When the least bit of C is greater than V, (x&C) must be greater than V
if it is not zero, so the comparison can be simplified.

Although this was suggested in Target/X86/README.txt, it benefits any
architecture with a directly testable form of AND.

Patch by Kevin Schoedel

llvm-svn: 170576

llvm-svn: 153741

llvm-svn: 147739

llvm-svn: 147247

promoting allocas to preferred alignments that exceed the natural
alignment. This avoids some potentially expensive dynamic stack realignments.

The natural stack alignment is set in target data strings via the "S<size>"
option. Size is in bits and must be a multiple of 8. The natural stack alignment
defaults to "unspecified" (represented by a zero value), and the "unspecified"
value does not prevent any alignment promotions. Target maintainers that care
about avoiding promotions should explicitly add the "S<size>" option to their
target data strings.

llvm-svn: 141599

llvm-svn: 141563

On x86 we can't encode an immediate LHS of a sub directly. If the RHS comes from a XOR with a constant we can
fold the negation into the xor and add one to the immediate of the sub. Then we can turn the sub into an add,
which can be commuted and encoded efficiently.

This code is generated for __builtin_clz and friends.

llvm-svn: 136167

llvm-svn: 135996

llvm-svn: 131455

llvm-svn: 130499

llvm-svn: 130385

llvm-svn: 129654

Luis Felipe Strano Moraes!

llvm-svn: 129558

llvm-svn: 129527

llvm-svn: 129012

Change the test to force a sign extension and expose the problem again.

llvm-svn: 129011

llvm-svn: 127786

but which is responsible for us doing really bad things to 256.bzip2.

llvm-svn: 126126

on Core 2 and Nehalem, so the code we generate is better than GCC's here.

llvm-svn: 126100

of a constant had a minor typo introduced when copying it from the book, which
caused it to favor negative approximations over positive approximations in many
cases. Positive approximations require fewer operations beyond the multiplication.

In the case of division by 3, we still generate code that is a single instruction
larger than GCC's code.

llvm-svn: 126097

llvm-svn: 126054

llvm-svn: 125832

llvm-svn: 122700

llvm-svn: 122560

llvm-svn: 122559

llvm-svn: 122495

  int test(unsigned long a, unsigned long b) { return -(a < b); }
compiles to
  _test:                              ## @test
    cmpq  %rsi, %rdi                  ## encoding: [0x48,0x39,0xf7]
    sbbl  %eax, %eax                  ## encoding: [0x19,0xc0]
    ret                               ## encoding: [0xc3]
instead of
  _test:                              ## @test
    xorl  %ecx, %ecx                  ## encoding: [0x31,0xc9]
    cmpq  %rsi, %rdi                  ## encoding: [0x48,0x39,0xf7]
    movl  $-1, %eax                   ## encoding: [0xb8,0xff,0xff,0xff,0xff]
    cmovael %ecx, %eax                ## encoding: [0x0f,0x43,0xc1]
    ret                               ## encoding: [0xc3]

llvm-svn: 122451

to catch cases where n != m with a shift.

llvm-svn: 121608

legalization time. Since at legalization time there is no mapping from
SDNode back to the corresponding LLVM instruction and the return
SDNode is target specific, this requires a target hook to check for
eligibility. Only x86 and ARM support this form of sibcall optimization
right now.
rdar://8707777

llvm-svn: 120501

llvm-svn: 119947

there are clearly no stores between the load and the store. This fixes
this miscompile reported as PR7833.

This breaks the test/CodeGen/X86/narrow_op-2.ll optimization, which is
safe, but awkward to prove safe. Move it to X86's README.txt.

llvm-svn: 112861

llvm-svn: 112411

for lowering without sse2.  Add a couple of new testcases.

Fixes a few libgomp tests and latent bugs.  Remove a few todos.

llvm-svn: 109078

llvm-svn: 105375

llvm-svn: 97450

code with lots of bitfields.

llvm-svn: 95809

llvm-svn: 93373

llvm-svn: 93372

llvm-svn: 92511

llvm-svn: 91823