Current Patches, Where to get them, How they affect speed/output

[burfadel]

Isn't SVN no longer used?

[LoRd_MuldeR]

Originally Posted by burfadel

Isn't SVN no longer used?

Nope. VideoLAN, which is also hosting x264, switched to git recently...

[burfadel]

Thats what I thought, and would explain the problem he's having!

[Wishbringer]

Originally Posted by burfadel

Thats what I thought, and would explain the problem he's having!

Thought he is talking about compiler svn repo...

Originally Posted by bob0r

@MythCreator

Did you compile gcc 4.3.0 or did you use some pre packages version?

...

Originally Posted by MythCreator

pre package.I can't connect to SVN and I don't know why.

[akupenguin]

Originally Posted by burfadel

On my C2D, your 798.modified GCC 4.3.0 versions is fractionally slower than bob0r's 798.modified version. By fractionally I mean around .05 - 0.1 fps! (consistently)

I can think of 3 possible reasons:
* Bssd is slower than unpatched (applies only with direct=auto).
* Different video content for fprofile. Yes, I have said that the content doesn't matter much, but that doesn't mean I'd discount the possibility of .05 fps. (Well, maybe I would. .05 out of what, .2? Speed differences are meaningless. Please don't ever post one again. The meaningful measure is speed ratio, preferably accompanied by the standard deviation of N trials. The base fps that the ratio is relative to (or the amount of the difference, interderivable) is optional and is the least important datum.)
* Different compiler. No, I do not have faith in newer gcc being consistently faster than old versions. (Consider that in a program as small as x264, even a single pessimized instruction in the wrong place could have a significant effect on total speed.)

[Sharktooth]

gcc 4.x was always slower than 3.4.x in my tests...

[bob0r]

@MythCreator

Doom9 forum shortens URLs, so copying plain text isn't going to work.

Here to make it easy for you:
http://x264.nl/x264.modified.txt

Also note the updated x264_2pass_vbv.7.diff patch.
grab it from the collected patches dir, apply like this:
patch -p1 < x264_patches/x264_2pass_vbv.7.diff

[MythCreator]

Originally Posted by bob0r

@MythCreator

Doom9 forum shortens URLs, so copying plain text isn't going to work.

Here to make it easy for you:
http://x264.nl/x264.modified.txt

Also note the updated x264_2pass_vbv.7.diff patch.
grab it from the collected patches dir, apply like this:
patch -p1 < x264_patches/x264_2pass_vbv.7.diff

Thanks~~~~

[MythCreator]

x264.798.modified.beta2.exe

General thread:
http://forum.doom9.org/showthread.php?t=130364

x264_aq_var.48.786.fixed.diff
http://forum.doom9.org/showthread.php?t=132760
x264.gaussian.cplxblur.01.diff
Dark Shikari: - gaussian cplxblur: gives a tiny improvement in 2pass ratecontrol
x264_me-prepass_DeathTheSheep.01.diff
http://forum.doom9.org/showthread.php?p=1093523
x264_2pass_vbv.7.diff
http://thread.gmane.org/gmane.comp.v...093/focus=3748
x264_hrd_pulldown.04_interlace.diff
- HRD and pulldown for HD compatibility, updated patch for interlacing
http://forum.doom9.org/showthread.ph...19#post1047919

Link to x264 patches collected: http://files.x264.nl/x264_patches/



make fprofiled by GCC 4.3.0 , just for test

[Inventive Software]

@clsid: If you still read this thread, can you provide an install of MinGW with GCC 3.4.6 please?

@everybody: What's the difference between fprofiled and "normal" builds?

[survivant001]

@MythCreator. I obtain unknown option -- nal-hrd with your latest build

[Inventive Software]

x264_hrd_pulldown.04_interlace.diff
- HRD and pulldown for HD compatibility, updated patch for interlacing
http://forum.doom9.org/showthread.ph...19#post1047919

That patch adds the option. If you don't know what it is, don't use it.

[survivant001]

Originally Posted by Inventive Software

That patch adds the option. If you don't know what it is, don't use it.

I know that, and Yes I need it, but it's not include in the 798beta2

[MythCreator]

Originally Posted by survivant001

I know that, and Yes I need it, but it's not include in the 798beta2

It's just my fault...I forget to add it

[MythCreator]

x264.798.modified.beta2.fixed.exe

General thread:
http://forum.doom9.org/showthread.php?t=130364

x264_aq_var.48.786.fixed.diff
http://forum.doom9.org/showthread.php?t=132760
x264.gaussian.cplxblur.01.diff
Dark Shikari: - gaussian cplxblur: gives a tiny improvement in 2pass ratecontrol
x264_me-prepass_DeathTheSheep.01.diff
http://forum.doom9.org/showthread.php?p=1093523
x264_2pass_vbv.7.diff
http://thread.gmane.org/gmane.comp.v...093/focus=3748
x264_hrd_pulldown.04_interlace.diff
- HRD and pulldown for HD compatibility, updated patch for interlacing
http://forum.doom9.org/showthread.ph...19#post1047919

Link to x264 patches collected: http://files.x264.nl/x264_patches/



make fprofiled by GCC 4.3.0 , just for test

[MythCreator]

x264.798.modified.final.exe

General thread:
http://forum.doom9.org/showthread.php?t=130364

x264_aq_var.48.786.fixed.diff
http://forum.doom9.org/showthread.php?t=132760
x264.gaussian.cplxblur.01.diff
Dark Shikari: - gaussian cplxblur: gives a tiny improvement in 2pass ratecontrol
x264_me-prepass_DeathTheSheep.01.diff
http://forum.doom9.org/showthread.php?p=1093523
x264_2pass_vbv.7.diff
http://thread.gmane.org/gmane.comp.v...093/focus=3748
x264_hrd_pulldown.04_interlace.diff
- HRD and pulldown for HD compatibility, updated patch for interlacing
http://forum.doom9.org/showthread.ph...19#post1047919
x264_bssd.diff

Link to x264 patches collected: http://files.x264.nl/x264_patches/

make fprofiled by GCC 4.3.0






BTW:Is thread pool still usefull?

[survivant001]

for me it is. I don't where else to get the latest build patched.

[buzzqw]

i agree with survivant

BHH

[MythCreator]

Originally Posted by survivant001

for me it is. I don't where else to get the latest build patched.

I mean, is the Thread Pool.diff still useful at this time?

[burfadel]

Well, the modified.final build of Mythcreator's is a fraction fractionally (...?!) slower than that of Bobor's modified 798. On a test I just did its was 99.998 percent (rounded up!) the performance of Bobor's modified build.

I'm just curious, if Mythcreator is willing to give it a try, purely as an experimental build whether he could try making a build with GCC 4.4.0. It is very much in testing stage, but snapshots can be downloaded by mirror sites. They're updated weekly according to the gcc site. One such site is:

http://gcc-ca.internet.bs/snapshots/

(of course then select the latest 4.4 folder at the bottom of the list)

Course could always build your own latest 4.4.0 snapshot from the GCC SVN
svn://gcc.gnu.org/svn/gcc/trunk
http://gcc.gnu.org/svn/gcc/trunk
^^ view contents

I realise you probably know all this already, I listed it in case anyone was wondering

GCC 4.4.0 supposedly has many optimisations that may finally allow it to be faster than GCC 3.4.6. Would be interesting to see anyway!