[commit: ghc] master: Comment to explain why we need to split proc points on x86/Darwin with -fPIC (65fec07)

Repository : ssh://darcs.haskell.org//srv/darcs/ghc

On branch  : master

http://hackage.haskell.org/trac/ghc/changeset/65fec07bfbf41846689e0b28bc94b168c8cacc84

>---------------------------------------------------------------

commit 65fec07bfbf41846689e0b28bc94b168c8cacc84
Author: Simon Marlow <[email protected]>
Date:   Wed Oct 24 12:03:34 2012 +0100

    Comment to explain why we need to split proc points on x86/Darwin with -fPIC

>---------------------------------------------------------------

 compiler/cmm/CmmPipeline.hs |   32 +++++++++++++++++++++++++++++++-
 1 files changed, 31 insertions(+), 1 deletions(-)

diff --git a/compiler/cmm/CmmPipeline.hs b/compiler/cmm/CmmPipeline.hs
index c523b13..aa8fa2c 100644
--- a/compiler/cmm/CmmPipeline.hs
+++ b/compiler/cmm/CmmPipeline.hs
@@ -179,11 +179,41 @@ cpsTop hsc_env proc =
         -- the entry point.
         splitting_proc_points = hscTarget dflags /= HscAsm
                              || not (tablesNextToCode dflags)
-                             || usingDarwinX86Pic
+                             || usingDarwinX86Pic -- Note [darwin-x86-pic]
         usingDarwinX86Pic = platformArch platform == ArchX86
                          && platformOS platform == OSDarwin
                          && gopt Opt_PIC dflags
 
+{- Note [darwin-x86-pic]
+
+On x86/Darwin, PIC is implemented by inserting a sequence like
+
+    call 1f
+ 1: popl %reg
+
+at the proc entry point, and then referring to labels as offsets from
+%reg.  If we don't split proc points, then we could have many entry
+points in a proc that would need this sequence, and each entry point
+would then get a different value for %reg.  If there are any join
+points, then at the join point we don't have a consistent value for
+%reg, so we don't know how to refer to labels.
+
+Hence, on x86/Darwin, we have to split proc points, and then each proc
+point will get its own PIC initialisation sequence.
+
+This isn't an issue on x86/ELF, where the sequence is
+
+    call 1f
+ 1: popl %reg
+    addl $_GLOBAL_OFFSET_TABLE_+(.-1b), %reg
+
+so %reg always has a consistent value: the address of
+_GLOBAL_OFFSET_TABLE_, regardless of which entry point we arrived via.
+
+-}
+
+
+
 runUniqSM :: UniqSM a -> IO a
 runUniqSM m = do
   us <- mkSplitUniqSupply 'u'