Commit Graph

16125 Commits

Author SHA1 Message Date
Evan Cheng
d99d68bcee Fix two-address pass's aggressive instruction commuting heuristics. It's meant
to catch cases like:
 %reg1024<def> = MOV r1
 %reg1025<def> = MOV r0
 %reg1026<def> = ADD %reg1024, %reg1025
 r0            = MOV %reg1026

By commuting ADD, it let coalescer eliminate all of the copies. However, there
was a bug in the heuristics where it ended up commuting the ADD in:

 %reg1024<def> = MOV r0
 %reg1025<def> = MOV 0
 %reg1026<def> = ADD %reg1024, %reg1025
 r0            = MOV %reg1026

That did no benefit but rather ensure the last MOV would not be coalesced.

rdar://11355268


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@156048 91177308-0d34-0410-b5e6-96231b3b80d8
2012-05-03 01:45:13 +00:00
Owen Anderson
062c0a5b58 Teach DAGCombine the same multiply-by-1.0 folding trick when doing FMAs, just like it now knows for FMULs.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@156029 91177308-0d34-0410-b5e6-96231b3b80d8
2012-05-02 22:17:40 +00:00
Owen Anderson
363e4b90c0 Teach DAG combine that multiplication by 1.0 can always be constant folded.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@156023 91177308-0d34-0410-b5e6-96231b3b80d8
2012-05-02 21:32:35 +00:00
Jim Grosbach
2727930ab4 ARM: Add missing two-operand VBIC aliases.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@156019 91177308-0d34-0410-b5e6-96231b3b80d8
2012-05-02 21:11:56 +00:00
Manman Ren
e2849851b2 Revert r155853
The commit is intended to fix rdar://10961709.
But it is the root cause of PR12720.
Revert it for now.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155992 91177308-0d34-0410-b5e6-96231b3b80d8
2012-05-02 15:24:32 +00:00
Bill Wendling
55e7098bbc The value held in the vector may be RAUW'ed by some of the canonicalization
methods. Use a weak value handle to keep up with this.
PR12245


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155984 91177308-0d34-0410-b5e6-96231b3b80d8
2012-05-02 09:59:45 +00:00
Richard Barton
0a552d611e Disallow YIELD and other allocated nop hints in pre-ARMv6 architectures.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155983 91177308-0d34-0410-b5e6-96231b3b80d8
2012-05-02 09:43:18 +00:00
Craig Topper
a9a568a79d Add support for selecting AVX2 vpshuflw and vpshufhw. Add decoding support for AsmPrinter.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155982 91177308-0d34-0410-b5e6-96231b3b80d8
2012-05-02 08:03:44 +00:00
Bill Wendling
95dd442041 Strip the pointer casts off of allocas so that the selection DAG can find them.
PR10799


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155954 91177308-0d34-0410-b5e6-96231b3b80d8
2012-05-01 22:50:45 +00:00
Jim Grosbach
54319e2a8c ARM: Add a few missing add->sub aliases w/ 'w' suffix.
Aliases for adding a negative immediate when using an explicit 'w'
suffix. E.g.,
        adds.w r2, #-16
        adds.w r2, r2, #-16
        addw r2, #-16
        addw r2, #-16
        addw r2, r2, #-16

rdar://11330769

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155946 91177308-0d34-0410-b5e6-96231b3b80d8
2012-05-01 21:17:34 +00:00
Jim Grosbach
94b590f8fa ARM: allow vanilla expressions for movw/movt.
Expressions for movw/movt don't always have an :upper16: or :lower16:
on them and that's ok. When they don't, it's just a plain [0-65536]
immediate result, effectively the same as a :lower16: variant kind.

rdar://10550147

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155941 91177308-0d34-0410-b5e6-96231b3b80d8
2012-05-01 20:43:21 +00:00
Jim Grosbach
686c01854e MC: Unknown assembler directives are now hard errors.
Previously, an unsupported/unknown assembler directive issued a warning.
That's generally unsafe, and inconsistent with the behaviour of pretty
much every system assembler. Now that the MC assemblers are mature
enough to be the default on multiple targets, it's reasonable to
issue errors for these.

For target or platform directives that need to stay warnings, we
should add explicit handlers for them in, e.g., ELFAsmParser.cpp,
DarwinAsmParser.cpp, et. al., and issue the warning there.

rdar://9246275

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155926 91177308-0d34-0410-b5e6-96231b3b80d8
2012-05-01 18:38:27 +00:00
Manman Ren
769ea2f93f X86: optimization for max-like struct
This patch will optimize the following cases on X86
(a > b) ? (a-b) : 0
(a >= b) ? (a-b) : 0
(b < a) ? (a-b) : 0
(b <= a) ? (a-b) : 0

FROM
movl    %edi, %ecx
subl    %esi, %ecx
cmpl    %edi, %esi
movl    $0, %eax
cmovll  %ecx, %eax
TO
xorl    %eax, %eax
subl    %esi, %edi
cmovll  %eax, %edi
movl    %edi, %eax

rdar: 10734411


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155919 91177308-0d34-0410-b5e6-96231b3b80d8
2012-05-01 17:16:15 +00:00
Alexey Samsonov
d07d06ceef X86: Use StackRegister instead of FrameRegister in getFrameIndexReference (to generate debug info for local variables) if stack needs realignment
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155917 91177308-0d34-0410-b5e6-96231b3b80d8
2012-05-01 15:16:06 +00:00
Jay Foad
4ab6fcaadc Regression test for PR2960.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155912 91177308-0d34-0410-b5e6-96231b3b80d8
2012-05-01 11:11:34 +00:00
Nick Lewycky
4056a73638 An instruction in a loop is not guaranteed to be executed just because the loop
has no exit blocks. Fixes PR12706!


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155884 91177308-0d34-0410-b5e6-96231b3b80d8
2012-05-01 04:03:01 +00:00
Lang Hames
973f72a29a Add support for llvm.arm.neon.vmull* intrinsics to InstCombine. Fixes
<rdar://problem/11291436>.

This is a second attempt at a fix for this, the first was r155468. Thanks
to Chandler, Bob and others for the feedback that helped me improve this.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155866 91177308-0d34-0410-b5e6-96231b3b80d8
2012-05-01 00:20:38 +00:00
Manman Ren
16a76519a5 X86: optimization for -(x != 0)
This patch will optimize -(x != 0) on X86
FROM 
cmpl	$0x01,%edi
sbbl	%eax,%eax
notl	%eax
TO
negl %edi
sbbl %eax %eax


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155853 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-30 22:51:25 +00:00
Manman Ren
1701105022 test/CodeGen/X86/select.ll: remove spaces
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155840 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-30 18:54:27 +00:00
Derek Schuff
ddc693bd22 Fix fastcc structure return with fast-isel on x86-32
On x86-32, structure return via sret lets the callee pop the hidden
pointer argument off the stack, which the caller then re-pushes.
However if the calling convention is fastcc, then a register is used
instead, and the caller should not adjust the stack. This is
implemented with a check of IsTailCallConvention
X86TargetLowering::LowerCall but is now checked properly in
X86FastISel::DoSelectCall.

(this time, actually commit what was reviewed!)



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155825 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-30 16:57:15 +00:00
Bob Wilson
ff73d8fef9 Don't introduce illegal types when creating vmull operations. <rdar://11324364>
ARM BUILD_VECTORs created after type legalization cannot use i8 or i16
operands, since those types are not legal.  Instead use i32 operands, which
will be implicitly truncated by the BUILD_VECTOR to match the element type.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155824 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-30 16:53:34 +00:00
Duncan Sands
5ff30e70f8 Just mark the sign bit as known zero, rather than any other irrelevant bits
known zero in the LHS.  Fixes PR12541.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155818 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-30 11:56:58 +00:00
Bill Wendling
bfbab99b58 Second attempt at PR12573:
Allow the "SplitCriticalEdge" function to split the edge to a landing pad. If
the pass is *sure* that it thinks it knows what it's doing, then it may go ahead
and specify that the landing pad can have its critical edge split. The loop
unswitch pass is one of these passes. It will split the critical edges of all
edges coming from a loop to a landing pad not within the loop. Doing so will
retain important loop analysis information, such as loop simplify.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155817 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-30 10:44:54 +00:00
Rafael Espindola
9719cf329b Make sure HoistInsertPosition finds a position that is dominated by all
inputs.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155809 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-30 03:53:06 +00:00
Andrew Trick
2674a4acdb Reapply 155668: Fix the SD scheduler to avoid gluing the same node twice.
This time, also fix the caller of AddGlue to properly handle
incomplete chains. AddGlue had failure modes, but shamefully hid them
from its caller. It's luck ran out.

Fixes rdar://11314175: BuildSchedUnits assert.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155749 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-28 01:03:23 +00:00
Jim Grosbach
a9cc08f24f ARM: Thumb add(sp plus register) asm constraints.
Make sure when parsing the Thumb1 sp+register ADD instruction that
the source and destination operands match. In thumb2, just use the
wide encoding if they don't. In Thumb1, issue a diagnostic.

rdar://11219154

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155748 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-27 23:51:36 +00:00
Derek Schuff
f3db6b855e Revert r155745
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155746 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-27 23:37:41 +00:00
Derek Schuff
9dc28b0722 Fix fastcc structure return with fast-isel on x86-32
On x86-32, structure return via sret lets the callee pop the hidden
pointer argument off the stack, which the caller then re-pushes.
However if the calling convention is fastcc, then a register is used
instead, and the caller should not adjust the stack. This is
implemented with a check of IsTailCallConvention
X86TargetLowering::LowerCall but is now checked properly in
X86FastISel::DoSelectCall.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155745 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-27 23:27:17 +00:00
Andrew Trick
0e47cfd5b6 Temporarily revert r155668: Fix the SD scheduler to avoid gluing.
This definitely caused regression with ARM -mno-thumb.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155743 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-27 22:55:59 +00:00
Chad Rosier
a73b6fc511 Add x86-specific DAG combine to simplify:
x == -y --> x+y == 0
 x != -y --> x+y != 0

On x86, the generated code goes from
   negl    %esi
   cmpl    %esi, %edi
   je    .LBB0_2
to
   addl    %esi, %edi
   je    .L4

This case is correctly handled for ARM with "cmn".

Patch by Manman Ren.
rdar://11245199
PR12545


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155739 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-27 22:33:25 +00:00
Evan Cheng
da71cca458 Make test less fragile.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155732 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-27 20:48:18 +00:00
Hal Finkel
e32e5440d6 Don't vectorize target-specific types (ppc_fp128, x86_fp80, etc.).
Target specific types should not be vectorized. As a practical matter,
these types are already register matched (at least in the x86 case),
and codegen does not always work correctly (at least in the ppc case,
and this is not worth fixing because ppc_fp128 is currently broken and
will probably go away soon).

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155729 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-27 19:34:00 +00:00
Lang Hames
7787800481 Fix the order of the operands in the llvm.fma intrinsic patterns for ARM,
<rdar://problem/11325085>.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155724 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-27 18:51:24 +00:00
Dan Gohman
03e091f0b5 Reapply r155682, making constant folding more consistent, with a fix to work
properly with how the code handles all-undef PHI nodes.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155721 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-27 17:50:22 +00:00
Richard Barton
04a09a461b Fix ARM assembly parsing for upper case condition codes on IT instructions.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155720 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-27 17:34:01 +00:00
Benjamin Kramer
71275b129b Missed some register numbers.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155706 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-27 12:21:46 +00:00
Benjamin Kramer
a356e9414c Update edis test for r155704.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155705 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-27 12:14:03 +00:00
Benjamin Kramer
17c836c4b5 X86: Don't emit conditional floating point moves on when targeting pre-pentiumpro architectures.
* Model FPSW (the FPU status word) as a register.
* Add ISel patterns for the FUCOM*, FNSTSW and SAHF instructions.
* During Legalize/Lowering, build a node sequence to transfer the comparison
result from FPSW into EFLAGS. If you're wondering about the right-shift: That's
an implicit sub-register extraction (%ax -> %ah) which is handled later on by
the instruction selector.

Fixes PR6679. Patch by Christoph Erhardt!


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155704 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-27 12:07:43 +00:00
NAKAMURA Takumi
d213ee7643 Revert r155682, "Use ConstantExpr::getExtractElement when constant-folding vectors"
It broke stage2 build. stage1/clang sometimes crashed.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155699 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-27 07:59:20 +00:00
Kostya Serebryany
e507922779 [tsan] Atomic support for ThreadSanitizer, patch by Dmitry Vyukov
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155698 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-27 07:31:53 +00:00
Craig Topper
76c5897eae Add mcpu to tests to prevent them from using AVX instructions on Sandy Bridge after r155618.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155696 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-27 07:11:58 +00:00
Evan Cheng
afb3b5ebe6 Implement a bastardized ABI.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155686 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-27 02:11:10 +00:00
Evan Cheng
97a454317a - thumbv6 shouldn't imply +thumb2. Cortex-M0 doesn't suppport 32-bit Thumb2
instructions.
- However, it does support dmb, dsb, isb, mrs, and msr.
rdar://11331541


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155685 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-27 01:27:19 +00:00
Dan Gohman
97b44f9b80 Use ConstantExpr::getExtractElement when constant-folding vectors
instead of getAggregateElement. This has the advantage of being
more consistent and allowing higher-level constant folding to
procede even if an inner extract element cannot be folded.

Make ConstantFoldInstruction call ConstantFoldConstantExpression
on the instruction's operands, making it more consistent with 
ConstantFoldConstantExpression itself. This makes sure that
ConstantExprs get TargetData-aware folding before being handed
off as operands for further folding.

This causes more expressions to be folded, but due to a known
shortcoming in constant folding, this currently has the side effect
of stripping a few more nuw and inbounds flags in the non-targetdata
side of constant-fold-gep.ll. This is mostly harmless.

This fixes rdar://11324230.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155682 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-27 00:54:36 +00:00
Chad Rosier
c1fc5e4464 Add instcombine patterns for the following transformations:
(x & y) | (x ^ y) -> x | y 
 (x & y) + (x ^ y) -> x | y 

Patch by Manman Ren.
rdar://10770603


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155674 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-26 23:29:14 +00:00
Andrew Trick
aec9240be2 Fix the SD scheduler to avoid gluing the same node twice.
DAGCombine strangeness may result in multiple loads from the same
offset. They both may try to glue themselves to another load. We could
insist that the redundant loads glue themselves to each other, but the
beter fix is to bail out from bad gluing at the time we detect it.

Fixes rdar://11314175: BuildSchedUnits assert.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155668 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-26 21:48:25 +00:00
Tim Northover
37abe8df4a Use VLD1 in NEON extenting-load patterns instead of VLDR.
On some cores it's a bad idea for performance to mix VFP and NEON instructions
and since these patterns are NEON anyway, the NEON load should be used.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155630 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-26 08:46:29 +00:00
Chandler Carruth
464bda3a16 Teach the reassociate pass to fold chains of multiplies with repeated
elements to minimize the number of multiplies required to compute the
final result. This uses a heuristic to attempt to form near-optimal
binary exponentiation-style multiply chains. While there are some cases
it misses, it seems to at least a decent job on a very diverse range of
inputs.

Initial benchmarks show no interesting regressions, and an 8%
improvement on SPASS. Let me know if any other interesting results (in
either direction) crop up!

Credit to Richard Smith for the core algorithm, and helping code the
patch itself.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155616 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-26 05:30:30 +00:00
Evan Cheng
cac31de146 Specify cpu to unbreak tests.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155604 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-26 01:38:10 +00:00
Evan Cheng
e67a4163f5 If triple is armv7 / thumbv7 and a CPU is specified, do not automatically assume
the feature set of v7a. This comes about if the user specifies something like
-arch armv7 -mcpu=cortex-m3. We shouldn't be generating instructions such as
uxtab in this case.

rdar://11318438


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155601 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-26 01:13:36 +00:00