This is done by creating a named shared memory region, unlinking it
and setting up a private (i.e. copy-on-write) mapping of that instead
of a regular anonymous mapping. I've experimented with regular
(sparse) files, but they can not be scaled to the size of MSan shadow
mapping, at least on Linux/X86_64 and ext3 fs.
Controlled by a common flag, decorate_proc_maps, disabled by default.
This patch has a few shortcomings:
* not all mappings are annotated, especially in TSan.
* our handling of memset() of shadow via mmap() puts small anonymous
mappings inside larger named mappings, which looks ugly and can, in
theory, hit the mapping number limit.
llvm-svn: 238621
Also, sized-delete with size_t == unsigned int.
These guys appear when building for Android with gnu-stl.
Fixes a number of ASan tests in that particular configuration.
llvm-svn: 238484
Python may not be /usr/bin/python on some systems. For example, on
FreeBSD it will be /usr/local/bin/python.
Reviewers: samsonov
Differential Revision: http://reviews.llvm.org/D9914
llvm-svn: 238428
CMake pre-3.1 has a bug where a shared library depending on a custom_target
does not really depend on it.
This commit implements a hilariously ugly workaround for this problem, see
code comments for more details.
llvm-svn: 238367
Fix 2 bugs in memory mapping setup:
- the invalid region at offset 0 was not protected because mmap at
address 0 fails with EPERM on most Linux systems. We did not
notice this because the check condition was flipped: the code was
checking that mprotect has failed. And the test that was supposed
to catch this was weakened by the mitigations in the mmap
interceptor.
- when running without origins, the origin shadow range was left
unprotected.
The new test ensures that mmap w/o MAP_FIXED always returns valid
application addresses.
llvm-svn: 238109
Windows does not use AAPCS, but rather AAPCS-VFP, and thus the functions which
are assumed to be AAPCS will cause invalid argument setup. Ensure that the
functions are marked as AAPCS.
llvm-svn: 238056
This reverts part of r237808.
The CMP0057 warnings came from an old development build of CMake that
nobody else has. We don't need the cruft.
llvm-svn: 237819
Fix a trivial instance of CMP0054 that came up on llvmdev.
The other warnings were CMP0057, which is about using the same file as a
MAIN_DEPENDENCY multiple times. The old behavior hasn't been a problem
yet, so I silenced the warning and filed PR23595 to document the issue
if someone cares.
llvm-svn: 237808
Without the --target flag, clang uses the mips64 triple which selects the n64 abi. We need to add --target=mips-linux-gnu, so that clang can select the correct abi for mips32r2.
Reviewers: dsanders, kcc, samsonov
Subscribers: llvm-commits, mohit.bhakkad, jaydeep
Differential Revision: http://reviews.llvm.org/D9249
llvm-svn: 237675
Current code tries to find the dynamic TLS header to the left of the
TLS block without checking that it's not a static TLS allocation.
llvm-svn: 237495
Follow-up to r237161; seems like we can't use aliases, but we
can do better than duplicating the bodies, especially when that
body, after inlining, isn't as small as it looks.
Better approaches welcome. Perhaps the best thing is just to have
an #ifndef __APPLE__ over the GNUEABI names, since they're not used
there.
llvm-svn: 237323
Mostly uninteresting, except:
- in __extendXfYf2, when checking if the number is normal, the old
code relied on the unsignedness of src_rep_t, which is a problem
when sizeof(src_rep_t) < sizeof(int): the result gets promoted to
int, the signedness of which breaks the comparison.
I added an explicit cast; it shouldn't affect other types.
- we can't pass __fp16, so src_t and src_rep_t are the same.
- the gnu_*_ieee symbols are simply duplicated definitions, as aliases
are problematic on mach-o (where only weak aliases are supported;
that's not what we want).
Differential Revision: http://reviews.llvm.org/D9693
llvm-svn: 237161
Mark longjmp tests as XFAIL because longjmp assembly for mips is not yet implemented.
Reviewers: dsanders, dvyukov, samsonov
Subscribers: llvm-commits, mohit.bhakkad, jaydeep
Differential Revision: http://reviews.llvm.org/D9526
llvm-svn: 236847
This test was failing on mips because mips addresses are 40-bit long.
Using regex for address solves this issue.
Reviewers: dsanders, kcc, samsonov
Subscribers: llvm-commits, mohit.bhakkad, jaydeep
Differential Revision: http://reviews.llvm.org/D9516
llvm-svn: 236844
Talking to John and Doug, we concluded that these functions from
stdatomic really do belong here in compiler-rt rather than in libc,
since the compiler owns stdatomic.h and these need to refer to
clang-specific builtins. Nonetheless, I've only added these on darwin
for now - other platforms should probably do the same unless their
libc does implement these functions.
llvm-svn: 236805
Summary:
Add a weak hook to be called from dfsan's custom memcmp.
The primary user will be lib/Fuzzer.
If this works well we'll add more hooks (strcmp, etc).
Test Plan: Will be covered by lib/Fuzzer tests.
Reviewers: pcc
Reviewed By: pcc
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D9541
llvm-svn: 236679
dl_iterate_phdr is somewhat broken in L (see the code for details).
We add runtime OS version detection and fallback to /proc/maps on L or earlier.
This fixes a number of ASan tests on L.
llvm-svn: 236628
Symbol for pc used in ucontext->uc_mcontext was wrong.
Reviewers: dsanders, kcc, samsonov
Subscribers: llvm-commits, mohit.bhakkad, jaydeep
Differential Revision: http://reviews.llvm.org/D9246
llvm-svn: 236572
They are not part of check-all :(
This change adds sized-delete operators to the version list, and disables the
hack that excluded versioned symbols from the dynamic list - this is not an
issue in this case.
llvm-svn: 236559
Summary:
The spec for these functions says that they should return 0 in this case but
this regressed in r234148. That revision essentially delegates the conversion
to the hardware, but that has different behavior on different platforms (e.g.
it is wrong on x86).
Also fix a typo in the name of __fixunsdfti
Test Plan: The existing unit tests now pass
Reviewers: joerg, howard.hinnant
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D9305
llvm-svn: 236319
Fix a couple of new tests that were reverted because they were causing
intermittent test failures since they were writing the same default
"default.profraw" file. Fixed by creating a unique directory and
running tests in that directory.
Patch by Teresa Johnson. Thanks!
llvm-svn: 236168
* Remove __ubsan_default_options, so that test would work on Darwin
* Fix unintentional undefined behavior in the code (missing return)
* Build the test with -fno-sanitize-recover to distinguish expected
failures and expected passes by return code.
llvm-svn: 236152
This partially reverts commit r236056, as the two tests were using the same
names and, when executed in parallel, were having intermittent failures.
llvm-svn: 236101
the profile output from the command line via -fprofile-instr-generate=<path>,
where the specified output path/file will be overridden by the
LLVM_PROFILE_FILE environment variable.
Several changes are made to the runtime to support this:
Add a new interface __llvm_profile_override_default_filename that will
set the profile output filename, but allows LLVM_PROFILE_FILE to override.
This is the interface used by the new option.
Refactor the pid-expansion done for LLVM_PROFILE_FILE into a separate
routine that can be shared by the various filename setting routines
(so that the filename from the option can also use the "%p" syntax).
Move the truncation into setFilename, and only truncate if there is a
new filename specified (to maintain support for appending to the same
profile file in the case of multiple shared objects built with profiling).
Move the handling for a NULL filename passed to __llvm_profile_set_filename and
__llvm_profile_override_default_filename into the new setFilenamePossiblyWithPid
routine. This now correctly resets the output file to default.profraw
instead of NULL.
The handling for a null LLVM_PROFILE_FILE (which should not reset) is done
by caller setFilenameFromEnvironment.
Patch by Teresa Johnson.
llvm-svn: 236055
We incorrectly replaced shadow slots
when the new value is not stronger than the old one.
The bug can lead to false negatives.
The bug was detected by Go race test suite:
https://github.com/golang/go/issues/10589
llvm-svn: 236008
Embed UBSan runtime into TSan and MSan runtimes in the same as we do
in ASan. Extend UBSan test suite to also run tests for these
combinations.
llvm-svn: 235954
For now tsan_cxx and msan_cxx contain only operator new/delete
replacements. In the future, when we add support for running UBSan+TSan
and UBSan+MSan, they will also contain bits ubsan_cxx runtime.
llvm-svn: 235928
Summary:
During the review of http://reviews.llvm.org/D9199 where I had originally
changed the debug_mapping.cc test to accept hexadecimal values, we realized
that SHADOW_SCALE and SHADOW_GRANULARITY ought to be printed as decimal values.
This patch makes that change.
This patch also adds a '0x' prefix to the SHADOW_OFFSET to make it clear that
it is hexadecimal while the other two are decimal.
Reviewers: kcc, timurrrr, samsonov
Reviewed By: timurrrr, samsonov
Subscribers: samsonov, llvm-commits, sagar
Differential Revision: http://reviews.llvm.org/D9224
llvm-svn: 235798
Summary:
Add an interface function which can be used to periodically trigger
leak detection in a long-running process.
NB: The meaning of the kIgnored tag has been changed to allow easy clean-up
between subsequent leak checks. Previously, this tag was applied to explicitly
ignored (i.e. with __lsan_disable() or __lsan_ignore_object()) chunks *and* any
chunks only reachable from those. With this change, it's only applied to
explicitly ignored chunks.
Reviewers: samsonov
Reviewed By: samsonov
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D9159
llvm-svn: 235728
Summary:
This makes their local declarations match their definitions for ARM targets,
where they have a different calling convention.
This really only affects functions that use floating point types (since the
runtime functions use soft-float, and some targets may default to hard-float)
but it seemed good to make it uniform and do the int-only ones too.
Differential Revision: http://reviews.llvm.org/D9062
llvm-svn: 235722
Summary:
On ppc64le Fedora 21, getpwnam_r("no-such-user", ...) returns ENOENT
instead of 0. Tolerate this in the test case.
Reviewers: eugenis
Reviewed By: eugenis
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D9233
llvm-svn: 235654
Summary:
Previously the CHECK directive for SHADOW_SCALE only matched decimal digits
causing it to match '7' on x86_64 instead of the whole value.
This fixes a failure on mips-linux-gnu targets where the leading digit is 'a'.
Reviewers: kcc, sagar, timurrrr
Reviewed By: timurrrr
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D9199
llvm-svn: 235594
float-cast-overflow handler doesn't have source location provided by the
compiler, but we still have *some* source location if we have a
symbolizer.
llvm-svn: 235567
Summary:
ubsan was correctly catching the undefined behaviour but lit's shell was
failing the test anyway because the exit code was non-zero as a result of the
undefined behaviour.
This fixes the test on a mips-linux-gnu target.
Reviewers: samsonov
Reviewed By: samsonov
Subscribers: samsonov, llvm-commits, rsmith, sagar
Differential Revision: http://reviews.llvm.org/D9155
llvm-svn: 235518
Summary:
This will allow us to enable it on the buildbot (and then enable a
sanitizer buildbot).
The problem is that gcc for Mips does not accept -m32/-m64 like clang does.
We therefore need to use the nearest equivalent -mips32r2/-mips64r2. We must
also specify -mabi=64 in the -mips64r2 case since gcc's default ABI has
traditionally been N32 whereas clang's is N64.
Reviewers: sagar
Reviewed By: sagar
Subscribers: llvm-commits, samsonov, kcc, mohit.bhakkad, Anand.Takale, sagar
Differential Revision: http://reviews.llvm.org/D8892
llvm-svn: 235299
Fixes https://llvm.org/bugs/show_bug.cgi?id=23235
If pthread_create is followed by pthread_detach,
the new thread may not acquire synchronize with
the parent thread.
llvm-svn: 235293
As with the other sanitizers, it is desirable to allow ubsan's output to be
redirected to somewhere other than stderr (and into per-process log files).
llvm-svn: 235277
Summary:
When running the compiler-rt testsuite on FreeBSD (it didn't matter which version), I always got the same error result:
FAILED: cd /home/dim/obj/llvm-234092-trunk-freebsd10-amd64-ninja-rel-1/projects/compiler-rt/lib && LLVM_CHECKOUT=/home/dim/src/llvm/trunk SILENT=1 TMPDIR= PYTHON_EXECUTABLE=/usr/local/bin/python2.7 COMPILER_RT=/home/dim/src/llvm/trunk/projects/compiler-rt /home/dim/src/llvm/trunk/projects/compiler-rt/lib/sanitizer_common/scripts/check_lint.sh
The following differences between the implemented custom wrappers and the tests have been found:
--- /tmp/tmp.vPFEefvclf 2015-04-04 17:11:44.996734563 +0200
+++ /tmp/tmp.NbIuvjxKEe 2015-04-04 17:11:45.044734647 +0200
@@ -1,52 +0,0 @@
-calloc
-clock_gettime
-ctime_r
-dfsan_set_write_callback
-dl_iterate_phdr
-dlopen
-fgets
-fstat
-get_current_dir_name
-getcwd
-gethostname
-getpwuid_r
-getrlimit
-getrusage
-gettimeofday
-inet_pton
-localtime_r
-memchr
-memcmp
-memcpy
-memset
-nanosleep
-poll
-pread
-pthread_create
-read
-sched_getaffinity
-select
-sigaction
-sigemptyset
-snprintf
-socketpair
-sprintf
-stat
-strcasecmp
-strchr
-strcmp
-strcpy
-strdup
-strlen
-strncasecmp
-strncmp
-strncpy
-strrchr
-strstr
-strtod
-strtol
-strtoll
-strtoul
-strtoull
-time
-write
After some investigation, it turns out to be due to the regex used for search test_ symbols in lib/dfsan/scripts/check_customer_wrappers.sh:
grep -E "^\\s*test_.*\(\);" ${DFSAN_CUSTOM_TESTS}
This uses \s to search for whitespace at start of line, but support for \s was only introduced in GNU grep v2.5.4-112-gf979ca0, while both FreeBSD and OSX only have grep 2.5.1 (the last version released under GPLv2).
I propose to change \s into [[:space:]], which is the more portable notation.
Test Plan: Run regression tests, and see that they work now. :)
Reviewers: pcc, samsonov, emaste
Reviewed By: emaste
Subscribers: llvm-commits, emaste
Differential Revision: http://reviews.llvm.org/D8832
llvm-svn: 234704
The two subclasses of Symbolizer now only define two virtual functions, PlatformDemangle and PlatformPrepareForSandboxing. Let's make these non-virtual and directly defined by individual platforms.
Reviewed at http://reviews.llvm.org/D8912
llvm-svn: 234690
The patch is generated using clang-tidy misc-use-override check.
This command was used:
tools/clang/tools/extra/clang-tidy/tool/run-clang-tidy.py \
-checks='-*,misc-use-override' -header-filter='llvm|clang' -j=32 -fix \
-format
llvm-svn: 234680
On Windows, we have to know if a memory to be protected is mapped or not.
On POSIX, Mprotect was semantically different from mprotect most people know.
llvm-svn: 234602
Tim Northover