Let me begin with the following: I downloaded the texmem branch from DRI CVS. It fails to compile, giving the error
./config/imake/imake -I./config/cf -s ./config/makedepend/Makefile.proto -f ./config/makedepend/Imakefile -DTOPDIR=../.. -DCURDIR=./config/makedepend ./config/imake/imake: No such file or directory ./config/imake/imake: Cannot exec /lib/cpp. Stop. ./config/imake/imake: Exit code 1. Stop. make[1]: *** [config/makedepend/Makefile.proto] Error 1 make[1]: Leaving directory `/usr/local/src/XFree86.workon/xc/xc' make: *** [World] Error 2 That said, I am using a DRI texmem CVS checkout from Nov 1., 2002. I checked the function I was looking at (RADEONScreenInit) and didn't notice any significant differences from the current version. I had a look at radeon_driver.c. The first thing that drew my attention was the the conditional that was prefixed by /* Xinerama has sync problem with DRI, disable it for now */ Does this imply that everything is set up already and there is some bug that must be ironed out? My problem, however, comes earlier in the file. DRI is disabled for me at the part that says "Static buffer allocation failed ...". For the initialization of each screen, the log file says it needs at least (about) 9 MB of memory. I have only 16 MB onboard, though the AGP aperture goes up to 64 MB (I'm using the texmem branch). Has the dynamic allocation of back and depth buffers been implemented yet? I don't think there is much else I can do if I don't have enough memory (or would it not be too difficult for me to implement this myself?). --Jonathan Thambidurai ------------------------------------------------------- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf _______________________________________________ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
