Remove intrinsic specific instructions for (V)SQRTPS/PD. Instead lower to target...
[oota-llvm.git] / lib / Target / X86 /
2012-12-29 Craig TopperRemove intrinsic specific instructions for (V)SQRTPS...
2012-12-29 Craig TopperMerge similar functionality using a nested switch.
2012-12-29 Craig TopperRemove intrinsic specific instructions for SSE/SSE2...
2012-12-29 Jakub StaszakSimplify code, no functionality change.
2012-12-28 Nadav RotemCostModel: initial checkin for code that estimates...
2012-12-28 Nadav Rotemwrap 80-col lines.
2012-12-28 Nadav RotemAVX: Move the ZEXT/ANYEXT DAGCo optimizations to the...
2012-12-27 Nadav RotemReverse the 'if' condition and reduce the indentation.
2012-12-27 Craig TopperMerge basic_sse12_fp_binop_p_int and basic_sse12_fp_bin...
2012-12-27 Nadav RotemAVX/AVX2: Move the SEXT lowering code from a target...
2012-12-27 Craig TopperMerge basic_sse12_fp_binop_p and basic_sse12_fp_binop_p...
2012-12-27 Nadav RotemOn AVX/AVX2 the type v8i1 is legalized to v8i16, which...
2012-12-27 Nadav RotemAVX/AVX2: Move the code that lowers vector-trunc from...
2012-12-27 Craig TopperAdd hasSideEffects=0 to some forms of ROUND, RCP, and...
2012-12-27 Craig TopperMove single letter 'P' prefix out of multiclass now...
2012-12-27 Craig TopperAdd hasSideEffects=0 to some shift and rotate instructi...
2012-12-27 Craig TopperMark the divide instructions as hasSideEffects=0.
2012-12-27 Craig TopperAdd hasSideEffects=0 to CMP*rr_REV.
2012-12-27 Craig TopperAdd mayLoad, mayStore, and hasSideEffects tags to BT...
2012-12-26 Craig TopperFix operands and encoding form for ARPL instruction...
2012-12-26 Craig TopperAdd hasSideEffects=0 to some atomic instructions.
2012-12-26 Craig TopperMark the AL/AX/EAX forms of the basic arithmetic operat...
2012-12-26 Craig TopperMark all the _REV instructions as not having side effec...
2012-12-26 Craig TopperRemove a special conditional setting of neverHasSideEff...
2012-12-26 Craig TopperMerge still more SSE/AVX instruction definitions.
2012-12-26 Craig TopperMerge more SSE/AVX instruction definitions.
2012-12-26 Craig TopperFix 80 column violation.
2012-12-26 Craig TopperFix class name in comment.
2012-12-26 Craig TopperMerge SSE/AVX PCMPEQ/PCMPGT instruction definitions.
2012-12-26 Craig TopperRemove 'v' from mnemonic to fix asm matching failures.
2012-12-26 Craig TopperUse an additional multiclass to merge the 128/256-bit...
2012-12-26 Nadav RotemReformat the docs.
2012-12-26 Craig TopperUse an additional multiclass to merge the 128/256-bit...
2012-12-26 Craig TopperMerge an AVX/SSE 256-bit and 128-bit multiclass.
2012-12-26 Craig TopperMark VANDNPD/VANDNPDS as not commutable.
2012-12-26 Craig TopperRemove alignment from a bunch more VEX encoded operatio...
2012-12-26 Craig TopperRemove alignment from folding table for VMOVUPD as...
2012-12-26 Craig TopperRemove alignment requirements from (V)EXTRACTPS. This...
2012-12-26 Craig TopperRemove alignment requirement from VCVTSS2SD in folding...
2012-12-25 Benjamin KramerX86: Shave off one shuffle from the pcmpeqq sequence...
2012-12-25 Benjamin KramerX86: Custom lower <2 x i64> eq and ne when SSE41 is...
2012-12-25 Nadav RotemVCVTSS2SD requires a strict alignment. Thanks Elena.
2012-12-24 Nick LewyckyQuiet gcc's -Wparenthesis warning. No functionality...
2012-12-24 Nadav RotemSome x86 instructions can load/store one of the operand...
2012-12-23 Nadav Rotemwhitespace
2012-12-23 Nadav RotemRename a function.
2012-12-23 Nadav RotemLoop Vectorizer: Update the cost model of scatter/gathe...
2012-12-22 Benjamin KramerX86: Turn mul of <4 x i32> into pmuludq when no SSE4...
2012-12-22 Benjamin KramerX86: Emit vector sext as shuffle + sra if vpmovsx is...
2012-12-21 Nadav RotemIn some cases, due to scheduling constraints we copy...
2012-12-21 Benjamin KramerX86: Match pmin/pmax as a target specific dag combine...
2012-12-21 Benjamin KramerX86: Match the SSE/AVX min/max vector ops using a custo...
2012-12-21 Nadav RotemAdd a missing "virtual" keyword.
2012-12-21 Nadav RotemImprove the X86 cost model for loads and stores.
2012-12-20 Jakob Stoklund OlesenAdd an MF argument to MI::copyImplicitOps().
2012-12-20 Roman DivackyRemove MCTargetAsmLexer and its derived classes now...
2012-12-20 Richard SmithFix use-before-construction of X86TargetLowering.
2012-12-19 Jim GrosbachMC: Add MCInstrDesc::mayAffectControlFlow() method.
2012-12-19 Jakob Stoklund OlesenRemove the explicit MachineInstrBuilder(MI) constructor.
2012-12-19 Roman DivackyRemove edis - the enhanced disassembler. Fixes PR14654.
2012-12-19 Paul RedmondTransform (x&C)>V into (x&C)!=0 where possible
2012-12-19 Patrik HagglundChange TargetLowering::getTypeForExtArgOrReturn to...
2012-12-19 Patrik HagglundChange TargetLowering::RegisterTypeForVT to contain...
2012-12-19 Patrik HagglundChange TargetLowering::findRepresentativeClass to take...
2012-12-19 NAKAMURA TakumiX86ISelLowering.cpp: Fix warnings. [-Wlogical-op-parent...
2012-12-19 Elena DemikhovskyOptimized load + SIGN_EXTEND patterns in the X86 backend.
2012-12-19 Bill WendlingRename the 'Attributes' class to 'Attribute'. It's...
2012-12-18 Jakub StaszakReverse order of checking SSE level when calculating...
2012-12-17 Craig TopperRemove EFLAGS from the BLSI/BLSMSK/BLSR patterns. The...
2012-12-17 Craig TopperSimplify BMI ANDN matching to use patterns instead...
2012-12-17 Craig TopperAdd rest of BMI/BMI2 instructions to the folding tables...
2012-12-17 Craig TopperRemove store forms of DEC/INC from isDefConvertible...
2012-12-15 Benjamin KramerX86: Add a couple of target-specific dag combines that...
2012-12-15 Chandler CarruthMake '-mtune=x86_64' assume fast unaligned memory accesses.
2012-12-14 Nadav RotemTypeLegalizer: Do not generate target specific nodes...
2012-12-13 Eli BenderskyFix a bogus comment
2012-12-12 Evan ChengSorry about the churn. One more change to getOptimalMem...
2012-12-12 Evan Cheng- Rename isLegalMemOpType to isSafeMemOpType. "Legal...
2012-12-12 Evan ChengAvoid using lossy load / stores for memcpy / memset...
2012-12-11 Patrik HagglundRevert EVT->MVT changes, r169836-169851, due to buildbo...
2012-12-11 Patrik HagglundChange TargetLowering::getTypeForExtArgOrReturn to...
2012-12-11 Patrik HagglundChange TargetLowering::RegisterTypeForVT to contain...
2012-12-11 Patrik HagglundChange TargetLowering::findRepresentativeClass to take...
2012-12-11 Chad RosierFall back to the selection dag isel to select tail...
2012-12-10 Evan ChengSome enhancements for memcpy / memset inline expansion.
2012-12-10 Chandler CarruthRevert "Make '-mtune=x86_64' assume fast unaligned...
2012-12-10 Chandler CarruthMake '-mtune=x86_64' assume fast unaligned memory accesses.
2012-12-10 Chandler CarruthFix a typo in my previous commit -- bloomfield is 0x1A...
2012-12-10 Chandler CarruthAddress a FIXME and update the fast unaligned memory...
2012-12-09 Shuxin Yang- Re-enable population count loop idiom recognization
2012-12-08 Chandler CarruthRevert the patches adding a popcount loop idiom recogni...
2012-12-07 Bill Wendlings/AttrListPtr/AttributeSet/g to better label what this...
2012-12-07 Nadav RotemWhen we use the BLEND instruction that uses the MSB...
2012-12-07 Nadav RotemX86: Prefer using VPSHUFD over VPERMIL because it has...
2012-12-06 Evan ChengReplace r169459 with something safer. Rather than havin...
2012-12-06 Jakub StaszakRemove unneeded function, since PR8156 was fixed over...
2012-12-06 Jakub StaszakSimplify code.
2012-12-06 Craig TopperRemove intrinsic specific instructions for (V)MOVQUmr...
2012-12-06 Craig TopperMark MOVDQ(A/U)rm as ReMaterializable. Mark all MOVDQ...
2012-12-06 Evan ChengLet targets provide hooks that compute known zero and...
next