The BLX instruction is encoded differently than the BL, because why not? In
[oota-llvm.git] / lib / Target / SystemZ / SystemZInstrFP.td
2010-06-21 Eric ChristopherRemove isTwoAddress from SystemZ.
2010-03-28 Chris LattnerImprove systemz to model cmp and ucmp nodes as returning
2010-02-27 Dan GohmanThe mayHaveSideEffects flag is no longer used.
2009-10-29 Dan GohmanRename usesCustomDAGSchedInserter to usesCustomInserter...
2009-10-28 Dan GohmanUpdate SystemZ to use PSW following the way x86 uses...
2009-08-21 Anton KorobeynikovAdd fcopysign instructions
2009-07-16 Anton KorobeynikovAll FP instructions have 12 bit memory displacement...
2009-07-16 Anton KorobeynikovFix instruction mnemonics for some fp_to_sint operations
2009-07-16 Anton KorobeynikovHandle bitconverts
2009-07-16 Anton KorobeynikovEmit proper rounding mode for fp_to_sint
2009-07-16 Anton KorobeynikovMake FP zero to be legal FP immediate via LOAD ZERO
2009-07-16 Anton KorobeynikovLoads are not two-address in any way
2009-07-16 Anton KorobeynikovAdd LOAD NEGATIVE instruction
2009-07-16 Anton KorobeynikovLOAD COMPLEMENT instruction is not really two-addr
2009-07-16 Anton KorobeynikovAdd multiple add/sub instructions
2009-07-16 Anton KorobeynikovProper FP extloads
2009-07-16 Anton KorobeynikovAdd proper PWS impdef's
2009-07-16 Anton KorobeynikovPropagate FP select_cc to dag inserters
2009-07-16 Anton KorobeynikovImplement fp_to_sint
2009-07-16 Anton KorobeynikovAdd fabs
2009-07-16 Anton KorobeynikovAdd fneg
2009-07-16 Anton KorobeynikovMore sint_to_fp stuff
2009-07-16 Anton KorobeynikovAdd bunch of FP instructions