Stripping in Chapter 6

Jonathan Oksman jonathan.oksman at
Mon Oct 31 16:44:11 PDT 2011

On Mon, Oct 31, 2011 at 6:43 PM, Bruce Dubbs <bruce.dubbs at> wrote:
> Are the numbers in K?  I suspect so.

I forgot to put a size flag when I got the stats from du, so they're
in the default amounts du returns.  I think that's 512 bytes, so half
kilobyte units.  Half the numbers below would be in kilobytes.  Sorry
about that... I thought something seemed off when I was getting the

> Is this for Chapter 5 or 6?

Chapter 6.  I ran it against my host system backup which has a couple
of extra things like which and gpm.  I didn't bother doing it against
chapter 5 after noticing how small of a difference there was between
--strip-debug and more excessive stripping flags.

> In any case, it seems the bottom line is that a user can save up to
> 113M.  For a 10G partition, that's about 1%.
> It seems that stripping would rarely provide significant improvements.

~55-56M factoring my little error above.  So yeah, the gain in space
isn't terribly impressive.  I always knew that the debugging symbols
ate up most of the symbol space, but now I don't think I'll bother
stripping unless I'm doing something with embedded, and even then
it'll probably only be --strip-debug at most.


More information about the lfs-dev mailing list