path: root/tests
Commit message (Collapse)AuthorAgeFilesLines
* tests/script-16.sh: mark as passing only for native ABISergei Trofimovich2020-05-311-1/+2
| | | | | | | | | | | | | All scripts assume that ran tools matck tested sandbox's ABI. Most scripts have a guard against ABI check, but script-16 was missing it. It's afollow-up commit to 24fd102c9976 ("check_syscall(): turn internal sandbox violation into denywrite") Reported-by: Michał Górny Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org> Closes: https://bugs.gentoo.org/590084 Signed-off-by: Michał Górny <mgorny@gentoo.org>
* check_syscall(): turn internal sandbox violation into denywriteSergei Trofimovich2020-05-312-0/+12
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In #590084 test suite performed to list files in a deleted directory: $ sandbox 'mkdir /tmp/zzz; cd /tmp/zzz; rmdir /tmp/zzz; ls' * sandbox-2.18/libsandbox/libsandbox.c:check_syscall():974: failure (No such file or directory): * ISE: opendir(.) abs_path: (null) res_path: (null) Another reproducer is to create file outside deleted directory relative to that directory: $ sandbox 'mkdir /tmp/zzz; cd /tmp/zzz; rmdir /tmp/zzz; touch ../foo' * sandbox-2.18/libsandbox/libsandbox.c:check_syscall():974: failure (No such file or directory): * ISE: open_wr(../foo) abs_path: (null) res_path: (null) sandbox can't validate safety of any of these operations as kernel does not provide a mechanism to resolve '.' back to an absolute path. As it's a rare condition let's turn it into a sandbox violation instead of internal sandbox error and link to the bug with details in the error message. Report after the change looks like: $ ./sandbox.sh 'mkdir /tmp/zzz; cd /tmp/zzz; rmdir /tmp/zzz; touch ../foo' * ACCESS DENIED: open_wr: '../foo' (from deleted directory, see https://bugs.gentoo.org/590084) * ACCESS DENIED: utimensat: '../foo' (from deleted directory, see https://bugs.gentoo.org/590084) touch: cannot touch '../foo': Permission denied Reported-by: Mike Gilbert Bug: https://bugs.gentoo.org/590084 Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org> Signed-off-by: Michał Górny <mgorny@gentoo.org>
* tests: disable utimensat-3 on *-linux-muslSergei Trofimovich2019-06-272-0/+12
| | | | | | | | | | | | | | | | | | | x86_64-gentoo-linux-musl fails a single test: 83: utimensat/3 FAILED (utimensat.at:3) The test checks if sandbox does not crash when utimensat(<filefd>, NULL, NULL, 0) is called. The behaviour is not specified by POSIX but glibc returns EINVAL for such a case. Thus the test behaves differently on varius libs. https://www.openwall.com/lists/musl/2019/06/25/1 has a conversation with musl upstream. The change restricts test down to glibc targets. Bug: https://bugs.gentoo.org/549108 Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org>
* exec*() wrappers: never mutate 'environ' of host processSergei Trofimovich2019-01-084-0/+27
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In bug #669702 gcc exposed sandbox bug where execv() wrapper changed 'environ' global variable underneath. A few GNU projects (pex_unix_exec_child in gcc and gdb) use the following idiom: for (;;) { vfork(); char ** save_environ = environ; // [1] if (child) { environ = child_environ; // [2] execv(payload); // [3] } if (parent) { environ = save_environ; // [4] ... waitpid(child, ...); } } Code above assumes that execv() does not mutate 'environ'. In case of #669702 sandbox's execv() wrapper at '[3]' mutated 'environ' and relocated it (via maloc()/free() internally). This caused '[4]' to point 'environ' fo freed location. The change fixes it in a following way: - execv() call now works more like execve() call by mutating external array and substitutes 'environ' only for a period of 'execv()' execution. - add basic execv()/'environ' corruption test Tested on: - linux/glibc-2.28 - linux/uclibc-ng-1.0.31 Reported-and-tested-by: Walther Reported-by: 0x6d6174@posteo.de Reported-by: Andrey Korolyov Bug: https://bugs.gentoo.org/669702 Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org>
* tests: Add a test for LD_PRELOAD non-preserving (SANDBOX_ON=0)Michał Górny2018-02-192-0/+22
* libsandbox: Fix path matching not to dumbly match prefixesMichał Górny2017-10-032-0/+21
| | | | | | Fix the path matching code to match prefixes component-wide rather than literally. This means that a path such as '/foo' will no longer match '/foobar' but only '/foo' and its subdirectories (if it is a directory).
* libsandbox: do not abort with a long name to opendirMart Raudsepp2017-09-262-0/+8
| | | | | | | | | | | | | | Add a pre-check for opendir that catches too long name arguments given to opendir, as it would get messed up and abort before it even gets to the open*() syscall (which would handle it correctly), due to opendir going through before_syscall/check_syscall, even though it isn't a true syscall and it getting cut to SB_PATH_MAX inbetween and getting confused somewhere. Test case added by Michał Górny <mgorny@gentoo.org>. Bug: https://bugs.gentoo.org/553092 Signed-off-by: Mart Raudsepp <leio@gentoo.org>
* libsandbox: whitelist renameat/symlinkat as symlink funcsMike Frysinger2017-03-106-0/+46
| | | | | | | These funcs don't deref their path args, so flag them as such. URL: https://bugs.gentoo.org/612202 Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* tests: make all shell scripts executableMike Frysinger2016-03-299-0/+0
| | | | Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* libsandbox: use ptrace on apps that interpose their own allocatorMike Frysinger2016-02-164-0/+52
| | | | | | | | | | | | | | | | | | | | | | | If an app installs its own memory allocator by overriding the internal glibc symbols, then we can easily hit a loop that cannot be broken: the dlsym functions can attempt to allocate memory, and sandbox relies on them to find the "real" functions. So when someone calls a symbol that the sandbox protects, we call dlsym, and that calls malloc, which calls back into the app, and their allocator might use another symbol such as open ... which is protected by the sandbox. So we hit the loop like: -> open -> libsandbox:open -> dlsym -> malloc -> open -> libsandbox:open -> dlsym -> malloc -> ... Change the exec checking logic to scan the ELF instead. If it exports these glibc symbols, then we have to assume it can trigger a loop, so scrub the sandbox environment to prevent us from being loaded. Then we use the out-of-process tracer (i.e. ptrace). This should generally be as robust anyways ... if it's not, that's a bug we want to fix as this is the same code used for static apps. URL: http://crbug.com/586444 Reported-by: Ryo Hashimoto <hashimoto@chromium.org> Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* tests: add test for overriding mmapMike Frysinger2016-02-166-0/+56
| | | | | | URL: http://bugs.gentoo.org/290249 Reported-by: Diego E. Pettenò <flameeyes@gentoo.org> Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* tests: check errno with more static testsMike Frysinger2015-12-203-3/+3
| | | | | | | This verifies the error code setting with ptrace logic -- if the ptrace code is broken, the errno will often be ENOSYS instead of EPERM. Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* libsandbox: do not unnecessarily dereference symlinksMike Frysinger2015-09-282-0/+31
| | | | | | | | | | | When the target uses a func that operates on a symlink, we should not dereference that symlink when trying to validate the call. It's both a waste of time and it subtly breaks code that checks atime updates. The act of reading symlinks is enough to cause their atime to change. URL: https://bugs.gentoo.org/415475 Reported-by: Marien Zwart <marienz@gentoo.org> Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* tests: add basic parsing of timespec fieldsMike Frysinger2015-09-283-2/+26
| | | | Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* tests: avoid overflowing exit codesMike Frysinger2015-09-201-1/+1
| | | | Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* libsandbox: do not abort when the target uses bad pointersMike Frysinger2015-09-205-1/+74
| | | | | | | | | | | If the target passes a bad pointer to the kernel, then trying to extract the data via ptrace will also throw an error. The tracing code should not abort though as there's no valid address to check, and kernel itself will return an error for us. Simply return and move on. URL: https://bugs.gentoo.org/560396 Reported-by: Jeroen Roovers <jer@gentoo.org> Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* libsandbox: fix handling of dangling symlinksMike Frysinger2015-09-203-0/+46
| | | | | | | | | | | | | | | | | Make sure we properly check the target of symlinks even when the target does not exist. This caused problems in two ways: (1) It allowed code to bypass checks by writing through a symlink that was in a good location but pointed to a bad (non-existent) location. (2) It caused code to be wrongly rejected when it tried writing to a symlink in a bad location but pointed to a good location. In order to get this behavior, we need to use the new gnulib helpers added in the previous commit. They include functions which can look up the targets of symlinks even when the final path doesn't exist. URL: https://bugs.gentoo.org/540828 Reported-by: Rick Farina <zerochaos@gentoo.org> Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* tests: add dynamic/static categoriesMike Frysinger2015-09-111-0/+4
| | | | | | | This way we can quickly execute the tests that run dynamic or static binaries. We leave scripts out as they're a bit of a special case. Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* tests: make sure we ignore stty exit statusMike Frysinger2015-09-111-1/+4
| | | | | | | | | | | | | | | If the stty step fails (for any reason really), the main testrunner will abort with a weird error message: $ make check ... /bin/sh './testsuite' AUTOTEST_PATH='src:tests' --jobs=`getconf _NPROCESSORS_ONLN || echo 1` testsuite: error: invalid content: atlocal ... Make sure we ignore stty's exit status, and we put a final comment/$? reset at the end of the script. Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* sandbox: use a non-shell var for the message pathMike Frysinger2013-03-031-1/+1
| | | | | | | | | | | | | | | | | By allowing the SANDBOX_MESSAGE_PATH var to be stored in the shell environment and then modified on the fly, we run into a fun edge case with the PM. When a phase has finished running, it saves the current environment. When the next phase runs, it loads the env from the previous run. Since the message path var can contain a pid, the previous run will no longer be valid. Since we want this to simply be a way for the active sandbox to pass information to the active libsandbox.so's, there's no need to use an env var that the shell can save/reload. As such, use a variable name that the shell will skip. Non-shell programs have no problem with this. Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* libsandbox: preserve more SANDBOX env varsMike Frysinger2013-02-234-2/+24
| | | | | | | | | | | | | | | | | While we took pains to preserve the LD_PRELOAD setting, this doesn't help us too much in practice. If a process is going out of its way to blow away LD_PRELOAD, chances are good it's blowing away all vars it doesn't know about. That means all of our SANDBOX_XXX settings. Since a preloaded libsandbox.so is useless w/out its SANDBOX_XXX env vars, make sure we preserve those as well. These changes also imply some behavioral differences from older versions. Previously, you could `unset` a sandbox var in order to disable it. That no longer works. If you wish to disable things, you have to explicitly set it to "". Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* libsandbox: clean up open file handles in parent tracing processMike Frysinger2012-06-235-0/+121
| | | | | | | | | | | | | | Currently, if a non-static app sets up a pipe (with cloexec enabled) and executes a static app, the handle to that pipe is left open in the parent process. This causes trouble when the parent is waiting for that to be closed immediately. Since none of the fds in the forked parent process matter to us, we can just go ahead and clean up all fds before we start tracing the child. URL: http://bugs.gentoo.org/364877 Reported-by: Victor Stinner <victor.stinner@haypocalc.com> Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* add a new message env varMike Frysinger2013-02-241-0/+1
| | | | | | | | | | | | | | | | | | | | | | | | This is used whenever sandbox wants to display an informational message. For example, early notification of a path violation, or debugging output. We can't just pop open an fd and pass that around as apps consider that leakage and will often break assumptions in terms of free fds. Or apps that start up and cleanse all of their open fds. So instead, we just pass around an env var that holds the full path to the file we want will write to. Since these messages are infrequent (compared to overall runtime), opening/writing/closing the path every time is fine. This also avoids all the problems associated with using external portage helpers for writing messages. A follow up commit will take care of the situation where apps (such as scons) attempt to also cleanse the env before forking. URL: http://bugs.gentoo.org/278761 URL: http://bugs.gentoo.org/431638 Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* libsandbox: reject "" paths with *at funcs before checking the dirfdMike Frysinger2012-12-246-0/+26
| | | | | | | | | | When it comes to processing errors, an empty path is checked before an invalid dirfd. Make sure sandbox matches that behavior for the random testsuites out there that look for this. URL: https://bugs.gentoo.org/346929 Reported-by: Marien Zwart <marienz@gentoo.org> Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* libsandbox: handle open(O_NOFOLLOW)Mike Frysinger2012-12-242-0/+11
| | | | | | | | | We don't check for O_NOFOLLOW in the open wrappers, so we end up returning the wrong error when operating on broken symlinks. URL: https://bugs.gentoo.org/413441 Reported-by: Marien Zwart <marienz@gentoo.org> Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* tests: expand usage stringsMike Frysinger2012-12-2429-36/+70
| | | | | | | Makes it easier to quickly figure out how to run a helper test without having to resort to existing usage or the code itself. Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* tests: disable tostop on controlling terminalMike Frysinger2012-11-261-0/+3
| | | | | | | If this flag is set, then the tests get all hung up. Clear it in case someone has it active on their terminal. Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* libsandbox: add x32 ABI supportv2.6Mike Frysinger2012-07-031-2/+2
| | | | | | | | | | | We can trace x32 when the host is x86_64 or x32, but x32 cannot trace x86_64 due to limitations in the kernel interface -- all pointers get truncated to 32bits. We'll have to add external ptrace helpers in the future to make this work, but for now, we'll just let x86_64 code run unchecked :(. URL: https://bugs.gentoo.org/394179 Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* precompile headers.h to speed up build slightlyMike Frysinger2012-07-031-1/+1
| | | | | | | Since all system headers are included by way of headers.h, we can pre-compile this to speed up the build up a bit. Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* sb_printf: expand feature set slightly and fix testingMike Frysinger2012-07-023-2/+16
| | | | | | | | | | This adds support for signed ll, unsigned z, l, and ll, hex l, and ll, ignores the # for hex output since this is what we do implicitly already. As for testing, looks like during the autogeneration of testsuite.list.at, the sb_printf test was lost. Restore it so it gets run again. Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* tests: enable color/parallel by defaultMike Frysinger2012-06-242-0/+3
| | | | Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* significantly overhaul output helpersMike Frysinger2012-06-231-0/+4
| | | | | | | | | | | | | | | | | | | | There are a few major points we want to hit here: - have all output from libsandbox go through portage helpers when we are in the portage environment so that output is properly logged - convert SB_E{info,warn,error} to sb_e{info,warn,error} to match style of other functions and cut down on confusion - move all abort/output helpers to libsbutil so it can be used in all source trees and not just by libsandbox - migrate all abort points to the centralized sb_ebort helper Unfortunately, it's not terribly easy to untangle these into separate patches, but hopefully this shouldn't be too messy as much of it is mechanical: move funcs between files, and change the name of funcs that get called. URL: http://bugs.gentoo.org/278761 Reported-by: Mounir Lamouri <volkmar@gentoo.org> Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* tests: always bubble up exit code regardless of log existenceMike Frysinger2012-06-2318-38/+18
| | | | | | | | | | | | To simplify testing, do not mung exit(0) to exit(1) just because the log file exists. In many of our tests, we will be doing things to generate a log file, but we explicitly test for exit values ourselves. This is also needed to make log file handling more resilient where we get the name at startup, but don't allow live env changes after that. The changing of the log name to sb.log on the fly no longer works. Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* tests: move disabling of sandbox verbose to common initMike Frysinger2012-06-235-4/+2
| | | | | | | | Since none of our tests care about the verbose output, move the disable to a common location so we don't have to do it on a more fine grained basis. Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* sandbox: drop beep supportMike Frysinger2012-06-231-1/+0
| | | | | | | | Almost no one has beep support turned on anymore, and ebeep in the main tree has been deprecated (meaning it wasn't found useful while building packages). So punt support for it from sandbox too. Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* tests: add static unlink testsMike Frysinger2012-03-064-0/+13
| | | | | | | | | Make sure that when we trace static apps, their bad syscalls don't get a chance to actually complete. URL: http://bugs.gentoo.org/406543 Reported-by: Marijn Schouten <hkbst@gentoo.org> Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* tests: note that testsuite.list.at is generatedMike Frysinger2012-03-051-1/+2
| | | | Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* tests: make `./testsuite #` workMike Frysinger2012-03-051-0/+4
| | | | | | | Without this, the test will fail because it can't find sandbox.sh which lives in the src dir. Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* tests: disable non-POSIX EBADFDMike Frysinger2011-07-081-1/+1
| | | | | | | POSIX doesn't specify EBADFD, and EBADF should cover us, and we don't really need it, so disable for now. Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* libsandbox: catch mktemp related funcsMike Frysinger2011-07-0429-1/+240
| | | | | | URL: http://bugs.gentoo.org/374059 Reported-by: Nick Bowler <nbowler@draconx.ca> Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* tests: autogenerate testsuite.atMike Frysinger2011-07-042-37/+6
| | | | Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* tests: fix flags parsing in open64Mike Frysinger2011-07-041-2/+1
| | | | | | The open test got this fix a while ago, but open64 was missed. Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* tests: use %#x instead of 0x%xMike Frysinger2011-07-043-3/+3
| | | | Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* tests: start futimesat tests based on utimensat onesMike Frysinger2011-01-015-0/+26
| | | | Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* tests: make defaults for fds sanerMike Frysinger2011-01-011-3/+16
| | | | | | | | | Let the default mode for files be 0777 rather than 0 so that the default creation of files actually works. Also make the flags part of a dirfd filename actually optional. Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* tests: add a bunch more feasible errno valuesMike Frysinger2011-01-011-0/+23
| | | | Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* tests: unify duplicated name/value lookup codeMike Frysinger2010-11-262-18/+22
| | | | Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* libsandbox: fix utimensat regression with NULL filenameMike Frysinger2010-11-262-0/+7
| | | | | | | | | | | The previous commit (libsandbox: handle dirfd in mkdir/open/unlink *at prechecks) unified some path checks while unifying the dirfd code, but prevented valid NULL paths from also being handled. Make sure we still handle that behavior, and add a test for it to prevent future regressions. URL: http://bugs.gentoo.org/346815 Reported-by: Jake Todd <jaketodd422@gmail.com> Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* tests: allow dirfds to specify mode tooMike Frysinger2010-11-261-17/+24
| | | | Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* tests: add another unlinkat test for long pathsMike Frysinger2010-11-232-0/+11
| | | | Signed-off-by: Mike Frysinger <vapier@gentoo.org>