With this pass, StrongPHIElim can compile very simple testcases correctly.  There's still a ways
to go before it's ready for prime time, though.

llvm-svn: 48719

llvm-svn: 48718

llvm-svn: 48717

Transform (zext (or (icmp), (icmp))) to (or (zext (cimp), (zext icmp))) if at least one of the (zext icmp) can be transformed to eliminate an icmp.

llvm-svn: 48715

llvm-svn: 48714

llvm-svn: 48713

llvm-svn: 48712

llvm-svn: 48711

llvm-svn: 48710

llvm-svn: 48709

llvm-svn: 48708

llvm-svn: 48707

llvm-svn: 48706

llvm-svn: 48705

llvm-svn: 48704

llvm-svn: 48703

of compiler used.

llvm-svn: 48702

llvm-svn: 48701

llvm-svn: 48700

on ideas mentioned in PR686.
Written by Mikhail Glushenkov and contributed by Codedgers, Inc.

Old llvmc will be removed soon after new one will have all its properties.

llvm-svn: 48699

by default.  Adding a %p works with this new version and is backwards
compatible, though it looks a little funny.

llvm-svn: 48698

llvm-svn: 48697

llvm-svn: 48696

Force stack alignment to 16 bytes on win targets.

llvm-svn: 48695

llvm-svn: 48694

llvm-svn: 48693

llvm-svn: 48692

llvm-svn: 48691

llvm-svn: 48690

The effect is that if a variable is uninitialized along a branch (but initialized along another), at merge points it is considered uninitialized.  Previously we had the opposite behavior.  The new behavior is more conservative, and more in line with gcc's behavior.

llvm-svn: 48689

llvm-svn: 48688

llvm-svn: 48687

llvm-svn: 48686

llvm-svn: 48685

llvm-svn: 48684

llvm-svn: 48683

llvm-svn: 48682

llvm-svn: 48681

This fires dozens of times across spec and multisource, but I don't know
if it actually speeds stuff up.  Hopefully the testers will show something
nice :)

llvm-svn: 48680

llvm-svn: 48679