/[path-sandbox]/trunk/ChangeLog
Gentoo

Contents of /trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 33 - (hide annotations) (download)
Wed Mar 9 07:34:30 2005 UTC (9 years, 4 months ago) by ferringb
File size: 14758 byte(s)
correct symbols.h building when libdir !+ "/lib"

1 ferringb 2 # ChangeLog for Path Sandbox
2     # Copyright 1999-2004 Gentoo Foundation; Distributed under the GPL v2
3     # $Header$
4    
5 ferringb 33 09 Mar 2005; Brian Harring <ferringb@gentoo.org> Makefile.am, configure.in:
6     Correct libc_version path detection, since it was screwing up if libdir !=
7     "/lib/".
8    
9     02 Mar 2005; Martin Schlemmer <azarah@gentoo.org> libsandbox.c:
10 azarah 32 Hack to make sure sandboxed process cannot remove a device node, bug #79836.
11    
12 ferringb 33 02 Mar 2005; Martin Schlemmer <azarah@gentoo.org> Makefile.am:
13 azarah 31 Fix symbols.in not added to dist.
14    
15 ferringb 33 02 Mar 2005; Martin Schlemmer <azarah@gentoo.org> Makefile.am, canonicalize.c,
16 azarah 30 getcwd.c, libsandbox.c, sandbox.c, sandbox.h, sandbox_futils.c:
17     White space fixes.
18    
19 ferringb 33 02 Mar 2005; Martin Schlemmer <azarah@gentoo.org> Makefile.am, canonicalize.c,
20 azarah 30 configure.in, getcwd.c, libsandbox.c, symbols.in:
21 azarah 29 Fix inverse test logic in canonicalize.c, use a strncpy. Fix gcc warning in
22     getcwd.c. Add symbols.in and logic to Makefile.am to generate symbol versions
23     for glibc and other libc's that use this. Update libsandbox.c to use these
24     symbol versions if available. Fix exec wrapper to re-export LD_PRELOAD if the
25     process unset it.
26    
27 ferringb 33 01 Mar 2005; Brian Harring <ferringb@gentoo.org> libsandbox.c:
28 ferringb 28 killed off _init and _fini in favor of
29     void __attribute__ ((constructor)) init_func and
30     void __attribute__ ((destructor)) closing_func. _(init|func) were deprecated.
31    
32 ferringb 26 06 Dec 2004; Brian Harring <ferringb@gentoo.org> Makefile.am, libsandbox.c,
33     canonicalize.c, getcwd.c: Fixed compilation *again*. Hopefully cvs is done
34     having the hick-ups.
35    
36 ferringb 24 04 Dec 2004; Brian Harring <ferringb@gentoo.org> libsandbox.c, getcwd.c,
37     Makefile.am: Fixed compilation.
38    
39 ferringb 18 01 Dec 2004; Brian Harring <ferringb@gentoo.org> aclocal.m4:
40     Gutted the bugger so it stops checking for a c++ and fortran compiler.
41     Do *not* regenerate aclocal.m4 for making a release until a better
42     solution is created.
43    
44 ferringb 16 20 Nov 2004; Brian Harring <ferringb@gentoo.org> Makefile.am, sandbox_futils.c:
45     Removal of more hardcoded paths.
46    
47 ferringb 13 20 Nov 2004; Brian Harring <ferringb@gentoo.org> Makefile.am, configure.in,
48     sandbox_futils.c: tweaks to install sandbox.bashrc, and use it.
49    
50     19 Nov 2004; Brian Harring <ferringb@gentoo.org>:
51 ferringb 6 Sandbox is now autotooled, create-localdecls needs to be killed and the code
52     shifted into configure.in. Currently builds *one* libsandbox.so- if multiple
53     are desired (-m64 and -m32 for amd64), the ebuild should do it (imo).
54     To get to a point of testing, automake && autoconf; created requisite files w/
55     a(utomake|clocal)-1.8, and autoconf 2.59. Installs to /usr/, instead of
56     /lib and /usr/lib/portage/bin.
57    
58 ferringb 2 14 Nov 2004; Brian Harring <ferringb@gentoo.org> libsandbox.c, sandbox.c:
59     closing out bug #70225, potential overflow of the sandbox_pids_file var.
60    
61     07 Nov 2004; Brian Harring <ferringb@gentoo.org> libsandbox.c: c99 standard,
62     (think it was at least) allows intermixing of code and data segments. bug #70351
63     should be fixed by this.
64    
65     03 Nov 2004; Brian Harring <ferringb@gentoo.org> libsandbox.c, sandbox_futils.c:
66     futils fix from bug #65201 via solar, and libsandbox log path checks via #69137
67    
68     02 Aug 2004; Nicholas Jones <carpaski@gentoo.org> libsandbox.c: Code from
69     Seth Robertson that tracked down all adjuct flags for read operations that
70     do not invoke a write operation.
71    
72     04 Apr 2004; Nicholas Jones <carpaski@gentoo.org> libsandbox.c, sandbox.c:
73     Another fix from jstubbs regarding a free() on a stack variable for the
74     environment -- tracking now prevents extraneous free()'s segfault.
75    
76     04 Apr 2004; Nicholas Jones <carpaski@gentoo.org> libsandbox.c, sandbox.c:
77     J. Stubbs tracked down a new bug where mkdir was failing to the patch on
78     the lstat in mkdir... it now only returns 0 or -1 as documented for mkdir.
79     Also remove the errno = ESUCCESS settings as documentation points out that
80     a library isn't allowed to do that.
81    
82     04 Apr 2004; Nicholas Jones <carpaski@gentoo.org> libsandbox.c: Added a
83     file_security_check() function to check random potential exploits on files
84     that sandbox is to load and read -- Normally sandboxpids.tmp. This fixes
85     the 'system-crippling' exploits (bug 21923) and catches a few other
86     potential problems.
87    
88     20 Mar 2004; Nicholas Jones <carpaski@gentoo.org> Makefile: Updates for
89     32/64 bit sandbox. Made CC and LD '?=' values to allow passed in CC to work.
90    
91     20 Mar 2004; Nicholas Jones <carpaski@gentoo.org> libsandbox.c:
92     bug 42048 -- Fixed the lstat/errno conditions for mkdir <caleb@g.o>.
93     Added the 64/32 bit sandbox patch for AMD64 bug 32963 <brad/azarah>.
94    
95     29 Feb 2004; Martin Schlemmer <azarah@gentoo.org> sandbox.c, sandbox_futils.c :
96     Fix permissions and group of pids file and logs. Permissions should be 0664
97     and group should be 'portage'. Bug #34260.
98    
99     28 Feb 2004; Martin Schlemmer <azarah@gentoo.org> libsandbox.c :
100     Besides a small cleanup, redo how we replace LD_PRELOAD in the environ passed
101     to the real execve (in our execve wrapper). Seems that on some arches (sparc
102     among others) do not allow us to tamper with the readonly copy passed to
103     execve, so pass our own copy of the environment. Bug #42290.
104    
105     11 Jan 2004; Nicholas Jones <carpaski@gentoo.org> create-decls:
106     Changed tail to head and added a notice about duration of glibc check.
107    
108     21 Dec 2003; Nicholas Jones <carpaski@gentoo.org> create-decls:
109     Changed the glibc subversion check to use /usr/bin/* instead of /bin/sh
110     as there isn't a guarentee that it is dynamic.
111    
112     02 Nov 2003; Martin Schlemmer <azarah@gentoo.org> libsandbox.c :
113     If 'file' passed to before_syscall(const char *func, const char *file) is
114     invalid, we should set errno to ENOENT, and not EINVAL. This should
115     close bug #32238.
116    
117     14 Oct 2003; Martin Schlemmer <azarah@gentoo.org> libsandbox.c :
118     Fix a bug that occurs mainly on 64bit arch, where the file passed to
119     the functions we wrap, is invalid, and then cause canonicalize to pass
120     garbage to before_syscall(), thanks to great detective work from
121     Andrea Luzzardi <al@sig11.org> (bug #29846).
122    
123     13 Oct 2003; Martin Schlemmer <azarah@gentoo.org> create-localdecls :
124     Add a uClibc detection patch from Peter S. Mazinger <ps.m@gmx.net>.
125    
126     13 Oct 2003; Martin Schlemmer <azarah@gentoo.org> libsandbox.c :
127     Fix a bug in libsandbox.c 's checking in the rename wrapper - it basically
128     only checked the destination patch, and not the source, so we could move
129     a protected file to a unprotected directory, and then delete/modify it.
130     Thanks to Andrea Luzzardi (scox) <al@sig11.org>, bug #30992, for this fix.
131    
132     12 Oct 2003; Nicholas Jones <carpaski@gentoo.org> sandbox.c :
133     Added python2.3 to the predict section/variable.
134    
135     28 Sep 2003; Martin Schlemmer <azarah@gentoo.org> libsandbox.c, sandbox.c,
136     sandbox.h, sandbox_futils.c :
137     Add support to set the pids file via SANDBOX_PIDS_FILE at startup. If
138     it is not set, it will revert to its old value.
139    
140     27 Sep 2003; Martin Schlemmer <azarah@gentoo.org> libsandbox.c :
141     Fix our mkdir wrapper to check if the dir exist, and return EEXIST if so,
142     rather than failing with a violation, bug #29748.
143    
144     27 Jul 2003; Martin Schlemmer <azarah@gentoo.org> libsandbox.c :
145     Fix canonicalize() to ignore calls with path = "".
146    
147     27 Jul 2003; Martin Schlemmer <azarah@gentoo.org> getcwd.c, libsandbox.c,
148     sandbox_futils.c, canonicalize.c :
149     Once again coreutils fails, as my systems had 2.5 kernel, the getcwd system
150     call handled strings larger than PATH_MAX (bug #21766). It however does not
151     work the same on 2.4 kernels.
152    
153     To fix, I added the posix implementation of getcwd() (from glibc cvs) that
154     do not need the system call. We use the default getcwd() function via a
155     wrapper (egetcwd), and then lstat the returned path. If lstat fails, it
156     means the current directory was removed, OR that the the system call for
157     getcwd failed (curious is that it do not fail and return NULL or set
158     errno, but rather just truncate the retured directory - usually from the
159     start), and if so, we use the generic getcwd() function (__egetcwd). Note
160     that we do not use the generic version all the time, as it calls lstat()
161     a great number of times, and performance degrade much.
162    
163     29 Jun 2003; Martin Schlemmer <azarah@gentoo.org> create-localdecls,
164     libsandbox.c :
165     Make sure SB_PATH_MAX will not wrap. Fix two possible memory leaks.
166    
167     22 Jun 2003; Martin Schlemmer <azarah@gentoo.org> libsandbox.c, canonicalize.c
168     create-localdecls :
169     When checking path names of files accessed, we need to canonicalize it, else
170     it may be a symlink in a 'write allowed' directory pointing to a file in a
171     directory we should not have write access to.
172    
173     With something like coreutils-5.0, we have two problems:
174     1) One of the tests checks if getcwd() can return a path longer than
175     PATH_MAX. This test then tries to create a dir which even while
176     created local (mkdir("conftest2")), it ends up being resolved with
177     a name that is much larger than PATH_MAX. The problem now is that
178     canonicalize() have undefined behaviour when the path was too long
179     (returned wrongly truncated paths, etc), and pass the wrong path to
180     before_syscall() (causing the bogus sandbox violations).
181     2) The ecanonicalize() function we used, along with the canonicalize()
182     function did not support longer than PATH_MAX. This is partly a
183     cause for 1), but the error checking (rather lack of it) of calls
184     to erealpath() in canonicalize() was the prime reason for 1).
185    
186     As we do not use this canonicalized name to call the function, we resolve this
187     by fixing canonicalize() to do better error checking, and ecanonicalize() as
188     well as all functions in libsandbox.c to use a PATH_MAX of 'PATH_MAX * 2'.
189     While they will resolve paths properly now, and can check if a write/read is
190     allowed, the functions called from the sandboxed environment will still work
191     as expected.
192    
193     This should resolve bug #21766.
194    
195     06 Apr 2003; Martin Schlemmer <azarah@gentoo.org> libsandbox.c :
196     For some reason sandbox fails with a 'open_wr' if you run 'locale -a' under
197     it (bug #16298).
198    
199     Problem is that for some reason locale fopen's locale.alias with mode "rm".
200    
201     -------------------------------------------------------
202     nosferatu root # grep fopen locale.log
203     fopen("/usr/share/locale/locale.alias", "rm"ACCESS DENIED open_wr: /usr/share/locale/locale.alias
204     nosferatu root #
205     --------------------------------------------------------
206    
207     I checked the source of locale, but it have fopen with mode 'r', so
208     not sure where the "rm" mode comes from. Anyhow, changed the check in
209     before_syscall_open_char() to also see mode "rm" as readonly.
210    
211     23 Feb 2003; Martin Schlemmer <azarah@gentoo.org> create-localdecls :
212    
213     Add glibc-2.3 support.
214    
215     22 Feb 2003; Martin Schlemmer <azarah@gentoo.org> sandbox.c :
216    
217     Some /etc/ld.so.preload fixes. Just changed the #if defines to cover all
218     operations releated to preload, as well as only try to modify ld.so.preload
219     if we can. Also modify to write the pid to /tmp/sandboxpids.tmp even when
220     not using ld.so.preload. Fix to not write this instance of sandbox's pid
221     to /tmp/sandboxpids.tmp on exit if this is not the last sandbox running.
222    
223     22 Feb 2003; Nicholas Jones <carpaski@gentoo.org> Makefile :
224    
225     Changed the LD to CC for hppa.
226    
227     22 Feb 2003; Nicholas Jones <carpaski@gentoo.org> create-localdecls :
228    
229     Killed the previous changes I made.
230    
231     17 Feb 2003; Nicholas Jones <carpaski@gentoo.org> create-localdecls :
232    
233     Added parisc to BROKEN_RTLD_ARCHLIST to see if it we can fix the relocation probs.
234    
235     09 Jan 2003; J Robert Ray <jrray@gentoo.org> sandbox.c :
236    
237     Don't segfault if $HOME isn't set, set $HOME to "/" instead. Fixes bug 10868.
238    
239     16 Dec 2002; Martin Schlemmer <azarah@gentoo.org> create-localdecls :
240    
241     Fix memory leak for mips, bug #12236. Thanks to Torgeir Hansen <torgeir@trenger.ro>
242     for this fix.
243    
244     4 Dec 2002; J Robert Ray <jrray@gentoo.org> sandbox.h sandbox_futils.c :
245    
246     sandbox_futils defined a dirname() function that was masking the same
247     function in glibc and was broken (e.g.: SANDBOX_DIR was being set to
248     '/usr/lib/portage/bi/'). Fixed function to return expected results and
249     renamed it to sb_dirname() to no longer mask the glibc function. Closes bug
250     11231.
251    
252     4 Dec 2002; Martin Schlemmer <azarah@gentoo.org> :
253    
254     Fix a segfault in libsandbox.c if canonicalize() was called with
255     first parameter = NULL.
256    
257     1 Sep 2002; Martin Schlemmer <azarah@gentoo.org> :
258    
259     Fix my braindead 'return 1;' in a void function. Updated sandbox.c,
260     cleanup() for this.
261    
262     Change cleanup() in sandbox.c not to exit with fail status if
263     the pidsfile is missing. We really should still display sandbox
264     violations if they occured.
265    
266     31 Aug 2002; Martin Schlemmer <azarah@gentoo.org> :
267    
268     Update cleanup() in sandbox.c to remove the PIDSFILE if this is
269     the last sandbox running.
270    
271     25 Aug 2002; Martin Schlemmer <azarah@gentoo.org> :
272    
273     Major cleanups to mainly libsandbox.c again.
274    
275     22 Aug 2002; Martin Schlemmer <azarah@gentoo.org> :
276    
277     Add copyrights to sandbox.h and sandbox_futils.h. If wrong, the
278     parties involved should please contact me so that we can fix it.
279    
280     Add opendir wrapper to libsandbox.c.
281    
282     21 Aug 2002; Martin Schlemmer <azarah@gentoo.org> :
283    
284     Do some more cleanups to ecanonicalize(), as it dropped filenames in
285     rare cases (after my symlink cleanups), and caused glibc to bork.
286     These fixes went into canonicalize.c.
287    
288     20 Aug 2002; Martin Schlemmer <azarah@gentoo.org> :
289    
290     Fix spawn_shell() and main() in sandbox.c to properly return fail
291     status.
292    
293     19 Aug 2002; Martin Schlemmer <azarah@gentoo.org> :
294    
295     The new canonicalize() function in libsandbox.c also resolved symlinks,
296     which caused on cleaning sandbox errors if the symlink pointed to a
297     file in the live root. Ripped out canonicalize() and realpath() from
298     glibc; removed the symlink stuff, and changed them to ecanonicalize()
299     and erealpath().
300    
301     18 Aug 2002; Martin Schlemmer <azarah@gentoo.org> :
302    
303     Ripped out all the wrappers, and implemented those of InstallWatch.
304     Losts of cleanups and bugfixes. Implement a execve that forces
305     $LIBSANDBOX in $LD_PRELOAD. We can now thus do away with the feared
306     /etc/ld.so.preload (*g*) ... Made the needed changes to sandbox.c,
307     sandbox.h and sandbox_futils.c. Rewrote the Makefile for most
308     parts; it now have an install target.
309    
310     Reformat the whole thing to look somewhat like the reworked sandbox.c
311     and new sandbox.h and sandbox_futils.c from:
312    
313     Brad House <brad@mainstreetsoftworks.com>.
314    
315     Additional Copyrights now due to the InstallWatch code:
316    
317     Copyright (C) 1998-9 Pancrazio `Ezio' de Mauro <p@demauro.net>

Properties

Name Value
svn:eol-style native
svn:keywords Author Date Id Revision

  ViewVC Help
Powered by ViewVC 1.1.20