llvm-svn: 193501

Differential Revision: http://llvm-reviews.chandlerc.com/D1984

llvm-svn: 193449

This moves away from creating the symbolizer object and initializing the
external symbolizer as separate steps.  Those steps now always take place
together.

Sanitizers with a legacy requirement to specify their own symbolizer path
should use InitSymbolizer to initialize the symbolizer with the desired
path, and GetSymbolizer to access the symbolizer.  Sanitizers with no
such requirement (e.g. UBSan) can use GetOrInitSymbolizer with no need for
initialization.

The symbolizer interface has been made thread-safe (as far as I can
tell) by protecting its member functions with mutexes.

Finally, the symbolizer interface no longer relies on weak externals, the
introduction of which was probably a mistake on my part.

Differential Revision: http://llvm-reviews.chandlerc.com/D1985

llvm-svn: 193448

llvm-svn: 193405

llvm-svn: 193308

llvm-svn: 193245

llvm-svn: 193229

Differential Revision: http://llvm-reviews.chandlerc.com/D1339

llvm-svn: 193060

llvm-svn: 192990

llvm-svn: 192980

llvm-svn: 192979

llvm-svn: 192973

llvm-svn: 192960

llvm-svn: 192956

llvm-svn: 192901

llvm-svn: 192892

Summary:
This fixes a deadlock which happens in lsan
on a large memalign-allocated chunk that resides in lsan's root set.

Reviewers: samsonov, earthdok

Reviewed By: earthdok

CC: llvm-commits

Differential Revision: http://llvm-reviews.chandlerc.com/D1957

llvm-svn: 192885

llvm-svn: 192870

llvm-svn: 192793

The same as with MSan, this test behavior depends on ASAN_OPTIONS, and we've
got a lit_test that covers this.

llvm-svn: 192777

llvm-svn: 192701

llvm-svn: 192685

llvm-svn: 192677

Treat the fake stack as live memory.

llvm-svn: 192593

Call AsanThread::Destroy() from a late-running TSD destructor.
Previously we called it before any user-registered TSD destructors, which caused
false positives in LeakSanitizer.

llvm-svn: 192585

Summary:
Out-of-bound access may touch not-yet allocated or already freed
and recycled from quarantine chunks. We should treat this situation as
a "free-range memory access" and avoid printing any data about that
irrelevant chunk (which may be inconsistent).

This should fix https://code.google.com/p/address-sanitizer/issues/detail?id=183

Reviewers: kcc

Reviewed By: kcc

CC: timurrrr, llvm-commits

Differential Revision: http://llvm-reviews.chandlerc.com/D1893

llvm-svn: 192581

llvm-svn: 192533

llvm-svn: 191943

llvm-svn: 191904

llvm-svn: 191903

[asan] introduce run-time flag uar_stack_size_log to control the size of FakeStack; don't crash when the fake stack is exhausted, move some code to .cc file

llvm-svn: 191510

Since the error is always reported by one of the GCD workers, this will help avoid premature program termination with exit code 0.

llvm-svn: 191294

llvm-svn: 191204

llvm-svn: 191190

llvm-svn: 191189

llvm-svn: 191188

llvm-svn: 191187

[asan] disable the detect_stack_use_after_return run-time flag by default and enable it explicitly in tests. This is done in preparation to enabling the -fsanitize=use-after-return compile-time flag by default when -fsanitize=address is present.

llvm-svn: 191184

That comment is repeated in another function below.

llvm-svn: 191085

llvm-svn: 191083