Commit Graph

187851 Commits

Author SHA1 Message Date
Will Newton 9efe625962 Add ARM relocations to ELFYAML
Tested with check-all with no regressions.

llvm-svn: 223112
2014-12-02 09:49:09 +00:00
Daniel Jasper 8c6e9ef676 clang-format: precedence-based indentation when breaking before operators.
Before:
  bool value = aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
               + aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
               + aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
               == aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
                  * bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb
                  + bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb
               && aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
                  * aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
                  > ccccccccccccccccccccccccccccccccccccccccc;

After:
  bool value = aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
                       + aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
                       + aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
                   == aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
                              * bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb
                          + bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb
               && aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
                          * aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
                      > ccccccccccccccccccccccccccccccccccccccccc;

Not particularly pretty, but can probably help to uncover bugs. And if this
bugs somebody, parentheses can help.

llvm-svn: 223111
2014-12-02 09:46:56 +00:00
Charlie Turner 15f91c5240 Emit Tag_ABI_FP_denormal correctly in fast-math mode.
The default ARM floating-point mode does not support IEEE 754 mode exactly. Of
relevance to this patch is that input denormals are flushed to zero. The way in
which they're flushed to zero depends on the architecture,

  * For VFPv2, it is implementation defined as to whether the sign of zero is
    preserved.
  * For VFPv3 and above, the sign of zero is always preserved when a denormal
    is flushed to zero.

When FP support has been disabled, the strategy taken by this patch is to
assume the software support will mirror the behaviour of the hardware support
for the target *if it existed*. That is, for architectures which can only have
VFPv2, it is assumed the software will flush to positive zero. For later
architectures it is assumed the software will flush to zero preserving sign.

Change-Id: Icc5928633ba222a4ba3ca8c0df44a440445865fd
llvm-svn: 223110
2014-12-02 08:22:29 +00:00
Sonam Kumari f2eacabd66 [signext.ll] Removal Of Duplicate Test Cases
Removed the duplicate test case existing in signext.ll file.

llvm-svn: 223109
2014-12-02 05:29:47 +00:00
Bob Wilson ec2580dec0 Re-apply "Revert r166370 and r166540 now that Xcode 4.6 has been available for a while."
This reverts commit r176892.

I had reverted this a while back to give Chromium more time to update, and
Nico says it should be OK now.

llvm-svn: 223108
2014-12-02 05:17:21 +00:00
NAKAMURA Takumi 8ca1b5a294 clang/test/Modules/malformed.cpp REQUIRES shell due to "cd".
llvm-svn: 223107
2014-12-02 03:55:16 +00:00
Hans Wennborg 76abe8ef98 CMake: make the regexes used for setting HOST_LINK_VERSION more forgiving (PR21268)
If the output of 'ld -v' didn't match the regexes, CMake would previously error
with a message like:

  CMake Error at tools/clang/CMakeLists.txt:269 (string):
    string sub-command REGEX, mode REPLACE regex "[^0-9]*([0-9.]*).*" matched
    an empty string.

llvm-svn: 223106
2014-12-02 03:08:38 +00:00
Hans Wennborg 0101b540db Try to fix the MSVC build.
llvm-svn: 223105
2014-12-02 02:13:09 +00:00
Nick Kledzik 07ba5121a1 [mach-o] Fix TrieEdge leak
In PR21682 Jean-Daliel Dupas found a leak in the trie builder and suggested
a fix was to use a list instead of SmallVector so that the list elements
could be allocated in the BumpPtrAllocator.

llvm-svn: 223104
2014-12-02 01:50:44 +00:00
Nick Kledzik a441b7b050 [mach-o] add support for arm64 compact unwind info
Tim previously added generic compact unwind processing and x86_64 support.
This patch adds arm64 support.

llvm-svn: 223103
2014-12-02 01:50:38 +00:00
Chandler Carruth bdf7e1dd4b Rely on fewer features of the 'env' command. Darwin only supports '-i'.
I'm explicitly setting LC_ALL=C somewhat for documentation, but
hopefully this also removes some host variation from the test results.

llvm-svn: 223102
2014-12-02 01:24:52 +00:00
Nick Lewycky ceb44dfd82 Fix variable used only in assertion.
llvm-svn: 223101
2014-12-02 01:09:56 +00:00
Chandler Carruth c2f9ea7847 Add a test that ensures the Clang driver behaves itself when the PATH
environment variable is changed to strange things out from under it.
Prior to r223099 in LLVM, these test cases would crash in various ways
(assert fails, stack exhaustion, etc.).

llvm-svn: 223100
2014-12-02 00:53:20 +00:00
Chandler Carruth ec8406d8f4 Fix several bugs in r221220's new program finding code.
In both the Unix and Windows variants, std::getenv was called and the
result passed directly to a function accepting a StringRef. This isn't
OK because it might return a null pointer and that causes the StringRef
constructor to assert (and generally produces crash-prone code if
asserts are disabled). Fix this by independently testing the result as
non-null prior to splitting things.

This in turn uncovered another bug in the Unix variant where it would
infinitely recurse if PATH="", or after this fix if PATH isn't set.
There is no need to recurse at all. Slightly re-arrange the code to make
it clear that we can just fixup the Paths argument based on the
environment if we find anything.

I don't know of a particularly useful way to test these routines in
LLVM. I'll commit a test to Clang that ensures that its driver correctly
handles various settings of PATH. However, I have no idea how to
correctly write a Windows test for the PATHEXT change. Any Windows
developers who could provide such a test, please have at. =D

Many thanks to Nick Lewycky and others for helping debug this. =/ It was
quite nasty for us to track down.

llvm-svn: 223099
2014-12-02 00:52:01 +00:00
Richard Smith 4742353414 Update and simplify to match Clang r223095.
llvm-svn: 223098
2014-12-02 00:32:02 +00:00
Bob Wilson 5f4346d538 Remove special case for aarch64 static vs. PIC code in iOS kernel code.
I added this check a while back but then made a note to myself that it
should be completely unnecessary since iOS always uses PIC code-gen for
aarch64. Since I could never come up with any reason why it would be
necessary, I'm just going to remove it and we'll see if anything breaks.
rdar://problem/13627985

llvm-svn: 223097
2014-12-02 00:27:35 +00:00
Jason Molenda 98b3fde8ac Fix the behavior of ${current-pc-arrow} in the default disassembly
format for the not-current-stack-frame.  This was causing
test/functionalities/inferior-assert to fail.

Also document the new additions to the format specifications used
in the disassembly-format changes to formats.html.

<rdar://problem/19102757> 

llvm-svn: 223096
2014-12-02 00:26:21 +00:00
Richard Smith 3c1a41ad99 [modules] Track how 'header' directives were written in module map files,
rather than trying to extract this information from the FileEntry after the
fact.

This has a number of beneficial effects. For instance, diagnostic messages for
failed module builds give a path relative to the "module root" rather than an
absolute file path, and the contents of the module includes file is no longer
dependent on what files the including TU happened to inspect prior to
triggering the module build.

llvm-svn: 223095
2014-12-02 00:08:08 +00:00
Alexey Samsonov 4cdf4eba74 Add override specifiers to Symbolizer methods. NFC.
llvm-svn: 223094
2014-12-01 23:47:51 +00:00
Hal Finkel afcd8dbbcf Simplify pointer comparisons involving memory allocation functions
System memory allocation functions, which are identified at the IR level by the
noalias attribute on the return value, must return a pointer into a memory region
disjoint from any other memory accessible to the caller. We can use this
property to simplify pointer comparisons between allocated memory and local
stack addresses and the addresses of global variables. Neither the stack nor
global variables can overlap with the region used by the memory allocator.

Fixes PR21556.

llvm-svn: 223093
2014-12-01 23:38:06 +00:00
Philip Reames 8e302cdb45 Try to fix a bot failure due to a variable used only in an assert.
Specifically, bot lld-x86_64-darwin13.  Resulting from change 223085.

llvm-svn: 223092
2014-12-01 23:27:45 +00:00
Oleksiy Vyalov a3ff6af484 Fix several test failures on Linux/FreeBSD caused by compiler configuration and invalid environment - make lldbtest.registerSharedLibrariesWithTarget to support multiple platforms.
http://reviews.llvm.org/D6392

llvm-svn: 223091
2014-12-01 23:21:18 +00:00
Nikola Smiljanic fce370eb52 Perform correct lookup when '__super' is used in class with dependent base.
llvm-svn: 223090
2014-12-01 23:15:01 +00:00
Zachary Turner e0d93f523b Enable enabling and disabling breakpoints on Windows.
llvm-svn: 223089
2014-12-01 23:13:52 +00:00
Zachary Turner e068d91824 Link test executables with LLD on Windows.
Clang generates DWARF debug info with section names longer
than 8 characters.  This is only supported by an extension
to PE/COFF which the MSVC linker doesn't implement.  So
trying to link DWARF-enabled object files with MSVC will
lead to corrupt debug information in the executable.

llvm-svn: 223088
2014-12-01 23:13:41 +00:00
Zachary Turner c7a17ed460 Only normalize FileSpec paths *after* resolving them.
Normalizing paths before resolving them can cause the path to
become denormalized after resolution.

llvm-svn: 223087
2014-12-01 23:13:32 +00:00
Zachary Turner 719f58c228 Make -fuse-ld=lld work properly on Windows.
Using lld on Windows requires calling link-lld.exe instead of
lld.exe.  This patch puts this knowledge into clang so that when
using the GCC style clang driver, it can properly delegate to
lld.

Differential Revision: http://reviews.llvm.org/D6428
Reviewed by: Reid Kleckner, Rui Ueyama

llvm-svn: 223086
2014-12-01 23:06:47 +00:00
Philip Reames 0365f1a376 [Statepoints 2/4] Statepoint infrastructure for garbage collection: MI & x86-64 Backend
This is the second patch in a small series.  This patch contains the MachineInstruction and x86-64 backend pieces required to lower Statepoints.  It does not include the code to actually generate the STATEPOINT machine instruction and as a result, the entire patch is currently dead code.  I will be submitting the SelectionDAG parts within the next 24-48 hours.  Since those pieces are by far the most complicated, I wanted to minimize the size of that patch.  That patch will include the tests which exercise the functionality in this patch.  The entire series can be seen as one combined whole in http://reviews.llvm.org/D5683.

The STATEPOINT psuedo node is generated after all gc values are explicitly spilled to stack slots.  The purpose of this node is to wrap an actual call instruction while recording the spill locations of the meta arguments used for garbage collection and other purposes.  The STATEPOINT is modeled as modifing all of those locations to prevent backend optimizations from forwarding the value from before the STATEPOINT to after the STATEPOINT.  (Doing so would break relocation semantics for collectors which wish to relocate roots.)

The implementation of STATEPOINT is closely modeled on PATCHPOINT.  Eventually, much of the code in this patch will be removed.  The long term plan is to merge the functionality provided by statepoints and patchpoints.  Merging their implementations in the backend is likely to be a good starting point.

Reviewed by: atrick, ributzka

llvm-svn: 223085
2014-12-01 22:52:56 +00:00
Enrico Granata e2068207be The register keyword is deprecated in C++11, so clang complains strongly about swig generating register declarations. Abuse the preprocessor to define the keyword away
llvm-svn: 223084
2014-12-01 22:51:03 +00:00
Greg Clayton afa91e339b lldb can deadlock when launched with an non-existing executable:
% lldb /bin/nonono
(lldb) target create "/bin/nonono"
error: unable to find executable for '/usr/bin/nonono'
<deadlock>

The problem was the initial commands 'target create "/bin/nonono"' were put into a pipe and the command interpreter was being run with:

void
CommandInterpreter::RunCommandInterpreter(bool auto_handle_events,
                                          bool spawn_thread,
                                          CommandInterpreterRunOptions &options)
{
    // Always re-create the command intepreter when we run it in case
    // any file handles have changed.
    bool force_create = true;
    m_debugger.PushIOHandler(GetIOHandler(force_create, &options));
    m_stopped_for_crash = false;
    
    if (auto_handle_events)
        m_debugger.StartEventHandlerThread();
    
    if (spawn_thread)
    {
        m_debugger.StartIOHandlerThread();
    }
    else
    {
        m_debugger.ExecuteIOHanders();
        
        if (auto_handle_events)
            m_debugger.StopEventHandlerThread();
    }
    
}

If "auto_handle_events" was set to true and "spawn_thread" was false, we would execute:

m_debugger.StartEventHandlerThread();
m_debugger.ExecuteIOHanders();
m_debugger.StopEventHandlerThread();


The problem was there was no synchonization in Debugger::StartEventHandlerThread() to ensure the event handler was listening to events and the the call to "m_debugger.StopEventHandlerThread()" would do:

void
Debugger::StopEventHandlerThread()
{
    if (m_event_handler_thread.IsJoinable())
    {
        GetCommandInterpreter().BroadcastEvent(CommandInterpreter::eBroadcastBitQuitCommandReceived);
        m_event_handler_thread.Join(nullptr);
    }
}

The problem was that the event thread might not be listening for the CommandInterpreter::eBroadcastBitQuitCommandReceived event yet.

The solution is to make sure the Debugger::DefaultEventHandler() is listening to events before we return from Debugger::StartEventHandlerThread(). Once we have this synchonization we remove the race condition.

This fixes radar:

<rdar://problem/19041192>

llvm-svn: 223083
2014-12-01 22:41:27 +00:00
Greg Clayton d90ac932d9 Added a new regular expression to the "_regexp-break" command ("b" by default):
(lldb) b /break here/

This will set a source level regular expression breakpoint on any text between the first '/' and the last '/'. The equivalent command will be:

(lldb) breakpoint set --source-pattern-regexp 'break here'

llvm-svn: 223082
2014-12-01 22:34:03 +00:00
Vince Harron 6eddf8df2a Added StringExtractor::DecodeHexU8 && GetHexBytesAvail
DecodeHexU8 returns a decoded hex character pair, returns -1 if a
valid hex pair is not available.

GetHexBytesAvail decodes all available hex pairs.

llvm-svn: 223081
2014-12-01 22:19:33 +00:00
Vince Harron e5a9eaaaed StringExtractor unit tests
Unit tests to cover StringExtractor in advance of minor new
functionality & minor refactor

llvm-svn: 223080
2014-12-01 22:10:15 +00:00
Reid Kleckner ee7cf84c8f Use nullptr to silence -Wsentinel when self-hosting on Windows
Richard rejected my Sema change to interpret an integer literal zero in
a varargs context as a null pointer, so -Wsentinel sees an integer
literal zero and fires off a warning. Only CodeGen currently knows that
it promotes integer literal zeroes in this context to pointer size on
Windows.  I didn't want to teach -Wsentinel about that compatibility
hack. Therefore, I'm migrating to C++11 nullptr.

llvm-svn: 223079
2014-12-01 22:02:27 +00:00
Philip Reames 337c4bd4ab [Statepoints 1/4] Statepoint infrastructure for garbage collection: IR Intrinsics
The statepoint intrinsics are intended to enable precise root tracking through the compiler as to support garbage collectors of all types. The addition of the statepoint intrinsics to LLVM should have no impact on the compilation of any program which does not contain them. There are no side tables created, no extra metadata, and no inhibited optimizations.

A statepoint works by transforming a call site (or safepoint poll site) into an explicit relocation operation. It is the frontend's responsibility (or eventually the safepoint insertion pass we've developed, but that's not part of this patch series) to ensure that any live pointer to a GC object is correctly added to the statepoint and explicitly relocated. The relocated value is just a normal SSA value (as seen by the optimizer), so merges of relocated and unrelocated values are just normal phis. The explicit relocation operation, the fact the statepoint is assumed to clobber all memory, and the optimizers standard semantics ensure that the relocations flow through IR optimizations correctly.

This is the first patch in a small series.  This patch contains only the IR parts; the documentation and backend support will be following separately.  The entire series can be seen as one combined whole in http://reviews.llvm.org/D5683.

Reviewed by: atrick, ributzka

llvm-svn: 223078
2014-12-01 21:18:12 +00:00
Jingyue Wu 5b62eb9b48 [NVPTX] Do not emit .weak symbols for NVPTX
Summary:
".weak" symbols cannot be consumed by ptxas (PR21685). This patch makes the
weak directive in MCAsmPrinter customizable, and disables emitting ".weak"
symbols for NVPTX.

Test Plan: weak-linkage.ll

Reviewers: jholewinski

Reviewed By: jholewinski

Subscribers: majnemer, jholewinski, llvm-commits

Differential Revision: http://reviews.llvm.org/D6455

llvm-svn: 223077
2014-12-01 21:16:17 +00:00
Reid Kleckner 35fc363ce8 Parse 'ghccc' in .ll files as the GHC convention (cc 10)
Previously we just used "cc 10" in the .ll files, but that isn't very
human readable.

llvm-svn: 223076
2014-12-01 21:04:44 +00:00
Ahmed Bougacha d0ce058f2c [AArch64] Don't combine "select (setcc i1 LHS, RHS), vL, vR".
r208210 introduced an optimization that improves the vector select
codegen by doing the setcc on vectors directly.
This is a problem they the setcc operands are i1s, because the
optimization would create vectors of i1, which aren't legal.

Part of PR21549.

Differential Revision: http://reviews.llvm.org/D6308

llvm-svn: 223075
2014-12-01 20:59:00 +00:00
Ahmed Bougacha 879463206e [AArch64] Fix v2i8->i16 bitcast legalization.
r213378 improved f16 bitcasts, so that they go directly through subregs,
instead of through the stack.  That code now causes an assertion failure
for bitcasts from other 16-bits types (most importantly v2i8).

Correct that by doing the custom lowering for i16 bitcasts only when the
input is an f16.

Part of PR21549.

Differential Revision: http://reviews.llvm.org/D6307

llvm-svn: 223074
2014-12-01 20:52:32 +00:00
Shankar Easwaran 7a64ec24d6 [Core] Remove function to not override RoundTripPasses.
RoundTripPasses should always be called in DEBUG mode if the environment
variable "LLD_RUN_ROUNDTRIP_TEST" is set.

Flavors should not be able to override this behavior.

llvm-svn: 223073
2014-12-01 20:28:54 +00:00
Peter Zotov 0d040f66a5 [OCaml] Move Llvm.clone_module to its own Llvm_transform_utils module.
This way most code won't link this (substantially large) library,
if compiled statically with LLVM.

llvm-svn: 223072
2014-12-01 19:50:39 +00:00
Peter Zotov b20073c63c [OCaml] [cmake] Add CMake buildsystem for OCaml.
Closes PR15325.

llvm-svn: 223071
2014-12-01 19:50:23 +00:00
JF Bastien 57148cbcbd libc++: add NaCl and PNaCl support for std::random_device
Summary:
The NaCl sandbox doesn't allow opening files under /dev, but it offers an API which provides the same capabilities. This is the same random device emulation that nacl_io performs for POSIX support, but nacl_io is an optional library so libc++ can't assume that device emulation will be performed. Note that NaCl only supports /dev/urandom, not /dev/random.

This patch also cleans up some of the preprocessor #endif, and fixes the test for Win32 (it accepts any token, and would therefore never throw regardless of the token provided).

Test Plan: ninja check-libcxx

Reviewers: dschuff, mclow.lists, danalbert

Subscribers: jfb, cfe-commits

Differential Revision: http://reviews.llvm.org/D6442

llvm-svn: 223068
2014-12-01 19:19:55 +00:00
Rafael Espindola 973b3612e4 Use a continue to reduce indentation and clang-format. NFC.
llvm-svn: 223067
2014-12-01 19:17:46 +00:00
Rafael Espindola 4d4c938770 Use a range loop. NFC.
llvm-svn: 223066
2014-12-01 19:08:07 +00:00
Richard Smith a634ff2194 Add missing 'break's, found by inspection. No functionality change; the
fallthrough happened to do the right thing in both cases.

llvm-svn: 223064
2014-12-01 18:59:10 +00:00
Pete Cooper 178e6548f5 Use C++ typed enums instead of 'unsigned char' for MCInst Kind. NFC.
This makes it much easier to see the value of operands in the debugger.

llvm-svn: 223060
2014-12-01 18:46:43 +00:00
Ahmed Bougacha fb6eeb74c5 [MachineVerifier] Accept a MBB with a single landing pad successor.
The MachineVerifier used to check that there was always exactly one
unconditional branch to a non-landingpad (normal) successor.
If that normal successor to an invoke BB is unreachable, it seems
reasonable to only have one successor, the landing pad.
On targets other than AArch64 (and on AArch64 with a different testcase),
the branch folder turns the branch to the landing pad into a fallthrough.
The MachineVerifier, which relies on AnalyzeBranch, is unable to check
the condition, and doesn't complain. However, it does in this specific
testcase, where the branch to the landing pad remained.
Make the MachineVerifier accept it.

llvm-svn: 223059
2014-12-01 18:43:53 +00:00
Rafael Espindola dd42b533e9 Drop SrcStructTypesSet. It is redundant.
At the only point in the code it is used, we haven't added any of the src types
to DstStructTypesSet yet.

llvm-svn: 223057
2014-12-01 18:42:18 +00:00
Nico Weber 4af55faa19 Add a test for devirtualization of virtual operator calls.
There was no test coverage for this before: Modifiying
EmitCXXOperatorMemberCallee() to not call CanDevirtualizeMemberFunctionCall()
didn't make any test fail.

llvm-svn: 223056
2014-12-01 17:48:04 +00:00