[project @ 2005-05-05 18:14:27 by sof]

ocResolve_PEi386():
    when fixing up REL32 relocations, _add_ displacement to value at the
    given offset. The existing value has so far been assumed to be zero
    (which we've asserted for), but curiously wxhaskell-0.9.4's wx.o
    contains lots of interesting non-zero values. Information / specifications
    are awfully thin on the ground as to how to precisely handle these
    relocations, but adding rather than overwriting seems to have a
    generally healthy effect; unable to crash wxhaskell-0.9.4 with a 6.4 build.

Merge to STABLE.

diff --git a/ghc/rts/Linker.c b/ghc/rts/Linker.c
index 5d887f4..11e111a 100644 (file)
--- a/ghc/rts/Linker.c
+++ b/ghc/rts/Linker.c
@@ -2222,9 +2222,15 @@ ocResolve_PEi386 ( ObjectCode* oc )
                   -- hence the constant 4.
                   Also I don't know if A should be added, but so
                   far it has always been zero.
+
+                 SOF 05/2005: 'A' (old contents of *pP) have been observed
+                 to contain values other than zero (the 'wx' object file
+                 that came with wxhaskell-0.9.4; dunno how it was compiled..).
+                 So, add displacement to old value instead of asserting
+                 A to be zero. Fixes wxhaskell-related crashes, and no other
+                 ill effects have been observed.
               */
-               ASSERT(A==0);
-               *pP = S - ((UInt32)pP) - 4;
+               *pP = S - ((UInt32)pP) - 4 + A;
                break;
             default:
                debugBelch("%s: unhandled PEi386 relocation type %d",