6502/llvm-6502
1
0
Fork 0
mirror of https://github.com/c64scene-ar/llvm-6502.git synced 2024-12-23 15:29:51 +00:00
Code Issues Projects Releases Wiki Activity
84c5eed15b
llvm-6502/test/CodeGen/ARM
History
Bob Wilson 84c5eed15b This patch combines several changes from Evan Cheng for rdar://8659675. 
Making use of VFP / NEON floating point multiply-accumulate / subtraction is
difficult on current ARM implementations for a few reasons.
1. Even though a single vmla has latency that is one cycle shorter than a pair
   of vmul + vadd, a RAW hazard during the first (4? on Cortex-a8) can cause
   additional pipeline stall. So it's frequently better to single codegen
   vmul + vadd.
2. A vmla folowed by a vmul, vmadd, or vsub causes the second fp instruction to
   stall for 4 cycles. We need to schedule them apart.
3. A vmla followed vmla is a special case. Obvious issuing back to back RAW
   vmla + vmla is very bad. But this isn't ideal either:
     vmul
     vadd
     vmla
   Instead, we want to expand the second vmla:
     vmla
     vmul
     vadd
   Even with the 4 cycle vmul stall, the second sequence is still 2 cycles
   faster.

Up to now, isel simply avoid codegen'ing fp vmla / vmls. This works well enough
but it isn't the optimial solution. This patch attempts to make it possible to
use vmla / vmls in cases where it is profitable.

A. Add missing isel predicates which cause vmla to be codegen'ed.
B. Make sure the fmul in (fadd (fmul)) has a single use. We don't want to
   compute a fmul and a fmla.
C. Add additional isel checks for vmla, avoid cases where vmla is feeding into
   fp instructions (except for the #3 exceptional case).
D. Add ARM hazard recognizer to model the vmla / vmls hazards.
E. Add a special pre-regalloc case to expand vmla / vmls when it's likely the
   vmla / vmls will trigger one of the special hazards.

Enable these fp vmlx codegen changes for Cortex-A9.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@129775 91177308-0d34-0410-b5e6-96231b3b80d8
2011-04-19 18:11:57 +00:00
..
2006-11-10-CycleInDAG.ll
2007-01-19-InfiniteLoop.ll Fix test. 2010-11-03 18:21:33 +00:00
2007-03-07-CombinerCrash.ll
2007-03-13-InstrSched.ll
2007-03-21-JoinIntervalsCrash.ll
2007-03-26-RegScavengerAssert.ll Correct bogus module triple specifications. 2010-08-30 10:48:29 +00:00
2007-03-27-RegScavengerAssert.ll
2007-03-30-RegScavengerAssert.ll
2007-04-02-RegScavengerAssert.ll
2007-04-03-PEIBug.ll
2007-04-03-UndefinedSymbol.ll
2007-04-30-CombinerCrash.ll
2007-05-03-BadPostIndexedLd.ll
2007-05-07-jumptoentry.ll
2007-05-07-tailmerge-1.ll Remove the -enable-sjlj-eh option, which doesn't do anything. 2010-05-02 15:36:26 +00:00
2007-05-09-tailmerge-2.ll Remove the -enable-sjlj-eh option, which doesn't do anything. 2010-05-02 15:36:26 +00:00
2007-05-14-InlineAsmCstCrash.ll
2007-05-14-RegScavengerAssert.ll
2007-05-22-tailmerge-3.ll Remove the -enable-sjlj-eh option, which doesn't do anything. 2010-05-02 15:36:26 +00:00
2007-05-23-BadPreIndexedStore.ll
2007-05-31-RegScavengerInfiniteLoop.ll Correct bogus module triple specifications. 2010-08-30 10:48:29 +00:00
2007-08-15-ReuseBug.ll
2008-02-04-LocalRegAllocBug.ll Remove the local register allocator. 2010-06-15 21:58:33 +00:00
2008-02-29-RegAllocLocal.ll Remove the local register allocator. 2010-06-15 21:58:33 +00:00
2008-03-05-SxtInRegBug.ll
2008-03-07-RegScavengerAssert.ll
2008-04-04-ScavengerAssert.ll
2008-04-10-ScavengerAssert.ll
2008-04-11-PHIofImpDef.ll
2008-05-19-LiveIntervalsBug.ll
2008-05-19-ScavengerAssert.ll
2008-07-17-Fdiv.ll
2008-07-24-CodeGenPrepCrash.ll
2008-08-07-AsmPrintBug.ll
2008-09-14-CoalescerBug.ll
2008-09-17-CoalescerBug.ll
2008-11-18-ScavengerAssert.ll
2009-02-16-SpillerBug.ll
2009-02-22-SoftenFloatVaArg.ll
2009-02-27-SpillerBug.ll
2009-03-07-SpillerBug.ll
2009-03-09-AddrModeBug.ll
2009-04-06-AsmModifier.ll
2009-04-08-AggregateAddr.ll
2009-04-08-FloatUndef.ll
2009-04-08-FREM.ll
2009-04-09-RegScavengerAsm.ll
2009-05-05-DAGCombineBug.ll Correct some bogus target triples. 2010-05-07 17:03:48 +00:00
2009-05-07-RegAllocLocal.ll Remove the local register allocator. 2010-06-15 21:58:33 +00:00
2009-05-11-CodePlacementCrash.ll
2009-05-18-InlineAsmMem.ll
2009-06-02-ISelCrash.ll
2009-06-04-MissingLiveIn.ll
2009-06-12-RegScavengerAssert.ll
2009-06-15-RegScavengerAssert.ll
2009-06-19-RegScavengerAssert.ll
2009-06-22-CoalescerBug.ll Remove arm_apcscc from the test files. It is the default and doing this 2010-06-17 15:18:27 +00:00
2009-06-30-RegScavengerAssert2.ll Remove arm_apcscc from the test files. It is the default and doing this 2010-06-17 15:18:27 +00:00
2009-06-30-RegScavengerAssert3.ll Remove arm_apcscc from the test files. It is the default and doing this 2010-06-17 15:18:27 +00:00
2009-06-30-RegScavengerAssert4.ll Remove arm_apcscc from the test files. It is the default and doing this 2010-06-17 15:18:27 +00:00
2009-06-30-RegScavengerAssert5.ll Remove arm_apcscc from the test files. It is the default and doing this 2010-06-17 15:18:27 +00:00
2009-06-30-RegScavengerAssert.ll Remove arm_apcscc from the test files. It is the default and doing this 2010-06-17 15:18:27 +00:00
2009-07-01-CommuteBug.ll Remove arm_apcscc from the test files. It is the default and doing this 2010-06-17 15:18:27 +00:00
2009-07-09-asm-p-constraint.ll
2009-07-18-RewriterBug.ll Remove arm_apcscc from the test files. It is the default and doing this 2010-06-17 15:18:27 +00:00
2009-07-22-ScavengerAssert.ll Remove arm_apcscc from the test files. It is the default and doing this 2010-06-17 15:18:27 +00:00
2009-07-22-SchedulerAssert.ll Remove arm_apcscc from the test files. It is the default and doing this 2010-06-17 15:18:27 +00:00
2009-07-29-VFP3Registers.ll Remove arm_apcscc from the test files. It is the default and doing this 2010-06-17 15:18:27 +00:00
2009-08-02-RegScavengerAssert-Neon.ll Remove arm_apcscc from the test files. It is the default and doing this 2010-06-17 15:18:27 +00:00
2009-08-04-RegScavengerAssert-2.ll Set the mtriple in some tests so that they use AAPCS. 2010-06-15 20:42:00 +00:00
2009-08-04-RegScavengerAssert.ll Set the mtriple in some tests so that they use AAPCS. 2010-06-15 20:42:00 +00:00
2009-08-15-RegScavenger-EarlyClobber.ll Set the mtriple in some tests so that they use AAPCS. 2010-06-15 20:42:00 +00:00
2009-08-15-RegScavengerAssert.ll Set the mtriple in some tests so that they use AAPCS. 2010-06-15 20:42:00 +00:00
2009-08-21-PostRAKill2.ll Remove arm_apcscc from the test files. It is the default and doing this 2010-06-17 15:18:27 +00:00
2009-08-21-PostRAKill3.ll Remove arm_apcscc from the test files. It is the default and doing this 2010-06-17 15:18:27 +00:00
2009-08-21-PostRAKill.ll Remove arm_apcscc from the test files. It is the default and doing this 2010-06-17 15:18:27 +00:00
2009-08-23-linkerprivate.ll Implement the "linker_private_weak" linkage type. This will be used for 2010-07-01 21:55:59 +00:00
2009-08-26-ScalarToVector.ll Remove arm_apcscc from the test files. It is the default and doing this 2010-06-17 15:18:27 +00:00
2009-08-27-ScalarToVector.ll Remove arm_apcscc from the test files. It is the default and doing this 2010-06-17 15:18:27 +00:00
2009-08-29-ExtractEltf32.ll Remove arm_apcscc from the test files. It is the default and doing this 2010-06-17 15:18:27 +00:00
2009-08-29-TooLongSplat.ll Remove arm_apcscc from the test files. It is the default and doing this 2010-06-17 15:18:27 +00:00
2009-08-31-LSDA-Name.ll Remove arm_apcscc from the test files. It is the default and doing this 2010-06-17 15:18:27 +00:00
2009-08-31-TwoRegShuffle.ll
2009-09-09-AllOnes.ll Remove arm_apcscc from the test files. It is the default and doing this 2010-06-17 15:18:27 +00:00
2009-09-09-fpcmp-ole.ll Use Unified Assembly Syntax for the ARM backend. 2009-11-09 00:11:35 +00:00
2009-09-10-postdec.ll
2009-09-13-InvalidSubreg.ll
2009-09-13-InvalidSuperReg.ll
2009-09-20-LiveIntervalsBug.ll
2009-09-21-LiveVariablesBug.ll
2009-09-22-LiveVariablesBug.ll
2009-09-23-LiveVariablesBug.ll
2009-09-24-spill-align.ll Remove arm_apcscc from the test files. It is the default and doing this 2010-06-17 15:18:27 +00:00
2009-09-27-CoalescerBug.ll
2009-09-28-LdStOptiBug.ll User proper libcall names & condcodes while compiling for ARM EABI. 2010-09-28 21:39:26 +00:00
2009-10-02-NEONSubregsBug.ll Fix PR5367. QPR_8 is the super regclass of DPR_8 and SPR_8. 2009-11-03 05:52:54 +00:00
2009-10-16-Scope.ll Move arch specific tests in arch specific directories. 2011-02-24 19:06:27 +00:00
2009-10-21-InvalidFNeg.ll Fix invalid for vector types fneg(bitconvert(x)) => bitconvert(x ^ sign) 2009-10-20 21:37:45 +00:00
2009-10-27-double-align.ll These tests no longer require linear scan because reserved register coalescing is now universal. 2011-04-05 21:40:41 +00:00
2009-10-30.ll Added sub-formats to the NeonI/NeonXI instructions to further refine the NEONFrm 2010-03-17 17:52:21 +00:00
2009-11-01-NeonMoves.ll Generalize the pre-coalescing of extract_subregs feeding reg_sequences, 2010-06-15 05:56:31 +00:00
2009-11-02-NegativeLane.ll Add NEON VLD1-dup instructions (load 1 element to all lanes). 2010-11-27 06:35:16 +00:00
2009-11-07-SubRegAsmPrinting.ll move fconst[sd] to UAL. <rdar://7414913> 2009-11-23 21:08:25 +00:00
2009-11-13-CoalescerCrash.ll Fix PR5410: LiveVariables lost subreg def: 2009-11-13 20:36:40 +00:00
2009-11-13-ScavengerAssert2.ll Fix PR5412: Fix an inverted check and another missing sub-register check. 2009-11-14 02:09:09 +00:00
2009-11-13-ScavengerAssert.ll Fix PR5411. Bug in UpdateKills. A reg def partially define its super-registers. 2009-11-13 23:16:41 +00:00
2009-11-13-VRRewriterCrash.ll Add radar number. 2009-11-14 02:11:32 +00:00
2009-11-30-LiveVariablesBug.ll Fix PR5614: parts of a physical register def may be killed the rest. 2009-12-01 00:44:45 +00:00
2009-12-02-vtrn-undef.ll Add a testcase for svn 111208. 2010-08-16 23:44:29 +00:00
2010-03-04-eabi-fp-spill.ll pr6478: The frame pointer spill frame index is only defined when there is a 2010-03-04 21:42:36 +00:00
2010-03-04-stm-undef-addr.ll pr6480: Don't try producing ld/st-multiple instructions when the address is 2010-03-04 21:04:38 +00:00
2010-03-18-ldm-rtrn.ll Properly pseudo-ize the ARM LDMIA_RET instruction. This has the nice side- 2011-03-11 22:51:41 +00:00
2010-04-07-DbgValueOtherTargets.ll If dbg_declare() or dbg_value() is not lowered by isel then emit DEBUG message instead of creating DBG_VALUE for undefined value in reg0. 2010-12-06 22:39:26 +00:00
2010-04-09-NeonSelect.ll Remove arm_apcscc from the test files. It is the default and doing this 2010-06-17 15:18:27 +00:00
2010-04-13-v2f64SplitArg.ll Remove arm_apcscc from the test files. It is the default and doing this 2010-06-17 15:18:27 +00:00
2010-04-14-SplitVector.ll Remove arm_apcscc from the test files. It is the default and doing this 2010-06-17 15:18:27 +00:00
2010-04-15-ScavengerDebugValue.ll Remove arm_apcscc from the test files. It is the default and doing this 2010-06-17 15:18:27 +00:00
2010-05-14-IllegalType.ll Allow TargetLowering::getRegClassFor() to be called on illegal types. Also 2010-05-15 02:18:07 +00:00
2010-05-17-FastAllocCrash.ll Avoid allocating the same physreg to multiple virtregs in one instruction. 2010-05-17 17:18:59 +00:00
2010-05-18-LocalAllocCrash.ll Remove arm_apcscc from the test files. It is the default and doing this 2010-06-17 15:18:27 +00:00
2010-05-18-PostIndexBug.ll don't test for codegen of 'store undef' 2011-04-09 02:31:26 +00:00
2010-05-19-Shuffles.ll Handle Neon v2f64 and v2i64 vector shuffles as register copies. 2010-05-20 18:39:53 +00:00
2010-05-20-NEONSpillCrash.ll Add alignment arguments to all the NEON load/store intrinsics. 2010-08-27 17:13:24 +00:00
2010-05-21-BuildVector.ll Add alignment arguments to all the NEON load/store intrinsics. 2010-08-27 17:13:24 +00:00
2010-06-11-vmovdrr-bitcast.ll Add alignment arguments to all the NEON load/store intrinsics. 2010-08-27 17:13:24 +00:00
2010-06-21-LdStMultipleBug.ll Fix PR7421: bug in kill transferring logic. It was ignoring loads / stores which have already been processed. 2010-06-21 21:21:14 +00:00
2010-06-21-nondarwin-tc.ll Temporarily disable tail calls on ARM to work around some linker problems. 2010-08-13 22:43:33 +00:00
2010-06-25-Thumb2ITInvalidIterator.ll Thumb2ITBlockPass: Fix a possible dereference of an invalid iterator. This was 2010-06-25 23:14:54 +00:00
2010-06-29-PartialRedefFastAlloc.ll Change register allocation order for ARM VFP and NEON registers to put the 2010-10-08 06:15:13 +00:00
2010-06-29-SubregImpDefs.ll Fix a register scavenger crash when dealing with undefined subregs. 2010-06-29 18:42:49 +00:00
2010-07-26-GlobalMerge.ll Currently EH lowering code expects typeinfo to be global only. 2010-07-26 18:45:39 +00:00
2010-08-04-EHCrash.ll Testcase for r110248. 2010-08-04 21:56:30 +00:00
2010-08-04-StackVariable.ll Move arch specific tests in arch specific directories. 2011-02-24 19:06:27 +00:00
2010-09-21-OptCmpBug.ll OptimizeCompareInstr should avoid iterating pass the beginning of the MBB when the 'and' instruction is after the comparison. 2010-09-21 23:49:07 +00:00
2010-09-29-mc-asm-header-test.ll Rework some .ARM.attribute work for improved gcc compatibility. 2011-02-07 00:49:53 +00:00
2010-10-19-mc-elf-objheader.ll Rework some .ARM.attribute work for improved gcc compatibility. 2011-02-07 00:49:53 +00:00
2010-10-25-ifcvt-ldm.ll Properly pseudo-ize the ARM LDMIA_RET instruction. This has the nice side- 2011-03-11 22:51:41 +00:00
2010-11-15-SpillEarlyClobber.ll Fix PR8612 in the standard spiller, take two. 2010-11-16 00:40:59 +00:00
2010-11-29-PrologueBug.ll Properly pseudo-ize the ARM LDMIA_RET instruction. This has the nice side- 2011-03-11 22:51:41 +00:00
2010-11-30-reloc-movt.ll Materialize GA addresses with movw + movt pairs for Darwin in PIC mode. e.g. 2011-01-17 08:03:18 +00:00
2010-12-07-PEIBug.ll Fix a bad prologue / epilogue codegen bug where the compiler would emit illegal 2010-12-07 23:08:38 +00:00
2010-12-08-tpsoft.ll ARM/MC/ELF TPsoft is now a proper pseudo inst. 2010-12-08 23:14:44 +00:00
2010-12-13-reloc-pic.ll Materialize GA addresses with movw + movt pairs for Darwin in PIC mode. e.g. 2011-01-17 08:03:18 +00:00
2010-12-15-elf-lcomm.ll Write the section table and the section data in the same order that 2011-03-20 18:44:20 +00:00
2010-12-17-LocalStackSlotCrash.ll During local stack slot allocation, the materializeFrameBaseRegister function 2010-12-17 23:09:14 +00:00
2011-01-19-MergedGlobalDbg.ll Enable GlobalMerge on darwin. 2011-03-23 23:34:19 +00:00
2011-02-04-AntidepMultidef.ll PostRA antidependence breaker unit test for rdar://8959122. 2011-02-08 17:41:12 +00:00
2011-02-07-AntidepClobber.ll PostRA antidependence breaker unit test for PR8986. 2011-02-08 17:42:05 +00:00
2011-03-10-DAGCombineCrash.ll Avoid replacing the value of a directly stored load with the stored value if the load is indexed. rdar://9117613. 2011-03-11 00:48:56 +00:00
2011-03-15-LdStMultipleBug.ll Do not form thumb2 ldrd / strd if the offset is by multiple of 4. rdar://9133587 2011-03-15 18:41:52 +00:00
2011-03-23-PeepholeBug.ll Cmp peephole optimization isn't always safe for signed arithmetics. 2011-03-23 22:52:04 +00:00
2011-04-07-schediv.ll Added a check in the preRA scheduler for potential interference on a 2011-04-07 19:54:57 +00:00
2011-04-11-MachineLICMBug.ll Look pass copies when determining whether hoisting would end up inserting more copies. rdar://9266679 2011-04-11 21:09:18 +00:00
2011-04-12-AlignBug.ll Fix an obvious problem with an alignment computation. AsmPrinter actually does 2011-04-13 09:02:43 +00:00
2011-04-12-FastRegAlloc.ll Fix a bug where we were counting the alias sets as completely used 2011-04-18 19:26:25 +00:00
2011-04-15-AndVFlagPeepholeBug.ll The AND instruction leaves the V flag unmodified, so it falls victim to the same 2011-04-15 20:45:00 +00:00
2011-04-15-RegisterCmpPeephole.ll Add ORR and EOR to the CMP peephole optimizer. It's hard to get isel to generate 2011-04-15 21:24:38 +00:00
addrmode.ll
aliases.ll emit jump table an alias ".set" directives through MCStreamer as 2010-01-26 21:53:08 +00:00
align.ll Increase ARM APCS preferred alignment for i64 and f64 from 32 bits to 64 bits. 2010-09-29 17:54:10 +00:00
alloca.ll In the presence of variable sized objects, allocate an emergency spill slot. 2010-07-09 20:27:06 +00:00
argaddr.ll
arguments2.ll
arguments3.ll Eliminate more uses of llvm-as and llvm-dis. 2009-09-09 00:09:15 +00:00
arguments4.ll
arguments5.ll
arguments6.ll
arguments7.ll Fix declarations in a few more tests. 2010-04-17 21:29:25 +00:00
arguments8.ll Fix declarations in a few more tests. 2010-04-17 21:29:25 +00:00
arguments_f64_backfill.ll Use Unified Assembly Syntax for the ARM backend. 2009-11-09 00:11:35 +00:00
arguments-nosplit-double.ll
arguments-nosplit-i64.ll
arguments.ll Fix ARM tests to be register allocator independent. 2011-03-31 22:14:03 +00:00
arm-and-tst-peephole.ll Fix ARM tests to be register allocator independent. 2011-03-31 22:14:03 +00:00
arm-asm.ll
arm-frameaddr.ll Remove the arm_aapcscc marker from the tests. It is the default 2010-06-15 19:04:29 +00:00
arm-negative-stride.ll Added sub-formats to the NeonI/NeonXI instructions to further refine the NEONFrm 2010-03-17 17:52:21 +00:00
arm-returnaddr.ll These tests no longer require linear scan because reserved register coalescing is now universal. 2011-04-05 21:40:41 +00:00
armv4.ll Remove the arm_aapcscc marker from the tests. It is the default 2010-06-15 19:04:29 +00:00
atomic-cmp.ll Fix an obvious typo which caused an isel assertion. rdar://8964854. 2011-02-07 18:50:47 +00:00
available_externally.ll available_externally (hidden or not) GVs are always accessed via stubs. rdar://9027648. 2011-02-22 06:58:34 +00:00
avoid-cpsr-rmw.ll Avoid some 's' 16-bit instruction which partially update CPSR 2011-04-19 18:11:49 +00:00
bfc.ll Fix some more failures by converting to FileCheck. 2009-10-27 05:50:28 +00:00
bfi.ll bfi A, (and B, C1), C2) -> bfi A, B, C2 iff C1 & C2 == C1. rdar://8458663 2010-12-14 03:22:07 +00:00
bfx.ll Update ARM DAGtoDAG for matching UBFX instruction for unsigned bitfield 2010-04-22 23:24:18 +00:00
bic.ll Forgot to alter RUN line when converting to FileCheck. 2009-11-22 13:09:48 +00:00
bits.ll FileCheck-ize 2010-09-17 21:46:16 +00:00
bswap-inline-asm.ll Recognize inline asm 'rev /bin/bash, ' as a bswap intrinsic call. 2011-01-08 01:24:27 +00:00
bx_fold.ll Properly pseudo-ize the ARM LDMIA_RET instruction. This has the nice side- 2011-03-11 22:51:41 +00:00
call_nolink.ll
call-tc.ll Clean up ARM tail calls a bit. They're pseudo-instructions for normal branches. 2011-03-15 00:30:40 +00:00
call.ll Remove a tail call, and move some CHECKs to the 2010-06-04 01:01:04 +00:00
carry.ll Fix some more failures by converting to FileCheck. 2009-10-27 05:50:28 +00:00
clz.ll FileCheck-ize a few tests. 2010-10-22 21:55:03 +00:00
code-placement.ll Properly pseudo-ize the ARM LDMIA_RET instruction. This has the nice side- 2011-03-11 22:51:41 +00:00
compare-call.ll Use Unified Assembly Syntax for the ARM backend. 2009-11-09 00:11:35 +00:00
constants.ll Fix test-llvm failures. 2011-04-05 18:41:40 +00:00
crash-greedy.ll Pick a conservative register class when creating a small live range for remat. 2011-03-31 03:54:44 +00:00
crash-O0.ll When using ADDri to get the address of a stack object, 255 is a conservative 2010-06-18 20:59:25 +00:00
crash.ll Simplify test and move into a generic "crash" ll file. 2010-10-09 00:29:04 +00:00
cse-libcalls.ll
ctors_dtors.ll Be nice to Xcore and the XMOS assembler and avoid quoting section names 2011-03-04 20:03:14 +00:00
ctz.ll Test case for r93758. 2010-01-19 00:35:20 +00:00
dg.exp
div.ll These tests are looking for library function names that 2010-11-17 21:57:32 +00:00
divmod.ll Fix divmod libcall lowering. Convert to {S|U}DIVREM first and then expand the node to a libcall. rdar://9280991 2011-04-16 03:08:26 +00:00
dyn-stackalloc.ll
extloadi1.ll
fabss.ll Avoid write-after-write issue hazards for Cortex-A9. 2011-04-19 18:11:45 +00:00
fadds.ll Avoid write-after-write issue hazards for Cortex-A9. 2011-04-19 18:11:45 +00:00
fast-isel-crash.ll Add a testcase for my last checkin. 2011-01-27 06:01:17 +00:00
fast-isel-pred.ll Fix this test up a bit. 2011-03-14 21:05:21 +00:00
fast-isel-static.ll Fix up testcase for previous commit. 2011-04-05 00:56:01 +00:00
fast-isel.ll Recommit this change and remove the failing part of the test - it didn't 2010-11-15 21:11:06 +00:00
fcopysign.ll Tighten test case a bit. 2011-04-19 06:14:45 +00:00
fdivs.ll Avoid write-after-write issue hazards for Cortex-A9. 2011-04-19 18:11:45 +00:00
fixunsdfdi.ll Use Unified Assembly Syntax for the ARM backend. 2009-11-09 00:11:35 +00:00
flag-crash.ll It's possible that a flag is added to the SDNode that points back to the 2010-06-24 22:00:37 +00:00
fmacs.ll This patch combines several changes from Evan Cheng for rdar://8659675. 2011-04-19 18:11:57 +00:00
fmdrr-fmrrd.ll
fmscs.ll Add some missing isel predicates on def : pat patterns to avoid generating VFP vmla / vmls (they cause stalls). Disabling them in isel is properly not a right solution, I'll look into a proper solution next. 2010-11-12 20:32:20 +00:00
fmuls.ll Avoid write-after-write issue hazards for Cortex-A9. 2011-04-19 18:11:45 +00:00
fnegs.ll Convert -enable-sched-cycles and -enable-sched-hazard to -disable 2011-01-21 05:51:33 +00:00
fnmacs.ll Add some missing isel predicates on def : pat patterns to avoid generating VFP vmla / vmls (they cause stalls). Disabling them in isel is properly not a right solution, I'll look into a proper solution next. 2010-11-12 20:32:20 +00:00
fnmscs.ll Fix ARM tests to be register allocator independent. 2011-03-31 22:14:03 +00:00
fnmul.ll Use Unified Assembly Syntax for the ARM backend. 2009-11-09 00:11:35 +00:00
fnmuls.ll Fix an unnecessary XFAIL 2010-09-02 20:19:24 +00:00
formal.ll
fp16.ll Fix typo 2010-03-14 18:42:52 +00:00
fp_convert.ll Avoid write-after-write issue hazards for Cortex-A9. 2011-04-19 18:11:45 +00:00
fp-arg-shuffle.ll Split a store of a VMOVDRR into two integer stores to avoid mixing NEON and ARM 2011-04-12 02:24:17 +00:00
fp.ll Fix test-llvm failures. 2011-04-05 18:41:40 +00:00
fparith.ll Use Unified Assembly Syntax for the ARM backend. 2009-11-09 00:11:35 +00:00
fpcmp_ueq.ll These tests are looking for library function names that 2010-11-17 21:57:32 +00:00
fpcmp-opt.ll LDRD/STRD instructions should print both Rt and Rt2 in the asm string. 2011-04-01 20:26:57 +00:00
fpcmp.ll Use Unified Assembly Syntax for the ARM backend. 2009-11-09 00:11:35 +00:00
fpconsts.ll Convert -enable-sched-cycles and -enable-sched-hazard to -disable 2011-01-21 05:51:33 +00:00
fpconv.ll These tests are looking for library function names that 2010-11-17 21:57:32 +00:00
fpmem.ll Use Unified Assembly Syntax for the ARM backend. 2009-11-09 00:11:35 +00:00
fpow.ll
fpowi.ll Correct bogus module triple specifications. 2010-08-30 10:48:29 +00:00
fptoint.ll Fix for bad FileCheck converts in revision 89584. 2009-11-22 12:50:05 +00:00
fsubs.ll switch the flag for using NEON for SP floating point to a subtarget 'feature'. 2010-03-25 23:47:34 +00:00
global-merge.ll Enable GlobalMerge on darwin. 2011-03-23 23:34:19 +00:00
globals.ll Start function numbering at 0. 2010-04-17 16:29:15 +00:00
hardfloat_neon.ll
hello.ll Fix epilogue codegen to avoid leaving the stack pointer in an invalid 2010-11-22 18:12:04 +00:00
hidden-vis-2.ll
hidden-vis-3.ll Start function numbering at 0. 2010-04-17 16:29:15 +00:00
hidden-vis.ll
iabs.ll convert to filecheck. 2010-02-08 23:47:34 +00:00
ifcvt1.ll
ifcvt2.ll Allow ARM if-converter to be run after post allocation scheduling. 2010-06-18 23:09:54 +00:00
ifcvt3.ll
ifcvt4.ll
ifcvt5.ll Properly pseudo-ize the ARM LDMIA_RET instruction. This has the nice side- 2011-03-11 22:51:41 +00:00
ifcvt6.ll Properly pseudo-ize the ARM LDMIA_RET instruction. This has the nice side- 2011-03-11 22:51:41 +00:00
ifcvt7.ll Properly pseudo-ize the ARM LDMIA_RET instruction. This has the nice side- 2011-03-11 22:51:41 +00:00
ifcvt8.ll Properly pseudo-ize the ARM LDMIA_RET instruction. This has the nice side- 2011-03-11 22:51:41 +00:00
ifcvt9.ll
ifcvt10.ll Properly pseudo-ize the ARM LDMIA_RET instruction. This has the nice side- 2011-03-11 22:51:41 +00:00
ifcvt11.ll Two sets of changes. Sorry they are intermingled. 2010-11-03 00:45:17 +00:00
illegal-vector-bitcast.ll
imm.ll
indirectbr.ll Fix ARM tests to be register allocator independent. 2011-03-31 22:14:03 +00:00
inlineasm2.ll
inlineasm3.ll Make tests register allocation independent again. 2011-04-19 00:14:43 +00:00
inlineasm-imm-arm.ll
inlineasm.ll llvm can't correctly support 'H', 'Q' and 'R' modifiers. Just mark it an error. 2010-05-27 22:08:38 +00:00
insn-sched1.ll Remove more tail calls. 2010-06-04 01:01:24 +00:00
int-to-fp.ll Add Neon SINT_TO_FP and UINT_TO_FP lowering from v4i16 to v4f32. Fixes 2011-03-29 21:41:55 +00:00
ispositive.ll Update tests to handle MC-inst instruction printing of shift operations. The 2010-09-17 21:58:46 +00:00
large-stack.ll
ldm.ll Properly pseudo-ize the ARM LDMIA_RET instruction. This has the nice side- 2011-03-11 22:51:41 +00:00
ldr_ext.ll
ldr_frame.ll
ldr_post.ll
ldr_pre.ll
ldr.ll Fix the rest of the ARM failures by converting them to FileCheck. 2009-10-27 06:16:45 +00:00
ldrd.ll LDRD/STRD instructions should print both Rt and Rt2 in the asm string. 2011-04-01 20:26:57 +00:00
ldst-f32-2-i32.ll Given a pair of floating point load and store, if there are no other uses of 2011-02-02 01:06:55 +00:00
load-global.ll Last round of fixes for movw + movt global address codegen. 2011-01-21 18:55:51 +00:00
load.ll
long_shift.ll Cmp peephole optimization isn't always safe for signed arithmetics. 2011-03-23 22:52:04 +00:00
long-setcc.ll
long.ll Fix test-llvm failures. 2011-04-05 18:41:40 +00:00
lsr-code-insertion.ll PR8921: LDM/POP do not support interworking prior to v5t. 2011-01-06 19:24:41 +00:00
lsr-on-unrolled-loops.ll remove a bunch of CHECK lines that aren't checking what 2011-04-09 06:31:06 +00:00
lsr-scale-addr-mode.ll
machine-cse-cmp.ll Re-apply 105308 with fix. 2010-06-04 23:28:13 +00:00
machine-licm.ll Last round of fixes for movw + movt global address codegen. 2011-01-21 18:55:51 +00:00
mem.ll
memcpy-inline.ll Recommit r129383. PreRA scheduler heuristic fixes: VRegCycle, TokenFactor latency. 2011-04-13 00:38:32 +00:00
memfunc.ll
mls.ll Convert a few tests to FileCheck for PR5307. 2009-11-22 11:45:44 +00:00
movt-movw-global.ll Materialize global addresses via movt/movw pair, this is always better 2009-11-24 00:44:37 +00:00
movt.ll Match more patterns to movt. 2009-10-21 08:15:52 +00:00
mul_const.ll Update tests to handle MC-inst instruction printing of shift operations. The 2010-09-17 21:58:46 +00:00
mul.ll
mulhi.ll
mult-alt-generic-arm.ll Inline asm mult-alt constraint tests. 2010-11-02 23:01:44 +00:00
mvn.ll
neon_arith1.ll
neon_div.ll Recommit r129383. PreRA scheduler heuristic fixes: VRegCycle, TokenFactor latency. 2011-04-13 00:38:32 +00:00
neon_ld1.ll Use Unified Assembly Syntax for the ARM backend. 2009-11-09 00:11:35 +00:00
neon_ld2.ll Use Unified Assembly Syntax for the ARM backend. 2009-11-09 00:11:35 +00:00
neon_minmax.ll Check for comparisons of +/- zero when optimizing less-than-or-equal and 2010-02-24 22:15:53 +00:00
neon_shift.ll Narrow right shifts need to encode their immediates differently from a normal 2011-03-01 01:00:59 +00:00
pack.ll Remove duplicate test. 2010-10-22 22:04:28 +00:00
peephole-bitcast.ll Fix ARM tests to be register allocator independent. 2011-03-31 22:14:03 +00:00
phi.ll move this test into the ARM test so that it is only run when the arm backend 2010-12-19 02:58:14 +00:00
pr3502.ll
prefetch.ll Add -mcpu=cortex-a9-mp. It's cortex-a9 with MP extension. rdar://8648637. 2011-04-19 18:11:52 +00:00
private.ll Delete useless trailing semicolons. 2010-01-05 17:55:26 +00:00
reg_sequence.ll Fix ARM tests to be register allocator independent. 2011-03-31 22:14:03 +00:00
ret0.ll
ret_arg1.ll
ret_arg2.ll
ret_arg3.ll
ret_arg4.ll
ret_arg5.ll
ret_f32_arg2.ll
ret_f32_arg5.ll
ret_f64_arg2.ll
ret_f64_arg_reg_split.ll
ret_f64_arg_split.ll
ret_f64_arg_stack.ll
ret_i64_arg2.ll
ret_i64_arg3.ll
ret_i64_arg_split.ll
ret_i128_arg2.ll
ret_void.ll
rev.ll Match a few more obvious patterns to revsh. rdar://9147637. 2011-03-18 21:52:42 +00:00
sbfx.ll add a simple dag combine to replace trivial shl+lshr with 2010-04-15 05:28:43 +00:00
section.ll Be nice to Xcore and the XMOS assembler and avoid quoting section names 2011-03-04 20:03:14 +00:00
select_xform.ll Fix test-llvm failures. 2011-04-05 18:41:40 +00:00
select-imm.ll Fix test-llvm failures. 2011-04-05 18:41:40 +00:00
select.ll Fix a compiler crash where a Glue value had multiple uses. Radar 9049552. 2011-03-08 01:17:20 +00:00
shifter_operand.ll Minor optimization. sign-ext/anyext of undef is still undef. 2011-03-14 18:15:55 +00:00
shuffle.ll Testcase for r127630. 2011-03-15 01:49:08 +00:00
smul.ll FileCheck'ify test. 2011-04-01 03:36:33 +00:00
spill-q.ll Add ARM patterns to match EXTRACT_SUBVECTOR nodes. 2011-01-07 04:59:04 +00:00
stack-frame.ll
stm.ll Teach the (non-MC) instruction printer to use the cannonical names for push/pop, 2010-09-17 22:36:38 +00:00
str_post.ll Fix the rest of the ARM failures by converting them to FileCheck. 2009-10-27 06:16:45 +00:00
str_pre-2.ll Fix ARM tests to be register allocator independent. 2011-03-31 22:14:03 +00:00
str_pre.ll
str_trunc.ll
sub.ll Teach the ARM peephole optimizer that RSB, RSC, ADC, and SBC can be used for folded comparisons, just like ADD and SUB. 2011-04-06 23:35:59 +00:00
sxt_rot.ll
t2-imm.ll Fix borken test 2010-08-30 23:41:49 +00:00
tail-opts.ll Last round of fixes for movw + movt global address codegen. 2011-01-21 18:55:51 +00:00
thread_pointer.ll
thumb1-varalloc.ll Fix ARM tests to be register allocator independent. 2011-03-31 22:14:03 +00:00
tls1.ll
tls2.ll Added sub-formats to the NeonI/NeonXI instructions to further refine the NEONFrm 2010-03-17 17:52:21 +00:00
tls3.ll
trap.ll Change -arm-trap-func= into a non-arm specific option. Now Intrinsic::trap is lowered into a call to the specified trap function at sdisel time. 2011-04-08 21:37:21 +00:00
trunc_ldr.ll
truncstore-dag-combine.ll
tst_teq.ll
uint64tof64.ll
umulo-32.ll If we can, lower the multiply part of a umulo/smulo call to a libcall 2011-01-20 00:29:24 +00:00
unaligned_load_store.ll Convert -enable-sched-cycles and -enable-sched-hazard to -disable 2011-01-21 05:51:33 +00:00
undef-sext.ll Minor optimization. sign-ext/anyext of undef is still undef. 2011-03-14 18:15:55 +00:00
unord.ll
uxt_rot.ll
uxtb.ll
va_arg.ll Recommit r129383. PreRA scheduler heuristic fixes: VRegCycle, TokenFactor latency. 2011-04-13 00:38:32 +00:00
vaba.ll Replace NEON vabdl, vaba, and vabal intrinsics with combinations of the 2010-09-03 01:35:08 +00:00
vabd.ll Replace NEON vabdl, vaba, and vabal intrinsics with combinations of the 2010-09-03 01:35:08 +00:00
vabs.ll
vadd.ll Remove NEON vaddl, vaddw, vsubl, and vsubw intrinsics. Instead, use llvm 2010-08-29 05:57:34 +00:00
vargs_align.ll
vargs.ll
vbits.ll Add missing colon. 2010-11-29 22:39:38 +00:00
vbsl-constant.ll Fix a typo in an ARM-specific DAG combine. This fixes <rdar://problem/9278274>. 2011-04-13 21:01:19 +00:00
vbsl.ll
vceq.ll Add support for ARM's specialized vector-compare-against-zero instructions. 2010-11-08 23:21:22 +00:00
vcge.ll Fix result type of Neon floating-point comparisons against zero. 2010-12-18 00:04:33 +00:00
vcgt.ll Fix ARM tests to be register allocator independent. 2011-03-31 22:14:03 +00:00
vcnt.ll
vcombine.ll Add testcases for PR8411 (vget_low and vget_high implemented as shuffles). 2011-01-07 06:44:14 +00:00
vcvt.ll Add Neon VCVT instructions for f32 <-> f16 conversions. 2010-12-15 22:14:12 +00:00
vdup.ll Add float patterns for Neon vld1-lane/dup and vst1-lane operations. 2010-12-10 22:13:32 +00:00
vector-DAGCombine.ll Prevent ARM DAG Combiner from doing an AND or OR combine on an illegal vector type (vectors of size 3). Also included test cases. 2011-04-07 15:24:20 +00:00
vext.ll Some minor cleanups based on feedback. 2011-03-15 20:47:26 +00:00
vfcmp.ll
vfp.ll Recommit r129383. PreRA scheduler heuristic fixes: VRegCycle, TokenFactor latency. 2011-04-13 00:38:32 +00:00
vget_lane.ll Add codegen patterns for VST1-lane instructions. Radar 8599955. 2010-11-03 16:24:53 +00:00
vhadd.ll
vhsub.ll
vicmp.ll
vld1.ll Do not lose mem_operands while lowering VLD / VST intrinsics. 2011-04-19 00:04:03 +00:00
vld2.ll Add codegen support for using post-increment NEON load/store instructions. 2011-02-07 17:43:21 +00:00
vld3.ll Fix ARM tests to be register allocator independent. 2011-03-31 22:14:03 +00:00
vld4.ll Add codegen support for using post-increment NEON load/store instructions. 2011-02-07 17:43:21 +00:00
vlddup.ll Add codegen support for using post-increment NEON load/store instructions. 2011-02-07 17:43:21 +00:00
vldlane.ll Do not lose mem_operands while lowering VLD / VST intrinsics. 2011-04-19 00:04:03 +00:00
vminmax.ll
vmla.ll Remove DAG combiner patch to fold vector splats. Instcombiner does it now. 2010-10-29 22:03:02 +00:00
vmls.ll Remove NEON vmull, vmlal, and vmlsl intrinsics, replacing them with multiply, 2010-09-01 23:50:19 +00:00
vmov.ll NEON does not support truncating vector stores. Radar 8598391. 2010-11-01 18:31:39 +00:00
vmul.ll Distribute (A + B) * C to (A * C) + (B * C) to make use of NEON multiplier 2011-03-31 19:38:48 +00:00
vneg.ll Fix tests to use fadd, fsub, and fmul, instead of add, sub, and mul, 2010-05-03 22:36:46 +00:00
vpadal.ll
vpadd.ll
vpminmax.ll
vqadd.ll
vqdmul.ll
vqshl.ll
vqshrn.ll
vqsub.ll
vrec.ll
vrev.ll SelectionDAG shuffle nodes do not allow operands with different numbers of 2010-10-27 20:38:28 +00:00
vshift.ll
vshiftins.ll
vshl.ll
vshll.ll
vshrn.ll
vsra.ll
vst1.ll Add codegen support for using post-increment NEON load/store instructions. 2011-02-07 17:43:21 +00:00
vst2.ll Add codegen support for using post-increment NEON load/store instructions. 2011-02-07 17:43:21 +00:00
vst3.ll Add codegen support for using post-increment NEON load/store instructions. 2011-02-07 17:43:21 +00:00
vst4.ll Add codegen support for using post-increment NEON load/store instructions. 2011-02-07 17:43:21 +00:00
vstlane.ll Add patterns to use post-increment addressing for Neon VST1-lane instructions. 2011-02-25 06:42:42 +00:00
vsub.ll Remove NEON vaddl, vaddw, vsubl, and vsubw intrinsics. Instead, use llvm 2010-08-29 05:57:34 +00:00
vtbl.ll
vtrn.ll Allow more cases of undef shuffle indices and add tests for them. 2010-08-17 05:54:34 +00:00
vuzp.ll Allow more cases of undef shuffle indices and add tests for them. 2010-08-17 05:54:34 +00:00
vzip.ll Allow more cases of undef shuffle indices and add tests for them. 2010-08-17 05:54:34 +00:00
weak2.ll
weak.ll