[ASan] Fix the log_path option to use different log files for parent and child processes after fork().

llvm-svn: 172828

llvm-svn: 172827

CMake: create AddCompilerRT module and implement convenience add_compiler_rt_object_library function

llvm-svn: 172826

llvm-svn: 172825

llvm-svn: 172824

llvm/test/CodeGen/X86/Atomics-64.ll: Tweak for 2nd RUN not to overwrite %t. It sometimes causes spurious failure on lit win32.

Feel free to prune or suppress each output.

llvm-svn: 172823

llvm-svn: 172822

llvm-svn: 172821

clang/test/CodeGen: Suppress a couple of tests on win32. It seems -fsanitize-blacklist doesn't accept DOSish pathnames.

llvm-svn: 172820

This patch prepares being able to test for and fix more problems (see
FIXME in the test for example).

Previously we would output unwrapped lines for preprocessor directives
at the point where we also parsed the hash token. Since often
projections only terminate (and thus output their own unwrapped line)
after peeking at the next token, this would lead to the formatter seeing
the preprocessor directives out-of-order (slightly earlier). To be able
to correctly identify lines to merge, the formatter needs a well-defined
order of unwrapped lines, which this patch introduces.

llvm-svn: 172819

llvm-svn: 172818

llvm-svn: 172817

llvm-svn: 172816

llvm-svn: 172815

Also add a missing include.

llvm-svn: 172814

llvm-svn: 172813

llvm-svn: 172812

llvm-svn: 172811

llvm-svn: 172810

llvm-svn: 172809

llvm-svn: 172808

[asan] fix two off-by-one errors that seem to affect only PowerPC because only there the stack top may be equal to the address space top. Noted by Andreas Schwab in http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55975#c11 . Also make swapcontext interceptor a bit more robust

llvm-svn: 172807

llvm-svn: 172806

llvm-svn: 172805

').' is likely part of a builder pattern statement.
This is based upon a patch developed by Nico Weber. Thank you!

Before:
int foo() {
  return llvm::StringSwitch<Reference::Kind>(name).StartsWith(
      ".eh_frame_hdr", ORDER_EH_FRAMEHDR).StartsWith(
      ".eh_frame", ORDER_EH_FRAME).StartsWith(".init", ORDER_INIT).StartsWith(
      ".fini", ORDER_FINI).StartsWith(".hash", ORDER_HASH).Default(ORDER_TEXT);
}

After:
int foo() {
  return llvm::StringSwitch<Reference::Kind>(name)
         .StartsWith(".eh_frame_hdr", ORDER_EH_FRAMEHDR)
         .StartsWith(".eh_frame", ORDER_EH_FRAME)
         .StartsWith(".init", ORDER_INIT).StartsWith(".fini", ORDER_FINI)
         .StartsWith(".hash", ORDER_HASH).Default(ORDER_TEXT);
}

Probably not ideal, but makes many cases much more readable.

The changes to overriding-ftemplate-comments.cpp don't seem better or
worse. We should address those soon.

llvm-svn: 172804

llvm-svn: 172802

Summary:
This allows unit tests for components that use Support/Debug.h to print
debug information from test runs by specifying -debug when running the
test.

CC: llvm-commits

Differential Revision: http://llvm-reviews.chandlerc.com/D301

llvm-svn: 172801

llvm-svn: 172800

Before:
  #include <a> // for x
  #include <a/b/c> // for yz
After:
  #include <a>     // for x
  #include <a/b/c> // for yz

llvm-svn: 172799

Before:
int a; // comment
int bbbbb; // comment

After:
int a;     // comment
int bbbbb; // comment

llvm-svn: 172798

Calculate vector element size more directly for VINSERTF128/VEXTRACTF128 immediate handling. Also use MVT since this only called on legal types during pattern matching.

llvm-svn: 172797

llvm-svn: 172795

llvm-svn: 172794

llvm-svn: 172793

llvm-svn: 172792

llvm-svn: 172791

Before:
switch (foo) {
case a: {
  int a = g();
  h(a);
}
  break;
}

Now:
switch (foo) {
case a: {
  int a = g();
  h(a);
} break;
}

llvm-svn: 172789

Check for less than 0 in shuffle mask instead of -1. It's more consistent with other code related to shuffles and easier to implement in compiled code.

llvm-svn: 172788

This doesn't work right with pointers to pointers, but that's likely just a
dupe of PR14884.

llvm-svn: 172785

llvm-svn: 172784