Summary:
https://reviews.llvm.org/D41121 broke the FreeBSD build due to that type not
being defined on FreeBSD. As far as I can tell, it is an int, but I do not have
a way to test the change.
Reviewers: alekseyshl, kparzysz
Reviewed By: kparzysz
Subscribers: kparzysz, emaste, kubamracek, krytarowski, #sanitizers, llvm-commits
Differential Revision: https://reviews.llvm.org/D41325
llvm-svn: 320931
Summary:
Hook on -DKERNEL_USE (which is also used in lib/builtins) to not import
strlen and not rely on write() being implemented with the stderr on fd 2.
With this, the only requirements to use this library are:
- "Good enough" std::atomic<void*> and std::atomic<int>
- abort() being implemented
- ubsan_message(const char*) being implemented
Reviewers: eugenis
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D39791
llvm-svn: 320831
Summary:
The initial implementation used an ASan like Chunk class that was deriving from
a Header class. Due to potential races, we ended up working with local copies
of the Header and never using the parent class fields. ScudoChunk was never
constructed but cast, and we were using `this` as the pointer needed for our
computations. This was meh.
So we refactored ScudoChunk to be now a series of static functions within the
namespace `__scudo::Chunk` that take a "user" pointer as first parameter (former
`this`). A compiled binary doesn't really change, but the code is more sensible.
Clang tends to inline all those small function (in -O2), but GCC left a few not
inlined, so we add the `INLINE` keyword to all.
Since we don't have `ScudoChunk` pointers anymore, a few variables were renamed
here and there to introduce a clearer distinction between a user pointer
(usually `Ptr`) and a backend pointer (`BackendPtr`).
Reviewers: alekseyshl, flowerhack
Reviewed By: alekseyshl
Subscribers: #sanitizers, llvm-commits
Differential Revision: https://reviews.llvm.org/D41200
llvm-svn: 320745
Summary:
This is the first mostly working version of the Sanitizer port to 32-bit Solaris/x86.
It is currently based on Solaris 11.4 Beta.
This part was initially developed inside libsanitizer in the GCC tree and should apply to
both. Subsequent parts will address changes to clang, the compiler-rt build system
and testsuite.
I'm not yet sure what the right patch granularity is: if it's profitable to split the patch
up, I'd like to get guidance on how to do so.
Most of the changes are probably straightforward with a few exceptions:
* The Solaris syscall interface isn't stable, undocumented and can change within an
OS release. The stable interface is the libc interface, which I'm using here, if possible
using the internal _-prefixed names.
* While the patch primarily target 32-bit x86, I've left a few sparc changes in. They
cannot currently be used with clang due to a backend limitation, but have worked
fine inside the gcc tree.
* Some functions (e.g. largefile versions of functions like open64) only exist in 32-bit
Solaris, so I've introduced a separate SANITIZER_SOLARIS32 to check for that.
The patch (with the subsequent ones to be submitted shortly) was tested
on i386-pc-solaris2.11. Only a few failures remain, some of them analyzed, some
still TBD:
AddressSanitizer-i386-sunos :: TestCases/Posix/concurrent_overflow.cc
AddressSanitizer-i386-sunos :: TestCases/init-order-atexit.cc
AddressSanitizer-i386-sunos :: TestCases/log-path_test.cc
AddressSanitizer-i386-sunos :: TestCases/malloc-no-intercept.c
AddressSanitizer-i386-sunos-dynamic :: TestCases/Posix/concurrent_overflow.cc
AddressSanitizer-i386-sunos-dynamic :: TestCases/Posix/start-deactivated.cc
AddressSanitizer-i386-sunos-dynamic :: TestCases/default_options.cc
AddressSanitizer-i386-sunos-dynamic :: TestCases/init-order-atexit.cc
AddressSanitizer-i386-sunos-dynamic :: TestCases/log-path_test.cc
AddressSanitizer-i386-sunos-dynamic :: TestCases/malloc-no-intercept.c
SanitizerCommon-Unit :: ./Sanitizer-i386-Test/MemoryMappingLayout.DumpListOfModules
SanitizerCommon-Unit :: ./Sanitizer-i386-Test/SanitizerCommon.PthreadDestructorIterations
Maybe this is good enough the get the ball rolling.
Reviewers: kcc, alekseyshl
Reviewed By: alekseyshl
Subscribers: srhines, jyknight, kubamracek, krytarowski, fedor.sergeev, llvm-commits, #sanitizers
Tags: #sanitizers
Differential Revision: https://reviews.llvm.org/D40898
llvm-svn: 320740
This includes a few nice bits of refactoring (e.g splitting out the
exclusive locking code into a common utility).
Hopefully the Windows support is fixed now.
Patch by Rainer Orth!
Differential Revision: https://reviews.llvm.org/D40944
llvm-svn: 320731
This includes a few nice bits of refactoring (e.g splitting out the
exclusive locking code into a common utility).
Patch by Rainer Orth!
Differential Revision: https://reviews.llvm.org/D40944
llvm-svn: 320726
Summary:
Before this change, XRay would conservatively patch sections of the code
one sled at a time. Upon testing/profiling, this turns out to take an
inordinate amount of time and cycles. For an instrumented clang binary,
the cycles spent both in the patching/unpatching routine constituted 4%
of the cycles -- this didn't count the time spent in the kernel while
performing the mprotect calls in quick succession.
With this change, we're coalescing the number of calls to mprotect from
being linear to the number of instrumentation points, to now being a
lower constant when patching all the sleds through `__xray_patch()` or
`__xray_unpatch()`. In the case of calling `__xray_patch_function()` or
`__xray_unpatch_function()` we're now doing an mprotect call once for
all the sleds for that function (reduction of at least 2x calls to
mprotect).
Reviewers: kpw, eizan
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D41153
llvm-svn: 320664
Summary:
Changes:
- Don't attempt to intercept GLIBC specific functions like __strtol_internal.
This is required to stop intercepting it as we leak dlerror(3) for dlsym(3)
that cannot manage to find a symbol.
- Correct interception of fstatat(2).
- Don't run a test for fgetgrent_r() that is missing on NetBSD.
- Correct link_map location (offset) in Obj_Entry on x86_64 and i386.
- Stop intercepting getpshared-like functions in pthread(3). This is feature
is not enabled by default on NetBSD as it's unfinished.
- Switch intercepting from UTMP to UTMPX functions.
Sponsored by <The NetBSD Foundation>
Reviewers: joerg, vitalybuka, eugenis, kcc
Reviewed By: vitalybuka
Subscribers: llvm-commits, srhines, kubamracek, #sanitizers
Tags: #sanitizers
Differential Revision: https://reviews.llvm.org/D41053
llvm-svn: 320663
This saves ~2 MB of dirty memory footprint. Can be a big deal on mobile devices especially when running multiple processes with ASan.
Differential Revision: https://reviews.llvm.org/D40627
llvm-svn: 320660
MADV_DONTNEED on Linux actually mark the pages as free to be overwritten with zeroes, but on Darwin and BSD, it's just an advisory flag (the OS cannot discard the content). We should use MADV_FREE on Darwin and BSD.
Differential Revision: https://reviews.llvm.org/D40666
llvm-svn: 320659
Summary:
Add dummy call of LLVMFuzzerTestOneInput to afl_driver before it starts
executing on actual inputs. Do this so that first time initialization
performed by LLVMFuzzerTestOneInput is not considered code covered by
a particular input.
Patch By: metzman
Reviewers: kcc, morehouse
Reviewed By: kcc
Subscribers: llvm-commits, Sanitizers
Differential Revision: https://reviews.llvm.org/D41193
llvm-svn: 320643
Summary:
The first and only function to start with allows to set the soft or hard RSS
limit at runtime. Add associated tests.
Reviewers: alekseyshl
Reviewed By: alekseyshl
Subscribers: mgorny, #sanitizers, llvm-commits
Differential Revision: https://reviews.llvm.org/D41128
llvm-svn: 320611
Summary:
See D40657 & D40679 for previous versions of this patch & description.
A couple of things were fixed here to have it not break some bots.
Weak symbols can't be used with `SANITIZER_GO` so the previous version was
breakin TsanGo. I set up some additional local tests and those pass now.
I changed the workaround for the glibc vDSO issue: `__progname` is initialized
after the vDSO and is actually public and of known type, unlike
`__vdso_clock_gettime`. This works better, and with all compilers.
The rest is the same.
Reviewers: alekseyshl
Reviewed By: alekseyshl
Subscribers: srhines, kubamracek, krytarowski, llvm-commits, #sanitizers
Differential Revision: https://reviews.llvm.org/D41121
llvm-svn: 320594
Summary:
getScudoChunk function is implicitly inlined for optimized builds on
clang, but not on gcc. It's a small enough function that it seems
sensible enough to just inline it by default.
Reviewers: cryptoad, alekseyshl
Reviewed By: cryptoad
Differential Revision: https://reviews.llvm.org/D41138
llvm-svn: 320592
This change makes XRay print the log file output only when the verbosity
level is higher than 0. It reduces the log spam in the default case when
we want XRay running silently, except when there are actual
fatal/serious errors.
We also update the documentation to show how to get the information
after the change to the default behaviour.
llvm-svn: 320550
Summary: This brings CPU overhead on bzip2 down from 5.5x to 2x.
Reviewers: kcc, alekseyshl
Subscribers: kubamracek, hiraditya, llvm-commits
Differential Revision: https://reviews.llvm.org/D41137
llvm-svn: 320538
This also slightly refactors the code that's checking the directory
presence which allows eliminating one unnecessary variable.
Differential Revision: https://reviews.llvm.org/D40637
llvm-svn: 320446
Summary:
Redo of D40657, which had the initial discussion. The initial code had to move
into a libcdep file, and things had to be shuffled accordingly.
`NanoTime` is a time sink when checking whether or not to release memory to
the OS. While reducing the amount of calls to said function is in the works,
another solution that was found to be beneficial was to use a timing function
that can leverage the vDSO.
We hit a couple of snags along the way, like the fact that the glibc crashes
when clock_gettime is called from a preinit_array, or the fact that
`__vdso_clock_gettime` is mangled (for security purposes) and can't be used
directly, and also that clock_gettime can be intercepted.
The proposed solution takes care of all this as far as I can tell, and
significantly improve performances and some Scudo load tests with memory
reclaiming enabled.
@mcgrathr: please feel free to follow up on
https://reviews.llvm.org/D40657#940857 here. I posted a reply at
https://reviews.llvm.org/D40657#940974.
Reviewers: alekseyshl, krytarowski, flowerhack, mcgrathr, kubamracek
Reviewed By: alekseyshl, krytarowski
Subscribers: #sanitizers, mcgrathr, srhines, llvm-commits, kubamracek
Differential Revision: https://reviews.llvm.org/D40679
llvm-svn: 320409
Summary:
Include <sys/tls.h> for:
- struct tls_tcb - thread control block structure
- __HAVE___LWP_GETTCB_FAST - __lwp_gettcb_fast() is available
- __HAVE___LWP_GETPRIVATE_FAST - __lwp_getprivate_fast() is available
- __HAVE_TLS_VARIANT_I - TLS Variant I for this architecture
- __HAVE_TLS_VARIANT_II - TLS Variant II for this architecture
Rename ThreadSelfSegbase() to ThreadSelfTlsTcb and switch it
to retrieve in a portable way TCB.
Switch ThreadSelf() to retrieve pthread from struct tcb_tls.
Use dl_iterate_phdr() to find out the size of TLS block of
the main program.
Correct the index of the TLS block of the main program
(dlpi_tls_modid); it's 1, not 2.
New NetBSD code is now CPU (NetBSD port) agnostic.
Stop sharing the same code with FreeBSD.
Sponsored by <The NetBSD Foundation>
Reviewers: dvyukov, joerg, vitalybuka
Reviewed By: vitalybuka
Subscribers: emaste, kubamracek, llvm-commits, #sanitizers
Tags: #sanitizers
Differential Revision: https://reviews.llvm.org/D40159
llvm-svn: 320234
Summary:
Reuse the Linux new mapping as it is.
Sponsored by <The NetBSD Foundation>
Reviewers: joerg, vitalybuka, eugenis
Reviewed By: vitalybuka
Subscribers: llvm-commits, #sanitizers
Tags: #sanitizers
Differential Revision: https://reviews.llvm.org/D41023
llvm-svn: 320218
Summary:
This patch adds the initial support for Fuchsia.
- LIBFUZZER_FUCHSIA is added as an OS type in FuzzerDefs.h
- Fuchsia is, by design, not POSIX compliant. However, it does use ELF and
supports common POSIX I/O functions. Thus, FuzzerExtFunctions.h and
FuzzerIO.h are implemented by extending the header guards in
FuzzerExtFunctionsWeak.cpp and FuzzerIOPosix.cpp to include
LIBFUZZER_FUCHSIA.
- The platform-specific portions of FuzzerUtil.h are implemented by
FuzzerUtilFuchsia.cpp, which makes use of exception ports, syscalls, and
the launchpad library.
- The experimental equivalence server is not currently supported, so
FuzzerShmem.h is implemented by stub methods in FuzzerShmemFuchsia.cpp.
Any future implementation will likely involve VMOs.
Tested with ASAN/SanCov on Fuchsia/x86-64 with the canonical toy fuzzer.
Patch By: aarongreen
Reviewers: kcc, morehouse, flowerhack, phosek
Reviewed By: kcc, phosek, Eugene.Zelenko
Subscribers: srhines, mgorny, Eugene.Zelenko
Differential Revision: https://reviews.llvm.org/D40974
llvm-svn: 320210
Summary:
It looks like clang was generating somewhat weird assembly with the current
code. `FromPrimary`, even though `const`, was replaced every time with the code
generated for `size <= SizeClassMap::kMaxSize` instead of using a variable or
register, and `FromPrimary` didn't induce `ClassId != 0` for the compiler, so a
dead branch was generated for `getActuallyAllocatedSize(Ptr, ClassId)` since
it's never called for `ClassId = 0` (Secondary backed allocations) [this one
was more wishful thinking on my side than anything else].
I rearranged the code bit so that the generated assembly is less clunky.
Also changed 2 whitespace inconsistencies that were bothering me.
Reviewers: alekseyshl, flowerhack
Reviewed By: flowerhack
Subscribers: llvm-commits, #sanitizers
Differential Revision: https://reviews.llvm.org/D40976
llvm-svn: 320160
In more recent Linux kernels with 47 bit VMAs the layout of virtual memory
for powerpc64 changed causing the address sanitizer to not work properly. This
patch adds support for 47 bit VMA kernels for powerpc64 and fixes up test
cases.
https://reviews.llvm.org/D40908
There is an associated patch for trunk.
Tested on several 4.x and 3.x kernel releases.
llvm-svn: 320110
Summary: With 3 Dictionary objects, each containing space of ~16k DictionaryEntry objects, the MutationDispatcher object is fairly memory heavy. On platforms with a lower default stack size, this can cause panics in FuzzerUnittest as those tests stack-allocate the MutationDispatcher. This may be especially problematic for platforms that do not (yet) have a way to programmatically change their stack size, aside from link-time flags. In general, it seems more prudent to use the heap for an object of this size.
Reviewers: kcc, morehouse
Reviewed By: kcc
Differential Revision: https://reviews.llvm.org/D40926
llvm-svn: 319988
Summary:
The NetBSD specific implementation of cxa_atexit() does not
preserve the 2nd argument if dso is equal to NULL.
Changes:
- Split paths of handling intercepted __cxa_atexit() and atexit(3).
This affects all supported Operating Systems.
- Add a local stack-like structure to hold the __cxa_atexit() context.
atexit(3) is documented in the C standard as calling callback from the
earliest to the oldest entry. This path also fixes potential ABI
problem of passing an argument to a function from the atexit(3)
callback mechanism.
- Allow usage of global vars with ctors in interceptors.
This allows to use Vector without automatic cleaning up the structures.
This code has been modeled after TSan implementation for the same functions.
Sponsored by <The NetBSD Foundation>
Reviewers: joerg, dvyukov, eugenis, vitalybuka, kcc
Reviewed By: vitalybuka
Subscribers: llvm-commits, #sanitizers
Tags: #sanitizers
Differential Revision: https://reviews.llvm.org/D40714
llvm-svn: 319967
Summary:
NetBSD renames symbols for historical and compat reasons.
Add required symbol renames in sanitizer_common_interceptors.inc:
- gettimeofday -> __gettimeofday50
- getrusage -> __getrusage50
- shmctl -> __shmctl50
Additionally handle sigaction symbol mangling.
Rename the function symbol in the file to SIGACTION_SYMNAME and define
it as __sigaction14 for NetBSD and sigaction for !NetBSD. We cannot use
simple renaming with the proprocessor, as there are valid fields named
sigaction and they must be left intact.
Sponsored by <The NetBSD Foundation>
Reviewers: joerg, eugenis, vitalybuka, dvyukov
Reviewed By: vitalybuka
Subscribers: kubamracek, llvm-commits, #sanitizers
Tags: #sanitizers
Differential Revision: https://reviews.llvm.org/D40766
llvm-svn: 319966
Summary:
Switching the rest of intercepted allocs to InternalAlloc (well, except
__libc_memalign) when current thread is 'in_symbolizer'. Symbolizer
might (and does) use allocation functions other than malloc/calloc/realloc.
posix_memalign is the one actually used, others switched just in case
(since the failure is obscure and not obvious to diagnose).
Reviewers: dvyukov
Subscribers: llvm-commits, kubamracek
Differential Revision: https://reviews.llvm.org/D40877
llvm-svn: 319929
Summary:
Unlike the rest of the sanitizer code, lib/interception uses native macros like __linux__
to check for specific targets instead of the common ones like SANITIZER_LINUX.
When working on the Solaris port of the sanitizers, the current style was found to not
only be inconsistent, but clumsy to use because the canonical way to check for Solaris
is to check for __sun__ && __svr4__ which is a mouthful.
Therefore, this patch switches to use SANITIZER_* macros instead.
Tested on x86_64-pc-linux-gnu.
Reviewers: kcc, vitalybuka
Reviewed By: vitalybuka
Subscribers: #sanitizers, srhines, krytarowski, llvm-commits, fedor.sergeev
Tags: #sanitizers
Differential Revision: https://reviews.llvm.org/D39798
llvm-svn: 319906
Summary:
This wasn't noticed: `RoundUpTo` doesn't produce a constant expression, so the
sizes were not constant either. Enforce them to be static const, replace
`RoundUpTo` by its expression. The compiler can now optimize the associated
computations accordingly.
Also looking at the produced assembly, `PageSize` was fetched multiple times
during `Allocate`, so keep a local value of it. As a result it's fetched once
and kept in a register.
Reviewers: alekseyshl, flowerhack
Reviewed By: alekseyshl
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D40862
llvm-svn: 319903
r319165 introduced a change to CMakeLists.txt for xray where the set of supported
architectures for XRay was iterated over, tested if they could be targeted then
passed to add_compiler_rt_object_libraries. However all targets were passed,
rather than the architecture that was just tested. For cases such as MIPS, where
mips and mips64 are supported, cmake would then test if mips64 could be targetted
resulting in an attempt to produce multiple identical logical target names, falling
afowl of CMP0002.
Reviewers: dberris
Differential Revision: https://reviews.llvm.org/D40890
llvm-svn: 319893
Summary:
NetBSD uses indirection symbol for a set of threading functions.
Add alias to handle __libc_thr_keycreate the same way as pthread_key_create.
Sponsored by <The NetBSD Foundation>
Reviewers: joerg, dvyukov, eugenis, vitalybuka
Reviewed By: vitalybuka
Subscribers: llvm-commits, #sanitizers
Tags: #sanitizers
Differential Revision: https://reviews.llvm.org/D40767
llvm-svn: 319868
Summary:
It was deemed that the salt in the chunk header didn't improve security
significantly (and could actually decrease it). The initial idea was that the
same chunk would different headers on different allocations, allowing for less
predictability. The issue is that gathering the same chunk header with different
salts can give information about the other "secrets" (cookie, pointer), and that
if an attacker leaks a header, they can reuse it anyway for that same chunk
anyway since we don't enforce the salt value.
So we get rid of the salt in the header. This means we also get rid of the
thread local Prng, and that we don't need a global Prng anymore as well. This
makes everything faster.
We reuse those 8 bits to store the `ClassId` of a chunk now (0 for a secondary
based allocation). This way, we get some additional speed gains:
- `ClassId` is computed outside of the locked block;
- `getActuallyAllocatedSize` doesn't need the `GetSizeClass` call;
- same for `deallocatePrimary`;
We add a sanity check at init for this new field (all sanity checks are moved
in their own function, `init` was getting crowded).
Reviewers: alekseyshl, flowerhack
Reviewed By: alekseyshl
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D40796
llvm-svn: 319791