and use this (as well as getCalledValue) to access the callee,
instead of {g|s}etOperand(0).

llvm-svn: 99084

llvm-svn: 99083

address register writeback.

llvm-svn: 99082

llvm-svn: 99081

rewrite the existing VLD3 and VLD4 instructions to use the same classes as
the others.

llvm-svn: 99080

llvm-svn: 99078

llvm-svn: 99077

load/stores with address register writeback, and use "odd" suffix to distinguish
instructions to access odd numbered registers (instead of "a" and "b").
No functional changes.

llvm-svn: 99066

writeback, and refactor the existing double-spaced VLD2 instructions.
These are only for the disassembler since codegen doesn't use them, at
least for now.

llvm-svn: 99065

llvm-svn: 99064

This avoids redundant diagnostics.

llvm-svn: 99063

llvm-svn: 99062

PIC16: Simplify code by using a std::set<std::string> instead of a sorted & uniqued std::list of leaked char*.

llvm-svn: 99061

We were inserting a value into a std::vector<> while iterating over
it, which could cause the underlying memory to get deallocated
and reallocated.  While not the best solution, use an llvm::ImmutableList
for now as it is safely supports insertions during iteration.

llvm-svn: 99058

target when both {MACOSX,IPHONEOS}_DEPLOYMENT_TARGET were set.

llvm-svn: 99055

llvm-svn: 99054

llvm-svn: 99053

llvm-svn: 99052

llvm-svn: 99051

 - This should be conservatively correct, we eventually should have target hooks for platforms that are less strict.

llvm-svn: 99050

--- Reverse-merging r98679 into 'lib/Target/ARM/ARMInstrVFP.td':
U    lib/Target/ARM/ARMInstrVFP.td

llvm-svn: 99049

load/store optimizer would incorrectly think that registers D26 and D28
were consecutive and would generate a VLDM instruction to load them.
The assembler was not convinced.

llvm-svn: 99043

expression computation in the wrong bit-width, and end up generating a totally
bogus array reference (_g0+8589934546).
 - This showed up on Prolangs/cdecl.

llvm-svn: 99042

 - Disable RTTI.
 - Disable use of __cxa_atexit.
 - Disable unwind tables.
 - Enable freestanding mode.

Also, honor -fhosted correctly.

<rdar://problem/7515383> C++ support: -fapple-kext not honored

llvm-svn: 99041

expression", statistical evidence indicates it has some kind of memory error.

llvm-svn: 99040

 - So much typing, so little gain...

Also, rename the __cxx_global_initialization function just to match llvm-gcc.

llvm-svn: 99039

llvm-svn: 99038

llvm-svn: 99037

so that the SCEVExpander doesn't retain a dangling pointer as its
insert position. The dangling pointer in this case wasn't ever used
to insert new instructions, but it was causing trouble with
SCEVExpander's code for automatically advancing its insert position
past debug intrinsics.

This fixes use-after-free errors that valgrind noticed in
test/Transforms/IndVarSimplify/2007-06-06-DeleteDanglesPtr.ll and
test/Transforms/IndVarSimplify/exit_value_tests.ll.

llvm-svn: 99036

llvm-svn: 99033

caller, then it is not safe to optimize the call into a sibcall since
the call result has to be popped off the x87 stack.

llvm-svn: 99032

llvm-svn: 99031

llvm-svn: 99026

llvm-svn: 99025

is not evaluated because the receiver is nil.

llvm-svn: 99024

<rdar://problem/7641151> clang must eat spaces after -l

llvm-svn: 99023

Also, simplify/fix SeparateArg::render with forced join.

llvm-svn: 99022

llvm-svn: 99021

one from emmintrin.h.

llvm-svn: 99020

llvm-svn: 99019