0.9.9.32:

	Fix hideously embarrassing ppc assembly bug in reg_LRA
	computation.
	... no longer go wrong if bit 15 of lra is set.

	(The symptoms from this have been reported many, many times:
	segmentation faults in the first triggered GC.  Kevin Rosenberg
	reported it first from my trawl on sbcl-devel, but I think it's
	been known for longer than that.  Previously it had been
	dismissed as gcc miscompilation problems, because the problem
	disappeared when using a different version of gcc, for any
	individual developer: in retrospect, the fact that it was our
	bug after all is pretty obvious from the fact that we were never
	able to characterize particular versions of gcc which were bad.)

diff --git a/NEWS b/NEWS
index 3b59dcb..0e50cc0 100644 (file)
--- a/NEWS
+++ b/NEWS
@@ -24,6 +24,12 @@ changes in sbcl-0.9.10 relative to sbcl-0.9.9:
   * fixed bug: metacircle resolution in cases where methods have slots
     added before the slots from STANDARD-METHOD.  (reported by Jean
     Bresson)
+  * fixed bug: the Power PC assembly code for calling into lisp
+    sometimes computed its return address wrongly (depending on the
+    15th bit of the address, fixed at link time).  This used to
+    manifest itself in a segmentation violation while building PCL.
+    (reported by Kevin Rosenberg, Eric Marsden, Lars Brinkhoff and
+    many others over the years)
 
 changes in sbcl-0.9.9 relative to sbcl-0.9.8:
   * new platform: experimental support for the Windows operating

diff --git a/src/runtime/ppc-assem.S b/src/runtime/ppc-assem.S
index 28fe473..b4c79af 100644 (file)
--- a/src/runtime/ppc-assem.S
+++ b/src/runtime/ppc-assem.S
@@ -339,7 +339,7 @@ x:
        addi reg_LRA,reg_LRA,lo16(lra)
 #else
        lis reg_LRA,lra@h
-       addi reg_LRA,reg_LRA,lra@l
+       ori reg_LRA,reg_LRA,lra@l
 #endif
        addi reg_LRA,reg_LRA,OTHER_POINTER_LOWTAG
 

diff --git a/version.lisp-expr b/version.lisp-expr
index d8c689e..29d8be7 100644 (file)
--- a/version.lisp-expr
+++ b/version.lisp-expr
@@ -17,4 +17,4 @@
 ;;; checkins which aren't released. (And occasionally for internal
 ;;; versions, especially for internal versions off the main CVS
 ;;; branch, it gets hairier, e.g. "0.pre7.14.flaky4.13".)
-"0.9.9.31"
+"0.9.9.32"