Optimize some common usage patterns of atomic built-ins __sync_add_and_fetch() and __sync_sub_and_fetch. 

When the return value is not used (i.e. only care about the value in the memory), x86 does not have to use add to implement these. Instead, it can use add, sub, inc, dec instructions with the "lock" prefix.

This is currently implemented using a bit of instruction selection trick. The issue is the target independent pattern produces one output and a chain and we want to map it into one that just output a chain. The current trick is to select it into a merge_values with the first definition being an implicit_def. The proper solution is to add new ISD opcodes for the no-output variant. DAG combiner can then transform the node before it gets to target node selection.

Problem #2 is we are adding a whole bunch of x86 atomic instructions when in fact these instructions are identical to the non-lock versions. We need a way to add target specific information to target nodes and have this information carried over to machine instructions. Asm printer (or JIT) can use this information to add the "lock" prefix.

llvm-svn: 77582

due to x86 encoding restrictions. This is currently off by default
because it may cause code quality regressions. This is for PR4572.

llvm-svn: 77565

llvm-svn: 77558

llvm-svn: 77553

   and convert code to using it, instead of having lots of things
   poke the isLookupPtrRegClass() method directly.

2. Make PointerLikeRegClass contain a 'kind' int, and store it in
   the existing regclass field of TargetOperandInfo when the
   isLookupPtrRegClass() predicate is set.  Make getRegClass pass
   this into TargetRegisterInfo::getPointerRegClass(), allowing
   targets to have multiple ptr_rc things.

llvm-svn: 77504

support multiple different pointer register classes.

llvm-svn: 77501

llvm-svn: 77480

llvm-svn: 77442

llvm-svn: 77441

llvm-svn: 77438

No functionality change.

llvm-svn: 77432

llvm-svn: 77415

Leopard.

llvm-svn: 77414

to ptest instruction plus setcc. Revamp ptest instruction. Add test.

llvm-svn: 77407

llvm-svn: 77404

llvm-svn: 77384

dealing with Data.

llvm-svn: 77372

New name is Metadata.h.

llvm-svn: 77370

llvm-svn: 77366

llvm-svn: 77362

mode, and "ld64" (the default linker) falls back to it in -static
mode.

llvm-svn: 77334

it is highly specific to the object file that will be generated in the end,
this introduces a new TargetLoweringObjectFile interface that is implemented
for each of ELF/MachO/COFF/Alpha/PIC16 and XCore.

Though still is still a brutal and ugly refactoring, this is a major step
towards goodness.

This patch also:
1. fixes a bunch of dangling pointer problems in the PIC16 backend.
2. disables the TargetLowering copy ctor which PIC16 was accidentally using.
3. gets us closer to xcore having its own crazy target section flags and
   pic16 not having to shadow sections with its own objects.
4. fixes wierdness where ELF targets would set CStringSection but not
   CStringSection_.  Factor the code better.
5. fixes some bugs in string lowering on ELF targets.

llvm-svn: 77294

 - My DFS traversal of LLVM is, at least for now, nearly complete! :)

llvm-svn: 77258

MCSections soon instead of Section for all targets, and we need something to
own them.

llvm-svn: 77252

llvm-svn: 77247

header even though there is only one COFF target.

llvm-svn: 77204

llvm-svn: 77197

llvm-svn: 77191

instead.

llvm-svn: 77186

instead and drive things based off of that.

llvm-svn: 77184

'unnamed' bss section, but some impls would want a named one.  Since
they don't have consistent behavior, just make each target do their
own thing, instead of doing something "sortof common" then having
targets change immutable objects later.

llvm-svn: 77165

llvm-svn: 77145

just use a smallstring instead.

llvm-svn: 77144

llvm-svn: 77138

group instead of a bunch of random unrelated ideas.  Provide predicates
to categorize a SectionKind into a group, and use them instead of 
getKind() throughout the code.

This also renames a ton of SectionKinds to be more consistent and
evocative, and adds a huge number of comments on the enums so that
I will hopefully be able to remember how this stuff works long from
now.

llvm-svn: 77129

classes, and migrate existing targets over.

llvm-svn: 77126

 - This was overkill and inconsistently implemented.

llvm-svn: 77114

Also, change MDString to use a StringRef.

llvm-svn: 77098

an enum.

llvm-svn: 77096

1. Spell SectionFlags::Writeable as "Writable".
2. Add predicates for deriving SectionFlags from SectionKinds.
3. Sink ELF-specific getSectionPrefixForUniqueGlobal impl into
   ELFTargetAsmInfo.
4. Fix SectionFlagsForGlobal to know that BSS/ThreadBSS has the
   BSS bit set (the real fix for PR4619).
5. Fix isSuitableForBSS to not put globals with explicit sections
   set in BSS (which was the reason #4 wasn't fixed earlier).
6. Remove my previous hack for PR4619.

llvm-svn: 77085