:> I'm not sure VFS will/should know 
anything about it.  Since
> fragmentation is a filesystem-specific thing, it seems to me this
> should all happen inside ffs; I wouldn't expect the VFS layer to know
> anything about it.
You are absolutely right. As I said, I did not really think about that
yet. Also I didn't have something in my code where this would be relevant.

> Actually, I suspect you'd get better crash tolerance out of
>
> fs_clean = false
> for each data move operation {
>    allocate new blocks
>    flush allocation bitmaps
>    copy data blocks
>    flush copied block contents
>    update metadata (inode, indirect blocks)
>    flush changed metadata
>    deallocate old blocks
>    flush allocation bitmaps
> }
> fs_clean = true
>
> As for fs_clean, I see little point in setting fs_clean to true after
> each copy only to immediately set it false again in almost all cases.
Jep, this is better than my approach. My idea was to save a whole 
(probably unnecessary)
fsck operation, when a crash happens in between two move operations. 
Which is pointless now, since the complete analysis part is done before 
anything is moved. Obviously I forgot to adapt the crash safety stuff, 
when I started I did not have an analysis step.

Thank you for sharing your thoughts. That was very helpful!

Manuel