aboutsummaryrefslogtreecommitdiff
path: root/lib/Target/ARM/ARM.td
AgeCommit message (Expand)Author
2011-07-07Rename attribute 'thumb' to a more descriptive 'thumb-mode'.Evan Cheng
2011-07-07Sink feature IsThumb into MC layer.Evan Cheng
2011-07-07Change some ARM subtarget features to be single bit yes/no in order to sink t...Evan Cheng
2011-07-07Factor ARM triple parsing out of ARMSubtarget. Another step towards making AR...Evan Cheng
2011-07-01ARMv7M vs. ARMv7E-M support.Jim Grosbach
2011-04-19This patch combines several changes from Evan Cheng for rdar://8659675.Bob Wilson
2011-04-19Add -mcpu=cortex-a9-mp. It's cortex-a9 with MP extension. rdar://8648637.Bob Wilson
2011-04-19Avoid some 's' 16-bit instruction which partially update CPSRBob Wilson
2011-03-31Distribute (A + B) * C to (A * C) + (B * C) to make use of NEON multiplierEvan Cheng
2010-12-15Add Neon VCVT instructions for f32 <-> f16 conversions.Bob Wilson
2010-12-05Code clean up.Evan Cheng
2010-12-05Making use of VFP / NEON floating point multiply-accumulate / subtraction isEvan Cheng
2010-11-12Add some missing isel predicates on def : pat patterns to avoid generating VF...Evan Cheng
2010-11-03Fix preload instruction isel. Only v7 supports pli, and only v7 with mp exten...Evan Cheng
2010-10-12PR8359: The ARM backend may end up allocating registers D16 to D31 whenBob Wilson
2010-09-30Nuke it from orbit. It's the only way to be sure.Jim Grosbach
2010-09-10Teach if-converter to be more careful with predicating instructions that wouldEvan Cheng
2010-08-1780 column cleanup.Jim Grosbach
2010-08-17fix emacs language spec's, patch by Edmund Grimley-Evans!Chris Lattner
2010-08-11cortex m4 has floating point support, but only single precision.Jim Grosbach
2010-08-11Report error if codegen tries to instantiate a ARM target when the cpu does s...Evan Cheng
2010-08-11ArchV7M implies HW division instructions.Evan Cheng
2010-08-11ArchV6T2, V7A, and V7M implies Thumb2; Archv7A implies NEON.Evan Cheng
2010-08-11Add ARM Archv6M and let it implies FeatureDB (having dmb, etc.)Evan Cheng
2010-08-11Add Cortex-M0 support. It's a ARMv6m device (no ARM mode) with some 32-bitEvan Cheng
2010-08-11- Add subtarget feature -mattr=+db which determine whether an ARM cpu has theEvan Cheng
2010-08-09Change -prefer-32bit-thumb to attribute -mattr=+32bit instead to disable more...Evan Cheng
2010-07-13Add an ARM "feature". Cortex-a8 fp comparison is very slow (> 20 cycles).Evan Cheng
2010-05-05Cleanup of ARMv7M support. Move hardware divide and Thumb2 extract/packJim Grosbach
2010-05-05Add initial support for ARMv7M subtarget and cortex-m3 cpu. Patch byJim Grosbach
2010-04-07Some bits of A9 scheduling: VFPAnton Korobeynikov
2010-04-05Replace TSFlagsFields and TSFlagsShifts with a simpler TSFlags field.Jakob Stoklund Olesen
2010-04-01vml[as] are slow on 1136jf-s also.Jim Grosbach
2010-03-25switch the flag for using NEON for SP floating point to a subtarget 'feature'.Jim Grosbach
2010-03-25need to fix 'make check' tests first. revert for a moment.Jim Grosbach
2010-03-25switch the flag for using NEON for SP floating point to a subtarget 'feature'Jim Grosbach
2010-03-25switch the use-vml[as] instructions flag to a subtarget 'feature'Jim Grosbach
2010-03-14Add substarget feature for FP16Anton Korobeynikov
2009-11-18Add ARMv6 itineraries.David Goodwin
2009-11-02Use NEON reg-reg moves, where profitable. This reduces "domain-cross" stalls,...Anton Korobeynikov
2009-10-01Remove neonfp attribute and instead set default based on CPU string. Add -arm...David Goodwin
2009-10-01Restore the -post-RA-scheduler flag as an override for the target specificati...David Goodwin
2009-09-30Remove -post-RA-schedule flag and add a TargetSubtarget method to enable post...David Goodwin
2009-09-23Checkpoint NEON scheduling itineraries.David Goodwin
2009-08-11Allow a zero cycle stage to reserve/require a FU without advancing the cycle ...David Goodwin
2009-08-07Make NEON single-precision FP support the default for cortex-a8 (again).David Goodwin
2009-08-05Disable NEON single-precision FP support for Cortex-A8, for now...David Goodwin
2009-08-05By default, for cortex-a8 use NEON for single-precision FP. David Goodwin
2009-08-04Initial support for single-precision FP using NEON. Added "neonfp" attribute ...David Goodwin
2009-07-21Add fake v7 itineraries for now.Evan Cheng