Minor logic simplification.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80453 91177308-0d34-0410-b5e6-96231b3b80d8

Add some comments.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80452 91177308-0d34-0410-b5e6-96231b3b80d8

Cleanup whitespace and indentation.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80451 91177308-0d34-0410-b5e6-96231b3b80d8

Remove some unused fields.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80450 91177308-0d34-0410-b5e6-96231b3b80d8

CMOV_GR8 clobbers EFLAGS when its expansion involves an xor to set
a register to 0. This fixes PR4814.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80445 91177308-0d34-0410-b5e6-96231b3b80d8

Inline empty destructor.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80431 91177308-0d34-0410-b5e6-96231b3b80d8

Fix warning about non-virtual destructor.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80429 91177308-0d34-0410-b5e6-96231b3b80d8

- Add target lowering methods to get the preferred format for the FDE and LSDA
  encodings.
- Make some of the values emitted by the FDEs dependent upon the pointer
  size. This is in line with how GCC does things. And it has the benefit of
  working for Darwin in 64-bit mode now.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80428 91177308-0d34-0410-b5e6-96231b3b80d8

Add a form of EOL which emits the text version of a DWARF format encoding. This
doesn't handle all values of the formatting. Those can be added as needed.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80427 91177308-0d34-0410-b5e6-96231b3b80d8

PR4795: Remove EEVT::isFP, isInt and isVec types used by TableGen's type
inferencing.  As far as I can tell, these are equivalent to the existing
MVT::fAny, iAny and vAny types, and having both of them makes it harder
to reason about and modify the type inferencing code.

The specific problem in PR4795 occurs when updating a vAny type to be fAny
or iAny, or vice versa.  Both iAny and fAny include vector types -- they
intersect with the set of types represented by vAny.  When merging them,
choose fAny/iAny to represent the intersection.  This is not perfect, since
fAny/iAny also include scalar types, but it is good enough for TableGen's
type inferencing.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80423 91177308-0d34-0410-b5e6-96231b3b80d8

Make the augmentation size and next set of bytes agree on size,
and make the reference pointer size as it should be.

Fixes an abort on a testcase derived from libunwind's personality
test in 64-bit.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80414 91177308-0d34-0410-b5e6-96231b3b80d8

Fix OProfile support after r80406 changed the DebugInfo interface from
GlobalVariables to MDNodes.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80411 91177308-0d34-0410-b5e6-96231b3b80d8

Another stab at fixing up register kill flags after post-RA scheduling.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80410 91177308-0d34-0410-b5e6-96231b3b80d8

Do not assert on too wide splats we don't support.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80409 91177308-0d34-0410-b5e6-96231b3b80d8

Add missed extract_element pattern

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80408 91177308-0d34-0410-b5e6-96231b3b80d8

Reapply 79977.
Use MDNodes to encode debug info in llvm IR.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80406 91177308-0d34-0410-b5e6-96231b3b80d8

Let Darwin linker auto-synthesize stubs and lazy-pointers. This deletes a bunch of nasty code in ARM asm printer.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80404 91177308-0d34-0410-b5e6-96231b3b80d8

Nuke trailing whitespace.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80401 91177308-0d34-0410-b5e6-96231b3b80d8

Fix creation of .bca libraries with EXPORTED_SYMBOLS_LIST, this was putting the
llvm-ld shell wrapper script in the archive, not the relinked object!

Also, rename the temp file to avoid conflicts.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80373 91177308-0d34-0410-b5e6-96231b3b80d8

Read Makefile.config before using LLVMGCC.
See http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090824/085828.html

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80372 91177308-0d34-0410-b5e6-96231b3b80d8

install-sh chmods to 0755 by default, and this causes 'git diff' to show
that all the Makefiles changed mode.
Fix this by tellint install-sh to chmod
only to 0644, these are not executable files after all!

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80371 91177308-0d34-0410-b5e6-96231b3b80d8

Short-term workaround for frame-related weirdness on win64.
Some other minor win64 fixes as well.

Patch by Michael Beck!

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80370 91177308-0d34-0410-b5e6-96231b3b80d8

rm needs -f

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80363 91177308-0d34-0410-b5e6-96231b3b80d8

Remove the llvmprof.out from the test output, otherwise running
make check in a non-clean directory causes it to fail (for example when running
make check twice), since execution counts will differ.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80362 91177308-0d34-0410-b5e6-96231b3b80d8

Preparation for Optimal Edge Profiling:
This implements the maximum spanning tree algorithm on CFGs according to
weights given by the ProfileEstimator. This is then used to implement Optimal
Edge Profiling.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80358 91177308-0d34-0410-b5e6-96231b3b80d8

Remove profiling output file because two consecutive runs of make check give
error.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80357 91177308-0d34-0410-b5e6-96231b3b80d8

Removed unnecessary file creation during test.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80356 91177308-0d34-0410-b5e6-96231b3b80d8

Pulled all tests into one test. Removed some redundant tests. Rename.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80355 91177308-0d34-0410-b5e6-96231b3b80d8

Fix -Asserts warning, round two.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80354 91177308-0d34-0410-b5e6-96231b3b80d8

llvm-mc: .lsym is more unsupported than unimplemented, pending a use case appearing.

Also, all one of the file level flags are implemented.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80352 91177308-0d34-0410-b5e6-96231b3b80d8

llvm-mc: Support .comm emission.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80351 91177308-0d34-0410-b5e6-96231b3b80d8

Print a nl before pic labels so they start at a new line. This makes assembly more readable.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80350 91177308-0d34-0410-b5e6-96231b3b80d8

Since all std::cout is gone, also remove iostream include.
(See http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20090824/085620.html)

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80349 91177308-0d34-0410-b5e6-96231b3b80d8

Readded test from r79615, this tests the complete profiling tool chain. Furhter
tests can test only parts of this system.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80348 91177308-0d34-0410-b5e6-96231b3b80d8

llvm-mc: Support .zerofill emission.
 - I'm still trying to figure out the cleanest way to implement this and match the assembler, currently there are some substantial differences.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80347 91177308-0d34-0410-b5e6-96231b3b80d8

llvm-mc: Tweak section alignment and size computation to match 'as' closer.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80345 91177308-0d34-0410-b5e6-96231b3b80d8

llvm-mc: Factor getSectionData out of SwitchSection.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80344 91177308-0d34-0410-b5e6-96231b3b80d8

llvm-mc: Emit .lcomm as .zerofill.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80343 91177308-0d34-0410-b5e6-96231b3b80d8

llvm-mc: Unique zero fill sections.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80342 91177308-0d34-0410-b5e6-96231b3b80d8

llvm-mc: Add const to EmitZeroFill section argument.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80341 91177308-0d34-0410-b5e6-96231b3b80d8

llvm-mc: Fix thinko in emitting values.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80340 91177308-0d34-0410-b5e6-96231b3b80d8

Add MathExtras.h OffsetToAlignment, like RoundUpToAlignment but returns the
offset to the next aligned integer.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80339 91177308-0d34-0410-b5e6-96231b3b80d8

Fix -Asserts warning.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80338 91177308-0d34-0410-b5e6-96231b3b80d8

finish a half formed thought :)

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80334 91177308-0d34-0410-b5e6-96231b3b80d8

Another NO_RUNTIME_LIBRARIES tweak...

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80331 91177308-0d34-0410-b5e6-96231b3b80d8

Tweak NO_RUNTIME_LIBS.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80330 91177308-0d34-0410-b5e6-96231b3b80d8

Mark Andersen's as experimental.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80328 91177308-0d34-0410-b5e6-96231b3b80d8

Fix PR3913, patch by Jakub Staszak!

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80327 91177308-0d34-0410-b5e6-96231b3b80d8

v4, v5 does not support sxtb / sxth.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80322 91177308-0d34-0410-b5e6-96231b3b80d8

Disable optional bindings for Apple-style builds.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80319 91177308-0d34-0410-b5e6-96231b3b80d8

Revert 76080.  This broke some powerpc cross compiles.
It also makes the llvmCore build dependent on whatever version of llvm-gcc
happens to be installed on the build machine.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80316 91177308-0d34-0410-b5e6-96231b3b80d8

Revert r80305, I forgot a dependent change.

--- Reverse-merging r80305 into '.':
U    tools/llvm-mc/AsmParser.cpp

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80309 91177308-0d34-0410-b5e6-96231b3b80d8

Closure is a very generic name. Use AppleBlock instead.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80307 91177308-0d34-0410-b5e6-96231b3b80d8

llvm-mc: Unique sections in .zerofill.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80305 91177308-0d34-0410-b5e6-96231b3b80d8

eliminate all 80-col violations that I have introduced in my recent checkins (and some others more)

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80304 91177308-0d34-0410-b5e6-96231b3b80d8

Don't build runtime libraries in an Apple style build.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80303 91177308-0d34-0410-b5e6-96231b3b80d8

Revert 80278 for now, it caused a lot of MIPS tests to fail

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80280 91177308-0d34-0410-b5e6-96231b3b80d8

Revamp our friend Mips :)
Add MO flags to simplify the printing of relocations.
Remove the support for printing large code model relocs (which
aren't supported anyway).

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80278 91177308-0d34-0410-b5e6-96231b3b80d8

Don't mark CMOV_GR8 as two-address, or commutable, since it's a pseudo.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80271 91177308-0d34-0410-b5e6-96231b3b80d8

Adjust the MachineBasicBlock verifier rules to be more
tolerant of blocks that end with "unreachable".

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80270 91177308-0d34-0410-b5e6-96231b3b80d8

Add FIXME for when we support more specific XMM registers.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80269 91177308-0d34-0410-b5e6-96231b3b80d8

Nuke trailing whitespace.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80268 91177308-0d34-0410-b5e6-96231b3b80d8

Use stripPointerCasts instead of doing the same manually.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80267 91177308-0d34-0410-b5e6-96231b3b80d8

Minor code simplification.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80266 91177308-0d34-0410-b5e6-96231b3b80d8

Teach getUnderlyingObject and skipPointerCasts about GlobalAliases.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80265 91177308-0d34-0410-b5e6-96231b3b80d8

Be somewhat more consistent about const qualifiers.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80264 91177308-0d34-0410-b5e6-96231b3b80d8

Global Aliases are not identifiable objects.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80263 91177308-0d34-0410-b5e6-96231b3b80d8

Handle TargetData with const.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80262 91177308-0d34-0410-b5e6-96231b3b80d8

Strip trailing whitespace from blank lines.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80259 91177308-0d34-0410-b5e6-96231b3b80d8

re-apply r80197, now that iterator.h is not mentioned any more

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80254 91177308-0d34-0410-b5e6-96231b3b80d8

Hopefully the final missing part :(
scalar_to_vector is fully legal now

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80251 91177308-0d34-0410-b5e6-96231b3b80d8

Forgot about actual change :)

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80250 91177308-0d34-0410-b5e6-96231b3b80d8

scalar_to_vector is fully legal now (implemented as subreg accesses)

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80249 91177308-0d34-0410-b5e6-96231b3b80d8

Remove an entry from ValueMap before removing elements.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80247 91177308-0d34-0410-b5e6-96231b3b80d8

Ok, sometimes it's profitable to turn scalar_to_vector stuff into subreg access.
Add a testcase.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80246 91177308-0d34-0410-b5e6-96231b3b80d8

Transform float scalar_to_vector into subreg accesses.
No idea whether this is profitable or not.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80245 91177308-0d34-0410-b5e6-96231b3b80d8

STRD and LDRD require ARMv5TE, not just ARMv5T.
See http://llvm.org/PR4687 for more info and links.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80244 91177308-0d34-0410-b5e6-96231b3b80d8

Inverse logic to increase portability.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80240 91177308-0d34-0410-b5e6-96231b3b80d8

To make mcc16 run correctly on mac.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80239 91177308-0d34-0410-b5e6-96231b3b80d8

llvm-mc/Mach-O: Add MCCodeEmitter support, for encoding instructions.
 - No relocations yet, of course.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80235 91177308-0d34-0410-b5e6-96231b3b80d8

llvm-mc/X86: Implement single instruction encoding interface for MC.
 - Note, this is a gigantic hack, with the sole purpose of unblocking further
   work on the assembler (its also possible to test the mathcer more completely
   now).

 - Despite being a hack, its actually good enough to work over all of 403.gcc
   (although some encodings are probably incorrect). This is a testament to the
   beauty of X86's MachineInstr, no doubt! ;)

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80234 91177308-0d34-0410-b5e6-96231b3b80d8

llvm-mc: Print encodings after the instruction, and only when we have an asm
printer.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80233 91177308-0d34-0410-b5e6-96231b3b80d8

X86: Mark EH_RETURN as code-gen-only.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80232 91177308-0d34-0410-b5e6-96231b3b80d8

Add {MCInst,MCOperand}::{print,dump}

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80231 91177308-0d34-0410-b5e6-96231b3b80d8

llvm-mc: Only show instruction encodings with --show-encoding.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80230 91177308-0d34-0410-b5e6-96231b3b80d8

the buildbots revealed one more breakage. fix. (why didn't I see this?)

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80227 91177308-0d34-0410-b5e6-96231b3b80d8

Clean up the minor mess I caused with removing iterator.h. I shall take care of 80-col violations and the FIXME later. (Thanks goodness that I live in another continent, so the monkeypox did not strike me :-)

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80224 91177308-0d34-0410-b5e6-96231b3b80d8

Implement a new optimization in the inliner: if inlining multiple
calls into a function and if the calls bring in arrays, try to merge
them together to reduce stack size.  For example, in the testcase
we'd previously end up with 4 allocas, now we end up with 2 allocas.

As described in the comments, this is not really the ideal solution
to this problem, but it is surprisingly effective.  For example, on
176.gcc, we end up eliminating 67 arrays at "gccas" time and another
24 at "llvm-ld" time.

One piece of concern that I didn't look into: at -O0 -g with
forced inlining this will almost certainly result in worse debug
info.  I think this is acceptable though given that this is a case
of "debugging optimized code", and we don't want debug info to
prevent the optimizer from doing things anyway.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80215 91177308-0d34-0410-b5e6-96231b3b80d8

the inliner shouldn't crash on this.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80214 91177308-0d34-0410-b5e6-96231b3b80d8

For now, only run MC tests if X86 is configured.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80213 91177308-0d34-0410-b5e6-96231b3b80d8

This is passing for PPC on Mac OS X.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80210 91177308-0d34-0410-b5e6-96231b3b80d8

Revert accidental commit.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80208 91177308-0d34-0410-b5e6-96231b3b80d8

unbreak the build, yay for symlinks + makefiles. :(

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80205 91177308-0d34-0410-b5e6-96231b3b80d8

reduce header #include'age

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80204 91177308-0d34-0410-b5e6-96231b3b80d8

enhance InlineFunction to be able to optionally return
a the list of static allocas that it inlined.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80203 91177308-0d34-0410-b5e6-96231b3b80d8

smallvectorize the list of returns built by CloneAndPruneFunctionInto.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80202 91177308-0d34-0410-b5e6-96231b3b80d8

remove CloneTrace, which appears to be dead since 2004.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80201 91177308-0d34-0410-b5e6-96231b3b80d8

reduce inlining factor some stuff out to a static helper function,
and other code cleanups.  No functionality change.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80199 91177308-0d34-0410-b5e6-96231b3b80d8

Revert accidental commit.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80198 91177308-0d34-0410-b5e6-96231b3b80d8

--- Reverse-merging r80147 into '.':
A    include/llvm/ADT/iterator.cmake
U    autoconf/configure.ac
--- Reverse-merging r80161 into '.':
U    cmake/config-ix.cmake
--- Reverse-merging r80171 into '.':
U    Makefile
--- Reverse-merging r80173 into '.':
U    configure
U    include/llvm/Config/config.h.in
--- Reverse-merging r80180 into '.':
A    include/llvm/ADT/iterator.h.in

Despite common miscomceptions, iterator.h is alive and well. It broke the build
bots for several hours. And yet no one bothered to look at them.

Gabor and Doug, please review your changes and make sure that they actually
build before resubmitting them.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80197 91177308-0d34-0410-b5e6-96231b3b80d8