This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: [new-regalloc] What is the status on current sources
- To: Michael Hayes <m dot hayes at elec dot canterbury dot ac dot nz>
- Subject: Re: [new-regalloc] What is the status on current sources
- From: Daniel Berlin <dberlin at redhat dot com>
- Date: Sat, 10 Feb 2001 17:47:41 -0500 (EST)
- cc: Daniel Berlin <dberlin at redhat dot com>, Michael Matz <matzmich at cs dot tu-berlin dot de>, Geert Bosch <bosch at gnat dot com>, "gcc at gcc dot gnu dot org" <gcc at gcc dot gnu dot org>
On Sun, 11 Feb 2001, Michael Hayes wrote:
> Daniel Berlin writes:
> >
> > a) Example bitvector of length 16.
> >
> > ...*...***.*****
> > compressed form: 0 3 1 3 3 1 5
> >
> > (Read it like this: start with 3 zeros, then 1 one,
> > then 3 zeros, then 3 ones, then 1 zero, then 5 ones.)
>
> This is a variant of run length encoding.
Right.
We just don't explicitly keep track of what numbers we are encoding,
except the start bit because we save space, and time that way (negate just
requires changing
the first number).
> How many bits did you allocate for each length element?
Right now the list of integers is a simple array of unsigned int, we
realloc it when necessary.
> Michael.
>