Commit Graph

2539 Commits

Author SHA1 Message Date
Chris Lattner
62b5772959 Implement folding of a bunch of binops with undef
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27863 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-20 05:39:12 +00:00
Chris Lattner
89a1b380a4 Simplify some code
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27846 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-19 23:17:50 +00:00
Chris Lattner
80edfb3af5 Fix handling of calls in functions that use vectors. This fixes a crash on
the code in GCC PR26546.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27780 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-17 22:10:08 +00:00
Chris Lattner
48d7c069c7 Add a MachineInstr::eraseFromParent convenience method.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27775 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-17 21:35:41 +00:00
Chris Lattner
8d5a894501 Codegen insertelement with constant insertion points as scalar_to_vector
and a shuffle.  For this:

void %test2(<4 x float>* %F, float %f) {
        %tmp = load <4 x float>* %F             ; <<4 x float>> [#uses=2]
        %tmp3 = add <4 x float> %tmp, %tmp              ; <<4 x float>> [#uses=1]
        %tmp2 = insertelement <4 x float> %tmp3, float %f, uint 2               ; <<4 x float>> [#uses=2]
        %tmp6 = add <4 x float> %tmp2, %tmp2            ; <<4 x float>> [#uses=1]
        store <4 x float> %tmp6, <4 x float>* %F
        ret void
}

we now get this on X86 (which will get better):

_test2:
        movl 4(%esp), %eax
        movaps (%eax), %xmm0
        addps %xmm0, %xmm0
        movaps %xmm0, %xmm1
        shufps $3, %xmm1, %xmm1
        movaps %xmm0, %xmm2
        shufps $1, %xmm2, %xmm2
        unpcklps %xmm1, %xmm2
        movss 8(%esp), %xmm1
        unpcklps %xmm1, %xmm0
        unpcklps %xmm2, %xmm0
        addps %xmm0, %xmm0
        movaps %xmm0, (%eax)
        ret

instead of:

_test2:
        subl $28, %esp
        movl 32(%esp), %eax
        movaps (%eax), %xmm0
        addps %xmm0, %xmm0
        movaps %xmm0, (%esp)
        movss 36(%esp), %xmm0
        movss %xmm0, 8(%esp)
        movaps (%esp), %xmm0
        addps %xmm0, %xmm0
        movaps %xmm0, (%eax)
        addl $28, %esp
        ret


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27765 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-17 19:21:01 +00:00
Chris Lattner
2efce0a589 Add support for promoting stores from one legal type to another, allowing us
to write one pattern for vector stores instead of 4.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27730 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-16 01:36:45 +00:00
Chris Lattner
547a16f273 Make these predicates return true for bit_convert(buildvector)'s as well as
buildvectors.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27723 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-15 23:38:00 +00:00
Chris Lattner
3824e50b1e Make this assertion better
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27695 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-14 06:08:35 +00:00
Reid Spencer
60d07eeffe Expand some code with temporary variables to rid ourselves of the warning
about "dereferencing type-punned pointer will break strict-aliasing rules"


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27671 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-13 18:29:58 +00:00
Evan Cheng
cc9876124e Promote vector AND, OR, and XOR
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27632 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-12 21:20:24 +00:00
Evan Cheng
41f6cbbeb2 Vector type promotion for ISD::LOAD and ISD::SELECT
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27606 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-12 16:33:18 +00:00
Chris Lattner
fdfded5588 Implement support for the formal_arguments node. To get this, targets shouldcustom legalize it and remove their XXXTargetLowering::LowerArguments overload
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27604 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-12 16:20:43 +00:00
Chris Lattner
9d956250f5 Don't memoize vloads in the load map! Don't memoize them anywhere here, let
getNode do it.  This fixes CodeGen/Generic/2006-04-11-vecload.ll


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27602 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-12 03:25:41 +00:00
Evan Cheng
98f8aeb45f Only get Tmp2 for cases where number of operands is > 1. Fixed return void.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27586 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-11 06:33:39 +00:00
Chris Lattner
b49e52c7e2 add some todos
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27580 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-11 02:00:08 +00:00
Chris Lattner
f87324e44d Add basic support for legalizing returns of vectors
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27578 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-11 01:31:51 +00:00
Jim Laskey
1069fbdd2f Use existing information.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27574 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-10 23:09:19 +00:00
Evan Cheng
db3c626eb9 Missing break
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27559 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-10 18:54:36 +00:00
Chris Lattner
b22e35a3c3 Add code generator support for VSELECT
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27542 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-08 22:22:57 +00:00
Chris Lattner
17614ea63d Canonicalize vvector_shuffle(x,x) -> vvector_shuffle(x,undef) to enable patterns
to match again :)


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27533 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-08 05:34:25 +00:00
Chris Lattner
3e104b1116 Codegen shufflevector as VVECTOR_SHUFFLE
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27529 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-08 04:15:24 +00:00
Chris Lattner
4ddd283f69 add a sanity check: LegalizeOp should return a value that is the same type
as its input.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27528 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-08 04:13:17 +00:00
Evan Cheng
f6bf87f2cc INSERT_VECTOR_ELT lowering bug:
store vector to $esp
  store element to $esp + sizeof(VT) * index
  load  vector from $esp
The bug is VT is the type of the vector element, not the type of the vector!


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27517 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-08 01:46:37 +00:00
Chris Lattner
67f1351498 Stub out shufflevector
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27514 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-08 01:19:25 +00:00
Jim Laskey
f36e2d3415 Remove section change in function end, preventing override of function's real
section.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27503 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-08 00:35:59 +00:00
Jim Laskey
6b92b8e50d Make sure that debug labels are defined within the same section and after the
entry point of a function.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27494 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-07 20:44:42 +00:00
Jim Laskey
4188699f80 Foundation for call frame information.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27491 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-07 16:34:46 +00:00
Evan Cheng
c04766a228 1. If both vector operands of a vector_shuffle are undef, turn it into an undef.
2. A shuffle mask element can also be an undef.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27472 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-06 23:20:43 +00:00
Chris Lattner
2e2ef95350 Make a vector live across blocks have the correct Vec type. This fixes
CodeGen/X86/2006-04-04-CrossBlockCrash.ll


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27436 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-05 06:54:42 +00:00
Evan Cheng
18dd6d0a2a Exapnd a VECTOR_SHUFFLE to a BUILD_VECTOR if target asks for it to be expanded
or custom lowering fails.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27432 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-05 06:07:11 +00:00
Chris Lattner
e44be60ee9 Do not create ZEXTLOAD's unless we are before legalize or the operation is
legal.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27402 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-04 17:39:18 +00:00
Chris Lattner
4352cc9e21 * Add supprot for SCALAR_TO_VECTOR operations where the input needs to be
promoted/expanded (e.g. SCALAR_TO_VECTOR from i8/i16 on PPC).
* Add support for targets to request that VECTOR_SHUFFLE nodes be promoted
  to a canonical type, for example, we only want v16i8 shuffles on PPC.
* Move isShuffleLegal out of TLI into Legalize.
* Teach isShuffleLegal to allow shuffles that need to be promoted.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27399 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-04 17:23:26 +00:00
Chris Lattner
08da55eeff Constant fold bitconvert(undef)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27391 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-04 01:02:22 +00:00
Chris Lattner
4672f71ac4 The stack alignment is now computed dynamically, just verify it is correct.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27380 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-03 21:39:57 +00:00
Chris Lattner
7dca8e2d73 Remove unused method
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27379 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-03 21:39:03 +00:00
Chris Lattner
97c2073270 Add a missing check, this fixes UnitTests/Vector/sumarray.c
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27375 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-03 17:29:28 +00:00
Chris Lattner
a4c5d8c6b4 Add a missing check, which broke a bunch of vector tests.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27374 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-03 17:21:50 +00:00
Andrew Lenharth
cf4fb61f6c back this out
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27367 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-03 03:16:50 +00:00
Andrew Lenharth
50a0d426e8 This should be a win of every arch
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27364 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-02 21:42:45 +00:00
Chris Lattner
1b5232a937 relax assertion
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27358 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-02 06:19:46 +00:00
Chris Lattner
1482b5fc7a Allow targets to compute masked bits for intrinsics.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27357 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-02 06:15:09 +00:00
Chris Lattner
350bec0fb9 Add a little dag combine to compile this:
int %AreSecondAndThirdElementsBothNegative(<4 x float>* %in) {
entry:
        %tmp1 = load <4 x float>* %in           ; <<4 x float>> [#uses=1]
        %tmp = tail call int %llvm.ppc.altivec.vcmpgefp.p( int 1, <4 x float> < float 0x7FF8000000000000, float 0.000000e+00, float 0.000000e+00, float 0x7FF8000000000000 >, <4 x float> %tmp1 )           ; <int> [#uses=1]
        %tmp = seteq int %tmp, 0                ; <bool> [#uses=1]
        %tmp3 = cast bool %tmp to int           ; <int> [#uses=1]
        ret int %tmp3
}

into this:

_AreSecondAndThirdElementsBothNegative:
        mfspr r2, 256
        oris r4, r2, 49152
        mtspr 256, r4
        li r4, lo16(LCPI1_0)
        lis r5, ha16(LCPI1_0)
        lvx v0, 0, r3
        lvx v1, r5, r4
        vcmpgefp. v0, v1, v0
        mfcr r3, 2
        rlwinm r3, r3, 27, 31, 31
        mtspr 256, r2
        blr

instead of this:

_AreSecondAndThirdElementsBothNegative:
        mfspr r2, 256
        oris r4, r2, 49152
        mtspr 256, r4
        li r4, lo16(LCPI1_0)
        lis r5, ha16(LCPI1_0)
        lvx v0, 0, r3
        lvx v1, r5, r4
        vcmpgefp. v0, v1, v0
        mfcr r3, 2
        rlwinm r3, r3, 27, 31, 31
        xori r3, r3, 1
        cntlzw r3, r3
        srwi r3, r3, 5
        mtspr 256, r2
        blr


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27356 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-02 06:11:11 +00:00
Chris Lattner
4aab2f4ab5 Implement promotion for EXTRACT_VECTOR_ELT, allowing v16i8 multiplies to work with PowerPC.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27349 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-02 05:06:04 +00:00
Chris Lattner
bc70cf8be9 Implement the Expand action for binary vector operations to break the binop
into elements and operate on each piece.  This allows generic vector integer
multiplies to work on PPC, though the generated code is horrible.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27347 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-02 03:57:31 +00:00
Chris Lattner
e58a780166 Intrinsics that just load from memory can be treated like loads: they don't
have to serialize against each other.  This allows us to schedule lvx's
across each other, for example.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27346 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-02 03:41:14 +00:00
Chris Lattner
edab1b9133 Constant fold all of the vector binops. This allows us to compile this:
"vector unsigned char mergeLowHigh = (vector unsigned char)
( 8, 9, 10, 11, 16, 17, 18, 19, 12, 13, 14, 15, 20, 21, 22, 23 );
vector unsigned char mergeHighLow = vec_xor( mergeLowHigh, vec_splat_u8(8));"

aka:

void %test2(<16 x sbyte>* %P) {
  store <16 x sbyte> cast (<4 x int> xor (<4 x int> cast (<16 x ubyte> < ubyte 8, ubyte 9, ubyte 10, ubyte 11, ubyte 16, ubyte 17, ubyte 18, ubyte 19, ubyte 12, ubyte 13, ubyte 14, ubyte 15, ubyte 20, ubyte 21, ubyte 22, ubyte 23 > to <4 x int>), <4 x int> cast (<16 x sbyte> < sbyte 8, sbyte 8, sbyte 8, sbyte 8, sbyte 8, sbyte 8, sbyte 8, sbyte 8, sbyte 8, sbyte 8, sbyte 8, sbyte 8, sbyte 8, sbyte 8, sbyte 8, sbyte 8 > to <4 x int>)) to <16 x sbyte>), <16 x sbyte> * %P
  ret void
}

into this:

_test2:
        mfspr r2, 256
        oris r4, r2, 32768
        mtspr 256, r4
        li r4, lo16(LCPI2_0)
        lis r5, ha16(LCPI2_0)
        lvx v0, r5, r4
        stvx v0, 0, r3
        mtspr 256, r2
        blr

instead of this:

_test2:
        mfspr r2, 256
        oris r4, r2, 49152
        mtspr 256, r4
        li r4, lo16(LCPI2_0)
        lis r5, ha16(LCPI2_0)
        vspltisb v0, 8
        lvx v1, r5, r4
        vxor v0, v1, v0
        stvx v0, 0, r3
        mtspr 256, r2
        blr

... which occurs here:
http://developer.apple.com/hardware/ve/calcspeed.html


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27343 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-02 03:25:57 +00:00
Chris Lattner
5e46a19ec8 Add a new -view-legalize-dags command line option
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27342 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-02 03:07:27 +00:00
Chris Lattner
6258fb2592 Implement constant folding of bit_convert of arbitrary constant vbuild_vector nodes.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27341 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-02 02:53:43 +00:00
Chris Lattner
5e08a11039 These entries already exist
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27340 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-02 02:51:27 +00:00
Chris Lattner
97d23335ad Add some missing node names
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27339 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-02 02:41:18 +00:00