Pullup ticket #3835 - requested by dholland

devel/electric-fence: build fix

Revisions pulled up:
- devel/electric-fence/Makefile                                 1.4
- devel/electric-fence/distinfo                                 1.2
- devel/electric-fence/patches/patch-efence_c                   1.1

---
   Module Name:	pkgsrc
   Committed By:	dholland
   Date:		Sat Jun 16 07:46:55 UTC 2012

   Modified Files:
   	pkgsrc/devel/electric-fence: Makefile distinfo
   Added Files:
   	pkgsrc/devel/electric-fence/patches: patch-efence_c

   Log Message:
   Add gcc memory barriers after the manner of emacs20's patch-bm, for
   the same reason: gcc "knows" that malloc has no side effects and
   reorders code around it, only it's wrong. Fixes SIGSEGV during build
   seen in some environments.

   Bump package revision as a precaution, because I don't understand why
   this sometimes doesn't fail and sometimes does with the same gcc
   version.

3 files changed, 50 insertions, 2 deletions

diff --git a/devel/electric-fence/Makefile b/devel/electric-fence/Makefile
index 440642ac632..55c89543e86 100644
--- a/devel/electric-fence/Makefile
+++ b/devel/electric-fence/Makefile
@@ -1,7 +1,8 @@
-# $NetBSD: Makefile,v 1.3 2010/11/13 21:08:56 shattered Exp $
+# $NetBSD: Makefile,v 1.3.12.1 2012/06/18 22:18:35 tron Exp $
 
 DISTNAME=		electric-fence_2.1.13-0.1
 PKGNAME=		electric-fence-2.1.13.0.1
+PKGREVISION=		1
 CATEGORIES=		devel
 MASTER_SITES=		http://perens.com/works/software/ElectricFence/
 
diff --git a/devel/electric-fence/distinfo b/devel/electric-fence/distinfo
index ea952a1a9fe..d1a6470b700 100644
--- a/devel/electric-fence/distinfo
+++ b/devel/electric-fence/distinfo
@@ -1,4 +1,4 @@
-$NetBSD: distinfo,v 1.1.1.1 2010/02/14 00:06:16 reed Exp $
+$NetBSD: distinfo,v 1.1.1.1.18.1 2012/06/18 22:18:35 tron Exp $
 
 SHA1 (electric-fence_2.1.13-0.1.tar.gz) = e6765bcb1543272040b806eea706fc7ae9b60524
 RMD160 (electric-fence_2.1.13-0.1.tar.gz) = 75e41de7bef263007f24a1053528959f9f7fe1fa
@@ -6,3 +6,4 @@ Size (electric-fence_2.1.13-0.1.tar.gz) = 29991 bytes
 SHA1 (patch-aa) = 9b2f1720a92c805d8d4b36631317493124c2f2d7
 SHA1 (patch-ab) = 64c503ce8d2ea9b55432d59a187e644af38e257b
 SHA1 (patch-ac) = 4c56eeea0b8e53cdf260b61aa8fdef16f601b186
+SHA1 (patch-efence_c) = 2d97cf8a643b2b1c63c0a039267f4cf63b1a854a
diff --git a/devel/electric-fence/patches/patch-efence_c b/devel/electric-fence/patches/patch-efence_c
new file mode 100644
index 00000000000..911c26ed24f
--- /dev/null
+++ b/devel/electric-fence/patches/patch-efence_c
@@ -0,0 +1,46 @@
+$NetBSD: patch-efence_c,v 1.1.2.2 2012/06/18 22:18:35 tron Exp $
+
+Add memory barriers for gcc around recursive calls to malloc and free
+with "internalUse" set to 1. gcc fervently believes that malloc and
+free do not interact with anything else, and so it freely reorders code
+around calls to malloc. This causes internalUse to not be set to 1
+during the recursive call, which results in an infinite recursion.
+
+Compare patch-bm in editors/emacs20.
+
+--- efence.c~	2002-02-19 22:10:46.000000000 +0000
++++ efence.c
+@@ -377,7 +377,16 @@ allocateMoreSlots(void)
+ 	noAllocationListProtection = 1;
+ 	internalUse = 1;
+ 
++#ifdef __GNUC__
++	__asm __volatile("":::"memory");
++#endif
++
+ 	newAllocation = malloc(newSize);
++
++#ifdef __GNUC__
++	__asm __volatile("":::"memory");
++#endif
++
+ 	memcpy(newAllocation, allocationList, allocationListSize);
+ 	memset(&(((char *)newAllocation)[allocationListSize]), 0, bytesPerPage);
+ 
+@@ -386,8 +395,16 @@ allocateMoreSlots(void)
+ 	slotCount += slotsPerPage;
+ 	unUsedSlots += slotsPerPage;
+ 
++#ifdef __GNUC__
++	__asm __volatile("":::"memory");
++#endif
++
+ 	free(oldAllocation);
+ 
++#ifdef __GNUC__
++	__asm __volatile("":::"memory");
++#endif
++
+ 	/*
+ 	 * Keep access to the allocation list open at this point, because
+ 	 * I am returning to memalign(), which needs that access.