[MPlayer-users] Greetings! NetBSD user.. muaaahaha I love pain!

Felix Buenemann atmosfear at users.sourceforge.net
Wed Mar 20 00:35:02 CET 2002


Am Die, 2002-03-19 um 19.32 schrieb S. D.:
> [Automatic answer: RTFM (read DOCS, FAQ), also read DOCS/bugreports.html]
> On Sunday 17 March 2002 17:25, Felix Buenemann wrote:
> >
> > > just fine, but warnings such as:
> > > 
> > > dct64_k7.c:12: warning: alignment of `x_plus_minus_3dnow' is greater than 
> > > maximum object file alignment. Using 4.
> > > 
> > > --show up all over the place. The resulting binary plays movies not too 
> > > badly, however.
> > I guess there are some .align statements over the asm sources that
> > should really be .balign, because on eg. non-elf binaries .align will be
> > alignment by bits and .balign will be alignment by bytes, but on elf
> > it's both alignment in bytes.
> 
> Actually the line in question is:
> 
> static unsigned long long int __attribute__((aligned(8))) x_plus_minus_3dnow 
> =0x8000000000000000ULL;
> 
> So it looks like C-style compiler directives, aren't those on byte boundaries 
> anyway? Also, 0x80000...0ULL is only four bytes long. Out of curiosity, why 
> is it trying to align on 8 bytes? Is there some kind of packed doubleword 
> thing going on? :) Point me at a URL, I'll go away, promise. :)
simd instructions are faster on aligned data or might crash if data is
wrongly aligned.
> 
> -sudog
> 
> _______________________________________________
> RTFM!!!  http://www.MPlayerHQ.hu/DOCS
> Search:  http://www.MPlayerHQ.hu/cgi-bin/htsearch
> http://mplayerhq.hu/mailman/listinfo/mplayer-users
-- 
Best Regards,
	Atmos
____________________________________________
- MPlayer Developer - http://mplayerhq.hu/ -
____________________________________________




More information about the MPlayer-users mailing list