llvm-svn: 46681

This case returns the value in ST(0) and then has to convert it to an SSE
register.  This causes significant codegen ugliness in some cases.  For 
example in the trivial fp-stack-direct-ret.ll testcase we used to generate:

_bar:
	subl	$28, %esp
	call	L_foo$stub
	fstpl	16(%esp)
	movsd	16(%esp), %xmm0
	movsd	%xmm0, 8(%esp)
	fldl	8(%esp)
	addl	$28, %esp
	ret

because we move the result of foo() into an XMM register, then have to
move it back for the return of bar.

Instead of hacking ever-more special cases into the call result lowering code
we take a much simpler approach: on x86-32, fp return is modeled as always 
returning into an f80 register which is then truncated to f32 or f64 as needed.
Similarly for a result, we model it as an extension to f80 + return.

This exposes the truncate and extensions to the dag combiner, allowing target
independent code to hack on them, eliminating them in this case.  This gives 
us this code for the example above:

_bar:
	subl	$12, %esp
	call	L_foo$stub
	addl	$12, %esp
	ret

The nasty aspect of this is that these conversions are not legal, but we want
the second pass of dag combiner (post-legalize) to be able to hack on them.
To handle this, we lie to legalize and say they are legal, then custom expand
them on entry to the isel pass (PreprocessForFPConvert).  This is gross, but
less gross than the code it is replacing :)

This also allows us to generate better code in several other cases.  For 
example on fp-stack-ret-conv.ll, we now generate:

_test:
	subl	$12, %esp
	call	L_foo$stub
	fstps	8(%esp)
	movl	16(%esp), %eax
	cvtss2sd	8(%esp), %xmm0
	movsd	%xmm0, (%eax)
	addl	$12, %esp
	ret

where before we produced (incidentally, the old bad code is identical to what
gcc produces):

_test:
	subl	$12, %esp
	call	L_foo$stub
	fstpl	(%esp)
	cvtsd2ss	(%esp), %xmm0
	cvtss2sd	%xmm0, %xmm0
	movl	16(%esp), %eax
	movsd	%xmm0, (%eax)
	addl	$12, %esp
	ret

Note that we generate slightly worse code on pr1505b.ll due to a scheduling 
deficiency that is unrelated to this patch.

llvm-svn: 46307

llvm-svn: 45859

x86 backend where instructions were not marked maystore/mayload, and perf issues where
instructions were not marked neverHasSideEffects.  It would be really nice if we could
write patterns for copy instructions.

I have audited all the x86 instructions down to MOVDQAmr.  The flags on others and on
other targets are probably not right in all cases, but no clients currently use this
info that are enabled by default.

llvm-svn: 45829

inferred from the instr patterns.

llvm-svn: 45824

llvm-svn: 45667

llvm-svn: 45418

llvm-svn: 45268

based what flag to set on whether it was already marked as
"isRematerializable". If there was a further check to determine if it's "really"
rematerializable, then I marked it as "mayHaveSideEffects" and created a check
in the X86 back-end similar to the remat one.

llvm-svn: 45132

X86CodeEmitter.cpp:378: failed assertion `0 && "Immediate size not set!"'

I *think* this is right, but Evan, please verify.  It also looks like
CMPSDrr and maybe others are missing this info.  Evan, plz investigate.

llvm-svn: 45074

llvm-svn: 45058

Implicit def instructions, e.g. X86::IMPLICIT_DEF_GR32, are always re-materializable and they should not be spilled.

llvm-svn: 44960

Remove a bogus optimization. It's not possible to do a move to low element to a <8 x i16> or <16 x i8> vector.

llvm-svn: 44669

sometimes emit "zero" and "all one" vectors multiple times,
for example:

_test2:
	pcmpeqd	%mm0, %mm0
	movq	%mm0, _M1
	pcmpeqd	%mm0, %mm0
	movq	%mm0, _M2
	ret

instead of:

_test2:
	pcmpeqd	%mm0, %mm0
	movq	%mm0, _M1
	movq	%mm0, _M2
	ret

This patch fixes this by always arranging for zero/one vectors
to be defined as v4i32 or v2i32 (SSE/MMX) instead of letting them be
any random type.  This ensures they get trivially CSE'd on the dag.
This fix is also important for LegalizeDAGTypes, as it gets unhappy
when the x86 backend wants BUILD_VECTOR(i64 0) to be legal even when
'i64' isn't legal.

This patch makes the following changes:

1) X86TargetLowering::LowerBUILD_VECTOR now lowers 0/1 vectors into
   their canonical types.
2) The now-dead patterns are removed from the SSE/MMX .td files.
3) All the patterns in the .td file that referred to immAllOnesV or
   immAllZerosV in the wrong form now use *_bc to match them with a
   bitcast wrapped around them.
4) X86DAGToDAGISel::SelectScalarSSELoad is generalized to handle 
   bitcast'd zero vectors, which simplifies the code actually.
5) getShuffleVectorZeroOrUndef is updated to generate a shuffle that
   is legal, instead of generating one that is illegal and expecting
   a later legalize pass to clean it up.
6) isZeroShuffle is generalized to handle bitcast of zeros.
7) several other minor tweaks.

This patch is definite goodness, but has the potential to cause random
code quality regressions.  Please be on the lookout for these and let 
me know if they happen.

llvm-svn: 44310

llvm-svn: 44204

CVTTPD2PI, CVTTPS2PI, CVTPI2PD, CVTPI2PS.

llvm-svn: 43523

for fastcc from X86CallingConv.td.  This means that nested functions
are not supported for calling convention 'fastcc'.

llvm-svn: 42934

llvm-svn: 42874

(vextract (v4f32 s2v (f32 load $addr)), 0) -> (f32 load $addr) 
(vextract (v4i32 bc (v4f32 s2v (f32 load $addr))), 0) -> (i32 load $addr)
Remove x86 specific patterns.

llvm-svn: 42677

llvm-svn: 42492

llvm-svn: 42459

Added support for new condition code modeling scheme (i.e. physical register dependency). These are a bunch of instructions that are duplicated so the x86 backend can support both the old and new schemes at the same time. They will be deleted after
all the kinks are worked out.

llvm-svn: 42285

keep f32 in SSE registers and f64 in x87.  This
is effectively a new codegen mode.
Change addLegalFPImmediate to permit float and
double variants to do different things.
Adjust callers.

llvm-svn: 42246

llvm-svn: 41962

llvm-svn: 41863

by flagging the associated instructions as being trivially rematerializable.

llvm-svn: 41775

llvm-svn: 41595

Make a 'memop' pattern just for them.

llvm-svn: 41017

llvm-svn: 40985

llvm-svn: 40982

Generalize isPSHUFDMask and add a unary SHUFPD pattern so that SHUFPD's
memory operand alignment can be tested as well, with a fix to avoid
breaking MMX's use of isPSHUFDMask.

llvm-svn: 40756

llvm-svn: 40754

X86InstrInfo::isReallyTriviallyReMaterializable knows how to handle
with the isReMaterializable flag so that it is given a chance to handle
them. Without hoisting constant-pool loads from loops this isn't very
visible, though it does keep CodeGen/X86/constant-pool-remat-0.ll from
making a copy of the constant pool on the stack.

llvm-svn: 40736

llvm-svn: 40691

mnemonics from their operands instead of single spaces. This makes the
assembly output a little more consistent with various other compilers
(f.e. GCC), and slightly easier to read. Also, update the regression
tests accordingly.

llvm-svn: 40648

Redo and generalize previously removed opt for pinsrw: (vextract (v4i32 bc (v4f32 s2v (f32 load ))), 0) -> (i32 load )

llvm-svn: 40628

Make the alignedload and alignedstore patterns always require 16-byte
alignment. This way when they are used in the "Fs" instructions, in which
a vector instruction is used for a scalar purpose, they can still require
the full vector alignment. And add a regression test for this.

llvm-svn: 40555

llvm-svn: 40547

llvm-svn: 40528

x86 target, replacing them with the new alignment attributes on memory
references.

llvm-svn: 40504