[mythtv] request: OSX mythfrontend and winmyth for CVS backend

Joseph A. Caputo jcaputo1 at comcast.net
Wed Feb 2 10:38:14 EST 2005


On Wednesday 02 February 2005 9:07, Joseph A. Caputo wrote:
> On Tuesday 01 February 2005 19:31, Jeremiah Morris wrote:
> > On Feb 1, 2005, at 2:46 PM, Joseph A. Caputo wrote:
> > 
> > > Now the flag I defined earlier has come back to
> > > bite me:
> > 
> > For now, you can comment out the line "#define ACCEL_DETECT" in 
> > libmythmpeg2's config.h, then you won't need that custom flag. I 
hope 
> > to do some configure work soon and make this easier, but I can't 
> > promise anything definite.
> 
> Thanks, I'll try that.

Nope, doesn't seem to work.  I get:

cc -c -pipe -w -O2 -fPIC -fno-common  -D_GNU_SOURCE 
-D_FILE_OFFSET_BITS=64 -DPREFIX=\"/usr/local\" -DQT_NO_DEBUG 
-DQT_THREAD_SUPPORT -DQT_SHARED -DQT_ACCESSIBILITY_SUPPORT 
-DQT_TABLET_SUPPORT -I/usr/local/lib/qt-mac-free-3.3.2/mkspecs/default 
-I. -I/usr/local/include -I/sw/include 
-I/usr/local/lib/qt-mac-free-3.3.2/include -o cpu_state.o cpu_state.c
{standard input}:19:stvx vector instruction is optional for the PowerPC 
(not allowed without -force_cpusubtype_ALL option)

repeated for many lines in that file.  All the other files compile OK.

I tracked down the problem to the following macros defined in 
cpu_state.c:

STVX0
STVX
LVX0
LVX

used in the static functions state_save_altivec() and 
state_restore_altivec().  If I #ifdef out the bodies of those functions 
(leaving the macros defined), the library compiles fine.  I'm just not 
exactly sure what the correct way to proceed is... if there is no 
altivec, is it OK to just noop those functions, and if so, what is the 
appropriate macro to check?

I imagine this will solve my remaining undefined symbols problems; 
hopefully I'll be able to try running mythfrontend soon and see how a 
G3 performs playing MPEG-4 Myth content.

-JAC


More information about the mythtv-dev mailing list