[MPlayer-users] mplayer compilation error

[David C. Diemer]

MMX stands for the [Intel] Multi-media extensions built into the Pentuim II 
MMX-233 and higher speed and class processors. It has special instructions 
to help graphics and sound process faster if written to the processor's 
special intructions.

But, hey!, it can't hurt in your case, to remove them. I don't know if the 
various audio and video codecs use those instruction sets but mplayer will 
specifically look for them and use them if you compile them in.

What this means is that you can expect your processor to run at least 5-10% 
more than an Intel with the instructions built-in (if the processor, as you 
suggested, has no MMX set). So your processor runs harder but that 
movie/video should still play.

Dave



----Original Message Follows----
From: "Dean S. Messing" <deanm at sharplabs.com>
Reply-To: mplayer-users at mplayerhq.hu
To: mplayer-users at mplayerhq.hu
Subject: Re: [MPlayer-users] mplayer compilation error
Date: Mon, 25 Nov 2002 14:28:48 -0800 (PST)

[Automatic answer: RTFM (read DOCS, FAQ), also read DOCS/bugreports.html]

  :: From: "David C. Diemer" <daviddiemer at hotmail.com>
  :: To: mplayer-users at mplayerhq.hu
  :: Subject: Re: [MPlayer-users] mplayer compilation error
  :: Date: Mon, 25 Nov 2002 16:29:54 -0500

Thanks David for your help.  I do appreciate it.

David C. Diemer writes:
  :: It's just possible that you may have to include the option
  :: '--disable-shared-pp', that is, disable shared post-processing. It's 
just a
  :: guess from the message about postproc. It's also possible that your 
kernel
  :: dosn't support the MMX registers in which case compile with the
  :: '--disable-mmx --disable-mmx2' switches (1 or the other or both).

What will I lose in functionality if I disable post-processing?
Regarding my kernel's MMX support, recall in my original post these lines:

CLFAGS= ./configure --prefix=/usr --datadir=/usr/share/mplayer
        --confdir=/etc/mplayer --libdir=/usr/lib --enable-largefiles
        --disable-runtime-cpudetection --enable-mmx --enable-mmx2
        --enable-3dnow --enable-3dnowex --enable-sse --enable-sse2
        --enable-fastmemcpy --enable-freetype --disable-sighandler
        --enable-gui --enable-i18n --language=en --enable-linux-devfs
        --disable-mpdvdkit --enable-lirc --enable-tv --enable-tv-v4l
        --enable-joystick --enable-gl --disable-svga --disable-dvb
        --enable-mga --enable-xmga --enable-fbdev --disable-directfb
        --disable-dvdnav --enable-mencoder --enable-shared-pp
        --with-extraincdir=./libvo --with-x11incdir=/usr/X11R6/include
        --with-xanimlibdir=/usr/X11R6/lib/xanim/mods
        --with-reallibdir=/usr/lib/RealPlayer8/Codecs
        --with-win32libdir=/usr/lib/win32
Detected operating system: Linux
Detected host architecture: i386
Checking for gcc version ... 3.2, ok
Checking for CPU vendor ... AuthenticAMD (6:6:2)
Checking for CPU type ...  AMD Athlon(tm) MP 1500+
Checking for GCC & CPU optimization abilities ... athlon-xp
Checking for kernel support of mmx ... yes
Checking for kernel support of mmx2 ... yes
Checking for kernel support of 3dnow ... yes
Checking for kernel support of 3dnowex ... yes
Checking for kernel support of sse ... yes
Checking for mtrr support ... yes
Checking for assembler (as 2.12.90.0.15) ... ok
<snip>

I assume that this indicates I have the needed support.  Or is there
something else.  Please note that this is an Athlon system, so
I am using the "mem=nopentium" boot option. I don't really
understand exactly what this does (except prevent system lock-ups)
but I recall that it has something to do with MMX.

Also, in one of my many many attempts to compile this things I did
disable the MMX/MMX2 and still got the errors.  They may have been in
a different place, but it was the same sort of error.  The place that
causes the error, by the way, is a swath of assembler code embedded in
the C-source code.  I wonder if this is an Athlon issue?

I'll try the --disable-shared-pp idea.

  :: I doubt it's a GCC problem as I've compiled
  :: mplayer under 2.95, 3.2, and 3.4
  :: without problems.

Holy Smokes!! is 3.4 out?  I just looked on Mandrake
Cooker and all they have is 3.2 dated Nov. 18 (which was
the last build date for that .src.rpm.  What happened
to 3.3?  Maybe this was just a typo?

Anyway, thanks.  I'm still waiting to hear from the fellow
who suggested it was gcc 3.2.  Maybe he knows something
I don't.

  ::
  :: Dave

Regards,
Dean

_______________________________________________
RTFM!!!  http://www.MPlayerHQ.hu/DOCS
Search:  http://www.MPlayerHQ.hu/cgi-bin/htsearch
http://mplayerhq.hu/mailman/listinfo/mplayer-users


_________________________________________________________________
The new MSN 8: advanced junk mail protection and 2 months FREE* 
http://join.msn.com/?page=features/junkmail