Re: Replacing bitmap.[ch]

[Daniel Berlin <dan at cgsoftware dot com>]

Richard Henderson <rth@redhat.com> writes:

> On Tue, Jul 24, 2001 at 05:12:39PM -0600, law@redhat.com wrote:
>>   > Removing NOT and IOR_COMPL make sense, especially if they aren't used.
>> But sbitmap_not is used in gcse & lcm
> 
> I thought he was just talking about removing
> them from bitmap.[ch], not sbitmap.[ch].
> 
Correct.
I'm also trying to provide most of the functionality of sbitmap, so
the new bitmaps *could* be used for things like GCSE, with increased
space efficiency, and little (or no) loss of speed.

For the purposes of computing ~(TRANS | COMP), we'd be better off
computing it as an sbitmap, and converting it.
The intermediates are dense, the result, not so much.
If it was, we'd be better off leaving as an sbitmap anyway, or
using an interval tree based packed set (which I have if we do
determine it's really dense).  
I can't think of an alternative formulation that would involve less
dense sets as intermediates.

> 
> r~

-- 
"Babies don't need a vacation, but I still see them at the
beach...  It pisses me off!  I'll go over to a little baby and
say, "What are you doing here?  You haven't worked a day in your
life!"
"-Steven Wright