llvm-6502/lib/Target/Sparc
Evan Cheng 276365dd4b Fix the ridiculous SubtargetFeatures API where it implicitly expects CPU name to
be the first encoded as the first feature. It then uses the CPU name to look up
features / scheduling itineray even though clients know full well the CPU name
being used to query these properties.

The fix is to just have the clients explictly pass the CPU name!


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134127 91177308-0d34-0410-b5e6-96231b3b80d8
2011-06-30 01:53:36 +00:00
..
TargetInfo
CMakeLists.txt Merge XXXGenRegisterNames.inc into XXXGenRegisterInfo.inc 2011-06-28 20:07:07 +00:00
DelaySlotFiller.cpp - Rename TargetInstrDesc, TargetOperandInfo to MCInstrDesc and MCOperandInfo and 2011-06-28 19:10:37 +00:00
FPMover.cpp
Makefile Merge XXXGenRegisterNames.inc into XXXGenRegisterInfo.inc 2011-06-28 20:07:07 +00:00
README.txt
Sparc.h Merge XXXGenRegisterNames.inc into XXXGenRegisterInfo.inc 2011-06-28 20:07:07 +00:00
Sparc.td
SparcAsmPrinter.cpp
SparcCallingConv.td Pass sret arguments through the stack instead of through registers in Sparc backend. It makes the code generated more compliant with the sparc32 ABI. 2011-01-22 13:05:16 +00:00
SparcFrameLowering.cpp Teach frame lowering to ignore debug values after the terminators. 2011-01-13 21:28:52 +00:00
SparcFrameLowering.h
SparcInstrFormats.td
SparcInstrInfo.cpp Move CallFrameSetupOpcode and CallFrameDestroyOpcode to TargetInstrInfo. 2011-06-28 21:14:33 +00:00
SparcInstrInfo.h Implement AnalyzeBranch in Sparc Backend. 2011-01-16 03:15:11 +00:00
SparcInstrInfo.td Generate correct Sparc32 ABI compliant code for functions that return a struct. 2011-02-21 03:42:44 +00:00
SparcISelDAGToDAG.cpp
SparcISelLowering.cpp Remove getRegClassForInlineAsmConstraint from sparc. 2011-06-29 18:53:10 +00:00
SparcISelLowering.h Remove getRegClassForInlineAsmConstraint from sparc. 2011-06-29 18:53:10 +00:00
SparcMachineFunctionInfo.h Pass sret arguments through the stack instead of through registers in Sparc backend. It makes the code generated more compliant with the sparc32 ABI. 2011-01-22 13:05:16 +00:00
SparcMCAsmInfo.cpp
SparcMCAsmInfo.h
SparcRegisterInfo.cpp Move CallFrameSetupOpcode and CallFrameDestroyOpcode to TargetInstrInfo. 2011-06-28 21:14:33 +00:00
SparcRegisterInfo.h Merge XXXGenRegisterDesc.inc XXXGenRegisterNames.inc XXXGenRegisterInfo.h.inc 2011-06-27 18:32:37 +00:00
SparcRegisterInfo.td Use set operations instead of plain lists to enumerate register classes. 2011-06-15 23:28:14 +00:00
SparcSelectionDAGInfo.cpp
SparcSelectionDAGInfo.h
SparcSubtarget.cpp Fix the ridiculous SubtargetFeatures API where it implicitly expects CPU name to 2011-06-30 01:53:36 +00:00
SparcSubtarget.h Fix the ridiculous SubtargetFeatures API where it implicitly expects CPU name to 2011-06-30 01:53:36 +00:00
SparcTargetMachine.cpp Fix the ridiculous SubtargetFeatures API where it implicitly expects CPU name to 2011-06-30 01:53:36 +00:00
SparcTargetMachine.h Fix the ridiculous SubtargetFeatures API where it implicitly expects CPU name to 2011-06-30 01:53:36 +00:00

To-do
-----

* Keep the address of the constant pool in a register instead of forming its
  address all of the time.
* We can fold small constant offsets into the %hi/%lo references to constant
  pool addresses as well.
* When in V9 mode, register allocate %icc[0-3].
* Add support for isel'ing UMUL_LOHI instead of marking it as Expand.
* Emit the 'Branch on Integer Register with Prediction' instructions.  It's
  not clear how to write a pattern for this though:

float %t1(int %a, int* %p) {
        %C = seteq int %a, 0
        br bool %C, label %T, label %F
T:
        store int 123, int* %p
        br label %F
F:
        ret float undef
}

codegens to this:

t1:
        save -96, %o6, %o6
1)      subcc %i0, 0, %l0
1)      bne .LBBt1_2    ! F
        nop
.LBBt1_1:       ! T
        or %g0, 123, %l0
        st %l0, [%i1]
.LBBt1_2:       ! F
        restore %g0, %g0, %g0
        retl
        nop

1) should be replaced with a brz in V9 mode.

* Same as above, but emit conditional move on register zero (p192) in V9 
  mode.  Testcase:

int %t1(int %a, int %b) {
        %C = seteq int %a, 0
        %D = select bool %C, int %a, int %b
        ret int %D
}

* Emit MULX/[SU]DIVX instructions in V9 mode instead of fiddling 
  with the Y register, if they are faster.

* Codegen bswap(load)/store(bswap) -> load/store ASI

* Implement frame pointer elimination, e.g. eliminate save/restore for 
  leaf fns.
* Fill delay slots

* Implement JIT support