Bootstrapping GCC

[Jeremy Huntwork]

Ken Moffat wrote:
>  I've long-since lost any lingering belief that 'upstream' _always_
> knows what is best, but my initial thought is that knowing the
> toolchain lets gcc recreate itself is a good thing.  But, I wonder
> if it is ever likely to fail at pass 2 without general failures in
> the next compile (i.e. does bootstrapping of itself highlight errors
> more quickly ?)

Well, consider, gcc is compiled once, then that resultant gcc compiles 
the next stage, and then that one the final stage. If the *bootstrap* 
fails, ie, in stage2 or stage3, (especially stage2) then that should 
indicate that our toolchain can't produce a sane and working gcc.

--
JH