[MPlayer-dev-eng] new Altivec patches and G3

[Alan Curry]

Tim writes the following:
>
>
>On Feb 19, 2006, at 2:56 PM, Alan Curry wrote:
>
>> If you use --enable-altivec in this case, you're lying to configure,
>> preventing it from doing the right thing... crashes are to be  
>> expected. Did
>> any past versions work if configured this way?
>
>Like I said, without --enable-altivec works just fine.  That's why I  
>was confused and did it with.  It doesn't even *try* to detect  
>altivec after the CPU detection realizes it has a 750.  The only  
>other argument I passed to configure was --enable-debug.

Andrea's bug report probably only applies to --enable-runtime-cpudetection

>> -	if [ `sysctl -n hw.vectorunit` -eq 1 -o \
>> +	if [ "`sysctl -n hw.vectorunit 2>/dev/null`" -eq 1 -o \
>>  	    "`sysctl -n hw.optional.altivec 2>/dev/null`" -eq 1 ]; then
>>  	    _altivec=yes
>>  	fi
>
>Nope, still the same error.

OK then, what is the output of
  sysctl -n hw.vectorunit
and
  sysctl -n hw.optional.altivec

Whoever wrote that section of the configure script expected them to be
numbers, and apparently they aren't. I'll bet that at least one of them
generates no output, or just an error message.

If I'm right, the -eq's should just be changed to ='s