RPCS3/llvm-mirror
1
0
Fork 0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-10-26 14:33:02 +02:00
Code Issues Projects Releases Wiki Activity
67,221 Commits 4 Branches 5 Tags 704 MiB
68135a728b
Commit Graph

43383 Commits

Author SHA1 Message Date
Jim Grosbach
1b1636f0cb ARM conditional mov encoding fix. 
llvm-svn: 119354
 2010-11-16 18:13:42 +00:00
Duncan Sands
74aeda71dd Have a few places that want to simplify phi nodes use SimplifyInstruction 
rather than calling hasConstantValue.  No intended functionality change.

llvm-svn: 119352
 2010-11-16 17:41:24 +00:00
Duncan Sands
825c7d7f79 In which I discover the existence of loops. Threading an operation 
over a phi node by applying it to each operand may be wrong if the
operation and the phi node are mutually interdependent (the testcase
has a simple example of this).  So only do this transform if it would
be correct to perform the operation in each predecessor of the block
containing the phi, i.e. if the other operands all dominate the phi.
This should fix the FFMPEG snow.c regression reported by İsmail Dönmez.

llvm-svn: 119347
 2010-11-16 12:16:38 +00:00
Rafael Espindola
47e7ef8c8f A bit more of gnu as compatibility when handling relocations with aliases. 
llvm-svn: 119328
 2010-11-16 04:11:46 +00:00
Bill Wendling
655c05147e L_bit doesn't work here. 
llvm-svn: 119325
 2010-11-16 02:20:22 +00:00
Bill Wendling
2ac47f0959 - Remove dead patterns. 
- Add encodings to the *LDMIA_RET instrs. Probably not needed...

llvm-svn: 119323
 2010-11-16 02:08:45 +00:00
Bill Wendling
64cb4dd72d vldm and vstm are mnemonics for vldmia and vstmia resp. 
llvm-svn: 119321
 2010-11-16 02:00:24 +00:00
Eric Christopher
954db96e84 Make the verifier a little quieter on instructions that it's probably 
(and likely) wrong about anyhow.

llvm-svn: 119320
 2010-11-16 01:58:21 +00:00
Chris Lattner
a47f6f03f9 Fix a bug I introduced in the ppc refactoring, which caused long 
branches to be emitted as:

 	bne cr0, 2
instead of:
 	bne cr0, $+8

llvm-svn: 119317
 2010-11-16 01:45:05 +00:00
Bill Wendling
b450d320ec Encode the multi-load/store instructions with their respective modes ('ia', 
'db', 'ib', 'da') instead of having that mode as a separate field in the
instruction. It's more convenient for the asm parser and much more readable for
humans.
<rdar://problem/8654088>

llvm-svn: 119310
 2010-11-16 01:16:36 +00:00
Michael J. Spencer
5c5cfac314 Object: Get rid of sys::Path, Triple, and cleanup namespace. 
llvm-svn: 119309
 2010-11-16 01:06:51 +00:00
Chris Lattner
af9558a178 add copy of comment to the code that will survive the mcjit'ization 
llvm-svn: 119308
 2010-11-16 00:57:32 +00:00
Chris Lattner
378215748c relax an assertion a bit, allowing the GPR argument of 
these instructions to be encoded with getMachineOpValue.
This unbreaks ExecutionEngine/2003-01-04-ArgumentBug.ll 
when running on a G5

llvm-svn: 119307
 2010-11-16 00:55:51 +00:00
Jakob Stoklund Olesen
5d9371952b Remember to resize SpillSlotToUsesMap when allocating an emergency spill slot. 
Use amazing new function call technology instead of writing identical code in
multiple places.

This fixes PR8604.

llvm-svn: 119306
 2010-11-16 00:41:01 +00:00
Jakob Stoklund Olesen
ea79975a68 Fix PR8612 in the standard spiller, take two. 
The live range of a register defined by an early clobber starts at the use slot,
not the def slot.

Except when it is an early clobber tied to a use operand. Then it starts at the
def slot like a standard def.

llvm-svn: 119305
 2010-11-16 00:40:59 +00:00
Owen Anderson
686fae59e6 Add Thumb2 encodings for mov and friends. 
llvm-svn: 119295
 2010-11-16 00:29:56 +00:00
Rafael Espindola
53a8ae85b7 Change the 11 byte nop to be a single instruction. 
llvm-svn: 119286
 2010-11-15 23:10:30 +00:00
Chris Lattner
24aff5bdbd fix a pasto that massively broke the ppc jit while the buildbots happened 
to be broken for other reasons

llvm-svn: 119283
 2010-11-15 22:50:50 +00:00
Jakob Stoklund Olesen
465c07cc66 Revert "Fix PR8612 in the standard spiller as well." 
This reverts r119183 which borke the buildbots.

llvm-svn: 119270
 2010-11-15 21:51:51 +00:00
Owen Anderson
3d47a3536c Attempt to provide encodings for some miscellaneous Thumb2 encodings. 
llvm-svn: 119187
 2010-11-15 21:30:39 +00:00
Evan Cheng
4afa3a6b1f Code clean up. The peephole pass should be the one updating the instruction 
iterator, not TII->OptimizeCompareInstr.

llvm-svn: 119186
 2010-11-15 21:20:45 +00:00
Owen Anderson
db1a4541c2 Provide Thumb2 encodings for sxtb and friends. 
llvm-svn: 119185
 2010-11-15 21:12:05 +00:00
Eric Christopher
22b01f0d76 Recommit this change and remove the failing part of the test - it didn't 
pass in the first place and was masked by earlier failures not warning
and aborting the block.

llvm-svn: 119184
 2010-11-15 21:11:06 +00:00
Jakob Stoklund Olesen
920856ab38 Fix PR8612 in the standard spiller as well. 
The live range of a register defined by an early clobber starts at the use slot,
not the def slot.

llvm-svn: 119183
 2010-11-15 20:55:53 +00:00
Jakob Stoklund Olesen
86fba16eb1 When spilling a register defined by an early clobber, make sure that the new 
live ranges for the spill register are also defined at the use slot instead of
the normal def slot.

This fixes PR8612 for the inline spiller. A use was being allocated to the same
register as a spilled early clobber def.

This problem exists in all the spillers. A fix for the standard spiller is
forthcoming.

llvm-svn: 119182
 2010-11-15 20:55:49 +00:00
Jim Grosbach
7aabb8c5ee ARM LDR_PRE/LDR_POST/STR_PRE/STR_POST (and the *B counterparts) binary encoding. 
llvm-svn: 119180
 2010-11-15 20:47:07 +00:00
Owen Anderson
0bb2efeb7d Add Thumb2 encodings for comparison and shift operators. 
llvm-svn: 119176
 2010-11-15 19:58:36 +00:00
Benjamin Kramer
8ee77c5320 Fix compiler warnigns. 
llvm-svn: 119175
 2010-11-15 19:20:50 +00:00
Owen Anderson
5ae5db8931 Add correct Thumb2 encodings for mvn and friends. 
llvm-svn: 119170
 2010-11-15 18:45:17 +00:00
Jim Grosbach
e0122f5d54 Add FIXMEs. 
llvm-svn: 119167
 2010-11-15 18:36:48 +00:00
Jim Grosbach
a7676f5079 Nuke redundant encoding bit set. 
llvm-svn: 119164
 2010-11-15 18:17:24 +00:00
Duncan Sands
ccdccb2776 Teach InstructionSimplify the trick of skipping incoming phi 
values that are equal to the phi itself.

llvm-svn: 119161
 2010-11-15 17:52:45 +00:00
Rafael Espindola
1f6440b1e2 Change MCExpr::EvaluateAsRelocatableImpl of variables to return the original 
variable if recursing fails to simplify it.

Factor AliasedSymbol to be a method of MCSymbol.

Update MCAssembler::EvaluateFixup to match the change in
EvaluateAsRelocatableImpl.

Remove the WeakRefExpr hack, as the object writer now sees the weakref with
no extra effort needed.

Nothing else is using MCTargetExpr, but keep it for now.

Now that the ELF writer sees relocations with aliases, handle

    .weak    foo2
foo2:
    .weak    bar2
    .set    bar2,foo2
    .quad    bar2

the same way gas does and produce a relocation with bar2.

llvm-svn: 119152
 2010-11-15 16:33:49 +00:00
Jason W Kim
8dc24d5066 Dovetail with Dan Dunbar's rework of ELFObjectWriter. 
Added 2 new subclasses - X86ELFObjectWriter and ARMELFObectWriter.
ARM and X86 require different code for RecordRelocation(), possibly others.

llvm-svn: 119149
 2010-11-15 16:18:39 +00:00
Rafael Espindola
f6c0121c93 Fix PR8565. 
This moves most of the isUsed logic to the MCSymbol itself. With this we
get a bit more relaxed about allowing definitions after uses: uses that
don't evaluate their argument immediately (jmp foo) are accepted.

ddunbar, this was the smallest compromise I could think of that lets us
accept gcc (and clang!) assembly.

llvm-svn: 119144
 2010-11-15 14:40:36 +00:00
Kalle Raiskila
917df2ee14 Improve code layout, mostly indentation. 
No functionality change.

llvm-svn: 119142
 2010-11-15 10:12:32 +00:00
Chris Lattner
c00f41ef21 Wire up primitive support in the assembler backend for writing .o files 
directly on the mac.  This is very early, doesn't support relocations and
has a terrible hack to avoid .machine from being printed, but despite
that it generates an bitwise-identical-to-cctools .o file for stuff like 
this:

  define i32 @test() nounwind { ret i32 42 }

I don't plan to continue pushing this forward, but if anyone else was
interested in doing it, it should be really straight-forward.

llvm-svn: 119136
 2010-11-15 08:49:58 +00:00
Chris Lattner
51793c5b92 split out an encoder for memri operands, allowing a relocation to be plopped 
into the immediate field.  This allows us to encode stuff like this:

        lbz r3, lo16(__ZL4init)(r4)     ; globalopt.cpp:5
                                        ; encoding: [0x88,0x64,A,A]
                                        ;   fixup A - offset: 0, value: lo16(__ZL4init), kind: fixup_ppc_lo16

        stw r3, lo16(__ZL1s)(r5)        ; globalopt.cpp:6
                                        ; encoding: [0x90,0x65,A,A]
                                        ;   fixup A - offset: 0, value: lo16(__ZL1s), kind: fixup_ppc_lo16

With this, we should have a completely function MCCodeEmitter for PPC, wewt.

llvm-svn: 119134
 2010-11-15 08:22:03 +00:00
Chris Lattner
96f8078924 add support for encoding the lo14 forms used for a few PPC64 addressing 
modes.  For example, we now get:

	ld r3, lo16(_G)(r3)             ; encoding: [0xe8,0x63,A,0bAAAAAA00]
                                        ;   fixup A - offset: 0, value: lo16(_G), kind: fixup_ppc_lo14

llvm-svn: 119133
 2010-11-15 08:02:41 +00:00
Chris Lattner
a028364bcd fix a regression with the new instprinter: we lost the ability to 
print DBG_VALUE instructions.  This should unbreak the llvm-gcc-powerpc-darwin9
buildbot.

llvm-svn: 119132
 2010-11-15 07:52:06 +00:00
Chris Lattner
ac5f6ff408 implement the start of support for lo16 and ha16, allowing us to get stuff like: 
lis r4, ha16(__ZL4init)         ; encoding: [0x3c,0x80,A,A]
                                        ;   fixup A - offset: 0, value: ha16(__ZL4init), kind: fixup_ppc_ha16

llvm-svn: 119127
 2010-11-15 06:33:39 +00:00
Chris Lattner
f04cb8ebad add a fixup for conditional branches, giving us output like this: 
beq cr0, LBB0_4                 ; encoding: [0x41,0x82,A,0bAAAAAA00]
                                        ;   fixup A - offset: 0, value: LBB0_4, kind: fixup_ppc_brcond14

llvm-svn: 119126
 2010-11-15 06:12:22 +00:00
Chris Lattner
8b6744a612 change direct branches to encode with the same encoding method 
as direct calls.  Change conditional branches to encode with
their own method, simplifying the JIT encoder and making room
for adding an mc fixup.

llvm-svn: 119125
 2010-11-15 06:09:35 +00:00
Chris Lattner
4795a876f6 eliminate a now-unneeded operand printer. 
llvm-svn: 119124
 2010-11-15 06:01:10 +00:00
Chris Lattner
37bebc344a split call operands out to their own encoding class, simplifying 
code in the JIT.  Use this to form the first fixup for the PPC backend,
giving us stuff like this:

	bl L_foo$stub ; encoding: [0b010010AA,A,A,0bAAAAAA01]
                                        ;   fixup A - offset: 0, value: L_foo$stub, kind: fixup_ppc_br24

llvm-svn: 119123
 2010-11-15 05:57:53 +00:00
Chris Lattner
3b5f2a871a correct the fixup comment printer to work on big endian platforms. 
llvm-svn: 119122
 2010-11-15 05:56:19 +00:00
Chris Lattner
6eb2a0b277 add proper encoding for MTCRF instead of using a hack. 
llvm-svn: 119121
 2010-11-15 05:19:25 +00:00
Chris Lattner
b2daeac125 add fields to the .td files unconditionally, simplifying tblgen a bit. 
Switch the ARM backend to use 'let' instead of 'set' with this change.

llvm-svn: 119120
 2010-11-15 05:19:05 +00:00
Chris Lattner
6f23a467b0 add basic encoding support for immediates and registers, allowing us 
to encode all of these instructions correctly (for example):

        mflr r0                         ; encoding: [0x7c,0x08,0x02,0xa6]
        stw r0, 8(r1)                   ; encoding: [0x90,0x01,0x00,0x08]
        stwu r1, -64(r1)                ; encoding: [0x94,0x21,0xff,0xc0]

llvm-svn: 119118
 2010-11-15 04:51:55 +00:00
Chris Lattner
ca7d8427c9 add a dummy entry to fix a build error 
llvm-svn: 119117
 2010-11-15 04:47:16 +00:00
Chris Lattner
d5a087b439 Implement a basic MCCodeEmitter for PPC. This doesn't handle 
fixups yet, and doesn't handle actually encoding operand values,
but this is enough for llc -show-mc-encoding to show the base
instruction encoding information, e.g.:

	mflr r0                         ; encoding: [0x7c,0x08,0x02,0xa6]
	stw r0, 8(r1)                   ; encoding: [0x90,0x00,0x00,0x00]
	stwu r1, -64(r1)                ; encoding: [0x94,0x00,0x00,0x00]
Ltmp0:
	lhz r4, 4(r3)                   ; encoding: [0xa0,0x00,0x00,0x00]
	cmplwi cr0, r4, 8               ; encoding: [0x28,0x00,0x00,0x00]
	beq cr0, LBB0_2                 ; encoding: [0x40,0x00,0x00,0x00]

llvm-svn: 119116
 2010-11-15 04:16:32 +00:00
Chris Lattner
d79d1b6297 dissolve some more hacks. 
llvm-svn: 119115
 2010-11-15 03:53:53 +00:00
Chris Lattner
3ee55a9b8d fix some fixme's, removing dead code. 
llvm-svn: 119114
 2010-11-15 03:51:13 +00:00
Chris Lattner
8fa38e3222 remove asmstrings (which can never be printed) from pseudo 
instructions, allowing is to eliminate some dead operand 
printing methods from the instprinter.

llvm-svn: 119113
 2010-11-15 03:48:58 +00:00
Chris Lattner
5f571f5a13 strength reduce TOC temp label generation, no functionality change. 
llvm-svn: 119112
 2010-11-15 03:42:54 +00:00
Chris Lattner
38e49f3f11 rip out a ton of old instruction printing junk now that the 
new instprinting logic is there.

llvm-svn: 119111
 2010-11-15 03:39:06 +00:00
Michael J. Spencer
b179ffafea Object: Add missing makefile. 
llvm-svn: 119110
 2010-11-15 03:33:14 +00:00
Evan Cheng
73e79b8a9c Make sure ARM multi load / store pass copies memoperands when forming ldrd / strd. pr8113. 
llvm-svn: 119109
 2010-11-15 03:30:30 +00:00
Chris Lattner
9ffced5b9d Turn on the new instprinter by default. 
The only change in the output is:

1) we get a better comment on mfcr, we get:
	mfcr r2                         ; cr2
  instead of:
 	mfcr r2 ; 32

2) we no longer emit $stub's on powerpc/leopard.  The Leopard
   linker autosynthesizes them.

llvm-svn: 119108
 2010-11-15 03:27:05 +00:00
Michael J. Spencer
23f31df887 Add LLVMObject Library. 
llvm-svn: 119107
 2010-11-15 03:21:41 +00:00
Chris Lattner
e50d695490 convert the operand bits into bitfields since they are all combinable in 
different ways.  Add $non_lazy_ptr support, and proper lowering for
global values.

Now all the ppc regression tests pass with the new instruction printer.

llvm-svn: 119106
 2010-11-15 03:13:19 +00:00
Chris Lattner
9a0a840839 add targetoperand flags for jump tables, constant pool and block address 
nodes to indicate when ha16/lo16 modifiers should be used.  This lets
us pass PowerPC/indirectbr.ll.

The one annoying thing about this patch is that the MCSymbolExpr isn't
expressive enough to represent ha16(label1-label2) which we need on
PowerPC.  I have a terrible hack in the meantime, but this will have
to be revisited at some point.

Last major conversion item left is global variable references.

llvm-svn: 119105
 2010-11-15 02:46:57 +00:00
Chris Lattner
7743db8f20 remove some extraneous quotes to make the new instprinter match. 
llvm-svn: 119104
 2010-11-15 02:43:46 +00:00
Chris Lattner
11e50c531a silence a ton of warnings from clang. 
llvm-svn: 119102
 2010-11-15 01:45:44 +00:00
Anton Korobeynikov
794259aec0 Attempt to unbreak cmake-based builds 
llvm-svn: 119098
 2010-11-15 00:48:12 +00:00
Anton Korobeynikov
76c52dcf44 First step of huge frame-related refactoring: move emit{Prologue,Epilogue} out of TargetRegisterInfo to TargetFrameInfo, which is definitely much better suitable place 
llvm-svn: 119097
 2010-11-15 00:06:54 +00:00
Anton Korobeynikov
33f85080b4 Whitespace cleanup 
llvm-svn: 119096
 2010-11-15 00:06:05 +00:00
Rafael Espindola
5572cb1aed Move the logic to decide with which symbol we produce a relocation (if any) to 
a central location. This also makes us a bit more compatible with gas.

llvm-svn: 119094
 2010-11-14 23:53:26 +00:00
Chris Lattner
3512e07ecf implement support for the MO_DARWIN_STUB TargetOperand flag, 
and have isel apply to to call operands as required.  This allows
us to get $stub suffixes on label references on ppc/tiger with the
new instprinter, fixing two tests.  Only 2 to go.

llvm-svn: 119093
 2010-11-14 23:42:06 +00:00
Chris Lattner
b347be831c tidy up, no functionality change. 
llvm-svn: 119092
 2010-11-14 23:32:42 +00:00
Chris Lattner
6af9d8969d with the picbase nonsense starting to be figured out, implement 
lowering support for MovePCtoLR[8].  Down to 4 failures again.

llvm-svn: 119090
 2010-11-14 22:56:43 +00:00
Chris Lattner
51168d6510 move the pic base symbol stuff up to MachineFunction 
since it is trivial and will be shared between ppc and x86.
This substantially simplifies the X86 backend also.

llvm-svn: 119089
 2010-11-14 22:48:15 +00:00
Chris Lattner
09ea2ce78f simplify getPICBaseSymbol a bit. 
llvm-svn: 119088
 2010-11-14 22:37:11 +00:00
Chris Lattner
fc626aa37d reimplement ppc asmprinter "toc" handling to use a VariantKind 
on the operand, required for .o file writing and fixing 
the PowerPC/mult-alt-generic-powerpc64.ll failure with the new
instprinter.

llvm-svn: 119087
 2010-11-14 22:22:59 +00:00
Chris Lattner
3c8d9ea286 lower PPC::MFCRpseud when transforming to MC, avoiding calling 
the aborting printSpecial() method.  This gets us to 8 failures.

llvm-svn: 119084
 2010-11-14 22:03:15 +00:00
Chris Lattner
36fad020cd make the stubbed-out printer methods abort instead of 
printing nothing.  This gets us back up to 24 failures.

llvm-svn: 119083
 2010-11-14 21:54:34 +00:00
Chris Lattner
331d966779 wire up a few more things, down to 4 test failures, all 
about handling $stub, lo/hi etc.

llvm-svn: 119082
 2010-11-14 21:51:37 +00:00
Chris Lattner
3959bd7fb4 properly wire up the instprinter to the ppc64 backend, down to 5 failures. 
llvm-svn: 119081
 2010-11-14 21:42:53 +00:00
Chris Lattner
27b65608d0 implement pretty printing support for the various pseudo 
ops the asmprinter supported, fixing PowerPC/rlwimi2.ll
among others.  Down to 20 failures.

llvm-svn: 119080
 2010-11-14 21:39:51 +00:00
Chris Lattner
437623476f Wire up symbol hi/lo printing. We don't print hi()/lo(), but this gets 
us further along.  Only 28 failures now.

llvm-svn: 119079
 2010-11-14 21:33:07 +00:00
Chris Lattner
b12ecaae5a implement basic support for symbol operand lowering, 
and printing support for call operands.  Down to 77 failures.

llvm-svn: 119078
 2010-11-14 21:20:46 +00:00
Chris Lattner
3bf85b2f4b trim #includes. 
llvm-svn: 119075
 2010-11-14 21:16:04 +00:00
Chris Lattner
32848508af switch PPC to a simplified MCInstLowering model. 
llvm-svn: 119074
 2010-11-14 21:12:33 +00:00
Rafael Espindola
59a93284d7 Fix another case of a .comm directive without a corresponding .type 
directive.

llvm-svn: 119073
 2010-11-14 21:11:16 +00:00
Chris Lattner
163fbca360 fix PPC.h to not pull in TargetMachine.h 
llvm-svn: 119072
 2010-11-14 21:09:28 +00:00
Chris Lattner
3a01d37b66 rename LowerToMCInst -> LowerARMMachineInstrToMCInst. 
llvm-svn: 119071
 2010-11-14 21:00:02 +00:00
Chris Lattner
4dac1b2742 even more simplifications. ARM MCInstLowering is now just 
a single function instead of a class.  It doesn't need the
complexity that X86 does.

llvm-svn: 119070
 2010-11-14 20:58:38 +00:00
Chris Lattner
3ebf1f315e more shrinkification 
llvm-svn: 119068
 2010-11-14 20:41:53 +00:00
Chris Lattner
758bc06aff more simplifications. 
llvm-svn: 119067
 2010-11-14 20:40:08 +00:00
Chris Lattner
2393da7c40 simplify and tidy up 
llvm-svn: 119066
 2010-11-14 20:31:06 +00:00
Chris Lattner
a6262cb13f implement basic support for memory operands and crbit operands, 
this fixes 3 more ppc tests.

llvm-svn: 119065
 2010-11-14 20:22:56 +00:00
Chris Lattner
3249cd7521 implement several trivial operand printers, reducing 
failures in CodeGen/PowerPC from 120 -> 117

llvm-svn: 119063
 2010-11-14 20:11:21 +00:00
Chris Lattner
8d769a7610 Implement support for printing register and immediate operands, 
add support for darwin vs aix syntax.  We now can print instructions
like this:

	add r3, r3, r4
	blr 

and (in aix mode):
	add 3, 3, 4
	blr 

llvm-svn: 119062
 2010-11-14 20:02:39 +00:00
Chris Lattner
e137166772 stub out PPCMCInstLowering, add a new option that uses it and the new 
instprinter when -enable-ppc-inst-printer is passed to llc.

llvm-svn: 119061
 2010-11-14 19:53:02 +00:00
Rafael Espindola
1884e6d475 Fix the type of a symbol created with .comm and no corresponding .type. 
llvm-svn: 119060
 2010-11-14 19:40:55 +00:00
Chris Lattner
d045c38bf6 stub out a powerpc MCInstPrinter implementation. 
llvm-svn: 119059
 2010-11-14 19:40:38 +00:00
Chris Lattner
cec3b226a4 move all the target's asmprinters into the main target. The piece 
that should be split out is the InstPrinter (if a target is mc'ized).
This change makes all the targets be consistent.

llvm-svn: 119056
 2010-11-14 18:43:56 +00:00
Duncan Sands
63e80e0593 If dom tree information is available, make it possible to pass 
it to get better phi node simplification.

llvm-svn: 119055
 2010-11-14 18:36:10 +00:00
Chris Lattner
38ca2c6f0f move PPCAsmPrinter into the main PPC library, like ARM and X86. 
llvm-svn: 119054
 2010-11-14 18:33:33 +00:00
Chris Lattner
7e03d38bd7 fix cmake 
llvm-svn: 119053
 2010-11-14 18:27:11 +00:00
Chris Lattner
4533444a49 remove JIT support from the Alpha backend. The JIT will be moving to MC, 
and the Alpha backend isn't MCized yet.  Approved by Andrew.

llvm-svn: 119051
 2010-11-14 18:24:41 +00:00
Chris Lattner
3e3e29381e random acts of tidiness. 
llvm-svn: 119049
 2010-11-14 18:09:50 +00:00
Duncan Sands
617030ad18 Teach InstructionSimplify about phi nodes. I chose to have it simply 
offload the work to hasConstantValue rather than do something more
complicated (such handling mutually recursive phis) because (1) it is
not clear it is worth it; and (2) if it is worth it, maybe such logic
would be better placed in hasConstantValue.  Adjust some GVN tests
which are now cleaned up much further (eg: all phi nodes are removed).

llvm-svn: 119043
 2010-11-14 13:30:18 +00:00
Duncan Sands
9a914e2b5e Boost the power of phi node constant folding slightly: if all 
operands are the phi node itself or undef, then return undef.
This logic already existed at a higher level so in practice it
shouldn't make the slightest difference.  Note that this code
could be replaced by a call to PN->hasConstantValue().  However
since we bail out the moment we see a non-constant operand, it
is more efficient to have a specialized version of that logic.

llvm-svn: 119041
 2010-11-14 12:53:18 +00:00
Duncan Sands
ab39985d42 Strip trailing whitespace. 
llvm-svn: 119038
 2010-11-14 11:23:23 +00:00
Owen Anderson
ad4f83df6a Second attempt at providing correct encodings for Thumb2 binary operators. 
llvm-svn: 119029
 2010-11-14 05:37:38 +00:00
Rafael Espindola
444d9c0b02 Handle a peculiar comdat case: Creating a section with an undefined 
signature symbol causes a local symbol to be created unless there is
some other use of the symbol.

llvm-svn: 119026
 2010-11-14 04:17:37 +00:00
Rafael Espindola
55e673b1c2 Simplify getSymbolIndexInSymbolTable by setting the actual index of 
the symbols.

llvm-svn: 119022
 2010-11-14 03:12:24 +00:00
Rafael Espindola
1b4c835948 Fix warning. 
llvm-svn: 119021
 2010-11-14 01:34:31 +00:00
Peter Collingbourne
4ec6b2dbe7 Recognise 32-bit ror-based bswap implementation used by uclibc 
llvm-svn: 119007
 2010-11-13 19:54:30 +00:00
Peter Collingbourne
6c3094234e Support ; as asm separator 
llvm-svn: 119006
 2010-11-13 19:54:23 +00:00
Duncan Sands
88fc6cd7fe Generalize the reassociation transform in SimplifyCommutative (now renamed to 
SimplifyAssociativeOrCommutative) "(A op C1) op C2" -> "A op (C1 op C2)",
which previously was only done if C1 and C2 were constants, to occur whenever
"C1 op C2" simplifies (a la InstructionSimplify).  Since the simplifying operand
combination can no longer be assumed to be the right-hand terms, consider all of
the possible permutations.  When compiling "gcc as one big file", transform 2
(i.e. using right-hand operands) fires about 4000 times but it has to be said
that most of the time the simplifying operands are both constants.  Transforms
3, 4 and 5 each fired once.  Transform 6, which is an existing transform that
I didn't change, never fired.  With this change, the testcase is now optimized
perfectly with one run of instcombine (previously it required instcombine +
reassociate + instcombine, and it may just have been luck that this worked).

llvm-svn: 119002
 2010-11-13 15:10:37 +00:00
Duncan Sands
6ff5e53731 Fix typo pointed out by Trevor Harmon. 
llvm-svn: 119001
 2010-11-13 12:16:27 +00:00
Bill Wendling
49dd03e223 Comment out the defms until they're activated. 
llvm-svn: 119000
 2010-11-13 11:20:05 +00:00
Bill Wendling
fadcb3cded Add uses of the *_ldst_multi multiclasses. These aren't used yet. 
llvm-svn: 118999
 2010-11-13 10:57:02 +00:00
Bill Wendling
184bc1368d Convert the modes to lower case. 
llvm-svn: 118998
 2010-11-13 10:43:34 +00:00
Bill Wendling
0a55425158 Minor cleanups: 
- Get the opcode once.
- Add a ParserMatchClass to reglist.

llvm-svn: 118997
 2010-11-13 10:40:19 +00:00
Bill Wendling
aa9ca6fcca Add *_ldst_mult multiclasses to the ARM back-end. These will be used in the 
future to separate out the ia, ib, da, db variants of the load/store multiple
instructions.

llvm-svn: 118995
 2010-11-13 09:09:38 +00:00
Daniel Dunbar
2ba9f79194 MC: Simplify Mach-O and ELF object writer implementations. 
- What was I thinking?????

llvm-svn: 118992
 2010-11-13 07:33:40 +00:00
Wesley Peck
4a2644f203 Fixed error and re-enabled MBlaze MC disassembler tests. 
llvm-svn: 118987
 2010-11-13 05:48:21 +00:00
Evan Cheng
239d9b439d Conditional moves are slightly more expensive than moves. 
llvm-svn: 118985
 2010-11-13 05:14:20 +00:00
Rafael Espindola
d2ff64abae Fix warning and add support for printing gnu_unique_object. 
llvm-svn: 118981
 2010-11-13 04:55:06 +00:00
Rafael Espindola
9cd0fd30e0 Parse and record the gnu_unique_object type. 
llvm-svn: 118980
 2010-11-13 04:51:02 +00:00
Rafael Espindola
00c5605efc Parse and remember discriminators in .loc line. I try to output them with 
another patch.
This lets us parse a bit more of the gcc 4.5 output.

llvm-svn: 118975
 2010-11-13 03:18:27 +00:00
Daniel Dunbar
fef193dc12 JIT: More nitty style tweakage, aka territory marking. 
llvm-svn: 118973
 2010-11-13 02:48:57 +00:00
Daniel Dunbar
1f695efa5e report_fatal_error: Simplify a possible ambiguity. 
llvm-svn: 118972
 2010-11-13 02:48:51 +00:00
Wesley Peck
c534d55751 1. Adding test cases for MBlaze MC disassembler. 
2. Fixing several errors in disassembler uncovered by test cases.
3. Fixing invalid encoding of PCMPEQ and PCMPNE uncovered by test cases.

llvm-svn: 118969
 2010-11-13 02:37:59 +00:00
Evan Cheng
a7d3c3d387 Add conditional move of large immediate. 
llvm-svn: 118968
 2010-11-13 02:25:14 +00:00
Jim Grosbach
cab8b12f7a Swap multiclass operand order for consistency with other patterns. 
llvm-svn: 118965
 2010-11-13 01:28:30 +00:00
Jim Grosbach
c8b3147ca5 Continue ARM indexed load refactoring. Multiclass for LDR{B} pre/post indexed 
instructions.

llvm-svn: 118963
 2010-11-13 01:07:20 +00:00
Rafael Espindola
e296fd654b Fix the encoding of negative line deltas. 
llvm-svn: 118962
 2010-11-13 01:06:27 +00:00
Daniel Dunbar
a34e9e580c Fix some more 80-col violas. 
llvm-svn: 118959
 2010-11-13 00:55:45 +00:00
Daniel Dunbar
2b1fd6592e Fix 80-col violation / non-sensicalness. 
llvm-svn: 118958
 2010-11-13 00:55:42 +00:00
Jim Grosbach
03eb1993ea More ARM load/store indexed refactoring. Also fix an incorrect IndexMode 
flag for the LDRT/STRT family instructions as a side effect.

llvm-svn: 118955
 2010-11-13 00:35:48 +00:00
Dan Gohman
a4aa69cf2a Revert r118917, which is implicated in the llvm-gcc-i386-linux-selfhost failure. 
llvm-svn: 118954
 2010-11-13 00:31:40 +00:00
Evan Cheng
a08372d4b4 Fix an obvious typo which inverted an immediate. 
llvm-svn: 118951
 2010-11-13 00:27:47 +00:00
Wesley Peck
28f124b047 1. Adding missing immediate mode asm parser test cases. 
2. Fixing improper immediate mode reverse subtract.

llvm-svn: 118948
 2010-11-13 00:03:46 +00:00
Eric Christopher
f6c7694c29 Temporarily revert this. 
llvm-svn: 118946
 2010-11-12 23:50:48 +00:00
Evan Cheng
1f5296d832 For pre-v6t2 targets, only select MOVi32imm if the immediate can be handled with movi + orr. 
llvm-svn: 118945
 2010-11-12 23:46:13 +00:00
Wesley Peck
f11db049fe Fixing improperly encoded reverse subtract instructions in MBlaze backend. 
llvm-svn: 118943
 2010-11-12 23:41:10 +00:00
Owen Anderson
053d9fb9b5 Revert r118939 while I work out why it broke some buildbots. 
llvm-svn: 118942
 2010-11-12 23:36:03 +00:00
Wesley Peck
be71f5db40 1. Finishing MBlaze MC asm parser test cases 
2. Parsing .word directive in MBlaze asm parser
3. Fixing hack where memory instructions reversed order of last two parameters
4. Fixing many improperly encoded instructions
5. Support parsing special instructions (MFS,MTS,etc.)
6. Removing unused functions from inst printer

llvm-svn: 118941
 2010-11-12 23:30:17 +00:00
Devang Patel
3dd764feda Remove DW_AT_start_scope support. It is incomplete and superseeded by location entries support. 
llvm-svn: 118940
 2010-11-12 23:20:42 +00:00
Owen Anderson
f8192cf0cc Attemt to provide correct encodings for Thumb2 binary operators. 
llvm-svn: 118939
 2010-11-12 23:18:11 +00:00
Evan Cheng
f3c75f91e9 Eliminate ARM::MOVi2pieces. Just use MOVi32imm and expand it to either movi+orr or movw+movt depending on the subtarget. 
llvm-svn: 118938
 2010-11-12 23:03:38 +00:00
Eric Christopher
b8ee838424 Make this happen for ARM like x86. Don't entirely bail out when 
an address is in a different block, get it into a register and go
from there.

llvm-svn: 118936
 2010-11-12 22:52:32 +00:00
Evan Cheng
19f018a1be Add conditional mvn instructions. 
llvm-svn: 118935
 2010-11-12 22:42:47 +00:00
Jim Grosbach
e0ba69b81d Zap a copy/paste-o bit of dead code. 
llvm-svn: 118926
 2010-11-12 21:29:10 +00:00
Jim Grosbach
eaaf8294a5 Refactor to parameterize some ARM load/store encoding patterns. Preparatory 
to splitting the load/store pre/post indexed instructions into [r, r] and
[r, imm] forms.

llvm-svn: 118925
 2010-11-12 21:28:15 +00:00
Owen Anderson
f1ffc8fdc9 First stab at providing correct Thumb2 encodings, start with adc. 
llvm-svn: 118924
 2010-11-12 21:12:40 +00:00
Duncan Sands
29593ddf27 Have GVN simplify instructions as it goes. For example, consider 
"%z = %x and %y".  If GVN can prove that %y equals %x, then it turns
this into "%z = %x and %x".  With the new code, %z will be replaced
with %x everywhere (and then deleted).  Previously %z would be value
numbered too, which is a waste of time.  Also, while a clever value
numbering algorithm would give %z the same value number as %x, our
current one doesn't do so (at least I don't think it does).  The new
logic has an essentially equivalent effect to what you would get if
%z was given the same value number as %x, i.e. it should make value
numbering smarter.  While there, get hold of target data once at the
start rather than a gazillion times all over the place.

llvm-svn: 118923
 2010-11-12 21:10:24 +00:00
Evan Cheng
b565d1acf9 Add some missing isel predicates on def : pat patterns to avoid generating VFP vmla / vmls (they cause stalls). Disabling them in isel is properly not a right solution, I'll look into a proper solution next. 
llvm-svn: 118922
 2010-11-12 20:32:20 +00:00
Jim Grosbach
160cc37f73 Kill more unused stuff. 
llvm-svn: 118921
 2010-11-12 19:27:45 +00:00
Benjamin Kramer
b4c0c4211e MCELF: Copy the symbol name only if we're going to modify it. 
llvm-svn: 118920
 2010-11-12 19:26:04 +00:00
Jim Grosbach
8bb507dd6b Remove unused class. 
llvm-svn: 118919
 2010-11-12 19:24:53 +00:00
Dan Gohman
8b86287a06 When the definition of an address value is in a different block 
from the user of the address, fall back to just using the
address in a register instead of bailing out of fast-isel
altogether.

llvm-svn: 118917
 2010-11-12 19:14:00 +00:00
Chris Lattner
2a70c4e38a accept lret as an alias for lretl, fixing the reopened part of PR8592 
llvm-svn: 118916
 2010-11-12 18:54:56 +00:00
Rafael Espindola
77cad43f26 Remove what looks like dead code in the production of debug lines. 
We only produce debug line information if we have seen a line directive, so
this code is dead. Also, if we want to be bug by bug compatible with
gas and sometimes produce "empty" .debug_line sections, this will
match the content produced by gas.

llvm-svn: 118914
 2010-11-12 18:41:26 +00:00
Andrew Trick
68d533cca4 typo (4th checkin for one fix) 
llvm-svn: 118913
 2010-11-12 18:36:03 +00:00
Jim Grosbach
f9d1b45754 Fill in the default predication bits for ARM unconditional branch. 
llvm-svn: 118907
 2010-11-12 18:13:26 +00:00
Jim Grosbach
cbab1ea04e Encoding for ARM LDRSB instructions. 
llvm-svn: 118905
 2010-11-12 17:52:59 +00:00
Andrew Trick
d0c919dcad Fixes PR8287: SD scheduling time. The fix is a failsafe that prevents 
catastrophic compilation time in the event of unreasonable LLVM
IR. Code quality is a separate issue--someone upstream needs to do a
better job of reducing to llvm.memcpy. If the situation can be reproduced with
any supported frontend, then it will be a separate bug.

llvm-svn: 118904
 2010-11-12 17:50:46 +00:00
Chris Lattner
612039e538 implement PR8592: empirically "lretq" is a "lret" with a rex.w prefix. 
llvm-svn: 118903
 2010-11-12 17:41:20 +00:00
Chris Lattner
bcdb696f6c tidy up. 
llvm-svn: 118896
 2010-11-12 17:24:29 +00:00
Rafael Espindola
dc54b2b285 gnu as support both % and @ before types, do the same. 
llvm-svn: 118893
 2010-11-12 15:47:08 +00:00
Dan Gohman
ec55b3ac75 Re-disable TBAA for now; it broke MultiSource/Applications/JM/lencod, 
at least.

llvm-svn: 118890
 2010-11-12 11:21:08 +00:00
Kalle Raiskila
f89d0d0389 Fix memory access lowering on SPU, adding 
support for the case where alignment<value size.

These cases were silently miscompiled before this patch.
Now they are overly verbose -especially storing is- and
any front-end should still avoid misaligned memory 
accesses as much as possible. The bit juggling algorithm
added here probably has some room for improvement still.

llvm-svn: 118889
 2010-11-12 10:14:03 +00:00
Eric Christopher
1ca7518097 Fix up a few more spots of addrmode2 (or not) changes that were 
missed. Update some comments accordingly.

Fixes rdar://8652289

llvm-svn: 118888
 2010-11-12 09:48:30 +00:00
Dan Gohman
f867bfd906 Enable TBAA. 
llvm-svn: 118884
 2010-11-12 06:20:01 +00:00
Lang Hames
b10e4b99e3 Fix some style issues in PBQP. Patch by David Blaikie. 
llvm-svn: 118883
 2010-11-12 05:47:21 +00:00
Dan Gohman
bc5a716f10 Enhance DSE to handle the case where a free call makes more than 
one store dead. This is especially noticeable in
SingleSource/Benchmarks/Shootout/objinst.

llvm-svn: 118875
 2010-11-12 02:19:17 +00:00
Dale Johannesen
659061fcd3 Remove possibly useful info from comment, per Chris. 
llvm-svn: 118865
 2010-11-12 00:43:18 +00:00
Bruno Cardoso Lopes
37f2439cbd Enable mips32 mul instruction. Patch by Akira Hatanaka <ahatanaka@mips.com> 
llvm-svn: 118864
 2010-11-12 00:38:32 +00:00
Jim Grosbach
f00a6faa17 Start of support for binary emit of 16-it Thumb instructions. 
llvm-svn: 118859
 2010-11-11 23:41:09 +00:00
Owen Anderson
85cf1b8ff5 Fill out support for Thumb2 encodings of NEON instructions. 
llvm-svn: 118854
 2010-11-11 23:12:55 +00:00
Wesley Peck
7974f4c201 The BRK instruction in the MicroBlaze is a branch-and-link. 
llvm-svn: 118848
 2010-11-11 22:21:08 +00:00
Nick Lewycky
1acee08e53 Doxygenify 
llvm-svn: 118846
 2010-11-11 21:51:44 +00:00
Dan Gohman
f3bf6591e0 Add helper functions for computing the Location of load, store, 
and vaarg instructions.

llvm-svn: 118845
 2010-11-11 21:50:19 +00:00
Wesley Peck
0e937ae80d Fix tblgen instruction errors exposed by MC asm parser tests 
Fix minimum 16-bit signed value error exposed by MC asm parser tests
Add initial MC asm parser tests for the MBlaze backend

llvm-svn: 118844
 2010-11-11 21:40:53 +00:00
Owen Anderson
0913dac245 Add correct Thumb2 encodings for NEON vst[1,2,3,4] and vld[1,2,3,4]. 
llvm-svn: 118843
 2010-11-11 21:36:43 +00:00
Dan Gohman
97ce60bfa4 Don't forget the TBAA info, if available. 
llvm-svn: 118842
 2010-11-11 21:27:26 +00:00
Dan Gohman
b20b7d2ec2 Factor out Instruction::isSafeToSpeculativelyExecute's code for 
testing for dereferenceable pointers into a helper function,
isDereferenceablePointer.  Teach it how to reason about GEPs
with simple non-zero indices.

Also eliminate ArgumentPromtion's IsAlwaysValidPointer,
which didn't check for weak externals or out of range gep
indices.

llvm-svn: 118840
 2010-11-11 21:23:25 +00:00
Eric Christopher
474a93ff76 Revert the accidental commit I made reverting the previous commit. 
llvm-svn: 118835
 2010-11-11 20:50:14 +00:00
Jim Grosbach
b48a272807 Trailing whitespace. 
llvm-svn: 118831
 2010-11-11 20:16:23 +00:00
Jim Grosbach
cc8e5dc0c6 ARM fixup encoding for direct call instructions (BL). 
llvm-svn: 118829
 2010-11-11 20:05:40 +00:00
Eric Christopher
e7f27cf66a Revert this temporarily. 
llvm-svn: 118827
 2010-11-11 19:47:02 +00:00
Eric Christopher
beb7a50acb Change the prologue and epilogue to use push/pop for the low ARM registers. 
llvm-svn: 118823
 2010-11-11 19:26:03 +00:00
Dan Gohman
b68f197748 Avoid calling alias on non-pointer values. 
llvm-svn: 118822
 2010-11-11 19:23:51 +00:00
Owen Anderson
f43f09f6f6 Add support for Thumb2 encodings of NEON data processing instructions, using the new PostEncoderMethod infrastructure. 
More tests to come.

llvm-svn: 118819
 2010-11-11 19:07:48 +00:00
Rafael Espindola
1725702868 Mark labels declared in tls sections as STT_TLS. This matches the behavior of 
gas.

llvm-svn: 118818
 2010-11-11 19:04:55 +00:00
Wesley Peck
9ad36509f8 Fixed some bugs in MBlaze asm parser that were introduced when removing OwningPtrs from the code. 
llvm-svn: 118807
 2010-11-11 18:41:33 +00:00
Chris Lattner
b3eff154cd add a note 
llvm-svn: 118806
 2010-11-11 18:23:57 +00:00
Rafael Espindola
8555d32c9e Initial comdat implementation. 
llvm-svn: 118805
 2010-11-11 18:13:52 +00:00
Dan Gohman
7c6a63aea3 TBAA-enable ArgumentPromotion. 
llvm-svn: 118804
 2010-11-11 18:09:32 +00:00
Dan Gohman
d88f1d63ac Add a FIXME comment. 
llvm-svn: 118803
 2010-11-11 18:08:43 +00:00
Jim Grosbach
abf0e10ea0 Encoding of destination fixup for ARM branch and conditional branch 
instructions.

llvm-svn: 118801
 2010-11-11 18:04:49 +00:00
Andrew Trick
5e77065d78 Check TRI->getReservedRegs because other allocators do it. Even though 
it makes no sense for allocation_order iterators to visit reserved regs.
The inline spiller depends on AliasAnalysis.
Manage the Query state to avoid uninitialized or stale results.

llvm-svn: 118800
 2010-11-11 17:46:29 +00:00
Rafael Espindola
f7b3bb84e8 Make AliasedSymbol able to handle MCTargetExpr. They can get here if 
a weakref is used with a VariantKind.

llvm-svn: 118798
 2010-11-11 17:24:43 +00:00
Chris Lattner
3514fa6b11 add pr# 
llvm-svn: 118797
 2010-11-11 17:17:56 +00:00
Jim Grosbach
cfc32e3b5a Encoding for ARM LDRSH_POST. 
llvm-svn: 118794
 2010-11-11 16:55:29 +00:00