summaryrefslogtreecommitdiffstats
path: root/lib/Target/X86/X86ISelLowering.cpp
Commit message (Expand)AuthorAgeFilesLines
* Added skeleton for inline asm multiple alternative constraint support.John Thompson2010-09-131-0/+28
* Minor change. Fix comments and remove unused and redundant codeBruno Cardoso Lopes2010-09-081-10/+3
* x86 vector shuffle lowering now relies only on target specificBruno Cardoso Lopes2010-09-081-6/+124
* Factor out some x86 vector shuffle rewriting and add comments about the direc...Bruno Cardoso Lopes2010-09-071-29/+59
* Move code around to prepare for moving some of the logic together to another ...Bruno Cardoso Lopes2010-09-071-3/+3
* Add an MVT::x86mmx type. It will take the place of all current MMX vector types.Bill Wendling2010-09-071-0/+3
* decouple MMX check from regular splat checks. Some refactoring is coming, and...Bruno Cardoso Lopes2010-09-071-4/+8
* Remove now useless check, because the code can be matched below, no need to l...Bruno Cardoso Lopes2010-09-071-1/+1
* Minor change. Since the checks are equivalent, use isMMXBruno Cardoso Lopes2010-09-071-1/+1
* Remove the last bit of isShuffleMaskLegal checks and improve the comment rega...Bruno Cardoso Lopes2010-09-041-5/+3
* make explicit that we not handle several mmx shufflesBruno Cardoso Lopes2010-09-041-7/+7
* Emit target specific nodes to handle palignr. Do not touch it for MMX version...Bruno Cardoso Lopes2010-09-041-2/+10
* Emit target specific nodes to handle splats starting at zero indiciesBruno Cardoso Lopes2010-09-041-1/+8
* Emit target specific nodes for isPSHUFHWMask and isPSHUFLWMaskBruno Cardoso Lopes2010-09-041-2/+10
* Emit target specific nodes for isSHUFPMaskBruno Cardoso Lopes2010-09-041-1/+10
* Previous isMOVLMask matching already emits targets nodes, remove checkBruno Cardoso Lopes2010-09-041-1/+0
* One more check from the original isShuffleMaskLegal goes awayBruno Cardoso Lopes2010-09-041-1/+0
* Remove a duplicated but useless check that i've inserted in the previous commit.Bruno Cardoso Lopes2010-09-041-9/+0
* Refactor some code and remove the extra checks for unpckl_undef and unpckh_undefBruno Cardoso Lopes2010-09-041-28/+25
* Remove check for unpckh maskBruno Cardoso Lopes2010-09-031-1/+0
* Remove check for unpckl maskBruno Cardoso Lopes2010-09-031-1/+0
* Inline isShuffleMaskLegal into LowerVECTOR_SHUFFLE, so we can startBruno Cardoso Lopes2010-09-031-4/+23
* Reapply considered harmfull part of rr112934 and r112942.Bruno Cardoso Lopes2010-09-031-12/+109
* Reintroduce a simple function refactoring done in r112934, also without any f...Bruno Cardoso Lopes2010-09-031-10/+12
* Reapply piecies of r112942 and r112934 which don't doBruno Cardoso Lopes2010-09-031-0/+36
* Reapply Fix commentBruno Cardoso Lopes2010-09-031-3/+3
* Revert r112934, "- Use specific nodes to match unpckl masks.", which introducedDaniel Dunbar2010-09-031-126/+16
* Revert r112938 "Fix comment", which depends on r112934, which introduced someDaniel Dunbar2010-09-031-3/+3
* Revert r112942, "Use punpckh and unpckh family of nodes instead of using unpckhDaniel Dunbar2010-09-031-23/+3
* Use punpckh and unpckh family of nodes instead of using unpckh mask pattern f...Bruno Cardoso Lopes2010-09-031-3/+23
* Fix commentBruno Cardoso Lopes2010-09-031-3/+3
* - Use specific nodes to match unpckl masks.Bruno Cardoso Lopes2010-09-031-16/+126
* Revert win64 changes. They seem to be incompleteAnton Korobeynikov2010-09-021-11/+2
* Properly allocate win64 shadow reg area.Anton Korobeynikov2010-09-021-2/+11
* Replace unpckl_undef and unpckh_undef matching with target specific opcodesBruno Cardoso Lopes2010-09-021-6/+39
* Move condition out to prepare for more matchingBruno Cardoso Lopes2010-09-021-4/+5
* Remove checking for isUNPCKL_v_undef_Mask, the specific node is already emitt...Bruno Cardoso Lopes2010-09-021-4/+2
* become more strict about when it's safe to use X86ISD::MOVLPSBruno Cardoso Lopes2010-09-021-5/+5
* Revert r112689, avoid those kind of checks cause they mess up with mmxBruno Cardoso Lopes2010-09-011-4/+4
* Use movlps, movlpd, movss and movsd specific nodes instead of pattern matchin...Bruno Cardoso Lopes2010-09-011-3/+72
* minor change, simplify some logicBruno Cardoso Lopes2010-09-011-4/+4
* Move some functions around so they can be used for some other to come functionBruno Cardoso Lopes2010-09-011-8/+8
* Use x86 specific MOVSLDUP node, add more patterns to match it and remove usel...Bruno Cardoso Lopes2010-08-311-3/+6
* Use x86 specific MOVSHDUP node and add more patterns to match itBruno Cardoso Lopes2010-08-311-2/+17
* Use MOVHLPS node instead of matching using movhlps and movhlps_undef pattern ...Bruno Cardoso Lopes2010-08-311-0/+21
* Use MOVLHPS and MOVHLPS x86 nodes whenever possible. Also remove some useless...Bruno Cardoso Lopes2010-08-311-11/+28
* Use X86ISD::MOVSS and MOVSD to represent the movl mask pattern, also fix the ...Bruno Cardoso Lopes2010-08-311-7/+14
* fix the buildvector->insertp[sd] logic to not always create a redundantChris Lattner2010-08-281-6/+12
* fix the BuildVector -> unpcklps logic to not do pointless shuffles Chris Lattner2010-08-281-1/+11
* improve comments in the unpcklps generating logic, introduceChris Lattner2010-08-281-11/+18