/[path-sandbox]/trunk/ChangeLog
Gentoo

Contents of /trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 114 - (show annotations) (download)
Thu Jun 9 16:53:13 2005 UTC (8 years, 10 months ago) by azarah
File size: 24076 byte(s)
Move symlink hack down a bit to try and minimize on the amount of lstat()
calls we do.

1 # ChangeLog for Path Sandbox
2 # Copyright 1999-2005 Gentoo Foundation; Distributed under the GPL v2
3 # $Header$
4
5 09 June 2005; Martin Schlemmer <azarah@gentoo.org> libsandbox.c:
6 Move symlink hack down a bit to try and minimize on the amount of lstat()
7 calls we do.
8
9 09 June 2005; Martin Schlemmer <azarah@gentoo.org> libsandbox.c, sandbox.c:
10 Add hack to allow writing to /proc/self/fd (or /dev/fd), bug #91516.
11
12 09 June 2005; Martin Schlemmer <azarah@gentoo.org> symbols.in, libsandbox.c:
13 Add wrapper for access() function, bug #85413.
14
15 09 June 2005; Martin Schlemmer <azarah@gentoo.org> getcwd.c:
16 Use generic getcwd() implementation from uclibc - should be more portable
17 and looks a bit cleaner.
18
19 09 June 2005; Martin Schlemmer <azarah@gentoo.org> getcwd.c, libsandbox.c:
20 Make sure our true_* pointers are initialized to NULL, and that we check for
21 all references that they are valid.
22
23 09 June 2005; Martin Schlemmer <azarah@gentoo.org> libsandbox.c:
24 Be default we will fail if the path name we try to canonicalize is too long.
25 This however could cause issues with some things (bug #94630 and #21766), so
26 if fail_nametoolong == 0, canonicalize() will return a null length string and
27 do not fail.
28
29 08 June 2005; Martin Schlemmer <azarah@gentoo.org> sandbox.c, sandbox_futils.c:
30 Do not abort if TMPDIR is not valid, but rather use '/tmp', bug #94360. Also
31 make sure we re-export the new TMPDIR environment variable.
32
33 08 June 2005; Martin Schlemmer <azarah@gentoo.org> libsandbox.c:
34 Fix incorrect free of non-malloc'd array, bug #92313 and #94020. Fix noted
35 by Marcus D. Hanwell <cryos@gentoo.org>.
36
37 08 June 2005; Martin Schlemmer <azarah@gentoo.org> sandbox.c:
38 Add /dev/console to write list, bug #38588.
39
40 * sandbox-1.2.8 (2005/05/13)
41
42 13 May 2005; Martin Schlemmer <azarah@gentoo.org> libsandbox.c, sandbox.c,
43 sandbox.h, sandbox_futils.c:
44 General cleanups.
45
46 13 May 2005; Martin Schlemmer <azarah@gentoo.org> libsandbox.c, sandbox.c,
47 sandbox.h:
48 Various LD_PRELOAD cleanups. Do not unset LD_PRELOAD for parent.
49
50 13 May 2005; Martin Schlemmer <azarah@gentoo.org> libsandbox.c, sandbox.c,
51 sandbox.h, sandbox_futils.c:
52 Modify get_sandbox_pids_file(), get_sandbox_log() and get_sandbox_debug_log()
53 to use TMPDIR if present in environment.
54
55 13 May 2005; Martin Schlemmer <azarah@gentoo.org> sandbox.c:
56 Remove sandbox_log_file from main() as its no longer used.
57
58 13 May 2005; Martin Schlemmer <azarah@gentoo.org> sandbox.c, sandbox.h,
59 sandbox_futils.c:
60 Add get_sandbox_debug_log(), and use it (add behaviour similar to SANDBOX_LOG
61 if already exported when sandbox started). Fix get_sandbox_log() and new
62 get_sandbox_debug_log() to not use already exported environment variables if
63 they have '/' in them. Use snprintf()'s instead of strncpy()'s. More
64 SB_PATH_MAX fixes.
65
66 * sandbox-1.2.7 (2005/05/12)
67
68 12 May 2005; Martin Schlemmer <azarah@gentoo.org> sandbox.c, sandbox.h,
69 sandbox_futils.c:
70 More path limit fixes. Declare SB_BUF_LEN global and use it where needed.
71
72 12 May 2005; Martin Schlemmer <azarah@gentoo.org> sandbox_futils.c:
73 Fix paths limited to 255 chars. Fix get_sandbox_dir() returning a string
74 with '(null)' in it if we did not call sandbox with absolute path.
75
76 12 May 2005; Martin Schlemmer <azarah@gentoo.org> sandbox.c:
77 Set SANDBOX_ON *before* doing the child's env stuff, else its not set
78 for the child.
79
80 12 May 2005; Martin Schlemmer <azarah@gentoo.org> sandbox.c:
81 Remove global preload_adaptable as it is no longer used.
82
83 12 May 2005; Martin Schlemmer <azarah@gentoo.org> sandbox.c:
84 Rewrite environment stuff to only be set when execve'ing the child process
85 to try and avoid issues like bug #91541 that causes sandbox to crash if
86 we set LD_PRELOAD sandbox side already.
87
88 11 May 2005; Martin Schlemmer <azarah@gentoo.org> sandbox.c:
89 Move print_sandbox_log() up to make things neater.
90
91 11 May 2005; Martin Schlemmer <azarah@gentoo.org> sandbox.c:
92 Remove load_preload_libs(), as its not used anymore.
93
94 11 May 2005; Martin Schlemmer <azarah@gentoo.org> sandbox.c, sandbox.h:
95 Remove NO_FORK stuff, as its not used, and 'strace -f' works just fine.
96
97 11 May 2005; Martin Schlemmer <azarah@gentoo.org> sandbox.c, sandbox.h:
98 Remove USE_SYSTEM_SHELL stuff, as it is not secure, and not in use.
99
100 11 May 2005; Martin Schlemmer <azarah@gentoo.org> sandbox.c, sandbox.h:
101 Remove ld.so.preload crap - we are not going to use it again.
102
103 10 May 2005; Martin Schlemmer <azarah@gentoo.org> sandbox_futils.c:
104 Fix typo in code that checks if we got valid group information, causing a
105 segmentation fault, bug #91637.
106
107 * sandbox-1.2.6 (2005/05/10)
108
109 10 May 2005; Martin Schlemmer <azarah@gentoo.org> sandbox.c:
110 Do not use LD_PRELOAD if it contains libtsocks.so, as it breaks sandbox
111 for some odd reason, bug #91541.
112
113 09 May 2005; Martin Schlemmer <azarah@gentoo.org> sandbox.c:
114 Fix typo (sizeof -> strlen).
115
116 08 May 2005; Brian Harring <ferringb@gentoo.org> libsandbox.c:
117 rewrote the sbcontext caching code so it accounts for env changes since lib
118 initialization.
119
120 05 May 2005; Martin Schlemmer <azarah@gentoo.org> configure.in, libctest.c:
121 We create libctest.c via configure, so no need to keep it around. Do some
122 cleanup related to libctest.c and libctest during configure.
123
124 04 May 2005; Martin Schlemmer <azarah@gentoo.org> libsandbox.c:
125 Add rename support of symlinks pointing to protected files/directories.
126
127 * sandbox-1.2.5 (2005/05/04)
128
129 04 May 2005; Martin Schlemmer <azarah@gentoo.org> libsandbox.c, sandbox.c,
130 sandbox.bashrc:
131 Do not reset already set LD_PRELOAD when starting sandbox. If LD_PRELOAD is
132 already set, init of the env vars fails for some reason, so do this later on,
133 and do not warn (bug #91431).
134
135 03 May 2005; Martin Schlemmer <azarah@gentoo.org> sandbox.c, sandbox.h,
136 sandbox.bashrc:
137 Fixup sandbox and sandbox.bashrc to call bash with the proper .bashrc.
138
139 * sandbox-1.2.4 (2005/05/03)
140
141 03 May 2005; Martin Schlemmer <azarah@gentoo.org> libsandbox.c:
142 Do not init the env entries with each call, as it creates too many calls to
143 lstat, etc. Should speedup things a bit, bug #91040.
144
145 03 May 2005; Martin Schlemmer <azarah@gentoo.org> sandbox.c:
146 Add /dev/pty to default write list. Noticed by Morfic.
147
148 02 May 2005; Mike Frysinger <vapier@gentoo.org> configure.in, localdecls.h,
149 sandbox.h:
150 uClibc doesn't support dlvsym() so add a configure check to make sure it doesn't
151 exist. Also update localdecls.h so BROKEN_RTLD_NEXT isn't defined in uClibc.
152
153 * sandbox-1.2.3 (2005/04/29)
154
155 29 Apr 2005; Martin Schlemmer <azarah@gentoo.org> configure.in:
156 Do not check for (*&#$(* CXX or F77.
157
158 29 Apr 2005; Martin Schlemmer <azarah@gentoo.org> libsandbox.c:
159 Do not append '/' to pathname in filter_path() if it already ends with it.
160
161 28 Apr 2005; Mike Frysinger <vapier@gentoo.org> Makefile.am, configure.in:
162 With az's help, clean up autotools to work with cross-compiling.
163
164 * sandbox-1.2.2 (2005/04/28)
165
166 28 Apr 2005; Martin Schlemmer <azarah@gentoo.org> libsandbox.c:
167 Only check for /dev/{null,zero} for unlink hack, else ricers using /dev/shm
168 have issues; bug #90592.
169
170 * sandbox-1.2.1 (2005/04/23)
171
172 23 Apr 2005; Martin Schlemmer <azarah@gentoo.org> Makefile.am, canonicalize.c,
173 getcwd.c, libsandbox.c, localdecls.h, sandbox.h, sandbox_futils.c:
174 Make sure all functions used in libsandbox.c is declared static. Define
175 SB_STATIC in localdecls.h for this. Include sandbox_futils.c rather than
176 linking with its object. Hopefully this will fix bug #90153.
177
178 * sandbox-1.2 (2005/04/23)
179
180 22 Mar 2005; Martin Schlemmer <azarah@gentoo.org> libsandbox.c:
181 Allow lchown a symlink in write-allowed path pointing to write-denied
182 target.
183
184 21 Mar 2005; Marius Mauch <genone@gentoo.org> libsandbox.c:
185 Also show resolved symlink names in the log.
186
187 14 Mar 2005; Martin Schlemmer <azarah@gentoo.org> Makefile.am, libsandbox.c:
188 Seems -nostdlib was the problem with the constructor/destructor - remove it
189 from Makefile.am, and change the constructor/destructor names again.
190
191 14 Mar 2005; Martin Schlemmer <azarah@gentoo.org> libsandbox.c:
192 Also rename the _init() and _fini() declarations.
193
194 14 Mar 2005; Martin Schlemmer <azarah@gentoo.org> libsandbox.c, sandbox.c:
195 Fixup the constructor/destructor function names again (they should be _init()
196 and _fini() it seems, and not being called caused sandbox_lib_path to be
197 unset, and thus breaking the execve() wrapper's LD_PRELOAD protection).
198 Add both the path in given SANDBOX_x variable, as well as its symlink
199 resolved path in init_env_entries(). Modify filter_path() to be able to
200 resolve paths without resolving symlinks, as well as to be able to resolve
201 symlinks. Fix a possible segfault in check_access(). Add symlink resolving
202 to check_access() resolving bug #31019. Add 'hack' for unlink, as the fix
203 for bug #31019 cause access violations if we try to remove a symlink that is
204 not in protected path, but points to a protected path. Fix a memory leak in
205 sandbox.c (sandbox_pids_file in main()). Fix the realpath() calls in main()
206 (sandbox.c) being unchecked. Fix the debug logname not having the pid in it
207 (pid_string was uninitialized). General syntax cleanups.
208
209 09 Mar 2005; Brian Harring <ferringb@gentoo.org> sandbox.c: Fixed the
210 infamous "pids file is not a regular file" w/out newline bug.
211
212 09 Mar 2005; Brian Harring <ferringb@gentoo.org> Makefile.am, configure.in:
213 Correct libc_version path detection, since it was screwing up if libdir !=
214 "/lib/".
215
216 02 Mar 2005; Martin Schlemmer <azarah@gentoo.org> libsandbox.c:
217 Hack to make sure sandboxed process cannot remove a device node, bug #79836.
218
219 02 Mar 2005; Martin Schlemmer <azarah@gentoo.org> Makefile.am:
220 Fix symbols.in not added to dist.
221
222 02 Mar 2005; Martin Schlemmer <azarah@gentoo.org> Makefile.am, canonicalize.c,
223 getcwd.c, libsandbox.c, sandbox.c, sandbox.h, sandbox_futils.c:
224 White space fixes.
225
226 02 Mar 2005; Martin Schlemmer <azarah@gentoo.org> Makefile.am, canonicalize.c,
227 configure.in, getcwd.c, libsandbox.c, symbols.in:
228 Fix inverse test logic in canonicalize.c, use a strncpy. Fix gcc warning in
229 getcwd.c. Add symbols.in and logic to Makefile.am to generate symbol versions
230 for glibc and other libc's that use this. Update libsandbox.c to use these
231 symbol versions if available. Fix exec wrapper to re-export LD_PRELOAD if the
232 process unset it.
233
234 01 Mar 2005; Brian Harring <ferringb@gentoo.org> libsandbox.c:
235 killed off _init and _fini in favor of
236 void __attribute__ ((constructor)) init_func and
237 void __attribute__ ((destructor)) closing_func. _(init|func) were deprecated.
238
239 06 Dec 2004; Brian Harring <ferringb@gentoo.org> Makefile.am, libsandbox.c,
240 canonicalize.c, getcwd.c: Fixed compilation *again*. Hopefully cvs is done
241 having the hick-ups.
242
243 04 Dec 2004; Brian Harring <ferringb@gentoo.org> libsandbox.c, getcwd.c,
244 Makefile.am: Fixed compilation.
245
246 01 Dec 2004; Brian Harring <ferringb@gentoo.org> aclocal.m4:
247 Gutted the bugger so it stops checking for a c++ and fortran compiler.
248 Do *not* regenerate aclocal.m4 for making a release until a better
249 solution is created.
250
251 20 Nov 2004; Brian Harring <ferringb@gentoo.org> Makefile.am, sandbox_futils.c:
252 Removal of more hardcoded paths.
253
254 20 Nov 2004; Brian Harring <ferringb@gentoo.org> Makefile.am, configure.in,
255 sandbox_futils.c: tweaks to install sandbox.bashrc, and use it.
256
257 19 Nov 2004; Brian Harring <ferringb@gentoo.org>:
258 Sandbox is now autotooled, create-localdecls needs to be killed and the code
259 shifted into configure.in. Currently builds *one* libsandbox.so- if multiple
260 are desired (-m64 and -m32 for amd64), the ebuild should do it (imo).
261 To get to a point of testing, automake && autoconf; created requisite files w/
262 a(utomake|clocal)-1.8, and autoconf 2.59. Installs to /usr/, instead of
263 /lib and /usr/lib/portage/bin.
264
265 14 Nov 2004; Brian Harring <ferringb@gentoo.org> libsandbox.c, sandbox.c:
266 closing out bug #70225, potential overflow of the sandbox_pids_file var.
267
268 07 Nov 2004; Brian Harring <ferringb@gentoo.org> libsandbox.c: c99 standard,
269 (think it was at least) allows intermixing of code and data segments. bug #70351
270 should be fixed by this.
271
272 03 Nov 2004; Brian Harring <ferringb@gentoo.org> libsandbox.c, sandbox_futils.c:
273 futils fix from bug #65201 via solar, and libsandbox log path checks via #69137
274
275 02 Aug 2004; Nicholas Jones <carpaski@gentoo.org> libsandbox.c: Code from
276 Seth Robertson that tracked down all adjuct flags for read operations that
277 do not invoke a write operation.
278
279 04 Apr 2004; Nicholas Jones <carpaski@gentoo.org> libsandbox.c, sandbox.c:
280 Another fix from jstubbs regarding a free() on a stack variable for the
281 environment -- tracking now prevents extraneous free()'s segfault.
282
283 04 Apr 2004; Nicholas Jones <carpaski@gentoo.org> libsandbox.c, sandbox.c:
284 J. Stubbs tracked down a new bug where mkdir was failing to the patch on
285 the lstat in mkdir... it now only returns 0 or -1 as documented for mkdir.
286 Also remove the errno = ESUCCESS settings as documentation points out that
287 a library isn't allowed to do that.
288
289 04 Apr 2004; Nicholas Jones <carpaski@gentoo.org> libsandbox.c: Added a
290 file_security_check() function to check random potential exploits on files
291 that sandbox is to load and read -- Normally sandboxpids.tmp. This fixes
292 the 'system-crippling' exploits (bug 21923) and catches a few other
293 potential problems.
294
295 20 Mar 2004; Nicholas Jones <carpaski@gentoo.org> Makefile: Updates for
296 32/64 bit sandbox. Made CC and LD '?=' values to allow passed in CC to work.
297
298 20 Mar 2004; Nicholas Jones <carpaski@gentoo.org> libsandbox.c:
299 bug 42048 -- Fixed the lstat/errno conditions for mkdir <caleb@g.o>.
300 Added the 64/32 bit sandbox patch for AMD64 bug 32963 <brad/azarah>.
301
302 29 Feb 2004; Martin Schlemmer <azarah@gentoo.org> sandbox.c, sandbox_futils.c :
303 Fix permissions and group of pids file and logs. Permissions should be 0664
304 and group should be 'portage'. Bug #34260.
305
306 28 Feb 2004; Martin Schlemmer <azarah@gentoo.org> libsandbox.c :
307 Besides a small cleanup, redo how we replace LD_PRELOAD in the environ passed
308 to the real execve (in our execve wrapper). Seems that on some arches (sparc
309 among others) do not allow us to tamper with the readonly copy passed to
310 execve, so pass our own copy of the environment. Bug #42290.
311
312 11 Jan 2004; Nicholas Jones <carpaski@gentoo.org> create-decls:
313 Changed tail to head and added a notice about duration of glibc check.
314
315 21 Dec 2003; Nicholas Jones <carpaski@gentoo.org> create-decls:
316 Changed the glibc subversion check to use /usr/bin/* instead of /bin/sh
317 as there isn't a guarentee that it is dynamic.
318
319 02 Nov 2003; Martin Schlemmer <azarah@gentoo.org> libsandbox.c :
320 If 'file' passed to before_syscall(const char *func, const char *file) is
321 invalid, we should set errno to ENOENT, and not EINVAL. This should
322 close bug #32238.
323
324 14 Oct 2003; Martin Schlemmer <azarah@gentoo.org> libsandbox.c :
325 Fix a bug that occurs mainly on 64bit arch, where the file passed to
326 the functions we wrap, is invalid, and then cause canonicalize to pass
327 garbage to before_syscall(), thanks to great detective work from
328 Andrea Luzzardi <al@sig11.org> (bug #29846).
329
330 13 Oct 2003; Martin Schlemmer <azarah@gentoo.org> create-localdecls :
331 Add a uClibc detection patch from Peter S. Mazinger <ps.m@gmx.net>.
332
333 13 Oct 2003; Martin Schlemmer <azarah@gentoo.org> libsandbox.c :
334 Fix a bug in libsandbox.c 's checking in the rename wrapper - it basically
335 only checked the destination patch, and not the source, so we could move
336 a protected file to a unprotected directory, and then delete/modify it.
337 Thanks to Andrea Luzzardi (scox) <al@sig11.org>, bug #30992, for this fix.
338
339 12 Oct 2003; Nicholas Jones <carpaski@gentoo.org> sandbox.c :
340 Added python2.3 to the predict section/variable.
341
342 28 Sep 2003; Martin Schlemmer <azarah@gentoo.org> libsandbox.c, sandbox.c,
343 sandbox.h, sandbox_futils.c :
344 Add support to set the pids file via SANDBOX_PIDS_FILE at startup. If
345 it is not set, it will revert to its old value.
346
347 27 Sep 2003; Martin Schlemmer <azarah@gentoo.org> libsandbox.c :
348 Fix our mkdir wrapper to check if the dir exist, and return EEXIST if so,
349 rather than failing with a violation, bug #29748.
350
351 27 Jul 2003; Martin Schlemmer <azarah@gentoo.org> libsandbox.c :
352 Fix canonicalize() to ignore calls with path = "".
353
354 27 Jul 2003; Martin Schlemmer <azarah@gentoo.org> getcwd.c, libsandbox.c,
355 sandbox_futils.c, canonicalize.c :
356 Once again coreutils fails, as my systems had 2.5 kernel, the getcwd system
357 call handled strings larger than PATH_MAX (bug #21766). It however does not
358 work the same on 2.4 kernels.
359
360 To fix, I added the posix implementation of getcwd() (from glibc cvs) that
361 do not need the system call. We use the default getcwd() function via a
362 wrapper (egetcwd), and then lstat the returned path. If lstat fails, it
363 means the current directory was removed, OR that the the system call for
364 getcwd failed (curious is that it do not fail and return NULL or set
365 errno, but rather just truncate the retured directory - usually from the
366 start), and if so, we use the generic getcwd() function (__egetcwd). Note
367 that we do not use the generic version all the time, as it calls lstat()
368 a great number of times, and performance degrade much.
369
370 29 Jun 2003; Martin Schlemmer <azarah@gentoo.org> create-localdecls,
371 libsandbox.c :
372 Make sure SB_PATH_MAX will not wrap. Fix two possible memory leaks.
373
374 22 Jun 2003; Martin Schlemmer <azarah@gentoo.org> libsandbox.c, canonicalize.c
375 create-localdecls :
376 When checking path names of files accessed, we need to canonicalize it, else
377 it may be a symlink in a 'write allowed' directory pointing to a file in a
378 directory we should not have write access to.
379
380 With something like coreutils-5.0, we have two problems:
381 1) One of the tests checks if getcwd() can return a path longer than
382 PATH_MAX. This test then tries to create a dir which even while
383 created local (mkdir("conftest2")), it ends up being resolved with
384 a name that is much larger than PATH_MAX. The problem now is that
385 canonicalize() have undefined behaviour when the path was too long
386 (returned wrongly truncated paths, etc), and pass the wrong path to
387 before_syscall() (causing the bogus sandbox violations).
388 2) The ecanonicalize() function we used, along with the canonicalize()
389 function did not support longer than PATH_MAX. This is partly a
390 cause for 1), but the error checking (rather lack of it) of calls
391 to erealpath() in canonicalize() was the prime reason for 1).
392
393 As we do not use this canonicalized name to call the function, we resolve this
394 by fixing canonicalize() to do better error checking, and ecanonicalize() as
395 well as all functions in libsandbox.c to use a PATH_MAX of 'PATH_MAX * 2'.
396 While they will resolve paths properly now, and can check if a write/read is
397 allowed, the functions called from the sandboxed environment will still work
398 as expected.
399
400 This should resolve bug #21766.
401
402 06 Apr 2003; Martin Schlemmer <azarah@gentoo.org> libsandbox.c :
403 For some reason sandbox fails with a 'open_wr' if you run 'locale -a' under
404 it (bug #16298).
405
406 Problem is that for some reason locale fopen's locale.alias with mode "rm".
407
408 -------------------------------------------------------
409 nosferatu root # grep fopen locale.log
410 fopen("/usr/share/locale/locale.alias", "rm"ACCESS DENIED open_wr: /usr/share/locale/locale.alias
411 nosferatu root #
412 --------------------------------------------------------
413
414 I checked the source of locale, but it have fopen with mode 'r', so
415 not sure where the "rm" mode comes from. Anyhow, changed the check in
416 before_syscall_open_char() to also see mode "rm" as readonly.
417
418 23 Feb 2003; Martin Schlemmer <azarah@gentoo.org> create-localdecls :
419
420 Add glibc-2.3 support.
421
422 22 Feb 2003; Martin Schlemmer <azarah@gentoo.org> sandbox.c :
423
424 Some /etc/ld.so.preload fixes. Just changed the #if defines to cover all
425 operations releated to preload, as well as only try to modify ld.so.preload
426 if we can. Also modify to write the pid to /tmp/sandboxpids.tmp even when
427 not using ld.so.preload. Fix to not write this instance of sandbox's pid
428 to /tmp/sandboxpids.tmp on exit if this is not the last sandbox running.
429
430 22 Feb 2003; Nicholas Jones <carpaski@gentoo.org> Makefile :
431
432 Changed the LD to CC for hppa.
433
434 22 Feb 2003; Nicholas Jones <carpaski@gentoo.org> create-localdecls :
435
436 Killed the previous changes I made.
437
438 17 Feb 2003; Nicholas Jones <carpaski@gentoo.org> create-localdecls :
439
440 Added parisc to BROKEN_RTLD_ARCHLIST to see if it we can fix the relocation probs.
441
442 09 Jan 2003; J Robert Ray <jrray@gentoo.org> sandbox.c :
443
444 Don't segfault if $HOME isn't set, set $HOME to "/" instead. Fixes bug 10868.
445
446 16 Dec 2002; Martin Schlemmer <azarah@gentoo.org> create-localdecls :
447
448 Fix memory leak for mips, bug #12236. Thanks to Torgeir Hansen <torgeir@trenger.ro>
449 for this fix.
450
451 4 Dec 2002; J Robert Ray <jrray@gentoo.org> sandbox.h sandbox_futils.c :
452
453 sandbox_futils defined a dirname() function that was masking the same
454 function in glibc and was broken (e.g.: SANDBOX_DIR was being set to
455 '/usr/lib/portage/bi/'). Fixed function to return expected results and
456 renamed it to sb_dirname() to no longer mask the glibc function. Closes bug
457 11231.
458
459 4 Dec 2002; Martin Schlemmer <azarah@gentoo.org> :
460
461 Fix a segfault in libsandbox.c if canonicalize() was called with
462 first parameter = NULL.
463
464 1 Sep 2002; Martin Schlemmer <azarah@gentoo.org> :
465
466 Fix my braindead 'return 1;' in a void function. Updated sandbox.c,
467 cleanup() for this.
468
469 Change cleanup() in sandbox.c not to exit with fail status if
470 the pidsfile is missing. We really should still display sandbox
471 violations if they occured.
472
473 31 Aug 2002; Martin Schlemmer <azarah@gentoo.org> :
474
475 Update cleanup() in sandbox.c to remove the PIDSFILE if this is
476 the last sandbox running.
477
478 25 Aug 2002; Martin Schlemmer <azarah@gentoo.org> :
479
480 Major cleanups to mainly libsandbox.c again.
481
482 22 Aug 2002; Martin Schlemmer <azarah@gentoo.org> :
483
484 Add copyrights to sandbox.h and sandbox_futils.h. If wrong, the
485 parties involved should please contact me so that we can fix it.
486
487 Add opendir wrapper to libsandbox.c.
488
489 21 Aug 2002; Martin Schlemmer <azarah@gentoo.org> :
490
491 Do some more cleanups to ecanonicalize(), as it dropped filenames in
492 rare cases (after my symlink cleanups), and caused glibc to bork.
493 These fixes went into canonicalize.c.
494
495 20 Aug 2002; Martin Schlemmer <azarah@gentoo.org> :
496
497 Fix spawn_shell() and main() in sandbox.c to properly return fail
498 status.
499
500 19 Aug 2002; Martin Schlemmer <azarah@gentoo.org> :
501
502 The new canonicalize() function in libsandbox.c also resolved symlinks,
503 which caused on cleaning sandbox errors if the symlink pointed to a
504 file in the live root. Ripped out canonicalize() and realpath() from
505 glibc; removed the symlink stuff, and changed them to ecanonicalize()
506 and erealpath().
507
508 18 Aug 2002; Martin Schlemmer <azarah@gentoo.org> :
509
510 Ripped out all the wrappers, and implemented those of InstallWatch.
511 Losts of cleanups and bugfixes. Implement a execve that forces
512 $LIBSANDBOX in $LD_PRELOAD. We can now thus do away with the feared
513 /etc/ld.so.preload (*g*) ... Made the needed changes to sandbox.c,
514 sandbox.h and sandbox_futils.c. Rewrote the Makefile for most
515 parts; it now have an install target.
516
517 Reformat the whole thing to look somewhat like the reworked sandbox.c
518 and new sandbox.h and sandbox_futils.c from:
519
520 Brad House <brad@mainstreetsoftworks.com>.
521
522 Additional Copyrights now due to the InstallWatch code:
523
524 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