app-arch/lbzip2: Backport fix from upstream git for Assertion failure, #436382

(Portage version: 2.2.0_alpha151/cvs/Linux x86_64, signed Manifest commit with key 8009D6F070EB7916)

4 files changed, 161 insertions, 10 deletions

diff --git a/app-arch/lbzip2/ChangeLog b/app-arch/lbzip2/ChangeLog
index 671a03b6090a..701029510612 100644
--- a/app-arch/lbzip2/ChangeLog
+++ b/app-arch/lbzip2/ChangeLog
@@ -1,6 +1,12 @@
 # ChangeLog for app-arch/lbzip2
-# Copyright 1999-2012 Gentoo Foundation; Distributed under the GPL v2
-# $Header: /var/cvsroot/gentoo-x86/app-arch/lbzip2/ChangeLog,v 1.40 2012/10/07 11:55:08 jlec Exp $
+# Copyright 1999-2013 Gentoo Foundation; Distributed under the GPL v2
+# $Header: /var/cvsroot/gentoo-x86/app-arch/lbzip2/ChangeLog,v 1.41 2013/01/12 19:46:09 jlec Exp $
+
+*lbzip2-2.2-r1 (12 Jan 2013)
+
+  12 Jan 2013; Justin Lecher <jlec@gentoo.org> +lbzip2-2.2-r1.ebuild,
+  +files/lbzip2-2.2-assertion.patch, metadata.xml:
+  Backport fix from upstream git for Assertion failure, #436382
 
   07 Oct 2012; Justin Lecher <jlec@gentoo.org> -lbzip2-0.23-r2.ebuild,
   -lbzip2-2.1.ebuild, -lbzip2-2.1-r2.ebuild, lbzip2-2.2.ebuild:
diff --git a/app-arch/lbzip2/files/lbzip2-2.2-assertion.patch b/app-arch/lbzip2/files/lbzip2-2.2-assertion.patch
new file mode 100644
index 000000000000..b4946175b947
--- /dev/null
+++ b/app-arch/lbzip2/files/lbzip2-2.2-assertion.patch
@@ -0,0 +1,107 @@
+From 57eeee36927f8a40ece1ca06c674e0bd56d0f21f Mon Sep 17 00:00:00 2001
+Message-Id: <57eeee36927f8a40ece1ca06c674e0bd56d0f21f.1358019732.git.jlec@gentoo.org>
+From: Mikolaj Izdebski <zurgunt@gmail.com>
+Date: Sat, 20 Oct 2012 18:37:17 +0200
+Subject: [PATCH] Fix assertion failure, closes #8
+
+src/encode.c (generate_initial_trees): Rewrite from scratch.
+---
+ src/encode.c | 61 +++++++++++++++++++++++++++++++++++++++++++++++-------------
+ 1 file changed, 48 insertions(+), 13 deletions(-)
+
+diff --git a/src/encode.c b/src/encode.c
+index 09cfacc..00a78dc 100644
+--- a/src/encode.c
++++ b/src/encode.c
+@@ -763,40 +763,75 @@ assign_codes(uint32_t C[], uint32_t L[], uint8_t B[], uint32_t n)
+ 
+ /* Create initial mapping of symbols to trees.
+ 
+-   The goal is to divide all as symbols [0,as) into nt equivalence classes
++   The goal is to divide all as symbols [0,as) into nt equivalence classes (EC)
+    [0,nt) such that standard deviation of symbol frequencies in classes is
+    minimal. We use a kind of a heuristic to achieve that. There might exist a
+    better way to achieve that, but this one seems to be good (and fast) enough.
+ 
+    If the symbol v belongs to the equivalence class t then set s->length[t][v]
+    to zero. Otherwise set it to 1.
+-
+-   TODO: This piece of code really needs some R&D...
+ */
+ static void
+ generate_initial_trees(struct encoder_state *s, unsigned nm, unsigned nt)
+ {
+-  unsigned a, b, c, t;
+-
+-  /* Equivalence classes are empty. */
++  unsigned a, b;   /* range [a,b) of symbols forming current EC */
++  unsigned freq;   /* symbol frequency */
++  unsigned cum;    /* cumulative frequency */
++  unsigned as;     /* effective alphabet size (alphabet size minus number
++                      of symbols with frequency equal to zero) */
++  unsigned t;      /* current tree */
++
++  /* Equivalence classes are initially empty. */
+   memset(s->length, 1, sizeof(s->length));
+ 
++  /* Determine effective alphabet size. */
++  as = 0;
++  for (a = 0, cum = 0; cum < nm; a++) {
++    freq = s->lookup[0][a];
++    cum += freq;
++    as += min(freq, 1);
++  }
++  assert(cum == nm);
++
++  /* Bound number of EC by number of symbols. Each EC is non-empty, so number
++     of symbol EC must be <= number of symbols. */
++  nt = min(nt, as);
++
+   /* For each equivalence class: */
+-  for (a = 0, t = 0; t < nt; t++) {
++  a = 0;
++  for (t = 0; nt > 0; t++, nt--) {
++    assert(nm > 0);
++    assert(as >= nt);
++
+     /* Find a range of symbols which total count is roughly proportional to one
+        nt-th of all values. */
+-    for (c = 0, b = a; c * (nt-t) < nm; b++)
+-      c += s->lookup[0][b];
+-    assert(a < b);
+-    if (a < b-1 && (2*c - s->lookup[0][b-1]) * (nt-t) > 2*nm) {
+-      c -= s->lookup[0][--b];
++    freq = s->lookup[0][a];
++    cum = freq;
++    as -= min(freq, 1);
++    b = a+1;
++    while (as > nt-1 && cum * nt < nm) {
++      freq = s->lookup[0][b];
++      cum += freq;
++      as -= min(freq, 1);
++      b++;
+     }
+-    nm -= c;
++    if (cum > freq && (2*cum - freq) * nt > 2*nm) {
++      cum -= freq;
++      as += min(freq, 1);
++      b--;
++    }
++    assert(a < b);
++    assert(cum > 0);
++    assert(cum <= nm);
++    assert(as >= nt-1);
++    Trace(("Tree %u: EC=[%3u,%3u), |EC|=%3u, cum=%6u", t, a, b, b-a, cum));
+ 
+     /* Now [a,b) is our range -- assign it to equivalence class t. */
+     bzero(&s->length[t][a], b - a);
+     a = b;
++    nm -= cum;
+   }
++  assert(as == 0);
+   assert(nm == 0);
+ }
+ 
+-- 
+1.8.1
+
diff --git a/app-arch/lbzip2/lbzip2-2.2-r1.ebuild b/app-arch/lbzip2/lbzip2-2.2-r1.ebuild
new file mode 100644
index 000000000000..766fa4061e69
--- /dev/null
+++ b/app-arch/lbzip2/lbzip2-2.2-r1.ebuild
@@ -0,0 +1,38 @@
+# Copyright 1999-2013 Gentoo Foundation
+# Distributed under the terms of the GNU General Public License v2
+# $Header: /var/cvsroot/gentoo-x86/app-arch/lbzip2/lbzip2-2.2-r1.ebuild,v 1.1 2013/01/12 19:46:09 jlec Exp $
+
+EAPI=4
+
+inherit autotools-utils
+
+DESCRIPTION="Parallel bzip2 utility"
+HOMEPAGE="https://github.com/kjn/lbzip2/"
+SRC_URI="mirror://github/kjn/${PN}/${P}.tar.gz"
+
+LICENSE="GPL-3"
+SLOT="0"
+KEYWORDS="~alpha ~amd64 ~arm ~hppa ~ia64 ~m68k ~mips ~ppc ~ppc64 ~s390 ~sh ~sparc ~x86 ~x86-fbsd ~amd64-linux ~x86-linux"
+IUSE="debug symlink"
+
+PATCHES=(
+	"${FILESDIR}"/${P}-s_isreg.patch
+	"${FILESDIR}"/${P}-assertion.patch
+	)
+
+RDEPEND="symlink? ( !app-arch/pbzip2[symlink] )"
+DEPEND=""
+
+src_configure() {
+	local myeconfargs=(
+		--disable-silent-rules
+		$(use_enable debug tracing)
+	)
+	autotools-utils_src_configure
+}
+
+src_install() {
+	autotools-utils_src_install
+
+	use symlink && dosym ${PN} /usr/bin/bzip2
+}
diff --git a/app-arch/lbzip2/metadata.xml b/app-arch/lbzip2/metadata.xml
index bf9b027e4bde..c53572127eee 100644
--- a/app-arch/lbzip2/metadata.xml
+++ b/app-arch/lbzip2/metadata.xml
@@ -1,14 +1,14 @@
 <?xml version="1.0" encoding="UTF-8"?>
 <!DOCTYPE pkgmetadata SYSTEM "http://www.gentoo.org/dtd/metadata.dtd">
 <pkgmetadata>
-	<maintainer>
-		<email>jlec@gentoo.org</email>
-	</maintainer>
-	<maintainer>
-		<email>mattst88@gentoo.org</email>
-		<name>Matt Turner</name>
-	</maintainer>
-	<longdescription lang="en">
+  <maintainer>
+    <email>jlec@gentoo.org</email>
+  </maintainer>
+  <maintainer>
+    <email>mattst88@gentoo.org</email>
+    <name>Matt Turner</name>
+  </maintainer>
+  <longdescription lang="en">
 	A multi-threaded bzip2/bunzip2 utility that employs multiple threads and an 
 	input-bound splitter even when decompressing .bz2 files created by standard 
 	bzip2