Source-Changes-HG archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

[src/trunk]: src/usr.bin/make/unit-tests make(1): add test for intentional ha...



details:   https://anonhg.NetBSD.org/src/rev/59d6407f96ae
branches:  trunk
changeset: 954500:59d6407f96ae
user:      rillig <rillig%NetBSD.org@localhost>
date:      Sat Sep 05 12:59:07 2020 +0000

description:
make(1): add test for intentional hash collision for variable names

Hash collisions may slow down make in certain special situations.  There
is no point though in maliciously triggering such a situation since
anyone who can inject values into makefiles can easily run shell
commands using the :!cmd! modifier or similar mechanisms.  Crafting
variable names just to slow down make is thus not an attack vector.

diffstat:

 usr.bin/make/unit-tests/Makefile   |   3 +-
 usr.bin/make/unit-tests/varname.mk |  41 +++++++++++++++++++++++++++++++++++--
 2 files changed, 40 insertions(+), 4 deletions(-)

diffs (67 lines):

diff -r 2d837253f5d4 -r 59d6407f96ae usr.bin/make/unit-tests/Makefile
--- a/usr.bin/make/unit-tests/Makefile  Sat Sep 05 12:19:25 2020 +0000
+++ b/usr.bin/make/unit-tests/Makefile  Sat Sep 05 12:59:07 2020 +0000
@@ -1,4 +1,4 @@
-# $NetBSD: Makefile,v 1.135 2020/09/05 06:36:40 rillig Exp $
+# $NetBSD: Makefile,v 1.136 2020/09/05 12:59:07 rillig Exp $
 #
 # Unit tests for make(1)
 #
@@ -393,6 +393,7 @@
                        ${TOOL_SED} -n -e '/^\#\*\*\* Suffixes/,/^\#\*/p'
 POSTPROC.vardebug=     ${TOOL_SED} -n -e '/:RELEVANT = yes/,/:RELEVANT = no/p'
 POSTPROC.varmod-match-escape= ${TOOL_SED} -n -e '/^Pattern/p'
+POSTPROC.varname=      ${TOOL_SED} -n -e '/^MAGIC/p' -e '/^ORDER_/p'
 POSTPROC.varname-dot-shell= \
                        awk '/\.SHELL/ || /^ParseReadLine/'
 POSTPROC.varname-empty=        ${TOOL_SED} -n -e '/^Var_Set/p' -e '/^out:/p'
diff -r 2d837253f5d4 -r 59d6407f96ae usr.bin/make/unit-tests/varname.mk
--- a/usr.bin/make/unit-tests/varname.mk        Sat Sep 05 12:19:25 2020 +0000
+++ b/usr.bin/make/unit-tests/varname.mk        Sat Sep 05 12:59:07 2020 +0000
@@ -1,8 +1,43 @@
-# $NetBSD: varname.mk,v 1.2 2020/08/16 14:25:16 rillig Exp $
+# $NetBSD: varname.mk,v 1.3 2020/09/05 12:59:07 rillig Exp $
 #
 # Tests for special variables, such as .MAKE or .PARSEDIR.
 
-# TODO: Implementation
+# These following MAGIC variables have the same hash code, at least with
+# the default hashing algorithm, which is the same as in Java.  The order
+# in which these variables are defined determines the order in which they
+# appear in the Hash_Table.  New entries are prepended to the bucket lists,
+# therefore this test numbers the values in descending order.
+
+.if defined(ORDER_01)
+
+MAGIC0a0a0a=   8
+MAGIC0a0a1B=   7
+MAGIC0a1B0a=   6
+MAGIC0a1B1B=   5
+MAGIC1B0a0a=   4
+MAGIC1B0a1B=   3
+MAGIC1B1B0a=   2
+MAGIC1B1B1B=   1
+
+all: # nothing
+
+.elif defined(ORDER_10)
+
+MAGIC1B1B1B=   8
+MAGIC1B1B0a=   7
+MAGIC1B0a1B=   6
+MAGIC1B0a0a=   5
+MAGIC0a1B1B=   4
+MAGIC0a1B0a=   3
+MAGIC0a0a1B=   2
+MAGIC0a0a0a=   1
+
+all: # nothing
+
+.else
 
 all:
-       @:;
+       @${.MAKE} -f ${MAKEFILE} -dg1 ORDER_01=yes
+       @${.MAKE} -f ${MAKEFILE} -dg1 ORDER_10=yes
+
+.endif



Home | Main Index | Thread Index | Old Index