Commit Graph

15353 Commits

Author SHA1 Message Date
Evan Cheng
abc346ceda Do away with CPU feature list. Just use CPUID to detect MMX, SSE, SSE2, SSE3, and 64-bit support.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30763 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-06 08:21:07 +00:00
Evan Cheng
88c1578011 It appears the inline asm in GetCpuIDAndInfo() may clobbers some registers if it isn't inlined (at < -O3). Force it to be inlined.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30762 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-06 07:50:56 +00:00
Chris Lattner
5504602836 MachineBasicBlock::splice was incorrectly updating parent pointers on
instructions.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30760 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-06 01:12:44 +00:00
Evan Cheng
786225adf0 Make use of getStore().
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30759 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-05 23:01:46 +00:00
Evan Cheng
ad071e1cd1 Add getStore() helper function to create ISD::STORE nodes.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30758 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-05 22:57:11 +00:00
Chris Lattner
9c78ecb511 Don't crash if an MBB doesn't have an LLVM BB
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30757 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-05 21:40:14 +00:00
Rafael Espindola
39b5a21259 use a const ref for passing the vector to ArgumentLayout
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30756 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-05 17:46:48 +00:00
Rafael Espindola
a284584352 implement a ArgumentLayout class to factor code common to LowerFORMAL_ARGUMENTS and LowerCALL
implement FMDRR
add support for f64 function arguments


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30754 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-05 16:48:49 +00:00
Jim Laskey
bc588b8bbf Alias analysis code clean ups.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30753 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-05 15:07:25 +00:00
Chris Lattner
867b99f845 add a new SimplifyDemandedVectorElts method, which works similarly to
SimplifyDemandedBits.  The idea is that some operations can be simplified if
not all of the computed elements are needed.  Some targets (like x86) have a
large number of intrinsics that operate on a single element, but pass other
elts through unmodified.  If those other elements are not needed, the
intrinsics can be simplified to scalar operations, and insertelement ops can
be removed.

This turns (f.e.):

ushort %Convert_sse(float %f) {
        %tmp = insertelement <4 x float> undef, float %f, uint 0                ; <<4 x float>> [#uses=1]
        %tmp10 = insertelement <4 x float> %tmp, float 0.000000e+00, uint 1             ; <<4 x float>> [#uses=1]
        %tmp11 = insertelement <4 x float> %tmp10, float 0.000000e+00, uint 2           ; <<4 x float>> [#uses=1]
        %tmp12 = insertelement <4 x float> %tmp11, float 0.000000e+00, uint 3           ; <<4 x float>> [#uses=1]
        %tmp28 = tail call <4 x float> %llvm.x86.sse.sub.ss( <4 x float> %tmp12, <4 x float> < float 1.000000e+00, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00 > )               ; <<4 x float>> [#uses=1]
        %tmp37 = tail call <4 x float> %llvm.x86.sse.mul.ss( <4 x float> %tmp28, <4 x float> < float 5.000000e-01, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00 > )               ; <<4 x float>> [#uses=1]
        %tmp48 = tail call <4 x float> %llvm.x86.sse.min.ss( <4 x float> %tmp37, <4 x float> < float 6.553500e+04, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00 > )               ; <<4 x float>> [#uses=1]
        %tmp59 = tail call <4 x float> %llvm.x86.sse.max.ss( <4 x float> %tmp48, <4 x float> zeroinitializer )          ; <<4 x float>> [#uses=1]
        %tmp = tail call int %llvm.x86.sse.cvttss2si( <4 x float> %tmp59 )              ; <int> [#uses=1]
        %tmp69 = cast int %tmp to ushort                ; <ushort> [#uses=1]
        ret ushort %tmp69
}

into:

ushort %Convert_sse(float %f) {
entry:
        %tmp28 = sub float %f, 1.000000e+00             ; <float> [#uses=1]
        %tmp37 = mul float %tmp28, 5.000000e-01         ; <float> [#uses=1]
        %tmp375 = insertelement <4 x float> undef, float %tmp37, uint 0         ; <<4 x float>> [#uses=1]
        %tmp48 = tail call <4 x float> %llvm.x86.sse.min.ss( <4 x float> %tmp375, <4 x float> < float 6.553500e+04, float undef, float undef, float undef > )           ; <<4 x float>> [#uses=1]
        %tmp59 = tail call <4 x float> %llvm.x86.sse.max.ss( <4 x float> %tmp48, <4 x float> < float 0.000000e+00, float undef, float undef, float undef > )            ; <<4 x float>> [#uses=1]
        %tmp = tail call int %llvm.x86.sse.cvttss2si( <4 x float> %tmp59 )              ; <int> [#uses=1]
        %tmp69 = cast int %tmp to ushort                ; <ushort> [#uses=1]
        ret ushort %tmp69
}

which improves codegen from:

_Convert_sse:
        movss LCPI1_0, %xmm0
        movss 4(%esp), %xmm1
        subss %xmm0, %xmm1
        movss LCPI1_1, %xmm0
        mulss %xmm0, %xmm1
        movss LCPI1_2, %xmm0
        minss %xmm0, %xmm1
        xorps %xmm0, %xmm0
        maxss %xmm0, %xmm1
        cvttss2si %xmm1, %eax
        andl $65535, %eax
        ret

to:

_Convert_sse:
        movss 4(%esp), %xmm0
        subss LCPI1_0, %xmm0
        mulss LCPI1_1, %xmm0
        movss LCPI1_2, %xmm1
        minss %xmm1, %xmm0
        xorps %xmm1, %xmm1
        maxss %xmm1, %xmm0
        cvttss2si %xmm0, %eax
        andl $65535, %eax
        ret


This is just a first step, it can be extended in many ways.  Testcase here:
Transforms/InstCombine/vec_demanded_elts.ll


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30752 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-05 06:55:50 +00:00
Chris Lattner
06a248c238 Add insertelement/extractelement helper ctors.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30750 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-05 06:24:58 +00:00
Chris Lattner
1907a7b37b Lower some min/max idioms to minss/maxss when unsafe fp math is enabled.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30748 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-05 04:11:26 +00:00
Chris Lattner
a66ba5f70b Don't bother setting JumpTableTextSection, it is about to disappear
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30745 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-05 03:13:59 +00:00
Chris Lattner
e08d6e915e Emit pic jumptables to the same section that the function is emitted to,
allowing label differences to work.  This fixes CodeGen/X86/pic_jumptable.ll


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30744 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-05 03:13:28 +00:00
Chris Lattner
1da31ee472 Pass the MachineFunction into EmitJumpTableInfo.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30742 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-05 03:01:21 +00:00
Chris Lattner
edad2b783f implement and use getSectionForFunction
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30741 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-05 02:51:36 +00:00
Chris Lattner
6f6f69950f Use getSectionForFunction.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30740 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-05 02:49:23 +00:00
Chris Lattner
29bd9e12d4 Use getSectionForFunction
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30739 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-05 02:48:40 +00:00
Chris Lattner
6e79629dab use getSectionForFunction to decide which section to emit code into
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30738 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-05 02:47:13 +00:00
Chris Lattner
afbfdeddff Implement getSectionForFunction, use it when printing function body.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30737 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-05 02:43:52 +00:00
Chris Lattner
52f0670470 move getSectionForFunction to AsmPrinter
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30736 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-05 02:42:47 +00:00
Chris Lattner
9b7ce7da82 Move getSectionForFunction to AsmPrinter, change it to return a string.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30735 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-05 02:42:20 +00:00
Chris Lattner
b56dcc453e implement DarwinTargetAsmInfo::getSectionForFunction, use it when outputting
function bodies


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30733 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-05 00:35:50 +00:00
Chris Lattner
f5b10ec509 Give TargetAsmInfo a virtual dtor, add a new getSectionForFunction method.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30732 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-05 00:35:16 +00:00
Chris Lattner
fea13d33e5 emit jump table before debug info
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30731 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-05 00:26:05 +00:00
Chris Lattner
37dfa02788 Always emit the jump table after the function so it's part of the same 'atom'
as the function body.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30730 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-05 00:24:46 +00:00
Chris Lattner
df20b96ffd getFilename/getDirectory shouldn't abort if the global has no init. This
can happen on bugpoint reduced testcases f.e..


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30729 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-04 23:06:26 +00:00
Evan Cheng
693163e74d Fix some typos that can cause a flag value to have more than one use.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30727 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-04 22:23:53 +00:00
Chris Lattner
c055a9191f Fix a static dtor issue
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30726 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-04 22:13:11 +00:00
Chris Lattner
90aa839c88 Fix more static dtor issues
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30725 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-04 21:52:35 +00:00
Chris Lattner
b336409673 Fix some more static dtor issues.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30724 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-04 21:49:37 +00:00
Evan Cheng
e111303408 Added option -disable-x86-shuffle-opti to disable X86 specific vector shuffle optimizations.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30723 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-04 18:33:38 +00:00
Evan Cheng
8e0055de8a Formating.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30722 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-04 18:33:00 +00:00
Jim Laskey
6ff23e5e84 More extensive alias analysis.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30721 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-04 16:53:27 +00:00
Jim Laskey
0c0feb9d5f More long term solution
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30720 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-04 10:40:15 +00:00
Chris Lattner
83e6c9925e Pattern match min/max nodes when we have sse. This implements
CodeGen/X86/scalar_sse_minmax.ll


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30719 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-04 06:57:07 +00:00
Chris Lattner
68e8b9e775 add a note :(
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30717 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-04 05:52:13 +00:00
Chris Lattner
3a15503c82 This case isn't implemented yet. It seems unlikely to be needed, but if it
ever is, we want to get an assert instead of silent bad codegen.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30716 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-04 04:58:58 +00:00
Jim Laskey
0f77fe5ea6 Work around for some problems with templates.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30715 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-04 01:43:13 +00:00
Evan Cheng
c548428c5d Combine ISD::EXTLOAD, ISD::SEXTLOAD, ISD::ZEXTLOAD into ISD::LOADX. Add an
extra operand to LOADX to specify the exact value extension type.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30714 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-04 00:56:09 +00:00
Chris Lattner
faf1daeb92 Use $( $| $) to represent alternatives in asm blocks instead of {|}. This
is needed to support targets where {|} aren't special symbols.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30712 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-03 23:27:09 +00:00
Evan Cheng
bf497a3a68 Fix an obvious typo.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30711 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-03 23:08:27 +00:00
Chris Lattner
f28bbda2c6 Bugfixes
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30709 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-03 20:19:23 +00:00
Chris Lattner
ee773ba72b Print the MBB ID # along with the bb tag in the -print-machine-instrs output.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30708 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-03 20:17:24 +00:00
Chris Lattner
e70cab0ca4 Provide a function that ensures MBB numbering is dense and inorder. This
can be used by MachineFunctionPasses who need this property.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30706 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-03 19:18:57 +00:00
Nick Lewycky
f938099aa1 Simplify logic further.
Ensure that we copy KnownProperties before calling visitBasicBlock, else
we may leak properties into blocks where they don't belong.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30705 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-03 17:36:01 +00:00
Rafael Espindola
cd71da5cf0 Implement floating point constants
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30704 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-03 17:27:58 +00:00
Chris Lattner
13bf6c1350 Fix PR933 and CodeGen/X86/2006-10-02-BoolRetCrash.ll
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30703 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-03 17:18:42 +00:00
Nick Lewycky
a73a654bb4 Simplify, now that predsimplify depends on break-crit-edges.
Fix SwitchInst where dest-block is the same as one of the cases.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30700 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-03 15:19:11 +00:00
Nick Lewycky
5c8c5d9ace Move break-crit-edges before the predicate simplifier. Allows us to
optimize in more cases.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30699 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-03 14:52:23 +00:00