Update of /cvsroot/mplayer/main/libvo In directory mail:/var/tmp.root/cvs-serv19941/libvo Modified Files: vo_svga.c Log Message: fixing rgb4 & bgr4 (2 pixels per byte) adding bg4b & rg4b (1 pixel per byte) Index: vo_svga.c =================================================================== RCS file: /cvsroot/mplayer/main/libvo/vo_svga.c,v retrieving revision 1.61 retrieving revision 1.62 diff -u -r1.61 -r1.62 --- vo_svga.c 25 Jan 2003 09:30:13 -0000 1.61 +++ vo_svga.c 30 Jan 2003 09:15:09 -0000 1.62 @@ -517,7 +517,10 @@ } #endif if (pformat == IMGFMT_YV12) { - yuv2rgb_init(bpp, MODE_RGB); + if(bpp==4) + yuv2rgb_init(bpp|128, MODE_RGB); + else + yuv2rgb_init(bpp, MODE_RGB); } x_pos = (WIDTH - maxw) / 2; @@ -748,7 +751,7 @@ case 8 : if ((format == IMGFMT_RGB8 ) || (format == IMGFMT_BGR8)) return ((bpp_avail & BPP_8 ) ? 1 : 0); break; - case 4 : if ((format == IMGFMT_RGB4 ) || (format == IMGFMT_BGR4)) + case 4 : if ((format == IMGFMT_RG4B ) || (format == IMGFMT_BG4B)) return ((bpp_avail & BPP_4 ) ? 1 : 0); break; case 1 : if ((format == IMGFMT_RGB1 ) || (format == IMGFMT_BGR1))
Modified Files: vo_svga.c Log Message: fixing rgb4 & bgr4 (2 pixels per byte) adding bg4b & rg4b (1 pixel per byte) Yes, now that you (A'rpi and Michael) fixed it, it won't segfault for me,
Michael Niedermayer CVS wrote: thanks! However, it didn't got faster :( Before: -vop scale=320:200 SwScaler: BICUBIC scaler, from Planar YV12 to BGR 4-bit using 3DNOW V: 10.7 268 19% 312% 0.0% 0 0 0% Now: -vop scale=320:200 SwScaler: BICUBIC scaler, from Planar YV12 to BGR 4-bit per byte using 3DNOW V: 18.0 450 21% 310% 0.0% 0 0 0% Now: -vop scale,crop=320:200 SwScaler: using unscaled Planar YV12 -> BGR 4-bit per byte special converter V: 18.0 450 21% 217% 0.0% 0 0 0% Now: -vop scale,crop=320:200 -dr SwScaler: using unscaled Planar YV12 -> BGR 4-bit per byte special converter V: 18.0 450 22% 215% 0.0% 0 0 0% Now: -vop scale SwScaler: using unscaled Planar YV12 -> BGR 4-bit per byte special converter VO: [svga] 352x288 => 352x288 BGR 4-bit per byte [fs] vo_svga: SVGAlib resolution: 640x350 4bpp - No video scaling V: 18.0 450 22% 636% 0.0% 0 0 0% -- Gabucino MPlayer Core Team "not sure how we will proceed here - xine's potential in the video processing field is imho so great that i certainly don't want to miss the chance to work into that direction." - Guenter, xine developer
Hi On Thursday 30 January 2003 14:07, Gabucino wrote:
Michael Niedermayer CVS wrote:
Modified Files: vo_svga.c Log Message: fixing rgb4 & bgr4 (2 pixels per byte) adding bg4b & rg4b (1 pixel per byte)
Yes, now that you (A'rpi and Michael) fixed it, it won't segfault for me, thanks!
However, it didn't got faster :(
Before: -vop scale=320:200 SwScaler: BICUBIC scaler, from Planar YV12 to BGR 4-bit using 3DNOW V: 10.7 268 19% 312% 0.0% 0 0 0% ^^^
Now: -vop scale=320:200 SwScaler: BICUBIC scaler, from Planar YV12 to BGR 4-bit per byte using 3DNOW V: 18.0 450 21% 310% 0.0% 0 0 0% ^^^ different number of frames? dunno if it is the problem but the used BGR4 code wasnt actually changed so this is quite mysterious otherwise, could u retry with a real -benchmark
[...] Michael
Michael Niedermayer wrote:
V: 10.7 268 19% 312% 0.0% 0 0 0% ^^^ 3DNOW V: 18.0 450 21% 310% 0.0% 0 0 0% ^^^ different number of frames? It's also different length of time tested (10.7 vs 18.0)
dunno if it is the problem but the used BGR4 code wasnt actually changed so this is quite mysterious otherwise, could u retry with a real -benchmark -vo svga -bpp 4 -noaspect -vop scale -nosound -benchmark -frames 200 -ss 20 -idx
Cannot init MMX colorspace transform No accelerated colorspace conversion found SwScaler: using unscaled Planar YV12 -> BGR 4-bit per byte special converter VO: [svga] 352x288 => 352x288 BGR 4-bit per byte [fs] vo_svga: SVGAlib resolution: 640x350 4bpp - No video scaling V: 28.0 700 20% 635% 0.0% 0 0 0% BENCHMARKs: VC: 1.614s VO: 50.858s A: 0.000s Sys: 1.069s = 53.542s BENCHMARK%: VC: 3.0148% VO: 94.9880% A: 0.0000% Sys: 1.9972% = 100.0000% -- Gabucino MPlayer Core Team
Hi,
Log Message: fixing rgb4 & bgr4 (2 pixels per byte) adding bg4b & rg4b (1 pixel per byte) Yes, now that you (A'rpi and Michael) fixed it, it won't segfault for me, thanks!
Michael and Ivan, not me
However, it didn't got faster :(
Don't expect Michael to optimize your ISA bus or EGA "GPU". A'rpi / Astral & ESP-team -- Developer of MPlayer, the Movie Player for Linux - http://www.MPlayerHQ.hu "However, many people beg for its inclusion in Debian. Why?" - Gabucino "Because having new software in Debian is good." - Josselin Mouette "Because having good software in Debian is new." - Gabucino
However, it didn't got faster :( Don't expect Michael to optimize your ISA bus or EGA "GPU". Thanks for flaming me, but, as svgalib can do 50% VO and vf_scale does 600%,
Arpi wrote: the problem does not lie with my bus, or graphics card. BTW, if Michael thinks he doesn't want to mess with this, _he_'ll tell me to fuck off. Not you. -- Gabucino MPlayer Core Team
Gabucino writes:
Arpi wrote:
However, it didn't got faster :( Don't expect Michael to optimize your ISA bus or EGA "GPU". Thanks for flaming me,
For someone so prone to flames you are very easily offended .. Diego
participants (5)
-
Arpi -
Diego Biurrun -
Gabucino -
Michael Niedermayer -
Michael Niedermayer CVS