Commit Graph

2279 Commits

Author SHA1 Message Date
Peter Collingbourne 6d4a7d3723 [nolibc] Begin moving sanitizer_common's libc-dependent code to a separate library
Introduce a new object library, RTSanitizerCommonLibc, which will contain
the subset of sanitizer_common with libc dependencies. RTSanitizerCommon
contains the remainder of sanitizer_common, and is intended to have no
libc dependencies. Begin moving code to RTSanitizerCommonLibc, starting
with sanitizer_common.cc, whose libc-dependent portion is moved to
sanitizer_common_libcdep.cc, the first member of the new library.

This split affects the CMake build only.  The makefile build continues
to produce the full sanitizer_common library.

llvm-svn: 182118
2013-05-17 16:17:19 +00:00
Evgeniy Stepanov bec53faad0 [sanitizer] Fix build with Android NDK r8e.
llvm-svn: 182110
2013-05-17 15:22:32 +00:00
Alexander Potapenko dd3c583211 [ASan] Make the regression test for time(NULL) Linux-specific, as it crashes on Darwin (and the problem used to occur on Linux).
Do not use zero-based shadow for the time() test.

llvm-svn: 182107
2013-05-17 15:11:26 +00:00
Evgeniy Stepanov 44ff0eb4b9 [sanitizer] Fix gcc build.
llvm-svn: 182094
2013-05-17 13:30:48 +00:00
Evgeniy Stepanov ff6728f2f8 [msan] Unpoison dlpi_name in dl_iterate_phdr interceptor.
llvm-svn: 182093
2013-05-17 12:51:13 +00:00
Dmitry Vyukov 9ff1023816 tsan: remove "printing report from signal handler. Can crash or hang" output
it was added to debug mysterious hangs, but it does not seem to happen anymore
now it only clutter up output

llvm-svn: 182089
2013-05-17 12:05:44 +00:00
Dmitry Vyukov c9af818515 tsan: introduce recursive mutex lock/unlock java interface
this is required to handle Object.Wait()

llvm-svn: 182088
2013-05-17 12:03:46 +00:00
Dmitry Vyukov 69c544d857 tsan: comment out debug output in test
llvm-svn: 182087
2013-05-17 11:54:37 +00:00
Alexander Potapenko 00b44ba778 [ASan] Fix allow_user_segv.cc on Darwin (32-bit build required a SIGBUS handler instead of the SIGSEGV one)
llvm-svn: 182080
2013-05-17 08:08:50 +00:00
Alexander Potapenko 12d09f3a3d [ASan] Fix init-order-dlopen.cc to pass under Mac OS.
llvm-svn: 182010
2013-05-16 14:51:01 +00:00
Alexander Potapenko 3917fe0d94 [ASan] Fix interface_symbols_darwin to pass a correct flag to nm.
llvm-svn: 182008
2013-05-16 13:49:12 +00:00
Kostya Serebryany 21cbaee6a9 [sanitizer] fix gcc build
llvm-svn: 182006
2013-05-16 13:24:31 +00:00
Kostya Serebryany aaf7275677 [asan] symbolize when possible, even if ASAN_SYMBOLIZER_PATH is not provided. On Linux this will use dl_iterate_phdr instead of /proc/self/maps, even if the symbolizer is not installed
llvm-svn: 182005
2013-05-16 13:04:29 +00:00
Evgeniy Stepanov 87c6ec7cfe [msan] Intercept dl_iterate_phdr.
llvm-svn: 182004
2013-05-16 13:00:25 +00:00
Sergey Matveev efb6172cb3 [sanitizer] Fix the region overflow condition in SanitizerAllocator64::PopulateFreeList().
llvm-svn: 182002
2013-05-16 12:58:34 +00:00
Kostya Serebryany 19eb46f4ca [sanitizer] fix a gcc warning
llvm-svn: 181992
2013-05-16 08:03:26 +00:00
Kostya Serebryany b7ca536888 [asan] fix ShadowGapTest on PowerPC64
llvm-svn: 181991
2013-05-16 07:54:28 +00:00
Kostya Serebryany a551aaa944 [sanitizer] fix the GetBlockBegin overflow bug while preserving the performance optimization (use 32-bit division when possible). Improve the benchmarks that checks for performance of GetBlockBegin/GetMetaData
llvm-svn: 181989
2013-05-16 07:11:16 +00:00
Kostya Serebryany ad9971d793 [sanitizer] added a test for a bug in allocator discovered by Sergey Matveev (uint32 overflow in GetBlockBegin)
llvm-svn: 181984
2013-05-16 05:22:50 +00:00
Bill Wendling c054086630 Try to improve performance by using a read/write buffer instead of I/O.
The calls to fwrite/fread can be very expensive. GCC avoids this by using a
buffer to read and write from the file, thus limiting the number of fwrite/fread
calls.

<rdar://problem/13466086>

llvm-svn: 181924
2013-05-15 21:31:22 +00:00
Kostya Serebryany 2b824f3575 [asan] disable BuiltinLongJmpTest on PowerPC
llvm-svn: 181890
2013-05-15 15:01:14 +00:00
Kostya Serebryany 2b42716213 [asan] fix powerpc build and one test; fix lint
llvm-svn: 181881
2013-05-15 12:36:29 +00:00
Sergey Matveev 2387687788 [sanitizer] Fix Android build.
llvm-svn: 181796
2013-05-14 15:22:39 +00:00
Sergey Matveev 61101ba419 [sanitizer] Fixed a bug in GetListOfModules.
llvm-svn: 181793
2013-05-14 14:48:58 +00:00
Sergey Matveev 4b603e5c50 [sanitizer] Filtering in GetListOfModules.
llvm-svn: 181791
2013-05-14 14:04:06 +00:00
Evgeniy Stepanov f1efb256f6 Add sanitizer syscall hooks to the tool's export lists.
llvm-svn: 181790
2013-05-14 13:55:21 +00:00
Sergey Matveev af05543c86 [sanitizer] LibraryNameIs in sanitizer_linux
llvm-svn: 181787
2013-05-14 13:24:46 +00:00
Sergey Matveev 14b56186ab [sanitizer] Minor fixes in sanitizer_common tests.
llvm-svn: 181786
2013-05-14 13:19:33 +00:00
Kostya Serebryany 9d987e34f7 [tsan] fix gcc build with -Wall
llvm-svn: 181763
2013-05-14 08:26:55 +00:00
Sylvestre Ledru 7b4baa91ce Fix a typo in the comment
llvm-svn: 181708
2013-05-13 14:53:09 +00:00
Sergey Matveev d9da20f56f [sanitizer] Generic sorting in sanitizer_common.
llvm-svn: 181698
2013-05-13 11:58:48 +00:00
Sergey Matveev 37432e815e [sanitizer] Fix StopTheWorld build on non-Android ARM.
Original patch by Abdoulaye Walsimou Gaye.

llvm-svn: 181697
2013-05-13 10:35:20 +00:00
Peter Collingbourne 53caf563a9 Try to fix Windows build too
llvm-svn: 181457
2013-05-08 18:15:01 +00:00
Peter Collingbourne 8e110ce57c Try to fix the mac buildbots
llvm-svn: 181440
2013-05-08 15:07:12 +00:00
Peter Collingbourne 6f4be19b57 [nolibc] Change internal syscall API to remove reliance on libc's errno.
This change moves to a model where the error value of a system call is
potentially contained in the return value itself rather than being
implicit in errno.  The helper function internal_iserror can be used
to extract the error value from a return value.  On platforms other
than Linux/x86_64 this still uses errno, but other platforms are free
to port their error handling to this new model.

Differential Revision: http://llvm-reviews.chandlerc.com/D756

llvm-svn: 181436
2013-05-08 14:43:49 +00:00
Sergey Matveev ffe38d267c Quick fix for ASan test build on Android.
llvm-svn: 181429
2013-05-08 14:11:12 +00:00
Sergey Matveev 0d8cc4a2d2 [sanitizer] Fix build breakage in Go TSan.
llvm-svn: 181428
2013-05-08 13:59:30 +00:00
Sergey Matveev af179b8d63 [sanitizer] Move GetStackTrace from ASan to sanitizer_common.
llvm-svn: 181424
2013-05-08 12:45:55 +00:00
Peter Collingbourne 5829a98c48 [nolibc] Disable the GCC stack protector when building sanitizer runtimes.
This is the first in a sequence of changes designed to eliminate the
libc dependency in sanitizer_common.  The main motivation for these
changes is to be able to provide an alternative for the current
interceptor-based technique for instrumenting functions in libc.
In this new technique, we compile libc with instrumentation.  This has
the potential advantages of being more accurate than interception and
reducing the amount of custom code required for each libc function.
As a side effect of this, we cannot depend on libc in the sanitizer
runtime due to mutual dependency issues.

This change disables the GCC stack protector, which introduces a libc
dependency and is enabled by default in Ubuntu.

Differential Revision: http://llvm-reviews.chandlerc.com/D755

llvm-svn: 181422
2013-05-08 10:39:05 +00:00
Sergey Matveev 3704453368 [sanitizer] Fixed InitTlsSize for Android.
llvm-svn: 181330
2013-05-07 16:29:26 +00:00
Sergey Matveev 88d0c664ef [sanitizer] Fixed includes in sanitizer_linux.cc for Android.
llvm-svn: 181327
2013-05-07 15:38:53 +00:00
Sergey Matveev 954c6ef10f [sanitizer] Move GetThreadStackAndTls from TSan to sanitizer_common.
Move this function to sanitizer_common because LSan uses it too. Also, fix a bug
where the TLS range reported for main thread was off by the size of the thread
descriptor from libc (TSan doesn't care much, but for LSan it's critical).

llvm-svn: 181322
2013-05-07 14:41:43 +00:00
Evgeniy Stepanov 517118e6b4 [sanitizer] Sanitizer __internal_*stat interface.
With this change, __internal_*stat always expect a "struct stat *" argument.
This avoids stat/stat64 caller-side confusion (sanitizer_common tests already
made this mistake), and allows the use of __internal_fstat() as a drop-in
replacement for libc's fstat().

llvm-svn: 181311
2013-05-07 12:47:04 +00:00
Sergey Matveev 6eff11e714 [msan] Common flags in MSan.
llvm-svn: 181194
2013-05-06 13:15:14 +00:00
Sergey Matveev 0c8ed9ce44 [asan] Common flags in ASan.
Some flags that are common to ASan/MSan/TSan/LSan have been moved to
sanitizer_common.

llvm-svn: 181193
2013-05-06 11:27:58 +00:00
Bill Wendling e575be4c0e I was wrong in my testing.
There isn't a speedup when using unbuffered I/O. It slows it down in fact.

llvm-svn: 181060
2013-05-03 22:06:41 +00:00
Bill Wendling 6ce98c49f4 Use unbuffered I/O. This reduces the runtime by about half. Our implementation is now only about 5 times slower than gcc's.
llvm-svn: 180980
2013-05-03 00:25:49 +00:00
Bob Wilson b1835b7947 Rewrite sw_vers makefile check to avoid error message on Linux.
When building compiler-rt on Linux, the "which sw_vers" check fails and
writes an error message into the build log. Apparently on Solaris "which"
writes the error message to stdout, so that the current test won't even
work properly. As far as I know sw_vers always lives in /usr/bin, so just
check for it there instead of using "which".

llvm-svn: 180792
2013-04-30 17:06:40 +00:00
Dmitry Vyukov 93b2cba03b asan: fix windows build
llvm-svn: 180788
2013-04-30 13:30:29 +00:00
Dmitry Vyukov db7d9e5db4 tsan: comment out debug output in test
llvm-svn: 180787
2013-04-30 13:09:31 +00:00