Formalize the notion that AVX and SSE are non-overlapping extensions from the compiler's point of view.  Per email discussion, we either want to always use VEX-prefixed instructions or never use them, and are taking "HasAVX" to mean "Always use VEX".  Passing -mattr=-avx,+sse42 should serve to restore legacy SSE support when desirable.

llvm-svn: 121439

Various fixes mostly relating to the User Settings stuff:

- Added new utility function to Arg, GetQuotedCommandString, which re-assembles
the args into a string, replacing quotes that were originally there.

- Modified user settings stuff to always show individual elements when printing out
arrays and dictionaries.

- Added more extensive help to 'settings set', explaining more about dictionaries
and arrays (including current dictionary syntax).

- Fixed bug in user settings  where quotes were being stripped and lost, so that
sometimes array or dictionary elements that ought to have been a single element
were being split up.

llvm-svn: 121438

Changing the ObjC find method implementation to use a ClangUtilityFunction inserted into the target.  Consolidate all the
logic for finding the target of a method dispatch into this function, insert & call it.  Gets calls to super, and all the
fixup & fixedup variants working properly.  Also gets the class from the object so that we step through KVO wrapper methods
into the actual user code.

llvm-svn: 121437

llvm-svn: 121436

llvm-svn: 121435

f:
        .cfi_startproc
        nop
        .cfi_endproc

assembled (on ELF).

llvm-svn: 121434

llvm-svn: 121433

llvm-svn: 121432

better to actually produce a decent set of completions by checking the
system include paths, but not today. Fixes PR8744.

llvm-svn: 121431

llvm-svn: 121430

llvm-svn: 121428

and ActOnVariableDeclarator
No functionality change. // rdar://8751949

llvm-svn: 121427

llvm-svn: 121426

llvm-svn: 121425

llvm-svn: 121424

Modify HandleCommand to not do any argument processing until it has determined whether or
not the command should take raw input, then handle & dispatch the arguments appropriately.

Also change the 'alias' command to be a command that takes raw input.  This is necessary to
allow aliases to be created for other commands that take raw input and might want to include
raw input in the alias itself.

Fix a bug in the aliasing mechanism when creating aliases for commands with 3-or-more words.

Raw input should now be properly handled by all the command and alias mechanisms.

llvm-svn: 121423

llvm-svn: 121421

llvm-svn: 121419

t_addrmode_s# address modes is used for ASM printing, not for encoding.
<rdar://problem/8745375>

llvm-svn: 121417

global code completions are disabled (e.g., because they are
cached). Also, make sure that forward-declared protocols are visited
when we look for all visible names within a declaration context.

Previously, we would end up with duplicate completions for protocols.

llvm-svn: 121416

llvm-svn: 121415

Use the new IsAligned fixup flag to improve fixup encodings for Thumb2 branches.  This is still not perfect,
but it gets many more of them correct than it did previously.

llvm-svn: 121414

second reparse.

llvm-svn: 121413

Necessary for byval support on ARM.  Radar 7662569.

llvm-svn: 121412

llvm-svn: 121411

llvm-svn: 121410

a bug in Apple GCC 4.0.

llvm-svn: 121409

Fix an issue in some Thumb fixups, where the effective PC address needs to be 4-byte aligned when calculating
the offset.  Add a new fixup flag to represent this, and use it for the one fixups that I have a testcase for needing
this.  It's quite likely that the other Thumb fixups will need this too, and to have their fixup encoding logic
adjusted accordingly.

llvm-svn: 121408

llvm-svn: 121404

This test intends to catch invalid use of ".byte 256" in output. Now, the assert in MC will be triggered in such cases.

llvm-svn: 121403

llvm-svn: 121402

the condition codes.  Where the ones that do have an 's' suffix and the ones
that don't don't have the suffix.  The trick is if MatchInstructionImpl() fails
we try again after adding a CCOut operand with the correct value and removing
the 's' if present.  Four simple test cases added for now, lots more to come.

llvm-svn: 121401

llvm-svn: 121399

t_brtarget to be more specific.

llvm-svn: 121398

The sensible thing would be to have these intrinsics take all quad-register
vector operands, but that's not what ARM did.  They made the last vector
operand always be a double-register type.  Since the lane number
must be a constant, the user can know which half of a quad-register contains
that lane, extract the high or low half of the vector, and adjust the lane
number accordingly.  The only advantage I can see for this is that it works
better when you want to multiply a quad-register value by a lane from a
double-register value, but I wouldn't have expected that to be the common
case.  Oh well -- at this point we just need to follow the spec.

llvm-svn: 121397

process launch now asks to kill the current process if it is alive, and if you affirm, does so for you.
Also added #pragma mark for the command objects defined in the file.

llvm-svn: 121396

llvm-svn: 121395

llvm-svn: 121393

Their suffixes are supposed to reflect the source operand element type,
not the destination element type.  Radar 8746481.

llvm-svn: 121392

Otherwise, a plain str/ldr should be used instead. Make sure we account for
that in prologue/epilogue code generation.
rdar://8745460

llvm-svn: 121391