- May 15, 2006
-
-
Sam Hocevar authored
-
Sam Hocevar authored
-
- Dec 14, 2002
-
-
(no author) authored
'MPEG2DEC_0_3_1'.
-
Michel Lespinasse authored
-
Michel Lespinasse authored
-
Michel Lespinasse authored
-
Michel Lespinasse authored
left where we had these.
-
Michel Lespinasse authored
work on HPPA too. I can not test on HPPA, david can you make a test package ? Of course life would be better if -prefer-non-pic actually worked on IA64 and HPPA...
-
- Dec 13, 2002
-
-
Michel Lespinasse authored
-
Michel Lespinasse authored
and the memalign() shortcut. make sure we test with ccmalloc-gcc as well. fix one of the memory leaks.
-
Michel Lespinasse authored
(bad HP-UX no donut)
-
- Dec 12, 2002
-
-
Michel Lespinasse authored
-
Michel Lespinasse authored
-
Michel Lespinasse authored
-
- Dec 10, 2002
-
-
Michel Lespinasse authored
The precision loss is not significant (tested with IEEE1180 conformance tests)
-
- Dec 09, 2002
-
-
Michel Lespinasse authored
-
Michel Lespinasse authored
-
Michel Lespinasse authored
-
Michel Lespinasse authored
-
Michel Lespinasse authored
-
- Dec 06, 2002
-
-
Michel Lespinasse authored
I got rid of it.
-
Michel Lespinasse authored
-
Michel Lespinasse authored
-
Michel Lespinasse authored
BUTTERFLY has a straightforward implementation or a 'smarter' implemetation, selectable at compile time.
-
Michel Lespinasse authored
I dont know what I'm doing.
-
Michel Lespinasse authored
(when decoding in C mode) In addition I think it makes the code slightly less unreadable.
-
- Dec 04, 2002
-
-
Michel Lespinasse authored
-
- Dec 03, 2002
-
-
Michel Lespinasse authored
-
- Dec 02, 2002
-
-
Michel Lespinasse authored
-
Michel Lespinasse authored
-
Michel Lespinasse authored
-
Michel Lespinasse authored
getting rid of libcpuaccel as well.
-
Michel Lespinasse authored
This should not be used much but I'm including for potential portability issues.
-
Michel Lespinasse authored
mpeg2_init() does not have an mm_accels argument anymore You can still set the accelerations using mpeg2_accel()
-
- Dec 01, 2002
-
-
Michel Lespinasse authored
-
Michel Lespinasse authored
AM_CFLAGS is in addition to CFLAGS - I did not know that so I ended up having flags show up in duplicate.
-
Michel Lespinasse authored
just a little more often. Some background here. Because of mpeg2 oddification, when there is only one DC coefficient (block[0]) transmitted, block[63] will still be 1 if block[0] is even. So we have to make sure that even in these cases, we end up producing the same result as the full IDCT code would. In a full-precision IDCT, the DC-only code is equivalent to the full IDCT whenever (block[0] & 7 != 4) - odd values are no problem since block[63] is zero then, and the values 0, 2, 6 are OK too because the rounding happens to work fine then. This also applies to the C IDCT code as it is precise enough too. Because the MMX code is not just as precise, it yelds a different result than the full-precision one when block[0] & 7 == 6. So I've been preserving this behaviour in my previous code by letting us go thru the slow path in that case. However the fast path is actually more precise, so it's OK to go there if I just update the test suite accordingly. We still go thru the slow path when (block[0] & 7) == 4, as it's actually more precise for that case.
-
- Nov 30, 2002
-
-
Michel Lespinasse authored
About 1% to 1.5% speedup on my celeron 400 system.
-
Michel Lespinasse authored
This speeds up libmpeg2 by 2.5 to 3% on my system
-
Michel Lespinasse authored
about 2.5% faster on my machine (when using plain-C mode)
-