Re: parallelized 'ld'?

[Andrew Cagney <ac131313 at redhat dot com>]

> An interesting idea, and one I, at least, never considered.  At one
> point the linker was mainly disk bound, which suggests that
> parallelization will not bring a big improvement.  However, I did that
> profiling 7 or 8 years ago, on Sun hardware running Solaris, and it is
> quite possible that on modern hardware the tradeoffs are very
> different.

The numbers have likely changed but the the tradeoffs haven't.    In 
fact, if anything, disks have got slower(1).  Consequently, I'd expect 
the traditional bottle necks of I/O bandwidth and memory usage(2) to 
still apply.

Andrew

(1) A vax had a 14ms latency disk, looking at a random modern disk it 
has 3ms latency - 5x faster.  CPUs on the other hand are 1000x faster.

(2) GDB needs to keep an eye on context switch / kernel overhead