Age | Commit message (Collapse) | Author |
|
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
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@320731 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
runtime to Solaris"
This reverts commit r320726. It looks like flock isn't available on
Windows:
http://lab.llvm.org:8011/builders/sanitizer-windows/builds/21317/steps/build%20compiler-rt/logs/stdio
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@320728 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
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
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@320726 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
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
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@320664 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
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
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@320663 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
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
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@320660 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
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
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@320659 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
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
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@320643 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
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
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@320611 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
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
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@320594 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
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
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@320592 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@320550 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
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
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@320538 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
max_len slower
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@320531 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
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
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@320446 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
Summary: D40679 broke a couple of builds, reverting while investigating.
Reviewers: alekseyshl
Reviewed By: alekseyshl
Subscribers: srhines, kubamracek, krytarowski, llvm-commits, #sanitizers
Differential Revision: https://reviews.llvm.org/D41088
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@320417 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
allocator [redo]
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
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@320409 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
use asan in one test to speed it up
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@320259 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
sometimes exploding the corpus size. This still needs more analysis and tuning
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@320237 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
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
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@320234 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
Summary:
Runtime library for HWASan, initial commit.
Does not randomize tags yet, does not handle stack or globals.
Reviewers: kcc, pcc, alekseyshl
Subscribers: srhines, kubamracek, dberris, mgorny, llvm-commits, krytarowski
Differential Revision: https://reviews.llvm.org/D40935
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@320231 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
Summary:
Switch NetBSD from GNU to the POSIX strerror_r(3) interceptor.
Sponsored by <The NetBSD Foundation>
Reviewers: joerg, eugenis, vitalybuka
Reviewed By: vitalybuka
Subscribers: srhines, kubamracek, llvm-commits, #sanitizers
Tags: #sanitizers
Differential Revision: https://reviews.llvm.org/D41019
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@320221 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
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
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@320218 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
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
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@320210 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
-experimental_len_control, call UpdateFeatureFrequency only if instructed by the flag
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@320205 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
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
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@320160 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@320110 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
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
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@319988 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@319984 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
"Correct atexit(3) support in MSan/NetBSD"
This causes failures on Linux.
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@319981 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
r319875 caused a sign comparison build failure. Explicitly set the
enum's type to be unsigned.
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@319977 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@319975 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
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
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@319967 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
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
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@319966 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
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
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@319929 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
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
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@319906 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
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
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@319903 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
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
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@319893 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@319872 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@319871 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
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
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@319868 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
Summary:
Disable for NetBSD missing functions missing in this OS:
- mempcpy,
- __libc_memalign,
- malloc_usable_size,
- stpcpy,
- gcvt,
- wmempcpy,
- fcvt.
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/D40764
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@319866 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@319844 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
Summary: Fixes https://github.com/google/sanitizers/issues/892
Reviewers: eugenis
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D40853
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@319843 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
"> file" must come before "2>&1" to have redirection occur correctly in
all cases. Fixes a regression on minimize_two_crashes.test.
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@319792 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
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
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@319791 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
Follow-up to D40828.
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@319764 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
Summary:
This change implements the basic mode filtering similar to what we do in
FDR mode. The implementation is slightly simpler in basic-mode filtering
because we have less details to remember, but the idea is the same. At a
high level, we do the following to decide when to filter function call
records:
- We maintain a per-thread "shadow stack" which keeps track of the
XRay instrumented functions we've encountered in a thread's
execution.
- We push an entry onto the stack when we enter an XRay instrumented
function, and note the CPU, TSC, and type of entry (whether we have
payload or not when entering).
- When we encounter an exit event, we determine whether the function
being exited is the same function we've entered recently, was
executing in the same CPU, and the delta of the recent TSC and the
recorded TSC at the top of the stack is less than the equivalent
amount of microseconds we're configured to ignore -- then we un-wind
the record offset an appropriate number of times (so we can
overwrite the records later).
We also support limiting the stack depth of the recorded functions,
so that we don't arbitrarily write deep function call stacks.
Reviewers: eizan, pelikan, kpw, dblaikie
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D40828
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@319762 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
Summary:
This change allows for registration of multiple logging implementations
through a central mechanism in XRay, mapping an implementation to a
"mode". Modes are strings that are used as keys to determine which
implementation to install through a single API. This mechanism allows
users to choose which implementation to install either from the
environment variable 'XRAY_OPTIONS' with the `xray_mode=` flag, or
programmatically using the `__xray_select_mode(...)` function.
Here, we introduce two API functions for the XRay logging:
__xray_log_register_mode(Mode, Impl): Associates an XRayLogImpl to a
string Mode. We can only have one implementation associated with a given
Mode.
__xray_log_select_mode(Mode): Finds the associated Impl for Mode and
installs it as if by calling `__xray_set_log_impl(...)`.
Along with these changes, we also deprecate the xray_naive_log and
xray_fdr_log flags and encourage users to instead use the xray_mode
flag.
Reviewers: kpw, dblaikie, eizan, pelikan
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D40703
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@319759 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
In this case const does nothing but trigger a warning.
git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@319685 91177308-0d34-0410-b5e6-96231b3b80d8
|