Re: Request: I/O request recording

[Bart Samwel]

Andrew Morton wrote:
> You could certainly do that.  Given disk block #N you need to search all
> files on the disk asking "who owns this block".  The FIBMAP ioctl can be
> used on most filesystems (ext2, ext3, others..) to find out which blocks a
> file is using.   See bmap.c in
>
> http://www.zip.com.au/~akpm/linux/patches/stuff/ext3-tools.tar.gz
>
> Unfortunately you cannot determine a directory's blocks in this way. 
> Ext3's directories live in the /dev/hda1 pagecache anyway.  ext2's
> directories each have their own pagecache.

OK, I've written something that does this (but only correctly for ext3). 
I've put it here:

http://www.xs4all.nl/~bsamwel/bootup_prefetch.tar.gz

I haven't had the opportunity to do good measurements, so I don't really 
know if it even increases performance. If anyone feels like benchmarking 
this, I'd be very happy to hear from you. I don't really expect 
performance increases, as the bootup scripts seem to have enough 
processing to do to keep the system busy even without disk I/O. I wonder 
if it might make a difference on a faster processor though, my system's 
kind of sluggish by today's standards.

-- Bart
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/