Repository : ssh://darcs.haskell.org//srv/darcs/ghc On branch : master
http://hackage.haskell.org/trac/ghc/changeset/2334112f46a02a15b4e7594462ffa476a300167b >--------------------------------------------------------------- commit 2334112f46a02a15b4e7594462ffa476a300167b Author: PHO <p...@cielonegro.org> Date: Sat Nov 24 20:07:27 2012 +0900 Fix broken -fPIC on Darwin/PPC (#7442) The workaround described in note [darwin-x86-pic] applies to Darwin/PPC too. >--------------------------------------------------------------- compiler/cmm/CmmPipeline.hs | 16 ++++++++++++---- 1 files changed, 12 insertions(+), 4 deletions(-) diff --git a/compiler/cmm/CmmPipeline.hs b/compiler/cmm/CmmPipeline.hs index 4e9a90a..159a31d 100644 --- a/compiler/cmm/CmmPipeline.hs +++ b/compiler/cmm/CmmPipeline.hs @@ -183,10 +183,12 @@ cpsTop hsc_env proc = -- the entry point. splitting_proc_points = hscTarget dflags /= HscAsm || not (tablesNextToCode dflags) - || usingDarwinX86Pic -- Note [darwin-x86-pic] - usingDarwinX86Pic = platformArch platform == ArchX86 - && platformOS platform == OSDarwin - && gopt Opt_PIC dflags + || usingInconsistentPicReg -- Note [darwin-x86-pic] + usingInconsistentPicReg = ( platformArch platform == ArchX86 || + platformArch platform == ArchPPC + ) + && platformOS platform == OSDarwin + && gopt Opt_PIC dflags {- Note [darwin-x86-pic] @@ -205,6 +207,12 @@ points, then at the join point we don't have a consistent value for Hence, on x86/Darwin, we have to split proc points, and then each proc point will get its own PIC initialisation sequence. +The situation is the same for ppc/Darwin. We use essentially the same +sequence to load the program counter onto reg: + + bcl 20,31,1f + 1: mflr reg + This isn't an issue on x86/ELF, where the sequence is call 1f _______________________________________________ Cvs-ghc mailing list Cvs-ghc@haskell.org http://www.haskell.org/mailman/listinfo/cvs-ghc
