| Commit message (Expand) | Author | Age | Files | Lines |
* | Use COPY in FastISel everywhere it is safe and trivial. | Jakob Stoklund Olesen | 2010-07-11 | 1 | -18/+8 |
* | Replace copyRegToReg with copyPhysReg for Mips. | Jakob Stoklund Olesen | 2010-07-11 | 2 | -56/+68 |
* | Replace copyRegToReg with copyPhysReg for Alpha. | Jakob Stoklund Olesen | 2010-07-11 | 2 | -28/+15 |
* | Use COPY in targets | Jakob Stoklund Olesen | 2010-07-10 | 3 | -22/+8 |
* | Don't emit st(0)/st(1) copies as FpMOV instructions. Use FpSET_ST? instead. | Jakob Stoklund Olesen | 2010-07-10 | 2 | -83/+65 |
* | Add parentheses yet again to satisfy GCC's warnings. | Chandler Carruth | 2010-07-10 | 1 | -3/+3 |
* | Reapply bottom-up fast-isel, with several fixes for x86-32: | Dan Gohman | 2010-07-10 | 8 | -94/+301 |
* | An x86 function returns a floating point value in st(0), and we must make sure | Jakob Stoklund Olesen | 2010-07-10 | 1 | -18/+26 |
* | Declare YMM subregisters in the right way! Thanks Jakob | Bruno Cardoso Lopes | 2010-07-09 | 1 | -2/+1 |
* | Add AVX 256-bit packed MOVNT variants | Bruno Cardoso Lopes | 2010-07-09 | 1 | -0/+22 |
* | Remember the *_TC opcodes for load/store | Jakob Stoklund Olesen | 2010-07-09 | 1 | -0/+4 |
* | Add AVX 256-bit unpack and interleave | Bruno Cardoso Lopes | 2010-07-09 | 1 | -0/+17 |
* | Automatically fold COPY instructions into stack load/store. | Jakob Stoklund Olesen | 2010-07-09 | 2 | -2/+2 |
* | Fix a few tests | Jakob Stoklund Olesen | 2010-07-09 | 1 | -1/+1 |
* | In the presence of variable sized objects, allocate an emergency spill slot. | Jim Grosbach | 2010-07-09 | 1 | -3/+10 |
* | Start the support for AVX instructions with 256-bit %ymm registers. A couple of | Bruno Cardoso Lopes | 2010-07-09 | 8 | -18/+99 |
* | --- Reverse-merging r107947 into '.': | Bob Wilson | 2010-07-09 | 5 | -205/+70 |
* | Merge VEX enums with other x86 enum forms. Also fix all checks of which VEX | Bruno Cardoso Lopes | 2010-07-09 | 2 | -22/+17 |
* | Fix the memoperand offsets in code generated for va_start. | Dan Gohman | 2010-07-09 | 1 | -3/+3 |
* | have the mc lowering process handle a few tail call forms, lowering them to | Chris Lattner | 2010-07-09 | 3 | -13/+19 |
* | Print "dregpair" NEON operands with a space between them, for readability and | Bob Wilson | 2010-07-09 | 1 | -1/+1 |
* | Re-apply bottom-up fast-isel, with fixes. Be very careful to avoid emitting | Dan Gohman | 2010-07-09 | 5 | -70/+205 |
* | Factor out x86 segment override prefix encoding, and also use it for VEX | Bruno Cardoso Lopes | 2010-07-09 | 1 | -18/+33 |
* | reject pseudo instructions early in the encoder. | Chris Lattner | 2010-07-09 | 2 | -11/+5 |
* | Remove trailing whitespaces from file | Bruno Cardoso Lopes | 2010-07-09 | 1 | -66/+66 |
* | Change LEA to have 5 operands for its memory operand, just | Chris Lattner | 2010-07-08 | 15 | -160/+98 |
* | add some long-overdue enums to refer to the parts of the 5-operand | Chris Lattner | 2010-07-08 | 6 | -44/+56 |
* | Remember the VR64 register class | Jakob Stoklund Olesen | 2010-07-08 | 1 | -0/+2 |
* | Rework segment prefix emission code to handle segments | Chris Lattner | 2010-07-08 | 1 | -47/+41 |
* | introduce a new X86II::getMemoryOperandNo method, which | Chris Lattner | 2010-07-08 | 1 | -12/+71 |
* | Switch SPU calling convention (function arguments) | Kalle Raiskila | 2010-07-08 | 4 | -118/+52 |
* | Check for FiniteOnlyFPMath as well. | Evan Cheng | 2010-07-08 | 1 | -1/+1 |
* | Teach the x86 floating point stackifier to handle COPY instructions. | Jakob Stoklund Olesen | 2010-07-08 | 1 | -1/+36 |
* | Implement X86InstrInfo::copyPhysReg | Jakob Stoklund Olesen | 2010-07-08 | 2 | -0/+64 |
* | The NEONPreAllocPass should never have to assign fixed registers anymore. | Bob Wilson | 2010-07-08 | 1 | -34/+1 |
* | For big-endian systems, VLD2/VST2 with 32-bit vector elements will swap the | Bob Wilson | 2010-07-08 | 1 | -2/+2 |
* | Clean up a comment. | Bob Wilson | 2010-07-08 | 1 | -5/+5 |
* | Convert EXTRACT_SUBREG to COPY when emitting machine instrs. | Jakob Stoklund Olesen | 2010-07-08 | 3 | -10/+8 |
* | Remove references to INSERT_SUBREG after de-SSA. | Jakob Stoklund Olesen | 2010-07-08 | 1 | -8/+6 |
* | Teach instcombine to transform | Benjamin Kramer | 2010-07-08 | 1 | -11/+0 |
* | A slight reworking of the custom patterns for x86-64 tpoff codegen and | Eric Christopher | 2010-07-08 | 1 | -9/+11 |
* | r107852 is only safe with -enable-unsafe-fp-math to account for +0.0 == -0.0. | Evan Cheng | 2010-07-08 | 1 | -3/+5 |
* | Optimize some vfp comparisons to integer ones. This patch implements the simp... | Evan Cheng | 2010-07-08 | 2 | -10/+48 |
* | Changes to ARM tail calls, mostly cosmetic. | Dale Johannesen | 2010-07-08 | 3 | -8/+20 |
* | Revert 107840 107839 107813 107804 107800 107797 107791. | Dan Gohman | 2010-07-08 | 5 | -193/+69 |
* | fix copies to/from GR8_ABCD_H even more | Jakob Stoklund Olesen | 2010-07-07 | 1 | -1/+3 |
* | grammar | Jim Grosbach | 2010-07-07 | 1 | -1/+1 |
* | Handle cases where the post-RA scheduler may move instructions between the | Jim Grosbach | 2010-07-07 | 1 | -6/+21 |
* | finish up support for callw: PR7195 | Chris Lattner | 2010-07-07 | 1 | -1/+1 |
* | Implement the major chunk of PR7195: support for 'callw' | Chris Lattner | 2010-07-07 | 6 | -11/+36 |