need the float->double part.

llvm-svn: 25452

llvm-svn: 25451

llvm-svn: 25450

llvm-svn: 25449

   boolean flag if we read a function prototype that needs upgrading.
2. Don't upgrade the CallInst instruction until after its been inserted
   into the basic block, and only if we know that we have seen an
   upgraded intrinsic function.

llvm-svn: 25448

in the case where it needs to cast back to a signed type.

llvm-svn: 25447

2. Have UpgradeInstrinicCall return an Instruction* instead of a CallInst*
   and return the needed CastInst* if the result of the upgraded call needs
   to be casted back to a signed type.

llvm-svn: 25446

need to be upgraded.

llvm-svn: 25445

from add / sub.

llvm-svn: 25444

llvm-svn: 25443

llvm-svn: 25442

of comparisons on the various type objects.

llvm-svn: 25441

llvm-svn: 25440

llvm-svn: 25439

(or (x >> c) | (y << (32 - c))) ==> (shrd x, y, c)
(or (x << c) | (y >> (32 - c))) ==> (shld x, y, c)

llvm-svn: 25438

in the pattern.

llvm-svn: 25437

llvm-svn: 25436

llvm-svn: 25435

llvm-svn: 25434

interface requires unsigned arguments.

llvm-svn: 25433

the Function and the CallInst: UpgradeCallsToIntrinsic(Function*). Also,
re-factor the AutoUpgrade implementation to eliminate some duplication of
code.

llvm-svn: 25432

floating point intrinsics.

llvm-svn: 25431

hypothetical future boog.

llvm-svn: 25430

unbreaks front-ends that don't use __main (like the new CFE).

llvm-svn: 25429

llvm-svn: 25428

llvm-svn: 25427

Now all constant folding in the code generator is in one place.

llvm-svn: 25426

support FSIN/FCOS nodes, do not lower sin/cos to them.

llvm-svn: 25425

library list as well.  This should help bugpoint.

llvm-svn: 25424

llvm-svn: 25423

llvm-svn: 25422

llvm-svn: 25421

llvm-svn: 25420

fixes: test/Regression/CodeGen/Generic/2005-01-18-SetUO-InfLoop.ll

llvm-svn: 25419

llvm-svn: 25418

support now works in gdb.

llvm-svn: 25417

llvm-svn: 25416

llvm-svn: 25415

llvm-svn: 25414

llvm-svn: 25413