Commit Graph

16361 Commits

Author SHA1 Message Date
Chris Lattner
62be1a71f4 make this code more aggressive about turning store fpimm into store int imm.
This is not sufficient to fix X86/store-fp-constant.ll


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32465 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-12 04:16:14 +00:00
Chris Lattner
a17f03980f teach scev to analyze X*4|1 like X*4+c. This allows us to produce:
LBB1_1: #bb
        movdqa (%esi), %xmm2
        movaps %xmm2, %xmm3
        punpcklbw %xmm0, %xmm3
        movaps %xmm3, %xmm4
        punpcklwd %xmm0, %xmm4
        cvtdq2ps %xmm4, %xmm4
        mulps %xmm1, %xmm4
        movaps %xmm4, (%edi)
        leal 1(,%eax,4), %ebx
        shll $4, %ebx
        punpckhwd %xmm0, %xmm3
        cvtdq2ps %xmm3, %xmm3
        mulps %xmm1, %xmm3
        movaps %xmm3, (%edx,%ebx)
        leal 2(,%eax,4), %ebx
        shll $4, %ebx
        punpckhbw %xmm0, %xmm2
        movaps %xmm2, %xmm3
        punpcklwd %xmm0, %xmm3
        cvtdq2ps %xmm3, %xmm3
        mulps %xmm1, %xmm3
        movaps %xmm3, (%edx,%ebx)
        leal 3(,%eax,4), %ebx
        shll $4, %ebx
        punpckhwd %xmm0, %xmm2
        cvtdq2ps %xmm2, %xmm2
        mulps %xmm1, %xmm2
        movaps %xmm2, (%edx,%ebx)
        addl $64, %edi
        incl %eax
        addl $16, %esi
        cmpl %ecx, %eax
        jne LBB1_1      #bb

instead of:

LBB1_1: #bb
        movdqa (%esi), %xmm2
        movaps %xmm2, %xmm3
        punpcklbw %xmm0, %xmm3
        movaps %xmm3, %xmm4
        punpcklwd %xmm0, %xmm4
        cvtdq2ps %xmm4, %xmm4
        mulps %xmm1, %xmm4
        movaps %xmm4, (%edi)
        leal 1(,%eax,4), %ebx
        shll $4, %ebx
        punpckhwd %xmm0, %xmm3
        cvtdq2ps %xmm3, %xmm3
        mulps %xmm1, %xmm3
        movaps %xmm3, (%edx,%ebx)
        leal 2(,%eax,4), %ebx
        shll $4, %ebx
        punpckhbw %xmm0, %xmm2
        movaps %xmm2, %xmm3
        punpcklwd %xmm0, %xmm3
        cvtdq2ps %xmm3, %xmm3
        mulps %xmm1, %xmm3
        movaps %xmm3, (%edx,%ebx)
        leal 3(,%eax,4), %ebx
        shll $4, %ebx
        punpckhwd %xmm0, %xmm2
        cvtdq2ps %xmm2, %xmm2
        mulps %xmm1, %xmm2
        movaps %xmm2, (%edx,%ebx)
        addl $64, %edi
        incl %eax
        addl $16, %esi
        cmpl %ecx, %eax
        jne LBB1_1      #bb

for a testcase.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32463 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-12 02:26:09 +00:00
Reid Spencer
86b89e5529 Tidy up a bit.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32462 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-12 01:32:02 +00:00
Reid Spencer
15f46d6c28 Change inferred cast creation calls to more specific cast creations.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32460 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-12 01:17:41 +00:00
Rafael Espindola
f64945d83c use MVN to handle small negative constants
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32459 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-12 01:03:11 +00:00
Reid Spencer
84f3eab017 Implement getIntegerCast and getFPCast for ConstantExpr. These are similar
to the createIntegerCast and createFPCast for CastInst instructions.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32457 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-12 00:51:07 +00:00
Reid Spencer
6d81a7d2ae Implement createIntegerCast and createFPCast factory methods for handling
integer and floating point cast creation. createIntegerCast generates
ZExt/SExt, BitCast or Trunc. createFPCast generates FPExt, Bitcast, or
FPTrunc.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32456 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-12 00:49:44 +00:00
Rafael Espindola
450856d01b add mvn
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32454 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-12 00:37:38 +00:00
Reid Spencer
22b36fb7eb Fix the BitCastUnion type for 32-bit targets.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32453 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-12 00:11:08 +00:00
Rafael Espindola
4dfab986ab add note
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32452 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-11 23:56:10 +00:00
Chris Lattner
57fc62c8d2 Another step forward in PPC64 JIT support: we now no-longer need stubs
emitted for external globals in PPC64-JIT-PIC mode (which is good because
we didn't handle them before!).

This also fixes a bug handling the picbase delta, which we would get wrong
in some cases.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32451 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-11 23:22:45 +00:00
Reid Spencer
a77fa7edbf Create the cast constant expression that was read instead of attempting
to infer the cast from its operand and type. This fixes:
test/Regression/Bytecode/2006-12-11-Cast-ConstExpr.ll


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32450 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-11 23:20:20 +00:00
Chris Lattner
442b9a6981 getInstrItineraryData shouldn't copy the itineraries
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32448 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-11 21:42:55 +00:00
Reid Spencer
8c4fed50e3 Fix constant folding of FP->int due to cut & paste error in last commit.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32447 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-11 21:27:28 +00:00
Reid Spencer
555a0b189c Implement correct bitcast of int<->float and long<->double by using a
union to perform the bitcast.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32444 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-11 20:39:15 +00:00
Evan Cheng
13acce3ef0 Re-apply changes that were backed out and fix a naughty typo.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32442 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-11 19:27:14 +00:00
John Criswell
2381f6f6c6 It seems the llvm::OStream class does not handle stream manipulators.
For now, just grab the stream and perform the output on it directly.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32441 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-11 19:15:36 +00:00
Chris Lattner
70a24b3141 Revert changes that broke oggenc on ppc
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32440 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-11 18:53:38 +00:00
Jim Laskey
2060a822fe Missing opcode.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32439 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-11 18:45:56 +00:00
Chris Lattner
14f440aae7 Implement correct constant folding of bitcast. This implements
Transforms/ConstProp/bitcast.ll and fixes
SingleSource/Regression/C/2003-10-12-GlobalVarInitializers


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32438 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-11 18:30:27 +00:00
Jim Laskey
18e2f4433e Layout proper frame for ppc64.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32436 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-11 18:10:54 +00:00
Evan Cheng
25ece66ff3 Don't convert store double C, Ptr to store long C, Ptr if i64 is not a legal type.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32434 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-11 17:25:19 +00:00
Evan Cheng
bbf1e5e2e9 f32 / f64 node is expanded to one i32 / i64 node.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32433 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-11 06:50:04 +00:00
Evan Cheng
b618230231 Clean up some bad code.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32432 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-11 06:25:26 +00:00
Nate Begeman
2cbba89347 Move something that should be in the dag combiner from the legalizer to the
dag combiner.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32431 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-11 02:23:46 +00:00
Nate Begeman
bfaaaa6e0f Properly mangles symbol table names
Supports constant pools
Supports relocations to jump tables
Supports relocations within the data segment (global = address of global)
Allocates memory in a non-hacky for all non-code objects.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32430 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-11 02:20:45 +00:00
Chris Lattner
36c2451ddb fix typo
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32428 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-11 02:16:58 +00:00
Chris Lattner
9f09fa2900 Update note, with the SROA change, we now produce:
_pairtest:
        movl 8(%esp), %eax
        movl 4(%esp), %ecx
        movd %eax, %xmm0
        movd %ecx, %xmm1
        addss %xmm0, %xmm1
        movl 12(%esp), %eax
        movss %xmm1, (%eax)
        ret

instead of:

_pairtest:
        subl $12, %esp
        movl 20(%esp), %eax
        movl %eax, 4(%esp)
        movl 16(%esp), %eax
        movl %eax, (%esp)
        movss (%esp), %xmm0
        addss 4(%esp), %xmm0
        movl 24(%esp), %eax
        movss %xmm0, (%eax)
        addl $12, %esp
        ret


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32427 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-11 01:20:25 +00:00
Chris Lattner
509f85140c trunc to integer, not to FP.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32426 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-11 01:17:00 +00:00
Chris Lattner
b4115ac8f3 this is done
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32424 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-11 01:01:03 +00:00
Chris Lattner
a5546fbd10 Evan implemented the machineinstr improvements.
The rot example works if the type is cast to ubyte.

Move sroa examples together and upgrade them to HEAD syntax.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32423 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-11 00:44:03 +00:00
Chris Lattner
1dbea1598c implement promotion of unions containing two packed types of the same width.
This implements Transforms/ScalarRepl/union-packed.ll


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32422 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-11 00:35:08 +00:00
Chris Lattner
82e8a8ffe4 clarify some comments, simplify some checks, fix:
Regression/Transforms/IndVarsSimplify/2006-12-10-BitCast.ll


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32420 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-11 00:12:31 +00:00
Chris Lattner
ae6a305094 * Eliminate calls to CastInst::createInferredCast.
* Add support for promoting unions with fp values in them.  This produces
   our new int<->fp bitcast instructions, implementing
   Transforms/ScalarRepl/union-fp-int.ll

As an example, this allows us to compile this:

union intfloat { int i; float f; };
float invsqrt(const float arg_x) {
    union intfloat x = { .f = arg_x };
    const float xhalf = arg_x * 0.5f;
    x.i = 0x5f3759df - (x.i >> 1);
    return x.f * (1.5f - xhalf * x.f * x.f);
}

into:

_invsqrt:
        movss 4(%esp), %xmm0
        movd %xmm0, %eax
        sarl %eax
        movl $1597463007, %ecx
        subl %eax, %ecx
        movd %ecx, %xmm1
        mulss LCPI1_0, %xmm0
        mulss %xmm1, %xmm0
        movss LCPI1_1, %xmm2
        mulss %xmm1, %xmm0
        subss %xmm0, %xmm2
        movl 8(%esp), %eax
        mulss %xmm2, %xmm1
        movss %xmm1, (%eax)
        ret

instead of:

_invsqrt:
        subl $4, %esp
        movss 8(%esp), %xmm0
        movss %xmm0, (%esp)
        movl (%esp), %eax
        movl $1597463007, %ecx
        sarl %eax
        subl %eax, %ecx
        movl %ecx, (%esp)
        mulss LCPI1_0, %xmm0
        movss (%esp), %xmm1
        mulss %xmm1, %xmm0
        mulss %xmm1, %xmm0
        movss LCPI1_1, %xmm2
        subss %xmm0, %xmm2
        mulss %xmm2, %xmm1
        movl 12(%esp), %eax
        movss %xmm1, (%eax)
        addl $4, %esp
        ret


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32418 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-10 23:56:50 +00:00
Anton Korobeynikov
d27a258d2d Cleaned setjmp/longjmp lowering interfaces. Now we're producing right
code (both asm & cbe) for Mingw32 target.
Removed autoconf checks for underscored versions of setjmp/longjmp.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32415 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-10 23:12:42 +00:00
Jim Laskey
0eadd73bd7 Reverting until finding the cause of secondary bugs.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32413 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-10 13:09:42 +00:00
Jim Laskey
630def54f4 __PPC64CompilationCallback code was allowing registers to be clobbered by stub.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32412 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-10 12:13:31 +00:00
Chris Lattner
87f28e87d7 fix PR1039 by making timing info be destroyed by llvm_shutdown, not by
static dtors.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32411 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-10 07:40:46 +00:00
Rafael Espindola
796cbd87b4 .align is in bits
.comm is in bytes
:-(


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32408 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-10 02:53:14 +00:00
Rafael Espindola
3e69a7e527 fix test/Regression/CodeGen/X86/weak.ll
if a variable has no initialization, I->getInitializer() will fail


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32407 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-09 23:14:08 +00:00
Evan Cheng
1a8f1fe676 Preliminary soft float support.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32394 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-09 02:42:38 +00:00
Evan Cheng
c9ab2f39ce Added option -soft-float to generate SW fp library calls instead of fp instructions.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32393 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-09 02:41:30 +00:00
Devang Patel
298feadd7a Analysis resolver now manages AnalysisImpls for the given patch.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32389 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-09 01:11:34 +00:00
Devang Patel
5b640e7edc Top level pass manager keeps track of other managers, so this can be
removed now.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32388 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-09 00:09:12 +00:00
Devang Patel
cd520b187a Use analysis resolver to find the info.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32387 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-09 00:07:38 +00:00
Devang Patel
c9a6293af7 Do not drop ImmutablePasses on the floor.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32386 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-08 23:57:43 +00:00
Devang Patel
145e83dc12 Set AnalysisResolver for the passes when they are inserted into
pass manager queuer.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32385 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-08 23:53:00 +00:00
Devang Patel
f3dc6d9f95 Add AnalysisResolver_New. It is a replacement for existing
AnalysisResolver.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32383 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-08 23:28:54 +00:00
Devang Patel
591b1b7526 Keep track of other pass managers, that are not directly managed by
top level manager. Use this info to findAnalysisPass.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32382 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-08 23:11:43 +00:00
Devang Patel
c475692c5e Implement top level FunctionPassManager::run(Function &F)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32381 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-08 22:57:48 +00:00