| Commit message (Collapse) | Author | Age | Files | Lines | |
|---|---|---|---|---|---|
| * | Change LegalFPImmediates to use APFloat. | Dale Johannesen | 2007-08-30 | 3 | -9/+9 |
| | | | | | | | | | | Add APFloat interfaces to ConstantFP, SelectionDAG. Fix integer bit in double->APFloat conversion. Convert LegalizeDAG to use APFloat interface in ConstantFPSDNode uses. llvm-svn: 41587 | ||||
| * | Move getX86RegNum into X86RegisterInfo and use it | Duncan Sands | 2007-08-29 | 6 | -88/+80 |
| | | | | | | | | | in the trampoline lowering. Lookup the jump and mov opcodes for the trampoline rather than hard coding them. llvm-svn: 41577 | ||||
| * | Added method to get Mips register numbers | Bruno Cardoso Lopes | 2007-08-28 | 3 | -61/+113 |
| | | | | | | | | Changed the stack frame layout, StackGrowsUp fits better to Mips strange stack. Stack offset calculation bug fixed! llvm-svn: 41529 | ||||
| * | Changed stack allocation On LowerFORMAL_ARGUMENTS. | Bruno Cardoso Lopes | 2007-08-28 | 1 | -20/+29 |
| | | | | | | | | Added comments about new stack allocation. Expand SelectCC for i32 results llvm-svn: 41527 | ||||
| * | Mask directive completed with CalleeSave info | Bruno Cardoso Lopes | 2007-08-28 | 2 | -73/+111 |
| | | | | | | | Comments for Mips directives added. llvm-svn: 41526 | ||||
| * | Added methods to record SPOffsets from LowerFORMAL_ARGUMENTS | Bruno Cardoso Lopes | 2007-08-28 | 1 | -3/+53 |
| | | | | | llvm-svn: 41525 | ||||
| * | Add a comment about using libc memset/memcpy or generating inline code. | Rafael Espindola | 2007-08-27 | 1 | -2/+5 |
| | | | | | llvm-svn: 41502 | ||||
| * | call libc memcpy/memset if array size is bigger then threshold. | Rafael Espindola | 2007-08-27 | 1 | -4/+4 |
| | | | | | | | | Coping 100MB array (after a warmup) shows that glibc 2.6.1 implementation on x86-64 (core 2) is 30% faster (from 0.270917s to 0.188079s) llvm-svn: 41479 | ||||
| * | rename isOperandValidForConstraint to LowerAsmOperandForConstraint, | Chris Lattner | 2007-08-25 | 4 | -36/+66 |
| | | | | | | | changing the interface to allow for future changes. llvm-svn: 41384 | ||||
| * | Disable EH generation until PPC works 100%. | Chris Lattner | 2007-08-24 | 1 | -1/+1 |
| | | | | | llvm-svn: 41360 | ||||
| * | add a note | Chris Lattner | 2007-08-24 | 1 | -0/+14 |
| | | | | | llvm-svn: 41359 | ||||
| * | add some notes on really poor codegen. | Chris Lattner | 2007-08-23 | 1 | -0/+44 |
| | | | | | llvm-svn: 41319 | ||||
| * | new example | Chris Lattner | 2007-08-23 | 1 | -0/+17 |
| | | | | | llvm-svn: 41318 | ||||
| * | Add the PCSymbol for Darwin x86 platforms. | Bill Wendling | 2007-08-22 | 1 | -0/+1 |
| | | | | | llvm-svn: 41284 | ||||
| * | InlineAsm asm support for integer registers added | Bruno Cardoso Lopes | 2007-08-21 | 2 | -0/+81 |
| | | | | | llvm-svn: 41225 | ||||
| * | Instruction Itinerary attribution fixed | Bruno Cardoso Lopes | 2007-08-21 | 1 | -0/+1 |
| | | | | | llvm-svn: 41224 | ||||
| * | Use only 1 knob to enable exceptions on Darwin :). | Anton Korobeynikov | 2007-08-21 | 1 | -0/+8 |
| | | | | | llvm-svn: 41208 | ||||
| * | Partial implementation of calling functions with byval arguments: | Rafael Espindola | 2007-08-20 | 1 | -1/+21 |
| | | | | | | | | *) The needed information is propagated to the DAG *) The X86-64 backend detects it and aborts llvm-svn: 41179 | ||||
| * | add a note | Chris Lattner | 2007-08-20 | 1 | -0/+18 |
| | | | | | llvm-svn: 41178 | ||||
| * | MipsHi now has ouput flag | Bruno Cardoso Lopes | 2007-08-18 | 1 | -91/+185 |
| | | | | | | | | | | | | | MipsAdd SDNode created to add support to an Add opcode which supports input flag Added an instruction itinerary to all instruction classes Added branches with zero cond codes Now call clobbers all non-callee saved registers Call w/ register support added Added DelaySlot to branch and load instructions Added patterns to handle all setcc, brcond/setcc and MipsAdd instructions llvm-svn: 41161 | ||||
| * | Fixed stack frame addressing bug | Bruno Cardoso Lopes | 2007-08-18 | 1 | -8/+8 |
| | | | | | llvm-svn: 41160 | ||||
| * | support for Schedule included on Mips.td | Bruno Cardoso Lopes | 2007-08-18 | 2 | -28/+15 |
| | | | | | llvm-svn: 41159 | ||||
| * | Removed LowerRETURADDR, fixed small bug into LowerRET, LowerGlobalAddress | Bruno Cardoso Lopes | 2007-08-18 | 2 | -17/+11 |
| | | | | | | | fixed to generate instructions (add, lui) glued! llvm-svn: 41158 | ||||
| * | Couple of small changes. Delay Slot handle header declared. | Bruno Cardoso Lopes | 2007-08-18 | 2 | -2/+3 |
| | | | | | | | Newline added after macros at function init on generated asm! llvm-svn: 41157 | ||||
| * | Added InstrItinClass support for instruction formats | Bruno Cardoso Lopes | 2007-08-18 | 1 | -9/+10 |
| | | | | | llvm-svn: 41156 | ||||
| * | Branch Analysis and InsertNoop inserted into header files | Bruno Cardoso Lopes | 2007-08-18 | 1 | -2/+36 |
| | | | | | llvm-svn: 41155 | ||||
| * | createMipsDelaySlotFillerPass added to mips codegen runtime | Bruno Cardoso Lopes | 2007-08-18 | 1 | -2/+2 |
| | | | | | llvm-svn: 41154 | ||||
| * | Added Branch Analysis support | Bruno Cardoso Lopes | 2007-08-18 | 1 | -5/+234 |
| | | | | | | | Added InsertNoop support llvm-svn: 41153 | ||||
| * | LowerRETURNADDR removed since it was wrong and does not have utility yet! | Bruno Cardoso Lopes | 2007-08-18 | 1 | -2/+3 |
| | | | | | | | MipsAdd opcode added llvm-svn: 41152 | ||||
| * | InstrItineraryData support on added. | Bruno Cardoso Lopes | 2007-08-18 | 1 | -2/+8 |
| | | | | | | | Added Mips3 ISA feature (needed when supporting R4000 machines) llvm-svn: 41151 | ||||
| * | A Pass to insert Nops on intructions with DelaySlot | Bruno Cardoso Lopes | 2007-08-18 | 1 | -0/+77 |
| | | | | | llvm-svn: 41150 | ||||
| * | Mips generic fallback instruction schedule support! | Bruno Cardoso Lopes | 2007-08-18 | 1 | -0/+63 |
| | | | | | llvm-svn: 41149 | ||||
| * | Move ReturnAddrIndex variable to X86MachineFunctionInfo structure. This fixed | Anton Korobeynikov | 2007-08-15 | 3 | -14/+24 |
| | | | | | | | hard to catch bugs with retaddr lowering llvm-svn: 41104 | ||||
| * | add a note. | Chris Lattner | 2007-08-15 | 1 | -0/+3 |
| | | | | | llvm-svn: 41103 | ||||
| * | Fix a typo pointd out by Maarten ter Huurne. | Evan Cheng | 2007-08-13 | 1 | -1/+1 |
| | | | | | llvm-svn: 41059 | ||||
| * | When x86 addresses matching exceeds its recursion limit, check to | Dan Gohman | 2007-08-13 | 1 | -6/+12 |
| | | | | | | | | see if the base register is already occupied before assuming it can be used. This fixes bogus code generation in the accompanying testcase. llvm-svn: 41049 | ||||
| * | Fix PR1607 | Chris Lattner | 2007-08-13 | 1 | -3/+3 |
| | | | | | llvm-svn: 41048 | ||||
| * | expand a note | Chris Lattner | 2007-08-11 | 1 | -3/+22 |
| | | | | | llvm-svn: 41021 | ||||
| * | With evan's explicit flag representation, hopefully we will finally be | Chris Lattner | 2007-08-11 | 1 | -1/+1 |
| | | | | | | | | | | able to 3-addressify away stuff like this: movl %ecx, %eax decl %eax llvm-svn: 41020 | ||||
| * | 64-bit SSSE3 ops that use MMX registers don't require 16-byte alignment. | Bill Wendling | 2007-08-11 | 1 | -6/+21 |
| | | | | | | | Make a 'memop' pattern just for them. llvm-svn: 41017 | ||||
| * | Use subregs to improve any_extend code generation when feasible. | Christopher Lamb | 2007-08-10 | 2 | -16/+33 |
| | | | | | llvm-svn: 41013 | ||||
| * | Increase efficiency of sign_extend_inreg by using subregisters for ↵ | Christopher Lamb | 2007-08-10 | 3 | -49/+89 |
| | | | | | | | truncation. As the README suggests sign_extend_subreg is selected to (sext(trunc)). llvm-svn: 41010 | ||||
| * | Edit README in light of previous LEA16 commit. | Christopher Lamb | 2007-08-10 | 1 | -3/+1 |
| | | | | | llvm-svn: 41009 | ||||
| * | Add 2-addr to 3-addr promotion code that allows 32-bit LEA to be used via ↵ | Christopher Lamb | 2007-08-10 | 1 | -10/+47 |
| | | | | | | | subregisters when 16-bit LEA is disabled. llvm-svn: 41007 | ||||
| * | propagate struct size and alignment of byval arguments to the DAG | Rafael Espindola | 2007-08-10 | 1 | -1/+6 |
| | | | | | llvm-svn: 40986 | ||||
| * | For kicks, I though it would be fun to use the correct opcode. | Bill Wendling | 2007-08-10 | 1 | -31/+32 |
| | | | | | llvm-svn: 40985 | ||||
| * | Adding SSSE3 intrinsics. | Bill Wendling | 2007-08-10 | 1 | -17/+284 |
| | | | | | llvm-svn: 40982 | ||||
| * | Temporarily backing out this change until we know why some dejagnu tests are ↵ | Evan Cheng | 2007-08-09 | 1 | -3/+2 |
| | | | | | | | failing. llvm-svn: 40973 | ||||
| * | divb / mulb outputs to ah. Under x86-64 it's not legal to read ah if the ↵ | Evan Cheng | 2007-08-09 | 2 | -4/+46 |
| | | | | | | | instruction requires a rex prefix (i.e. outputs to r8b, etc.). So issue shift right by 8 on AX and then truncate it to 8 bits instead. llvm-svn: 40972 | ||||
| * | GR16_ sub-register class should be GR8_, not GR8. That is, it should only be ↵ | Evan Cheng | 2007-08-09 | 1 | -3/+4 |
| | | | | | | | 8-bit registers in 32-bit mode. Ditto for GR32_. llvm-svn: 40970 | ||||

