| 2009-04-17 |
Rafael Espindola | For general dynamic TLS access we must use |
tree | commitdiff |
| 2009-04-17 |
Bill Wendling | Revert r69335 and r69336. They were causing build failures. |
tree | commitdiff |
| 2009-04-17 |
Dan Gohman | MOV8rr_NOREX is a "Move" instruction. This doesn't... |
tree | commitdiff |
| 2009-04-17 |
Dan Gohman | Don't use MOV8rr_NOREX on x86-32. It doesn't actually... |
tree | commitdiff |
| 2009-04-16 |
Rafael Espindola | fix PR3995. A scale must be 1, 2, 4 or 8. |
tree | commitdiff |
| 2009-04-15 |
Dan Gohman | Fix 80-column violations. |
tree | commitdiff |
| 2009-04-15 |
Dan Gohman | Add a folding table entry for MOV8rr_NOREX. |
tree | commitdiff |
| 2009-04-15 |
Dan Gohman | Fix X86MachineFunctionInfo's doxygen comment. |
tree | commitdiff |
| 2009-04-15 |
Dan Gohman | Do for GR16_NOREX what r69049 did for GR8_NOREX, to... |
tree | commitdiff |
| 2009-04-15 |
Dan Gohman | Add a new MOV8rr_NOREX, and make X86's copyRegToReg... |
tree | commitdiff |
| 2009-04-15 |
Dan Gohman | GR8_NOREX can contain the H registers, since they don... |
tree | commitdiff |
| 2009-04-14 |
Dan Gohman | For the h-register addressing-mode trick, use the corre... |
tree | commitdiff |
| 2009-04-14 |
Evan Cheng | Some of GR8_NOREX registers are only available in 64... |
tree | commitdiff |
| 2009-04-13 |
Dan Gohman | Rename COPY_TO_SUBCLASS to COPY_TO_REGCLASS, and generalize |
tree | commitdiff |
| 2009-04-13 |
Devang Patel | Reapply 68847. |
tree | commitdiff |
| 2009-04-13 |
Dan Gohman | Implement x86 h-register extract support. |
tree | commitdiff |
| 2009-04-13 |
Dan Gohman | Remove x86's special-case handling for ISD::TRUNCATE and |
tree | commitdiff |
| 2009-04-13 |
Dan Gohman | Fix copy+pastos in comments. |
tree | commitdiff |
| 2009-04-13 |
Dan Gohman | List the l registers before h registers, for consistency. |
tree | commitdiff |
| 2009-04-13 |
Dan Gohman | Use X86::SUBREG_8BIT instead of hard-coding the equival... |
tree | commitdiff |
| 2009-04-13 |
Dan Gohman | Add a comment about MOVSX64rr8. |
tree | commitdiff |
| 2009-04-13 |
Dan Gohman | Fix another hard-coded constant to use X86AddrNumOperands. |
tree | commitdiff |
| 2009-04-13 |
Rafael Espindola | X86-64 TLS support for local exec and initial exec. |
tree | commitdiff |
| 2009-04-12 |
Rafael Espindola | In X86DAGToDAGISel::MatchWrapper, if base or index... |
tree | commitdiff |
| 2009-04-12 |
Rafael Espindola | refactor some code into X86DAGToDAGISel::MatchWrapper |
tree | commitdiff |
| 2009-04-12 |
Chris Lattner | fix a cross-block fastisel crash handling overflow... |
tree | commitdiff |
| 2009-04-12 |
Chris Lattner | simplify code by using IntrinsicInst. |
tree | commitdiff |
| 2009-04-12 |
Chris Lattner | Add new TargetInstrDesc::hasImplicitUseOfPhysReg and |
tree | commitdiff |
| 2009-04-11 |
Dan Gohman | Revert r68847. It breaks the build on non-Darwin target... |
tree | commitdiff |
| 2009-04-11 |
Devang Patel | Keep track of inlined functions and their locations... |
tree | commitdiff |
| 2009-04-10 |
Rafael Espindola | Don't fold a load if the other operand is a TLS address. |
tree | commitdiff |
| 2009-04-10 |
Chris Lattner | a few fixes to "addrspace(256) is reference offset... |
tree | commitdiff |
| 2009-04-09 |
Dan Gohman | Remove the obsolete SelectionDAG::getNodeValueTypes... |
tree | commitdiff |
| 2009-04-09 |
Chris Lattner | Fix code size computation on x86-64, patch by Zoltan... |
tree | commitdiff |
| 2009-04-09 |
Dan Gohman | Fix grammaros in comments. |
tree | commitdiff |
| 2009-04-08 |
Rafael Espindola | Re-apply 68552. |
tree | commitdiff |
| 2009-04-08 |
Rafael Espindola | Avoid a hard coded constant. |
tree | commitdiff |
| 2009-04-08 |
Dan Gohman | Implement support for using modeling implicit-zero... |
tree | commitdiff |
| 2009-04-07 |
Bill Wendling | Temporarily revert r68552. This was causing a failure... |
tree | commitdiff |
| 2009-04-07 |
Rafael Espindola | Reduce code duplication on the TLS implementation. |
tree | commitdiff |
| 2009-04-03 |
Mon P Wang | Added a x86 dag combine to increase the chances to... |
tree | commitdiff |
| 2009-04-01 |
Chris Lattner | silence warning in release-asserts build. |
tree | commitdiff |
| 2009-03-31 |
Evan Cheng | i128 shift libcalls are not available on x86. |
tree | commitdiff |
| 2009-03-31 |
Dan Gohman | Reapply 68073, with fixes. EH Landing-pad basic blocks... |
tree | commitdiff |
| 2009-03-31 |
Rafael Espindola | remove unused arguments. |
tree | commitdiff |
| 2009-03-31 |
Bill Wendling | Really temporarily revert r68073. |
tree | commitdiff |
| 2009-03-31 |
Bill Wendling | Oy! When reverting r68073, I added in experimental... |
tree | commitdiff |
| 2009-03-31 |
Bill Wendling | Revert r68073. It's causing a failure in the Apple... |
tree | commitdiff |
| 2009-03-31 |
Evan Cheng | X86 address mode isel tweak. If the base of the address... |
tree | commitdiff |
| 2009-03-30 |
Dan Gohman | Except in asm-verbose mode, avoid printing labels for... |
tree | commitdiff |
| 2009-03-30 |
Evan Cheng | When optimzing a mul by immediate into two, the resulti... |
tree | commitdiff |
| 2009-03-30 |
Anton Korobeynikov | Do not propagate ELF-specific stuff (data.rel) into... |
tree | commitdiff |
| 2009-03-30 |
Anton Korobeynikov | Add data.rel stuff |
tree | commitdiff |
| 2009-03-28 |
Rafael Espindola | Use array_lengthof |
tree | commitdiff |
| 2009-03-28 |
Rafael Espindola | Have only one definition of X86AddrNumOperands. |
tree | commitdiff |
| 2009-03-28 |
Rafael Espindola | Make code a bit less brittle by no hardcoding the number |
tree | commitdiff |
| 2009-03-28 |
Evan Cheng | Optimize some 64-bit multiplication by constants into... |
tree | commitdiff |
| 2009-03-27 |
Rafael Espindola | Avoid hardcoding that X86 addresses have 4 operands. |
tree | commitdiff |
| 2009-03-27 |
Rafael Espindola | Use less hard coded constants to make the code less... |
tree | commitdiff |
| 2009-03-27 |
Rafael Espindola | I am trying to add a segment to the X86 addresses match... |
tree | commitdiff |
| 2009-03-26 |
Evan Cheng | -no-implicit-float means explicit fp operations are... |
tree | commitdiff |
| 2009-03-26 |
Bill Wendling | Pull transform from target-dependent code into target... |
tree | commitdiff |
| 2009-03-26 |
Bill Wendling | Match this pattern so that we can generate simpler... |
tree | commitdiff |
| 2009-03-26 |
Bill Wendling | Doxygen-ify comments. |
tree | commitdiff |
| 2009-03-25 |
Evan Cheng | CodeGen still defaults to non-verbose asm, but llc... |
tree | commitdiff |
| 2009-03-25 |
Evan Cheng | Don't print global names twice with -asm-verbose. |
tree | commitdiff |
| 2009-03-24 |
Dan Gohman | I was convinced that it's ok to allow a second i8 retur... |
tree | commitdiff |
| 2009-03-24 |
Evan Cheng | Do not emit comments unless -asm-verbose. |
tree | commitdiff |
| 2009-03-23 |
Dan Gohman | Correct some comments. Operand numbers start at 0. |
tree | commitdiff |
| 2009-03-23 |
Evan Cheng | Model inline asm constraint which ties an input to... |
tree | commitdiff |
| 2009-03-23 |
Dan Gohman | Fix a grammaro in a comment that Bill noticed. |
tree | commitdiff |
| 2009-03-23 |
Dan Gohman | Add comments explaining why there's only one register for |
tree | commitdiff |
| 2009-03-19 |
Nick Lewycky | Remove strange extra semicolons. |
tree | commitdiff |
| 2009-03-18 |
Chris Lattner | Disable the "call to immediate" optimization on x86... |
tree | commitdiff |
| 2009-03-17 |
Dan Gohman | Recognize bswapl as bswap too. |
tree | commitdiff |
| 2009-03-17 |
Dan Gohman | Recognize "bswapq" as an alternate spelling for the... |
tree | commitdiff |
| 2009-03-14 |
Dan Gohman | Use %rip-relative addressing on x86-64 whenever practic... |
tree | commitdiff |
| 2009-03-14 |
Dan Gohman | Don't forego folding of loads into 64-bit adds when... |
tree | commitdiff |
| 2009-03-13 |
Dan Gohman | Improve FastISel's handling of truncates to i1, and... |
tree | commitdiff |
| 2009-03-13 |
Dan Gohman | Fix FastISel's assumption that i1 values are always... |
tree | commitdiff |
| 2009-03-13 |
Rafael Espindola | add 8 and 16 bit TLS moves. |
tree | commitdiff |
| 2009-03-13 |
Rafael Espindola | Improve sext and zext of TLS variables. |
tree | commitdiff |
| 2009-03-13 |
Chris Lattner | generalize this code so that fast isel handles integer... |
tree | commitdiff |
| 2009-03-13 |
Bill Wendling | These instructions have special lowering that may lower... |
tree | commitdiff |
| 2009-03-13 |
Evan Cheng | Fix some significant problems with constant pools that... |
tree | commitdiff |
| 2009-03-13 |
Chris Lattner | generalize the previous code to use the full generality... |
tree | commitdiff |
| 2009-03-13 |
Chris Lattner | optimize the case of cond ? 42 : 41 and friends. This... |
tree | commitdiff |
| 2009-03-13 |
Dan Gohman | Enhance address-mode folding of ISD::ADD to handle... |
tree | commitdiff |
| 2009-03-12 |
Evan Cheng | Re-apply 66024 with fixes: 1. Fixed indirect call to... |
tree | commitdiff |
| 2009-03-12 |
Chris Lattner | Move 3 "(add (select cc, 0, c), x) -> (select cc, x... |
tree | commitdiff |
| 2009-03-12 |
Chris Lattner | improve comment. |
tree | commitdiff |
| 2009-03-12 |
Evan Cheng | On x86, if the only use of a i64 load is a i64 store... |
tree | commitdiff |
| 2009-03-11 |
Dan Gohman | Revert r66024. The JIT encoding for CALLpcrel32 is... |
tree | commitdiff |
| 2009-03-11 |
Rafael Espindola | optimize i8 and i16 tls values. |
tree | commitdiff |
| 2009-03-11 |
Bill Wendling | Add a -no-implicit-float flag. This acts like -soft... |
tree | commitdiff |
| 2009-03-11 |
Duncan Sands | It makes no sense to have a ODR version of common |
tree | commitdiff |
| 2009-03-11 |
Mon P Wang | For yonah, fix a vector shuffle case for v16i8 where... |
tree | commitdiff |
| 2009-03-11 |
Mon P Wang | Fixed a v8i16 shuffle case that should generate a pshuf... |
tree | commitdiff |
| 2009-03-11 |
Chris Lattner | formatting change, reduce indentation. No functionalit... |
tree | commitdiff |
| 2009-03-10 |
Dan Gohman | Add more information to the EFLAGS note. |
tree | commitdiff |
| next |
projects / oota-llvm.git / history