It has been enabled by default for a while, it was only there to allow
performance comparisons.

llvm-svn: 139501

SplitKit always computes a complement live range to cover the places
where the original live range was live, but no explicit region has been
allocated.

Currently, the complement live range is created to be as small as
possible - it never overlaps any of the regions.  This minimizes
register pressure, but if the complement is going to be spilled anyway,
that is not very important.  The spiller will eliminate redundant
spills, and hoist others by making the spill slot live range overlap
some of the regions created by splitting.  Stack slots are cheap.

This patch adds the interface to enable spill modes in SplitKit.  In
spill mode, SplitKit will assume that the complement is going to spill,
so it will allow it to overlap regions in order to avoid back-copies.
By doing some of the spiller's work early, the complement live range
becomes simpler.  In some cases, it can become much simpler because no
extra PHI-defs are required.  This will speed up both splitting and
spilling.

This is only the interface to enable spill modes, no implementation yet.

llvm-svn: 139500

llvm-svn: 139499

llvm-svn: 139498

declaration may be the first declaration, we want the ability to that
declaration to be marked module-private.

llvm-svn: 139497

Only predefine the OBJC_ZEROCOST_EXCEPTIONS macro if Objective-C
exceptions are turned on. Fixes PR10910.

llvm-svn: 139496

This information is required if we want LDWCP to be hoisted out of loops.

llvm-svn: 139495

llvm-svn: 139494

llvm-svn: 139493

Fixes PR10898. The warning should be silent when there are parenthesis
around the condition expression.

llvm-svn: 139492

llvm-svn: 139491

llvm-svn: 139490

Added a new log category for DWARF called "aranges" to log the parsing
of address ranges.

llvm-svn: 139489

and avoid returning a pointer to the current object. In the new 
"operator bool" implementation, check the filename object first 
since many times we have FileSpec objects with a filename, yet no
directory.

llvm-svn: 139488

of how ConstString objects work, and removed the duplicate and out of date
comments that were in the cpp file.

llvm-svn: 139487

Fix disassembling of one of the register/register forms of MOVUPS/MOVUPD/MOVAPS/MOVAPD/MOVSS/MOVSD and their VEX equivalents. Fixes PR10877.

llvm-svn: 139486

llvm-svn: 139485

Fix disassembling of PAUSE instruction. Fixes PR10900. Also fixed NOP disassembling to ignore OpSize and REX.W.

llvm-svn: 139484

llvm-svn: 139481

llvm-svn: 139480

Undo the changes from r139285 which added custom lowering to vselect.
Add tablegen lowering for vselect.

llvm-svn: 139479

changes that were just submitted.

llvm-svn: 139478

more efficiently when it contains a large number of items. Since
the map is actually a vector of "const char *" and type T values,
it will double in size every time you append to it. The extra
added functions allow the collection to be sized to fit the data
after all entries have been appended, and lookups by name or by
regex have been built in to the class to allow efficient lookup.

llvm-svn: 139477

name is "lldb". So currently when you startup any application and you
have not specified that you would like to skip loading init files through
the API or from "lldb" options, then LLDB will try and load:

"~/.lldbinit-%s" where %s the basename of your program
"~/.lldbinit"

Then LLDB will load any program specified on the command line and then
source the "./.llbinit" file for any temporary debug session specific
commands.

I want this feature because I have thread and frame formats that do
ANSI color codes that I only want to load when running in a terminal
which is when I am running the "lldb" command line program.

llvm-svn: 139476

llvm-svn: 139474

test case having instancetype. Fix in rewriter is unrelated to
using of instancetype. Test case uses other feature not yet
supported in the rewriter. There is more work to do, but this
is an ongoing task and not urgent at this time.

llvm-svn: 139473

llvm-svn: 139472

llvm-svn: 139470

Use a more portable heuristic for deciding when to emit a single
atomic store;  it's possible that I've lost information here, but
I'm not sure how much of the logic before was intentionally arch-specific
and how much was just not quite consistent.

llvm-svn: 139468

A little bit of cleanup; set watch_mode to eWatchInvalid at the OptionParsingStarting() lifecycle point.

llvm-svn: 139467

llvm-svn: 139466

Correctly referring to the null pointer as 'null' not the macro 'NULL' in the boolean conversion diagnostic message.

llvm-svn: 139465

llvm-svn: 139464

llvm-svn: 139463

llvm-svn: 139462

llvm-svn: 139461

llvm-svn: 139460

llvm-svn: 139459

r139454 activates an assert in a case where we were doing the right thing anyway.  Make that explicit, and un-XFAIL the testcase.

llvm-svn: 139458

a test failure in CodeGen/palignr.c, which has been marked
XFAIL for the time being.  A bug has been filed at PR10901
for this issue.

llvm-svn: 139457