Instruction selection priority fixes to remove the XMM/XMMInt/orAVX predicates. Another commit will remove orAVX functions from X86SubTarget.

llvm-svn: 147841

llvm-svn: 147805

AsmParser holds info specific to target parser.
AsmParserVariant holds info specific to asm variants supported by the target.

llvm-svn: 147787

this substraction will result in small negative numbers at worst which
become very large positive numbers on assignment and are thus caught by
the <=4 check on the next line. The >0 check clearly intended to catch
these as negative numbers.

Spotted by inspection, and impossible to trigger given the shift widths
that can be used.

llvm-svn: 147773

Remove AVX hack in X86Subtarget. AVX/AVX2 are now treated as an SSE level. Predicate functions have been altered to maintain previous names and behavior.

llvm-svn: 147770

llvm-svn: 147769

Reorder a bunch of patterns to put the AVX version first thus giving it priority over the SSE version. Another step towards trying to remove the AVX hack that disables SSE from X86Subtarget.

llvm-svn: 147768

Clean up patterns for MOVNT*. Not sure why there were floating point types on MOVNTPS and MOVNTDQ. And v4i64 was completely missing.

llvm-svn: 147767

Mark MOVNTI as being supported in SSE2 OR AVX mode. This instruction has no AVX equivalent so we should use the SSE version.

llvm-svn: 147766

Move SSE2 logical operations PAND/POR/PXOR/PANDN above SSE1 logical operations ANDPS/ORPS/XORPS/ANDNPS. This fixes a pattern ordering issue that meant that the SSE2 instructions could never be directly selected since the SSE1 patterns would always match first. This is largely moot with the ExeDepsFix pass, but I'm trying to audit for all such ordering issues.

llvm-svn: 147765

Change some places that were checking for AVX OR SSE1/2 to use hasXMM/hasXMMInt instead. Also fix one place that checked SSE3, but accidentally excluded AVX to use hasSSE3orAVX. This is a step towards removing the AVX hack from the X86Subtarget.h

llvm-svn: 147764

Don't disable MMX support when AVX is enabled. Fix predicates for MMX instructions that were added along with SSE instructions to check for AVX in addition to SSE level.

llvm-svn: 147762

llvm-svn: 147758

llvm-svn: 147748

llvm-svn: 147739

llvm-svn: 147738

llvm-svn: 147734

Fixes rdar://10614894

llvm-svn: 147704

llvm-svn: 147602

Peephole optimization of ptest-conditioned branch in X86 arch. Performs instruction combining of sequences generated by ptestz/ptestc intrinsics to ptest+jcc pair for SSE and AVX.

Testing: passed 'make check' including LIT tests for all sequences being handled (both SSE and AVX)

Reviewers: Evan Cheng, David Blaikie, Bruno Lopes, Elena Demikhovsky, Chad Rosier, Anton Korobeynikov
llvm-svn: 147601

This small bit of ASM code is sufficient to do what the old algorithm did:

     movq       %rax,  %xmm0
     punpckldq  (c0),  %xmm0  // c0: (uint4){ 0x43300000U, 0x45300000U, 0U, 0U }
     subpd      (c1),  %xmm0  // c1: (double2){ 0x1.0p52, 0x1.0p52 * 0x1.0p32 }
   #ifdef __SSE3__
     haddpd   %xmm0, %xmm0          
   #else
     pshufd   $0x4e, %xmm0, %xmm1 
     addpd    %xmm1, %xmm0
   #endif

It's arguably faster. One caveat, the 'haddpd' instruction isn't very fast on
all processors.
<rdar://problem/7719814>

llvm-svn: 147593

llvm-svn: 147553

(x > y) ? x : y
=>
(x >= y) ? x : y

So for something like
(x - y) > 0 : (x - y) ? 0
It will be
(x - y) >= 0 : (x - y) ? 0

This makes is possible to test sign-bit and eliminate a comparison against
zero. e.g.
subl   %esi, %edi
testl  %edi, %edi
movl   $0, %eax
cmovgl %edi, %eax
=>
xorl   %eax, %eax
subl   %esi, $edi
cmovsl %eax, %edi

rdar://10633221

llvm-svn: 147512

llvm-svn: 147495

llvm-svn: 147485

then a vxorps + vinsertf128 pair if the original vector came from a load.
rdar://10594409

llvm-svn: 147481

llvm-svn: 147453

Miscellaneous shuffle lowering cleanup. No functional changes. Primarily converting the indexing loops to unsigned to be consistent across functions.

llvm-svn: 147430

Make CanXFormVExtractWithShuffleIntoLoad reject loads with multiple uses. Also make it return false if there's not even a load at all. This makes the code better match the code in DAGCombiner that it tries to match. These two changes prevent some cases where vector_shuffles were making it to instruction selection and causing the older shuffle selection code to be triggered. Also needed to fix a bad pattern that this change exposed. This is the first step towards getting rid of the old shuffle selection support. No test cases yet because there's no way to tell whether a shuffle was handled in the legalize stage or at instruction selection.

llvm-svn: 147428

Optimize the sequence blend(sign_extend(x)) to blend(shl(x)) since SSE blend instructions only look at the highest bit.

llvm-svn: 147426

llvm-svn: 147411

Fix sfence, lfence, mfence, and clflush to be able to be selected when AVX is enabled. Fix monitor and mwait to require SSE3 or AVX, previously they worked even if SSE3 was disabled. Make prefetch instructions not set the execution domain since they don't use XMM registers.

llvm-svn: 147409

llvm-svn: 147404

llvm-svn: 147394

llvm-svn: 147393

Fix typo in a SHUFPD and VSHUFPD pattern that prevented SHUFPD/VSHUFPD with a load from being selected.

llvm-svn: 147392

Make FMA4 imply AVX so that YMM registers would be available. Necessitates removing from Bulldozer CPU types since it would enable AVX code generation implicitly. Also make SSE4A imply SSE3. Without some level of SSE implied, XMM registers wouldn't be legal.

llvm-svn: 147369

llvm-svn: 147368

llvm-svn: 147367

Separate the concept of having memory access in operand 4 from the concept of having the W bit set for XOP instructons. Removes ORing W-bits in the encoder and will similarly simplify the disassembler implementation.

llvm-svn: 147366