Re: [MiNT] Proposal for SLB extension

["Konrad M. Kokoszkiewicz" <draco@obta.uw.edu.pl>]

Hi Joerg,

> JR>However it has the drawback that for every process that accesses an
> JR>SLB, the kernel needs to allocate additional memory.
> Yes, but IMO that shouldn't be a problem as it's only a few bytes.

No, at least 4 kilobytes, see my other mail.

> The drawback of Konrad's proposal is that SLBs which are not aware of his
> extension will be executed with the overhead he wants to reduce. Not to

So what? Currently all the libs are executed with the overhead. Even after
my proposal will have been implemented, you still can keep the old
behaviour just keeping the flag fiels in the SLB header equal to zero.
Just when you want to execute the library function with less overhead, you
set the flags to 1 keeping in mind the basepage address is not passed to
libs functions then. Moreover, such a library (with flags = 1) will still
work properly on MagiC and older versions of MiNT. So where is the
problem??

> mention that his proposal creates two kinds of SLB, which is IMO unneccessary
> and will tend to produce at least irritations

See above. If one does not accept the fact tha there are 'two types' of
libs, one can ignore that and keep flags = 0. Such a lib will still work,
just function calls will be slower.

Switch off the complicator, please :)

--
Konrad M.Kokoszkiewicz
mail: draco@atari.org
http://draco.atari.org

** Ea natura multitudinis est,
** aut servit humiliter, aut superbe dominatur (Liv. XXIV,25)
*************************************************************
** Taka to juz natura pospolstwa, ze albo sluzy ono unizenie,
** albo bezczelnie sie panoszy.