Re: Bug#24217: bigloo not stripped
Rob Browning <rlb@cs.utexas.edu> writes:
> Steve Dunham <dunham@cse.msu.edu> writes:
>
> > > C) Depending on the size of the binary, not stripping it does more
> > > than just take up disk space, it takes up extra swap and ram when
> > > running, and potentially increases memory fragmentation and
> > > decreases cache locality)
> >
> > This doesn't make sense to me. How does not stripping the binary make
> > it take up more RAM and swap?
> Given your explanation, it wouldn't. Even after a reasonable amount
> of time, I'm still not entirely used to the fact that I'm now using an
> OS that often does the right thing :>
:) Yup. Certain other OSes exhibit some rather inexplicable behavior:
My brother is running a Lotus Notes server on a machine running NT 4.0
server. Whenever he gets a javascript error in Netscape (running on
the same machine), the Notes server dies.
> I had assumed that the debugging symbols would be intermingled with
> the code. In such a situation, you get deleterious effects on the
> memory heirarchy (especially wrt cache locality). If things are
> organized as you say, then that's not an issue...
They are seperate, "objdump --headers" will list them.
ojbdump says the sections labelled ".stab" and ".stabstr" are the
debugging sections.
IIRC, you do get some intermingling of extra code if you compile for
profiling, but not debugging.
Steve
dunham@cps.msu.edu
--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Reply to: