Re: [ft-devel] ttinterp.c optimization

[Alexei Podtelezhnikov]

On Wed, Aug 27, 2014 at 12:42 AM, Werner LEMBERG <address@hidden> wrote:
>
>> Furthermore, it turns out that MPS instruction is *very* *very*
>> common.  Why do we call CURRENT_Ppem here?  The comment suggest that
>> the decision was arbitrary and, perhaps, wasteful.  I see a huge
>> number of trivial FT_MulFix's by 1.0 still.
>
> Measuring ppem and pt values are done along the projection vector, so
> they *must* be recomputed each time since this value can change any
> time.  However, it certainly makes sense to catch factor `1.0'
> explicitly instead of performing a multiplication.  So please
> optimize :-)

Never mind my MPS nonsense due to a gdb miss. It is MPPEM which is
called a lot. I wonder if it makes sense to cache the value and update
it only when it potentially changes by other instructions. What do you
think? What can change it?