[FFmpeg-devel] pre discussion around Blackfin dct_quantize_bfin routine

[Reimar Doeffinger]

Hello,
On Tue, Jun 12, 2007 at 02:15:36PM -0400, Marc Hoffman wrote:
> On 6/12/07, Reimar Doeffinger <Reimar.Doeffinger at stud.uni-karlsruhe.de> wrote:
> > Btw.: Do you know of debian or SuSE packages for the compilers and stuff
> > needed for blackfin? As probably everyone knows by now, my main (gentoo)
> > development box (the laptop I also had at LinuxTag) is probably dead
> > forever so I can't do things the easy way right now :-(
> 
> Sorry to here about your dead machine Reimar, do you have something
> else to work on right now?

Yep, the great choice between a P2 and an Athlon 800, the later with
occasional lockups ;-). But I already ordered everything to build a new
machine. Unfortunately Acer can no longer deleiver replacement graphics
cards for the Aspire 1523. Proves that passively cooled and crappily
manufactured graphics cards break very often...

[...]
> > can the compiler handle "=d" (t0) ?
> > getting rid of the seperate lo in some way removes one possibility for
> > the compiler to mess up...
> >
> 
> YEP THATS IT! just curious if r1 has really been allocated or not
> tho..  I was walking through print_insn inside the compiler and %H0
> just selects the successor register i.e. REGNO+1.  I'm not sure if
> this really does what we want it to.
> 
>   asm volatile ("%0=cycles; %H0=cycles2;" : "=d" (t0));

I was more thinking of

asm volatile ("%0=cycles; %1=cycles2;" : "=d" (t0), "=d"(hi));
t0 |= hi << 32;
return t0;

But probably this results in the same mess as the first try.
I probably should also mention that sometimes the code if the standalone
function is quite a bit more messy (useless copying of registers) than
that of the actual inlined code (again, purely based on Atmel AVR
experience).

Greetings,
Reimar Doeffinger