File apparmor-parser-fix-two-x-transition-conflict-bugs of Package apparmor-parser

Overview Repositories Revisions Requests Users Attributes Meta

File apparmor-parser-fix-two-x-transition-conflict-bugs of Package apparmor-parser (Revision 3454aa4f0129b3e512160d4586969a10)

Currently displaying revision 3454aa4f0129b3e512160d4586969a10 , Show latest

From: John Johansen <john.johansen@canonical.com>
Subject: apparmor-parser: Fix two x transition conflict bugs.
References: bnc#662928 lpn#lp693082

The is_merged_x_consistend macro was incorrect in that is tested for
USER_EXEC_TYPE to determine if there was an x transition.  This fails
for unconfined execs so an unconfined exec would not correctly conflict
with another exec type.

The dfa match flag table for xtransitions was not large enough and not
indexed properly for pux, and cux transitions.  The index calculation did
not take into account the pux flag so that pux and px aliased to the same
location and cux and cx aliased to the same location.

This would result in the first rule being processed defining what the
transition type was for all following rules of the type following.  So
if a px transition was processed first all pux, transitions in the profile
would be treated pux.

Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Jeff Mahoney <jeffm@suse.com>
---
 immunix.h                           |    4 
 libapparmor_re/regexp.y             |    4 
 tst/Makefile                        |   13 ++-
 tst/gen-xtrans.pl                   |  152 ++++++++++++++++++++++++++++++++++++
 tst/simple_tests/generated_x/readme |    2 
 5 files changed, 169 insertions(+), 6 deletions(-)

--- a/immunix.h
+++ b/immunix.h
@@ -148,12 +148,12 @@ enum pattern_t {
 #include <stdio.h>
 static inline int is_merged_x_consistent(int a, int b)
 {
-	if ((a & AA_USER_EXEC_TYPE) && (b & AA_USER_EXEC_TYPE) &&
+	if ((a & AA_USER_EXEC) && (b & AA_USER_EXEC) &&
 	    ((a & AA_USER_EXEC_TYPE) != (b & AA_USER_EXEC_TYPE)))
 { fprintf(stderr, "failed user merge 0x%x 0x%x\n", a, b);
 		return 0;
 }
-	if ((a & AA_OTHER_EXEC_TYPE) && (b & AA_OTHER_EXEC_TYPE) &&
+	if ((a & AA_OTHER_EXEC) && (b & AA_OTHER_EXEC) &&
 	    ((a & AA_OTHER_EXEC_TYPE) != (b & AA_OTHER_EXEC_TYPE)))
 { fprintf(stderr, "failed other merge 0x%x 0x%x\n", a, b);
 		return 0;
--- a/libapparmor_re/regexp.y
+++ b/libapparmor_re/regexp.y
@@ -2098,8 +2098,8 @@ extern "C" int aare_add_rule_vec(aare_ru
 	flip_tree(tree);
 
 
-/* 0x3f == 4 bits x mods + 1 bit unsafe mask + 1 bit ix, after shift */
-#define EXTRACT_X_INDEX(perm, shift) (((perm) >> (shift + 8)) & 0x3f)
+/* 0x7f == 4 bits x mods + 1 bit unsafe mask + 1 bit ix, + 1 pux after shift */
+#define EXTRACT_X_INDEX(perm, shift) (((perm) >> (shift + 7)) & 0x7f)
 
 //if (perms & ALL_AA_EXEC_TYPE && (!perms & AA_EXEC_BITS))
 //	fprintf(stderr, "adding X rule without MAY_EXEC: 0x%x %s\n", perms, rulev[0]);
--- a/tst/Makefile
+++ b/tst/Makefile
@@ -10,9 +10,18 @@ endif
 
 all: tests
 
-.PHONY: tests
-tests: ../apparmor_parser
+.PHONY: tests gen_xtrans
+
+tests:  gen_xtrans parser_sanity
+
+parser_sanity: ../apparmor_parser
 	$(Q)${PROVE} ${PROVE_ARG} ${TESTS}
 
+gen_xtrans:
+	perl ./gen-xtrans.pl
+
 ../apparmor_parser:
 	make -C .. apparmor_parser
+
+clean:
+	rm -f simple_tests/generated_x/*
--- /dev/null
+++ b/tst/gen-xtrans.pl
@@ -0,0 +1,154 @@
+#!/usr/bin/perl
+
+use strict;
+use Locale::gettext;
+use POSIX;
+
+setlocale(LC_MESSAGES, "");
+
+my $prefix="simple_tests/";
+
+my @trans_types = ("p", "P", "c", "C", "u", "i");
+#my @modifiers = ("i", "u");
+# Apparmor 2.3 has no qualified 'u' mode.
+my @modifiers = ("i");
+my %trans_modifiers = (
+    "p" => \@modifiers,
+    "P" => \@modifiers,
+    "c" => \@modifiers,
+    "C" => \@modifiers,
+    );
+
+my @targets = ("", "target", "target2");
+my @null_target = ("");
+
+my %named_trans = (
+    "p" => \@targets,
+    "P" => \@targets,
+    "c" => \@targets,
+    "C" => \@targets,
+    "u" => \@null_target,
+    "i" => \@null_target,
+    );
+
+# audit qualifier disabled for now it really shouldn't affect the conflict
+# test but it may be worth checking every once in awhile
+#my @qualifiers = ("", "owner", "audit", "audit owner");
+my @qualifiers = ("", "owner");
+
+my $count = 0;
+
+gen_conflicting_x();
+gen_overlap_re_exact();
+gen_dominate_re_re();
+gen_ambiguous_re_re();
+
+print "Generated $count xtransition interaction tests\n";
+
+sub gen_list {
+    my @output;
+    foreach my $trans (@trans_types) {
+	if ($trans_modifiers{$trans}) {
+	    foreach my $mod (@{$trans_modifiers{$trans}}) {
+		push @output, "${trans}${mod}x";
+	    }
+	}
+	push @output, "${trans}x";
+    }
+    return @output;
+}
+
+sub print_rule($$$$) {
+    my ($file, $name, $perm, $target) = @_;
+    print $file "\t${name} ${perm}";
+    if ($target ne "") {
+	print $file " -> $target";
+    }
+    print $file ",\n";
+}
+
+sub gen_file($$$$$$$$) {
+    my ($name, $xres, $rule1, $perm1, $target1, $rule2, $perm2, $target2) = @_;
+
+#    print "$xres $rule1 $perm1 $target1 $rule2 $perm2 $target2\n";
+
+    my $file;
+    unless (open $file, ">$name") {
+	print("couldn't open $name\n");
+	exit 1;
+    }
+
+    print $file "#\n";
+    print $file "#=DESCRIPTION ${name}\n";
+    print $file "#=EXRESULT ${xres}\n";
+    print $file "#\n";
+    print $file "/usr/bin/foo {\n";
+    print_rule($file, $rule1, $perm1, $target1);
+    print_rule($file, $rule2, $perm2, $target2);
+    print $file "}";
+    close($file);
+
+    $count++;
+}
+
+#NOTE: currently we don't do px to cx, or cx to px conversion
+#      so
+# /foo {
+#    /* px -> /foo//bar,
+#    /* cx -> bar,
+#
+# will conflict
+#
+#NOTE: conflict tests don't tests leading permissions or using unsafe keywords
+#      It is assumed that there are extra tests to verify 1 to 1 coorispondance
+sub gen_files($$$$) {
+    my ($name, $rule1, $rule2, $default) = @_;
+
+    my @perms = gen_list();
+
+#    print "@perms\n";
+
+    foreach my $i (@perms) {
+	foreach my $t (@{$named_trans{substr($i, 0, 1)}}) {
+	    foreach my $q (@qualifiers) {
+		foreach my $j (@perms) {
+		    foreach my $u (@{$named_trans{substr($j, 0, 1)}}) {
+			foreach my $r (@qualifiers) {
+			    my $file="${prefix}/${name}-$q$i$t-$r$j$u.sd";
+#		    print "$file\n";
+
+		    #override failures when transitions are the same
+			    my $xres = ${default};
+			    if ($i eq $j && $t eq $u) {
+				$xres = "PASS";
+			    }
+
+
+#		    print "foo $xres $rule1 $i $t $rule2 $j $u\n";
+			    gen_file($file, $xres, "$q $rule1", $i, $t, "$r $rule2", $j, $u);
+			}
+		    }
+		}
+	    }
+	}
+    }
+
+}
+
+sub gen_conflicting_x {
+    gen_files("conflict", "/bin/cat", "/bin/cat", "FAIL");
+}
+
+sub gen_overlap_re_exact {
+
+    gen_files("exact", "/bin/cat", "/bin/*", "PASS");
+}
+
+# we currently don't support this, once supported change to "PASS"
+sub gen_dominate_re_re {
+    gen_files("dominate", "/bin/*", "/bin/**", "FAIL");
+}
+
+sub gen_ambiguous_re_re {
+    gen_files("ambiguous", "/bin/a*", "/bin/*b", "FAIL");
+}
--- /dev/null
+++ b/tst/simple_tests/generated_x/readme
@@ -0,0 +1,2 @@
+Directory for auto generated x-transition tests
+

Places

File apparmor-parser-fix-two-x-transition-conflict-bugs of Package apparmor-parser (Revision 3454aa4f0129b3e512160d4586969a10)

Places