llvm-svn: 49543

Use of implicit_def is not part of live interval. Create empty intervals for the uses when the live interval is being spilled.

llvm-svn: 49542

llvm-svn: 49524

of calls and less aggressive with non-readnone calls.

llvm-svn: 49516

llvm-svn: 49513

llvm-svn: 49512

llvm-svn: 49504

wrong order.

llvm-svn: 49499

llvm-svn: 49496

in addition to integer expressions. Rewrite GetOrEnforceKnownAlignment
as a ComputeMaskedBits problem, moving all of its special alignment
knowledge to ComputeMaskedBits as low-zero-bits knowledge.

Also, teach ComputeMaskedBits a few basic things about Mul and PHI
instructions.

This improves ComputeMaskedBits-based simplifications in a few cases,
but more noticeably it significantly improves instcombine's alignment
detection for loads, stores, and memory intrinsics.

llvm-svn: 49492

llvm-svn: 49491

llvm-svn: 49469

llvm-svn: 49466

llvm-svn: 49465

MOVZQI2PQIrr.  This would be better handled as a dag combine 
(with the goal of eliminating the bitconvert) but I don't know
how to do that safely.  Thoughts welcome.

llvm-svn: 49463

Teach branch folding pass about implicit_def instructions. Unfortunately we can't just eliminate them since register scavenger expects every register use to be defined. However, we can delete them when there are no intra-block uses.  Carefully removing some implicit def's which enable more blocks to be optimized away.

llvm-svn: 49461

This is not safe for all inputs.

llvm-svn: 49458

- Added insert_subreg coalescing support.

llvm-svn: 49448

llvm-svn: 49446

is needed for the x86-64-ABI handling of structs that contain floating-point
members that are returned by value.

llvm-svn: 49441

llvm-svn: 49440

figuring out the suffix to use.  implement pow(2,x) -> exp2(x).

llvm-svn: 49437

long double and simplify the code.

llvm-svn: 49435

llvm-svn: 49430

GVN and into its own pass.

llvm-svn: 49419

llvm-svn: 49418

particular value but variable type.

llvm-svn: 49416

llvm-svn: 49415

start migrating code over to use it.

llvm-svn: 49413

add support for  pow(x, 2.0) -> x*x.

llvm-svn: 49411

llvm-svn: 49389

attribute for a function.

llvm-svn: 49373

llvm-svn: 49367

correctly when unwind info is being generated.

llvm-svn: 49366

Corresponds to -fno-unwind-tables (usually default in gcc).

llvm-svn: 49361

llvm-svn: 49354

llvm-svn: 49352

review feedback from Chris quite a while ago. No functionality
change.

llvm-svn: 49348

in both time and memory savings for GVN.  For example, one testcase went from 10.5s to 6s with
this patch.

llvm-svn: 49345

llvm-svn: 49339