maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
150,605 Commits 23 Branches 0 Tags 2.2 GiB
Commit Graph

6708 Commits

Author SHA1 Message Date
Ahmed Bougacha f1ed334d55 Add a way to define the bit range covered by a SubRegIndex. 
NOTE: If this broke your out-of-tree backend, in *RegisterInfo.td, change
the instances of SubRegIndex that have a comps template arg to use the
ComposedSubRegIndex class instead.

In TableGen land, this adds Size and Offset attributes to SubRegIndex,
and the ComposedSubRegIndex class, for which the Size and Offset are
computed by TableGen. This also adds an accessor in MCRegisterInfo, and
Size/Offsets for the X86 and ARM subreg indices.

llvm-svn: 183020
 2013-05-31 17:08:36 +00:00
Tim Northover 4d14144024 ARM: permit upper-case BE/LE on setend instruction 
Patch by Amaury de la Vieuville.

llvm-svn: 183012
 2013-05-31 15:58:45 +00:00
Tim Northover 4173e29a98 ARM: add fstmx and fldmx instructions for assembly 
These instructions are deprecated oddities, but we still need to be able to
disassemble (and reassemble) them if and when they're encountered.

Patch by Amaury de la Vieuville.

llvm-svn: 183011
 2013-05-31 15:55:51 +00:00
Tim Northover 1bb672da81 ARM: fix VEXT encoding corner case 
The disassembly of VEXT instructions was too lax in the bits checked. This
fixes the case where the instruction affects Q-registers but a misaligned lane
was specified (should be UNDEFINED).

Patch by Amaury de la Vieuville

llvm-svn: 183003
 2013-05-31 13:47:25 +00:00
Rafael Espindola 99bd2ae479 Revert r182937 and r182877. 
r182877 broke MCJIT tests on ARM and r182937 was working around another failure
by r182877.

This should make the ARM bots green.

llvm-svn: 182960
 2013-05-30 20:37:52 +00:00
Andrew Trick ad6d08ac6f Order CALLSEQ_START and CALLSEQ_END nodes. 
Fixes PR16146: gdb.base__call-ar-st.exp fails after
pre-RA-sched=source fixes.

Patch by Xiaoyi Guo!

This also fixes an unsupported dbg.value test case. Codegen was
previously incorrect but the test was passing by luck.

llvm-svn: 182885
 2013-05-29 22:03:55 +00:00
JF Bastien f60e0e44ca Enable FastISel on ARM for Linux and NaCl 
FastISel was only enabled for iOS ARM and Thumb2, this patch enables it
for ARM (not Thumb2) on Linux and NaCl.

Thumb2 support needs a bit more work, mainly around register class
restrictions.

The patch punts to SelectionDAG when doing TLS relocation on non-Darwin
targets. I will fix this and other FastISel-to-SelectionDAG failures in
a separate patch.

The patch also forces FastISel to retain frame pointers: iOS always
keeps them for backtracking (so emitted code won't change because of
this), but Linux was getting much worse code that was incorrect when
using big frames (such as test-suite's lencod). I'll also fix this in a
later patch, it will probably require a peephole so that FastISel
doesn't rematerialize frame pointers back-to-back.

The test changes are straightforward, similar to:
  http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20130513/174279.html
They also add a vararg test that got dropped in that change.

I ran all of test-suite on A15 hardware with --optimize-option=-O0 and
all the tests pass.

llvm-svn: 182877
 2013-05-29 20:38:10 +00:00
JF Bastien 13969d0ab6 Tidy some register classes for ARM and Thumb 
Tidy up three places where the register class for ARM and Thumb wasn't
restrictive enough:
 - No PC dest for reg-reg add/orr/sub.
 - No PC dest for shifts.
 - No PC or SP for Thumb2 reg-imm add.

I encountered this while combining FastISel with
-verify-machineinstrs. These instructions defined registers whose
classes weren't restrictive enough, and the uses failed
verification. They're also undefined in the ISA, or would produce code
that FastISel wouldn't want. This doesn't fix the register class
narrowing issue (where uses should restrict definitions), and isn't
thorough, but it's a small step in the right direction.

llvm-svn: 182863
 2013-05-29 15:45:47 +00:00
Andrew Trick ef9de2a739 Track IR ordering of SelectionDAG nodes 2/4. 
Change SelectionDAG::getXXXNode() interfaces as well as call sites of
these functions to pass in SDLoc instead of DebugLoc.

llvm-svn: 182703
 2013-05-25 02:42:55 +00:00
Quentin Colombet f482805c28 Follow up of the introduction of MCSymbolizer. 
- Ressurect old MCDisassemble API to soften transition.
- Extend MCTargetDesc to set target specific symbolizer.

llvm-svn: 182688
 2013-05-24 22:51:52 +00:00
Michael J. Spencer df1ecbd734 Replace Count{Leading,Trailing}Zeros_{32,64} with count{Leading,Trailing}Zeros. 
llvm-svn: 182680
 2013-05-24 22:23:49 +00:00
Benjamin Kramer 534d3a4670 Remove the Copied parameter from MemoryObject::readBytes. 
There was exactly one caller using this API right, the others were relying on
specific behavior of the default implementation. Since it's too hard to use it
right just remove it and standardize on the default behavior.

Defines away PR16132.

llvm-svn: 182636
 2013-05-24 10:54:58 +00:00
Ahmed Bougacha aa79068157 MC: Disassembled CFG reconstruction. 
This patch builds on some existing code to do CFG reconstruction from
a disassembled binary:
- MCModule represents the binary, and has a list of MCAtoms.
- MCAtom represents either disassembled instructions (MCTextAtom), or
  contiguous data (MCDataAtom), and covers a specific range of addresses.
- MCBasicBlock and MCFunction form the reconstructed CFG. An MCBB is
  backed by an MCTextAtom, and has the usual successors/predecessors.
- MCObjectDisassembler creates a module from an ObjectFile using a
  disassembler. It first builds an atom for each section. It can also
  construct the CFG, and this splits the text atoms into basic blocks.

MCModule and MCAtom were only sketched out; MCFunction and MCBB were
implemented under the experimental "-cfg" llvm-objdump -macho option.
This cleans them up for further use; llvm-objdump -d -cfg now generates
graphviz files for each function found in the binary.

In the future, MCObjectDisassembler may be the right place to do
"intelligent" disassembly: for example, handling constant islands is just
a matter of splitting the atom, using information that may be available
in the ObjectFile. Also, better initial atom formation than just using
sections is possible using symbols (and things like Mach-O's
function_starts load command).

This brings two minor regressions in llvm-objdump -macho -cfg:
- The printing of a relocation's referenced symbol.
- An annotation on loop BBs, i.e., which are their own successor.

Relocation printing is replaced by the MCSymbolizer; the basic CFG
annotation will be superseded by more related functionality.

llvm-svn: 182628
 2013-05-24 01:07:04 +00:00
Ahmed Bougacha ad1084de84 Add MCSymbolizer for symbolic/annotated disassembly. 
This is a basic first step towards symbolization of disassembled
instructions. This used to be done using externally provided (C API)
callbacks. This patch introduces:
- the MCSymbolizer class, that mimics the same functions that were used
  in the X86 and ARM disassemblers to symbolize immediate operands and
  to annotate loads based off PC (for things like c string literals).
- the MCExternalSymbolizer class, which implements the old C API.
- the MCRelocationInfo class, which provides a way for targets to
  translate relocations (either object::RelocationRef, or disassembler
  C API VariantKinds) to MCExprs.
- the MCObjectSymbolizer class, which does symbolization using what it
  finds in an object::ObjectFile. This makes simple symbolization (with
  no fancy relocation stuff) work for all object formats!
- x86-64 Mach-O and ELF MCRelocationInfos.
- A basic ARM Mach-O MCRelocationInfo, that provides just enough to
  support the C API VariantKinds.

Most of what works in otool (the only user of the old symbolization API
that I know of) for x86-64 symbolic disassembly (-tvV) works, namely:
- symbol references: call _foo; jmp 15 <_foo+50>
- relocations:       call _foo-_bar; call _foo-4
- __cf?string:       leaq 193(%rip), %rax ## literal pool for "hello"
Stub support is the main missing part (because libObject doesn't know,
among other things, about mach-o indirect symbols).

As for the MCSymbolizer API, instead of relying on the disassemblers
to call the tryAdding* methods, maybe this could be done automagically
using InstrInfo? For instance, even though PC-relative LEAs are used
to get the address of string literals in a typical Mach-O file, a MOV
would be used in an ELF file. And right now, the explicit symbolization
only recognizes PC-relative LEAs. InstrInfo should have already have
most of what is needed to know what to symbolize, so this can
definitely be improved.

I'd also like to remove object::RelocationRef::getValueString (it seems
only used by relocation printing in objdump), as simply printing the
created MCExpr is definitely enough (and cleaner than string concats).

llvm-svn: 182625
 2013-05-24 00:39:57 +00:00
Tim Northover bc93308489 ARM: implement @llvm.readcyclecounter intrinsic 
This implements the @llvm.readcyclecounter intrinsic as the specific
MRC instruction specified in the ARM manuals for CPUs with the Power
Management extensions.

Older CPUs had slightly different methods which may also have to be
implemented eventually, but this should cover all v7 cases.

rdar://problem/13939186

llvm-svn: 182603
 2013-05-23 19:11:20 +00:00
Tim Northover cedd48183f ARM: Add Performance Monitor Extensions feature 
Performance monitors, including a basic cycle counter, are an official
extension in the ARMv7 specification. This adds support for enabling and
disabling them, orthogonally from CPU selection.

rdar://problem/13939186

llvm-svn: 182602
 2013-05-23 19:11:14 +00:00
Chad Rosier abdb1d69ab Simplify logic now that r182490 is in place. No functional change intended. 
llvm-svn: 182531
 2013-05-22 23:17:36 +00:00
Mihai Popa f41e3f56a5 VSTn instructions have a number of encoding constraints which are not implemented. I have added these using wrapper methods around the original custom decoder (incidentally - this is a huge poorly written method that should be cleaned up. I have left it as is since the changes would be much to hard to review). 
llvm-svn: 182281
 2013-05-20 14:57:05 +00:00
Mihai Popa dcf0922720 Q registers are encoded in fields of the same length as D registers. As Q registers are half as many, the ARM reference manual mandates the least significant bit to be zeroed out. Failure to do so should result in an undefined instruction. With this change test/MC/Disassembler/ARM/invalid-VQADD-arm.txt is passing (removed XFAIL). 
llvm-svn: 182279
 2013-05-20 14:42:43 +00:00
Stepan Dyatkovskiy d0e34a200f PR15868 fix. 
Introduction:
In case when stack alignment is 8 and GPRs parameter part size is not N*8:
we add padding to GPRs part, so part's last byte must be recovered at
address K*8-1.
We need to do it, since remained (stack) part of parameter starts from
address K*8, and we need to "attach" "GPRs head" without gaps to it:

Stack:
|---- 8 bytes block ----| |---- 8 bytes block ----| |---- 8 bytes...
[ [padding] [GPRs head] ] [ ------ Tail passed via stack  ------ ...

FIX:
Note, once we added padding we need to correct *all* Arg offsets that are going
after padded one. That's why we need this fix: Arg offsets were never corrected
before this patch. See new test-cases included in patch.

We also don't need to insert padding for byval parameters that are stored in GPRs
only. We need pad only last byval parameter and only in case it outsides GPRs
and stack alignment = 8.
Though, stack area, allocated for recovered byval params, must satisfy
"Size mod 8 = 0" restriction.

This patch reduces stack usage for some cases:
We can reduce ArgRegsSaveArea since inner N*4 bytes sized byval params my be
"packed" with alignment 4 in some cases.

llvm-svn: 182237
 2013-05-20 08:01:34 +00:00
Benjamin Kramer 8bad66e586 Replace some bit operations with simpler ones. No functionality change. 
llvm-svn: 182226
 2013-05-19 22:01:57 +00:00
Matt Arsenault 75865923c9 Add LLVMContext argument to getSetCCResultType 
llvm-svn: 182180
 2013-05-18 00:21:46 +00:00
JF Bastien 97b08c404c Support unaligned load/store on more ARM targets 
This patch matches GCC behavior: the code used to only allow unaligned
load/store on ARM for v6+ Darwin, it will now allow unaligned load/store
for v6+ Darwin as well as for v7+ on Linux and NaCl.

The distinction is made because v6 doesn't guarantee support (but LLVM
assumes that Apple controls hardware+kernel and therefore have
conformant v6 CPUs), whereas v7 does provide this guarantee (and
Linux/NaCl behave sanely).

The patch keeps the -arm-strict-align command line option, and adds
-arm-no-strict-align. They behave similarly to GCC's -mstrict-align and
-mnostrict-align.

I originally encountered this discrepancy in FastIsel tests which expect
unaligned load/store generation. Overall this should slightly improve
performance in most cases because of reduced I$ pressure.

llvm-svn: 182175
 2013-05-17 23:49:01 +00:00
Derek Schuff 36f00d9f02 Revert "Support unaligned load/store on more ARM targets" 
This reverts r181898.

llvm-svn: 181944
 2013-05-15 23:07:43 +00:00
Derek Schuff 72ddaba785 Support unaligned load/store on more ARM targets 
This patch matches GCC behavior: the code used to only allow unaligned
load/store on ARM for v6+ Darwin, it will now allow unaligned load/store for
v6+ Darwin as well as for v7+ on other targets.

The distinction is made because v6 doesn't guarantee support (but LLVM assumes
that Apple controls hardware+kernel and therefore have conformant v6 CPUs),
whereas v7 does provide this guarantee (and Linux behaves sanely).

Overall this should slightly improve performance in most cases because of
reduced I$ pressure.

Patch by JF Bastien

llvm-svn: 181897
 2013-05-15 16:08:30 +00:00
Arnold Schwaighofer af85f6083a ARM ISel: Don't create illegal types during LowerMUL 
The transformation happening here is that we want to turn a
"mul(ext(X), ext(X))" into a "vmull(X, X)", stripping off the extension. We have
to make sure that X still has a valid vector type - possibly recreate an
extension to a smaller type. In case of a extload of a memory type smaller than
64 bit we used create a ext(load()). The problem with doing this - instead of
recreating an extload - is that an illegal type is exposed.

This patch fixes this by creating extloads instead of ext(load()) sequences.

Fixes PR15970.

radar://13871383

llvm-svn: 181842
 2013-05-14 22:33:24 +00:00
Mihai Popa dc1764c5a4 The purpose of the patch is to fix the syntax of ARM mrc and mrc2 instructions when they are used to write to the APSR. In this case, the destination operand should be APSR_nzcv, and the encoding of the target should be 0b1111 (same as for PC). In pre-UAL syntax, this form used the PC register as a textual target. This is still allowed for backward compatibility. 
llvm-svn: 181705
 2013-05-13 14:10:04 +00:00
Lang Hames 67c09b3f88 Correctly preserve the input chain for potential tailcall nodes whose 
return values are bitcasts.

The chain had previously been being clobbered with the entry node to
the dag, which sometimes caused other code in the function to be
erroneously deleted when tailcall optimization kicked in.

<rdar://problem/13827621>

llvm-svn: 181696
 2013-05-13 10:21:19 +00:00
Rafael Espindola 227144c23c Remove the MachineMove class. 
It was just a less powerful and more confusing version of
MCCFIInstruction. A side effect is that, since MCCFIInstruction uses
dwarf register numbers, calls to getDwarfRegNum are pushed out, which
should allow further simplifications.

I left the MachineModuleInfo::addFrameMove interface unchanged since
this patch was already fairly big.

llvm-svn: 181680
 2013-05-13 01:16:13 +00:00
Rafael Espindola 140a837acd Remove unused argument. 
llvm-svn: 181618
 2013-05-10 18:16:59 +00:00
Logan Chien 4ea23b56c5 Implement AsmParser for ARM unwind directives. 
This commit implements the AsmParser for fnstart, fnend,
cantunwind, personality, handlerdata, pad, setfp, save, and
vsave directives.

This commit fixes some minor issue in the ARMELFStreamer:

* The switch back to corresponding section after the .fnend
  directive.

* Emit the unwind opcode while processing .fnend directive
  if there is no .handlerdata directive.

* Emit the unwind opcode to .ARM.extab while processing
  .handlerdata even if .personality directive does not exist.

llvm-svn: 181603
 2013-05-10 16:17:24 +00:00
Stepan Dyatkovskiy 2703bcaad3 For r181148: fixed warning 'enumeral and non-enumeral type in conditional expression'. 
llvm-svn: 181437
 2013-05-08 14:51:27 +00:00
Evan Cheng 9fad6352d4 ARM AnalyzeBranch should conservatively return true when it sees a predicated 
indirect branch at the end of the BB. Otherwise if-converter, branch folding
pass may incorrectly update its successor info if it consider BB as fallthrough
to the next BB.

rdar://13782395

llvm-svn: 181161
 2013-05-05 18:06:32 +00:00
Stepan Dyatkovskiy 8c02c98259 For ARM backend, fixed "byval" attribute support. 
Now even the small structures could be passed within byval (small enough
to be stored in GPRs).
In regression tests next function prototypes are checked:

PR15293:
  %artz = type { i32 }
  define void @foo(%artz* byval %s)
  define void @foo2(%artz* byval %s, i32 %p, %artz* byval %s2)
foo: "s" stored in R0
foo2: "s" stored in R0, "s2" stored in R2.

Next AAPCS rules are checked:
5.5 Parameters Passing, C.4 and C.5,
"ParamSize" is parameter size in 32bit words:
-- NSAA != 0, NCRN < R4 and NCRN+ParamSize > R4.
   Parameter should be sent to the stack; NCRN := R4.
-- NSAA != 0, and NCRN < R4, NCRN+ParamSize < R4.
   Parameter stored in GPRs; NCRN += ParamSize.

llvm-svn: 181148
 2013-05-05 07:48:36 +00:00
Dmitri Gribenko 3238fb7595 Add ArrayRef constructor from None, and do the cleanups that this constructor enables 
Patch by Robert Wilhelm.

llvm-svn: 181138
 2013-05-05 00:40:33 +00:00
Amara Emerson d9104c0359 Revert r181009. 
llvm-svn: 181079
 2013-05-03 23:57:17 +00:00
Amara Emerson 2f54d9fe10 Add support for reading ARM ELF build attributes. 
Build attribute sections can now be read if they exist via ELFObjectFile, and
the llvm-readobj tool has been extended with an option to dump this information
if requested. Regression tests are also included which exercise these features.

Also update the docs with a fixed ARM ABI link and a new link to the Addenda
which provides the build attributes specification.

llvm-svn: 181009
 2013-05-03 11:36:35 +00:00
Rafael Espindola 789a1c8a23 Text files should not be marked executable. 
Patch by Oliver Pinter.

llvm-svn: 180797
 2013-04-30 19:06:15 +00:00
Mihai Popa af22d91af0 s tightens up the encoding description for ARM post-indexed ldr instructions. All instructions in this class have bit 4 cleared. It turns out that there is a test case for this, but it was marked XFAIL. 
llvm-svn: 180778
 2013-04-30 09:00:12 +00:00
Stepan Dyatkovskiy f5aa83dbb0 Refactoring patch. 
1. VarArgStyleRegisters: functionality that emits "store" instructions for byval regs moved out into separated method "StoreByValRegs". Before this patch VarArgStyleRegisters had confused use-cases. It was used for both variadic functions and for regular functions with byval parameters. In last case it created new stack-frame and registered it as VarArg frame, that is wrong.

This patch replaces VarArgsStyleRegisters usage for byval parameters with StoreByValRegs method.

2. In ARMMachineFunctionInfo, "get/setVarArgsRegSaveSize" was renamed to "get/setArgRegsSaveSize". By the same reason. Sometimes it was used for variadic functions, and sometimes for byval parameters in regular functions. Actually, this property means the size of registers, that keeps arguments, and thats why it was renamed.

3. In ARMISelLowering.cpp, ARMTargetLowering class, in methods computeRegArea and StoreByValRegs, VARegXXXXXX was renamed to ArgRegsXXXXXX still by the same reasons.

llvm-svn: 180774
 2013-04-30 07:19:58 +00:00
Quentin Colombet a83d5e9f91 ARM: Fix encoding of hint instruction for Thumb. 
"hint" space for Thumb actually overlaps the encoding space of the CPS
instruction. In actuality, hints can be defined as CPS instructions where imod
and M bits are all nil.

Handle decoding of permitted nop-compatible hints (i.e. nop, yield, wfi, wfe,
sev) in DecodeT2CPSInstruction.

This commit adds a proper diagnostic message for Imm0_4 and updates all tests.

Patch by Mihail Popa <Mihail.Popa@arm.com>.

llvm-svn: 180617
 2013-04-26 17:54:54 +00:00
Benjamin Kramer ae81474a38 ARM/NEON: Pattern match vector integer abs to vabs. 
llvm-svn: 180604
 2013-04-26 15:00:57 +00:00
Arnold Schwaighofer 9881dcf2f2 ARM cost model: Integer div and rem is lowered to a function call 
Reflect this in the cost model. I observed this in MiBench/consumer-lame.

radar://13354716

llvm-svn: 180576
 2013-04-25 21:16:18 +00:00
Stephen Lin 8118e0b588 Add more tests for r179925 to verify correct handling of signext/zeroext; strengthen condition check to require actual MVT::i32 virtual register types, just in case (no actual functionality change) 
llvm-svn: 180138
 2013-04-23 19:42:25 +00:00
Stephen Lin 4eedb29b05 Lowercase "is" boolean variable prefix for consistency within function, no functionality change. 
llvm-svn: 180136
 2013-04-23 19:30:12 +00:00
Eric Christopher cc2cfe426d No really, don't store anything to this since it's unconditionally 
set below.

llvm-svn: 180015
 2013-04-22 14:11:25 +00:00
Eric Christopher 6647fb2c60 Remove variable store that is never read. 
llvm-svn: 180014
 2013-04-22 13:51:44 +00:00
Stepan Dyatkovskiy f80f9513ce Fix for 5.5 Parameter Passing --> Stage C: 
-- C.4 and C.5 statements, when NSAA is not equal to SP.
 -- C.1.cp statement for VA functions. Note: There are no VFP CPRCs in a
    variadic procedure.

Before this patch "NSAA != 0" means "don't use GPRs anymore ". But there are
some exceptions in AAPCS.
1. For non VA function: allocate all VFP regs for CPRC. When all VFPs are allocated
   CPRCs would be sent to stack, while non CPRCs may be still allocated in GRPs.
2. Check that for VA functions all params uses GPRs and then stack.
   No exceptions, no CPRCs here.

llvm-svn: 180011
 2013-04-22 13:06:52 +00:00
Jim Grosbach 563983c8a3 Legalize vector truncates by parts rather than just splitting. 
Rather than just splitting the input type and hoping for the best, apply
a bit more cleverness. Just splitting the types until the source is
legal often leads to an illegal result time, which is then widened and a
scalarization step is introduced which leads to truly horrible code
generation. With the loop vectorizer, these sorts of operations are much
more common, and so it's worth extra effort to do them well.

Add a legalization hook for the operands of a TRUNCATE node, which will
be encountered after the result type has been legalized, but if the
operand type is still illegal. If simple splitting of both types
ends up with the result type of each half still being legal, just
do that (v16i16 -> v16i8 on ARM, for example). If, however, that would
result in an illegal result type (v8i32 -> v8i8 on ARM, for example),
we can get more clever with power-two vectors. Specifically,
split the input type, but also widen the result element size, then
concatenate the halves and truncate again.  For example on ARM,
To perform a "%res = v8i8 trunc v8i32 %in" we transform to:
  %inlo = v4i32 extract_subvector %in, 0
  %inhi = v4i32 extract_subvector %in, 4
  %lo16 = v4i16 trunc v4i32 %inlo
  %hi16 = v4i16 trunc v4i32 %inhi
  %in16 = v8i16 concat_vectors v4i16 %lo16, v4i16 %hi16
  %res = v8i8 trunc v8i16 %in16

This allows instruction selection to generate three VMOVN instructions
instead of a sequences of moves, stores and loads.

Update the ARMTargetTransformInfo to take this improved legalization
into account.

Consider the simplified IR:

define <16 x i8> @test1(<16 x i32>* %ap) {
  %a = load <16 x i32>* %ap
  %tmp = trunc <16 x i32> %a to <16 x i8>
  ret <16 x i8> %tmp
}

define <8 x i8> @test2(<8 x i32>* %ap) {
  %a = load <8 x i32>* %ap
  %tmp = trunc <8 x i32> %a to <8 x i8>
  ret <8 x i8> %tmp
}

Previously, we would generate the truly hideous:
	.syntax unified
	.section	__TEXT,__text,regular,pure_instructions
	.globl	_test1
	.align	2
_test1:                                 @ @test1
@ BB#0:
	push	{r7}
	mov	r7, sp
	sub	sp, sp, #20
	bic	sp, sp, #7
	add	r1, r0, #48
	add	r2, r0, #32
	vld1.64	{d24, d25}, [r0:128]
	vld1.64	{d16, d17}, [r1:128]
	vld1.64	{d18, d19}, [r2:128]
	add	r1, r0, #16
	vmovn.i32	d22, q8
	vld1.64	{d16, d17}, [r1:128]
	vmovn.i32	d20, q9
	vmovn.i32	d18, q12
	vmov.u16	r0, d22[3]
	strb	r0, [sp, #15]
	vmov.u16	r0, d22[2]
	strb	r0, [sp, #14]
	vmov.u16	r0, d22[1]
	strb	r0, [sp, #13]
	vmov.u16	r0, d22[0]
	vmovn.i32	d16, q8
	strb	r0, [sp, #12]
	vmov.u16	r0, d20[3]
	strb	r0, [sp, #11]
	vmov.u16	r0, d20[2]
	strb	r0, [sp, #10]
	vmov.u16	r0, d20[1]
	strb	r0, [sp, #9]
	vmov.u16	r0, d20[0]
	strb	r0, [sp, #8]
	vmov.u16	r0, d18[3]
	strb	r0, [sp, #3]
	vmov.u16	r0, d18[2]
	strb	r0, [sp, #2]
	vmov.u16	r0, d18[1]
	strb	r0, [sp, #1]
	vmov.u16	r0, d18[0]
	strb	r0, [sp]
	vmov.u16	r0, d16[3]
	strb	r0, [sp, #7]
	vmov.u16	r0, d16[2]
	strb	r0, [sp, #6]
	vmov.u16	r0, d16[1]
	strb	r0, [sp, #5]
	vmov.u16	r0, d16[0]
	strb	r0, [sp, #4]
	vldmia	sp, {d16, d17}
	vmov	r0, r1, d16
	vmov	r2, r3, d17
	mov	sp, r7
	pop	{r7}
	bx	lr

	.globl	_test2
	.align	2
_test2:                                 @ @test2
@ BB#0:
	push	{r7}
	mov	r7, sp
	sub	sp, sp, #12
	bic	sp, sp, #7
	vld1.64	{d16, d17}, [r0:128]
	add	r0, r0, #16
	vld1.64	{d20, d21}, [r0:128]
	vmovn.i32	d18, q8
	vmov.u16	r0, d18[3]
	vmovn.i32	d16, q10
	strb	r0, [sp, #3]
	vmov.u16	r0, d18[2]
	strb	r0, [sp, #2]
	vmov.u16	r0, d18[1]
	strb	r0, [sp, #1]
	vmov.u16	r0, d18[0]
	strb	r0, [sp]
	vmov.u16	r0, d16[3]
	strb	r0, [sp, #7]
	vmov.u16	r0, d16[2]
	strb	r0, [sp, #6]
	vmov.u16	r0, d16[1]
	strb	r0, [sp, #5]
	vmov.u16	r0, d16[0]
	strb	r0, [sp, #4]
	ldm	sp, {r0, r1}
	mov	sp, r7
	pop	{r7}
	bx	lr

Now, however, we generate the much more straightforward:
	.syntax unified
	.section	__TEXT,__text,regular,pure_instructions
	.globl	_test1
	.align	2
_test1:                                 @ @test1
@ BB#0:
	add	r1, r0, #48
	add	r2, r0, #32
	vld1.64	{d20, d21}, [r0:128]
	vld1.64	{d16, d17}, [r1:128]
	add	r1, r0, #16
	vld1.64	{d18, d19}, [r2:128]
	vld1.64	{d22, d23}, [r1:128]
	vmovn.i32	d17, q8
	vmovn.i32	d16, q9
	vmovn.i32	d18, q10
	vmovn.i32	d19, q11
	vmovn.i16	d17, q8
	vmovn.i16	d16, q9
	vmov	r0, r1, d16
	vmov	r2, r3, d17
	bx	lr

	.globl	_test2
	.align	2
_test2:                                 @ @test2
@ BB#0:
	vld1.64	{d16, d17}, [r0:128]
	add	r0, r0, #16
	vld1.64	{d18, d19}, [r0:128]
	vmovn.i32	d16, q8
	vmovn.i32	d17, q9
	vmovn.i16	d16, q8
	vmov	r0, r1, d16
	bx	lr

llvm-svn: 179989
 2013-04-21 23:47:41 +00:00
Tim Northover 798697d662 ARM: Use ldrd/strd to spill 64-bit pairs when available. 
This allows common sp-offsets to be part of the instruction and is
probably faster on modern CPUs too.

llvm-svn: 179977
 2013-04-21 11:57:07 +00:00