out the canonical form (A8.6.98) instead of the pseudo-instruction as provided
via MOVs.

DBG_ARM_DISASM=YES llvm-mc -triple=arm-unknown-unknown --disassemble
0xc0 0x00 0xa0 0xe1
Opcode=29 Name=ASR Format=ARM_FORMAT_LDMISCFRM
 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10  9  8  7  6  5  4  3  2  1  0 
-------------------------------------------------------------------------------------------------
| 1: 1: 1: 0| 0: 0: 0: 1| 1: 0: 1: 0| 0: 0: 0: 0| 0: 0: 0: 0| 0: 0: 0: 0| 1: 1: 0: 0| 0: 0: 0: 0|
-------------------------------------------------------------------------------------------------

	asr	r0, r0, #1

llvm-svn: 96654

This hopefulyl should unbreak EH on PPC/Darwin.

llvm-svn: 96637

for ARM to just check if a function has a FP to determine if it's safe
to simplify the stack adjustment pseudo ops prior to eliminating frame
indices. Allow targets to override the default behavior and does so for ARM
and Thumb2.

llvm-svn: 96634

llvm-svn: 96619

comes out as comments but will eventually generate DWARF.

llvm-svn: 96601

llvm-svn: 96591

llvm-svn: 96589

llvm-svn: 96574

Radar 7461718.

llvm-svn: 96572

of AI3ldsbpo, AI3ldhpo, and AI3ldshpo in ARMInstrFormats.td in the process.

llvm-svn: 96565

Transform br (xor (x, y)) -> br (x != y)
Transform br (xor (xor (x,y), 1)) -> br (x == y)
Also normalize (and (X, 1) == / != 1 -> (and (X, 1)) != / == 0 to match to "test on x86" and "tst on arm"

llvm-svn: 96556

llvm-svn: 96540

since it has no pattern, there's not much point in distinguishing an "N2VS"
class for intrinsics anyway.

llvm-svn: 96525

A8.6.30

llvm-svn: 96523

* Use "S" abbreviation for scalar single FP registers in class and pattern
names, instead of keeping the "D" (for "double") abbreviation and tacking on
an "s" elsewhere in the name.
* Move the scalar single FP register classes and patterns to be more
consistent with other definitions in the file.
* Rename "VNEGf32d" definition to "VNEGfd" for consistency.
* Deleted the N2VDIntsPat pattern; N2VSPat is good enough.

llvm-svn: 96521

B6.1.8 RFE Return From Exception loads the PC and the CPSR from the word at the
specified address and the following word respectively.

llvm-svn: 96519

llvm-svn: 96504

"Fix and issue in SparcAsmPrinter where multiple identical .LLGETPCHn symbols could be emitted in the same file (it was uniqued by block number, but not by function number). "  Patch by Nathan Keynes!

llvm-svn: 96495

and add a sparc implementation that knows about delay slots.  Patch by
Nathan Keynes!

llvm-svn: 96492

llvm-svn: 96490

cloned functions.

llvm-svn: 96485

with mangled names).

llvm-svn: 96465

Renamed PIC16FrameOverlay namespace to PIC16OVERLAY.
Renamed PIC16FrameOverlay class to PIC16Overlay.

llvm-svn: 96463

A8.6.18  BFI - Bitfield insert (Encoding A1)

llvm-svn: 96462

reverse engineering what they are.

llvm-svn: 96456

Hopefully, this will fix the remaining issues seen there.

llvm-svn: 96454

llvm-svn: 96440

It's not clear why this is really required, but it was explicitly
added in r48808 with no real explanation or rdar #.

llvm-svn: 96438

This pass is supposed to be run on the linked .bc module.
It traveses the module call graph twice. Once starting from the main function
and marking each reached function as "ML". Again, starting from the ISR
and cloning any reachable function that was marked as "ML". After cloning
the function, it remaps all the call sites in IL functions to call the
cloned functions. 

Currently only marking is being done.

llvm-svn: 96435

64 bits, fixing a variety of problems.

llvm-svn: 96421

indentation.  No functional changes.

llvm-svn: 96418

llvm-svn: 96410

into a roundss intrinsic, producing a cyclic dag.  The root cause
of this is badness handling ComplexPattern nodes in the old dagisel
that I noticed through inspection.  Eliminate a copy of the of the
code that handled ComplexPatterns by making EmitChildMatchCode call
into EmitMatchCode.

llvm-svn: 96408

build failures due to my fix for pr6111.

llvm-svn: 96402

llvm-svn: 96401

llvm-svn: 96399

llvm-svn: 96393

If there exists a use of a build_vector that's the bitwise complement of the mask,
then transform the node to
(and (xor x, (build_vector -1,-1,-1,-1)), (build_vector ~c1,~c2,~c3,~c4)).

Since this transformation is only useful when 1) the given build_vector will
become a load from constpool, and 2) (and (xor x -1), y) matches to a single
instruction, I decided this is appropriate as a x86 specific transformation.
rdar://7323335

llvm-svn: 96389

llvm-svn: 96388

Add support for emitting non-temporal stores for DAGs marked
non-temporal.  Fix from r96241 for botched encoding of MOVNTDQ.

Add documentation for !nontemporal metadata.

Add a simpler movnt testcase.

llvm-svn: 96386