On Fri, 31 Jan 2020 19:33:52 +0100 Charlene Wendling <juliana...@posteo.jp> wrote:
> I've finished the build with your diff on my PowerBook G4 A1138, where > it took 29 hours, and so i tested: *Thank you!* I killed my macppc build because 1 C++ file (Parser.cpp) got stuck for hours trying to use over 800M swap on my iMac G3's internal hard disk. (It has 512M RAM.) I have now restarted the build after disabling swapencrypt and switching to swap on NFS (trying to borrow RAM from my amd64 desktop), but I might still be days away from completing the build. My now-broken PowerBook G4 (PowerBook5,4; A1095?) took about 28 hours to build clang 7.0.1, but that was with 1G RAM, and running its MPC7447A cpu at apm -L half speed (666 MHz, the default setting); I didn't like the loud fan at full speed (1333 MHz). Your RAM amount and cpu speed (if your PowerBook has dfs(4)) may affect your build time. > $ DISPLAY=:0 ./xcbtest > conn = 0xcb5ee000 > cookie.sequence = 1 > reply = 0xaccbbd20 This looks good to me. > I had to bring back games/galois to ports-gcc because it was broken at > startup with ports-clang, so i tried to build it with your clang; it's > playable, there are still segfaults when messing with options, and i > forgot to build it with debug infos on. I could if asked to, but this > is already an improvement! If I can finish my ports-clang build, I might try galois on my iMac. (No X server right now, but I might ssh -X from amd64.) --George
