llvm-svn: 98137

llvm-svn: 98132

for darwin/x86 to be completely mcized.

llvm-svn: 98130

llvm-svn: 98125

llvm-svn: 98124

1. Be careful with cse "cheap" expressions. e.g. constant materialization. Only cse them when the common expression is local or in a direct predecessor. We don't want cse of cheap instruction causing other expressions to be spilled.
2. Watch out for the case where the expression doesn't itself uses a virtual register. e.g. lea of frame object. If the common expression itself is used by copies (common for passing addresses to function calls), don't perform the cse. Since these expressions do not use a register, it creates a live range but doesn't close any, we want to be very careful with increasing register pressure.

Note these are heuristics so machine cse doesn't make register allocator unhappy. Once we have proper live range splitting and re-materialization support in place, these should be evaluated again.

Now machine cse is almost always a win on llvm nightly tests on x86 and x86_64.

llvm-svn: 98121

dead IsPCRel argument.

llvm-svn: 98117

need to be MCized, but the last debug info thing are LEB and
cygwin specific (which the MC api doesn't support yet) and
one specific form of EmitReference which I'll tackle next.

llvm-svn: 98116

llvm-svn: 98109

llvm-svn: 98108

llvm-svn: 98107

llvm-svn: 98106

method.  With this, comments should end up on the same lines as the .byte
directives (for example) and we now get no output with:

$ llc CodeGen/X86/2009-02-12-DebugInfoVLA.ll -o - -filetype=null -asm-verbose

woot.

llvm-svn: 98105

llvm-svn: 98104

off "EOL".

llvm-svn: 98102

significant debug info testcases are now all going through
MCStreamer, though they print a lot of extraneous newlines to "O".

llvm-svn: 98101

is preparatory to having PEI's scavenged frame index value reuse logic
properly distinguish types of frame values (e.g., whether the value is
stack-pointer relative or frame-pointer relative).

No functionality change.

llvm-svn: 98086

llvm-svn: 98048

Now that setStartLabel takes an MCSymbol, we can de-ID'ize
beginScope and RecordSourceLine.

llvm-svn: 98047

failure remaining on mainline.

llvm-svn: 98046

llvm-svn: 98045

llvm-svn: 98044

  coalescer) handle sub-register classes.
- Add heuristics to avoid non-profitable cse. Given the current lack of live
  range splitting, avoid cse when an expression has PHI use and the would be
  new use is in a BB where the expression wasn't already being used.

llvm-svn: 98043

failures.

llvm-svn: 98039

multibyte opcodes and in the case of multiple scopes.

llvm-svn: 98036

beginScope and RecordSourceLine.

llvm-svn: 98035

now that the dependence on ID is removed from MMI.

llvm-svn: 98034

happens, the start/end of a scope can never be the same.

llvm-svn: 98032

and add a FIXME about how we are eventually going to zap this
lookup table once mc world domination is complete.

llvm-svn: 98031

llvm-svn: 98025

physreg becomes ridiculously high.

std::upper_bound may be log(N), but for sufficiently large live intervals, it
becomes log(N)*cachemiss = a long long time.

This patch improves coalescer time by 4500x for a function with 20000
function calls. The generated code is different, but not significantly worse -
the allocator hints are almost as good as physreg coalescing anyway.

llvm-svn: 98023

llvm-svn: 98017

recomputation of the labels.

llvm-svn: 98016

llvm-svn: 98015

the one special case into EmitSectionOffset.  MCize
the non-special case in EmitSectionOffset.

llvm-svn: 98014

llvm-svn: 98013

used for 'flavor'.

llvm-svn: 98012

we can eliminate "flavor".

llvm-svn: 98011

MCize it.

llvm-svn: 98010

Don't waste time trying to CSE labels, phis, inline asm. Definitely avoid cse implicit-def for obvious performance reason.

llvm-svn: 98009