- Dec 13, 2009
-
-
Eli Friedman authored
llvm-svn: 91230
-
- Dec 12, 2009
-
-
Eli Friedman authored
merging x >u 5 and x <s 20 because it's impossible to implement. llvm-svn: 91228
-
Evan Cheng authored
llvm-svn: 91223
-
Anton Korobeynikov authored
No testcase yet - it seems we're exposing generic codegen bugs. llvm-svn: 91221
-
Evan Cheng authored
llvm-svn: 91220
-
Evan Cheng authored
llvm-svn: 91219
-
Jim Grosbach authored
just issues an error for the moment. The front end won't yet generate these intrinsics for ARM, so this is behind the scenes until complete. llvm-svn: 91200
-
Anton Korobeynikov authored
Based on the patch by Brian Lucas! llvm-svn: 91175
-
- Dec 11, 2009
-
-
Dan Gohman authored
llvm-svn: 91158
-
Jim Grosbach authored
memory barrier instructions by definition have side effects. This prevents the post-RA scheduler from moving them around. llvm-svn: 91150
-
Anton Korobeynikov authored
This is used in some weird cases like general dynamic TLS model. This fixes PR5723 llvm-svn: 91144
-
Johnny Chen authored
llvm-svn: 91143
-
Jim Grosbach authored
llvm-svn: 91140
-
Evan Cheng authored
llvm-svn: 91104
-
Jim Grosbach authored
llvm-svn: 91090
-
- Dec 10, 2009
-
-
Jim Grosbach authored
llvm-svn: 91053
-
Jim Grosbach authored
Add memory barrier intrinsic support for ARM. Moving towards adding the atomic operations intrinsics. llvm-svn: 91003
-
- Dec 09, 2009
-
-
Evan Cheng authored
vector_shuffle (scalar_to_vector (i32 load (ptr + 4))), undef, <0, 0, 0, 0> => vector_shuffle (v4i32 load ptr), undef, <1, 1, 1, 1> iff ptr is 16-byte aligned (or can be made into 16-byte aligned). llvm-svn: 90984
-
Evan Cheng authored
llvm-svn: 90925
-
Evan Cheng authored
Move isConsecutiveLoad to SelectionDAG. It's not target dependent and it's primary used by selectdag passes. llvm-svn: 90922
-
Evan Cheng authored
- Also support the 'q' NEON registers asm code. llvm-svn: 90894
-
- Dec 08, 2009
-
-
Anton Korobeynikov authored
Reduce (cmp 0, and_su (foo, bar)) into (bit foo, bar). This saves extra instruction. Patch inspired by Brian Lucas! llvm-svn: 90819
-
- Dec 07, 2009
-
-
Dan Gohman authored
current form, it is too expensive in compile time. llvm-svn: 90781
-
Anton Korobeynikov authored
llvm-svn: 90742
-
Anton Korobeynikov authored
llvm-svn: 90740
-
Anton Korobeynikov authored
llvm-svn: 90739
-
Anton Korobeynikov authored
llvm-svn: 90737
-
- Dec 06, 2009
-
-
Anton Korobeynikov authored
in "bic sp, sp, #15" leads to unpredicatble behaviour in Thumb2 mode. Emit the following code instead: mov r4, sp bic r4, r4, #15 mov sp, r4 llvm-svn: 90724
-
- Dec 05, 2009
-
-
Bill Wendling authored
llvm-svn: 90656
-
Bill Wendling authored
- Privatize a typedef. - Call the InvalidateEntry when refining a type. llvm-svn: 90655
-
Bill Wendling authored
llvm-svn: 90640
-
Bill Wendling authored
llvm-svn: 90639
-
Dan Gohman authored
MachineBasicBlock::canFallThrough(), which is target-independent and more thorough. llvm-svn: 90634
-
David Greene authored
Remove an unneeded include. llvm-svn: 90625
-
- Dec 04, 2009
-
-
David Greene authored
Have hasLoad/StoreFrom/ToStackSlot return the relevant MachineMemOperand. llvm-svn: 90608
-
Bill Wendling authored
llvm-svn: 90588
-
- Dec 03, 2009
-
-
Jim Grosbach authored
llvm-svn: 90490
-
Chris Lattner authored
llvm-svn: 90429
-
Chris Lattner authored
llvm-svn: 90428
-
Chris Lattner authored
llvm-svn: 90419
-