llvm-svn: 123061

them into the loop preheader, eliminating silly instructions like
"icmp i32 0, 100" in fixed tripcount loops.  This also better exposes the 
bigger problem with loop rotate that I'd like to fix: once this has been
folded, the duplicated conditional branch *often* turns into an uncond branch.

Not aggressively handling this is pessimizing later loop optimizations 
somethin' fierce by making "dominates all exit blocks" checks fail.

llvm-svn: 123060

llvm-svn: 123059

1. Take a flags argument instead of a bool.  This makes
   it more clear to the reader what it is used for.
2. Add a flag that says that "remapping a value not in the
   map is ok".
3. Reimplement MapValue to share a bunch of code and be a lot
   more efficient.  For lookup failures, don't drop null values
   into the map.
4. Using the new flag a bunch of code can vaporize in LinkModules
   and LoopUnswitch, kill it.

No functionality change.

llvm-svn: 123058

map from ValueMapper.h (giving us access to its utilities)
and add a fastpath in the loop rotation code, avoiding expensive
ssa updator manipulation for values with nothing to update.

llvm-svn: 123057

temporaries with no-return destructors. The CFG now properly supports
temporaries and implicit destructors which both makes this kludge no
longer work, and conveniently removes the need for it.

Turn on CFG handling of implicit destructors and initializers. Several
ad-hoc benchmarks don't indicate any measurable performance impact from
growing the CFG, and it fixes real correctness problems with warnings.

As a result of turning on these CFG elements, we started to tickle an
inf-loop in the unreachable code logic used for warnings. The fix is
trivial.

llvm-svn: 123056

prefix in a printf format string is matched
with the appropriate conversion specifier.

llvm-svn: 123055

prefix to format conversions (POSIX extension).

llvm-svn: 123054

new "hexname" key for the "key:value;" duple that is part of the packet. This
allows for thread names to contain special characters such as $ # : ; + -

Debugserver now detects if the thread name contains special characters and
sends the chars in hex format if needed.

llvm-svn: 123053

frames, UnwindLLDB could create the incorrect RegisterContext for a given
stack frame because it was using the frame index (not the concrete frame
index). This was causing crashes when doing backtraces through the 
SBFrame::GetFP() because a NULL register context was being returned for the
deepest stack frame.

llvm-svn: 123052

llvm-svn: 123051

llvm-svn: 123049

llvm-svn: 123048

was being searched and sorted using a shared pointer as the value which means
the pointer value was what was being searched for. This means that anytime
you did a stack backtrace, the collection of FuncUnwinders doubled and then
the array or shared pointer got sorted (by pointer value), so you had an ever
increasing collection of shared pointer where a match was never found. This
means we had a ton of duplicates in this table and would cause issues after
one had been debugging for a long time.

llvm-svn: 123045

Instead encode llvm IR level property "HasSideEffects" in an operand (shared
with IsAlignStack). Added MachineInstrs::hasUnmodeledSideEffects() to check
the operand when the instruction is an INLINEASM.

This allows memory instructions to be moved around INLINEASM instructions.

llvm-svn: 123044

This was suggested by Edmund Grimley Evans in pr8411.

llvm-svn: 123043

llvm-svn: 123042

llvm-svn: 123041

llvm-svn: 123040

llvm-svn: 123039

This means avoid using uint32_t. This patch reverts r112200 and fixes original  problem by fixing argument type in lto.cpp.

llvm-svn: 123038

by any means, but something to stress test our unwinder with 260,000+ frames
on a standard darwin thread.

llvm-svn: 123037

llvm-svn: 123036

Patch by Tim Northover.

llvm-svn: 123035

X = sext x; x >s c ? X : C+1 --> X = sext x; X <s C+1 ? C+1 : X
X = sext x; x <s c ? X : C-1 --> X = sext x; X >s C-1 ? C-1 : X
X = zext x; x >u c ? X : C+1 --> X = zext x; X <u C+1 ? C+1 : X
X = zext x; x <u c ? X : C-1 --> X = zext x; X >u C-1 ? C-1 : X
X = sext x; x >u c ? X : C+1 --> X = sext x; X <u C+1 ? C+1 : X
X = sext x; x <u c ? X : C-1 --> X = sext x; X >u C-1 ? C-1 : X

Instead of calculating this with mixed types promote all to the
larger type. This enables scalar evolution to analyze this
expression. PR8866

llvm-svn: 123034

llvm-svn: 123033

Enable live debug variables pass.

llvm-svn: 123032

llvm-svn: 123031

llvm-svn: 123030

command to do the disassembly.  Instead, use the Python API.

llvm-svn: 123029

instead.

llvm-svn: 123028

every BranchInst has a fixed number of operands.

llvm-svn: 123027

llvm-svn: 123026

llvm-svn: 123025

function class template. 

llvm-svn: 123024

to reject this code, but at least clang doesn't
crash anymore.

Crash reported in PR 8880.

llvm-svn: 123017

tuple class template. This implementation is boosted directly from the
variadic templates proposal. N2080.

Note that one section is #ifdef'd out. I'll implement that aspect of
template argument deduction next.

llvm-svn: 123016

Revert r122955. It seems using movups to lower memcpy can cause massive regression (even on Nehalem) in edge cases. I also didn't see any real performance benefit.

llvm-svn: 123015

TreeTransform version of TransformExprs() rather than explicit loop,
so that we expand pack expansions properly. Test cast coming soon...

llvm-svn: 123014

function parameter pack expansions.

llvm-svn: 123007