aboutsummaryrefslogtreecommitdiff
path: root/lib
AgeCommit message (Collapse)Author
2010-12-05Code clean up.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120965 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-05Remove an unused variable.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120964 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-05Some cleanup before I start committing some incremental progress onCameron Zwarich
StrongPHIElimination. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120961 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-05Making use of VFP / NEON floating point multiply-accumulate / subtraction isEvan Cheng
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. Work in progress, only A+B are enabled. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120960 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-05Remove the PHIElimination.h header, as it is no longer needed.Cameron Zwarich
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120959 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-05Fix PR 4170 by having ExtractValueInst::getIndexedType() reject ↵Frits van Bommel
out-of-bounds indexing. Also add asserts that the indices are valid in InsertValueInst::init(). ExtractValueInst already asserts when constructed with invalid indices. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120956 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-05I forgot to actually remove the FindCopyInsertPoint() declaration fromCameron Zwarich
PHIElimination.h. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120953 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-05Remove the SplitCriticalEdge() method declaration from PHIElimination.h. At oneCameron Zwarich
time, this method existed, but now PHIElimination uses the method of the same name on MachineBasicBlock. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120952 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-05Move the FindCopyInsertPoint method of PHIElimination to a new standaloneCameron Zwarich
function so that it can be shared with StrongPHIElimination. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120951 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-05Refactor jump threading.Frits van Bommel
Should have no functional change other than the order of two transformations that are mutually-exclusive and the exact formatting of debug output. Internally, it now stores the ConstantInt*s as Constant*s, and actual undef values instead of nulls. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120946 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-05Remove trailing whitespace.Frits van Bommel
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120945 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-05Teach SimplifyCFG to turnFrits van Bommel
(indirectbr (select cond, blockaddress(@fn, BlockA), blockaddress(@fn, BlockB))) into (br cond, BlockA, BlockB). git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120943 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-05Teach X86ISelLowering that the second result of X86ISD::UMUL is a flagsChris Lattner
result. This allows us to compile: void *test12(long count) { return new int[count]; } into: test12: movl $4, %ecx movq %rdi, %rax mulq %rcx movq $-1, %rdi cmovnoq %rax, %rdi jmp __Znam ## TAILCALL instead of: test12: movl $4, %ecx movq %rdi, %rax mulq %rcx seto %cl testb %cl, %cl movq $-1, %rdi cmoveq %rax, %rdi jmp __Znam Of course it would be even better if the regalloc inverted the cmov to 'cmovoq', which would eliminate the need for the 'movq %rdi, %rax'. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120936 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-05it turns out that when ".with.overflow" intrinsics were added to the X86Chris Lattner
backend that they were all implemented except umul. This one fell back to the default implementation that did a hi/lo multiply and compared the top. Fix this to check the overflow flag that the 'mul' instruction sets, so we can avoid an explicit test. Now we compile: void *func(long count) { return new int[count]; } into: __Z4funcl: ## @_Z4funcl movl $4, %ecx ## encoding: [0xb9,0x04,0x00,0x00,0x00] movq %rdi, %rax ## encoding: [0x48,0x89,0xf8] mulq %rcx ## encoding: [0x48,0xf7,0xe1] seto %cl ## encoding: [0x0f,0x90,0xc1] testb %cl, %cl ## encoding: [0x84,0xc9] movq $-1, %rdi ## encoding: [0x48,0xc7,0xc7,0xff,0xff,0xff,0xff] cmoveq %rax, %rdi ## encoding: [0x48,0x0f,0x44,0xf8] jmp __Znam ## TAILCALL instead of: __Z4funcl: ## @_Z4funcl movl $4, %ecx ## encoding: [0xb9,0x04,0x00,0x00,0x00] movq %rdi, %rax ## encoding: [0x48,0x89,0xf8] mulq %rcx ## encoding: [0x48,0xf7,0xe1] testq %rdx, %rdx ## encoding: [0x48,0x85,0xd2] movq $-1, %rdi ## encoding: [0x48,0xc7,0xc7,0xff,0xff,0xff,0xff] cmoveq %rax, %rdi ## encoding: [0x48,0x0f,0x44,0xf8] jmp __Znam ## TAILCALL Other than the silly seto+test, this is using the o bit directly, so it's going in the right direction. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120935 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-05generalize the previous check to handle -1 on either side of the Chris Lattner
select, inserting a not to compensate. Add a missing isZero check that I lost somehow. This improves codegen of: void *func(long count) { return new int[count]; } from: __Z4funcl: ## @_Z4funcl movl $4, %ecx ## encoding: [0xb9,0x04,0x00,0x00,0x00] movq %rdi, %rax ## encoding: [0x48,0x89,0xf8] mulq %rcx ## encoding: [0x48,0xf7,0xe1] testq %rdx, %rdx ## encoding: [0x48,0x85,0xd2] movq $-1, %rdi ## encoding: [0x48,0xc7,0xc7,0xff,0xff,0xff,0xff] cmoveq %rax, %rdi ## encoding: [0x48,0x0f,0x44,0xf8] jmp __Znam ## TAILCALL ## encoding: [0xeb,A] to: __Z4funcl: ## @_Z4funcl movl $4, %ecx ## encoding: [0xb9,0x04,0x00,0x00,0x00] movq %rdi, %rax ## encoding: [0x48,0x89,0xf8] mulq %rcx ## encoding: [0x48,0xf7,0xe1] cmpq $1, %rdx ## encoding: [0x48,0x83,0xfa,0x01] sbbq %rdi, %rdi ## encoding: [0x48,0x19,0xff] notq %rdi ## encoding: [0x48,0xf7,0xd7] orq %rax, %rdi ## encoding: [0x48,0x09,0xc7] jmp __Znam ## TAILCALL ## encoding: [0xeb,A] git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120932 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-05Improve an integer select optimization in two ways:Chris Lattner
1. generalize (select (x == 0), -1, 0) -> (sign_bit (x - 1)) to: (select (x == 0), -1, y) -> (sign_bit (x - 1)) | y 2. Handle the identical pattern that happens with !=: (select (x != 0), y, -1) -> (sign_bit (x - 1)) | y cmov is often high latency and can't fold immediates or memory operands. For example for (x == 0) ? -1 : 1, before we got: < testb %sil, %sil < movl $-1, %ecx < movl $1, %eax < cmovel %ecx, %eax now we get: > cmpb $1, %sil > sbbl %eax, %eax > orl $1, %eax git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120929 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-04Initialize HasPOPCNT.Bill Wendling
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120923 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-04Once the layout is done we don't need to keep updating which fragments areRafael Espindola
valid. Addresses will not change. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120921 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-04Remember the contents of leb and dwarfline fragments when relaxing. This avoidsRafael Espindola
having to evaluate the expression again when writing. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120920 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-04Remove PHIElimination's private copy of SkipPHIsAndLabels.Cameron Zwarich
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120918 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-04Add patterns for the x86 popcnt instruction.Benjamin Kramer
- Also adds a new POPCNT subtarget feature that is currently enabled if the target supports SSE4.2 (nehalem) or SSE4A (barcelona). git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120917 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-04Silence 'may be used uninitialized in this function' warnings. Static analysisBill Wendling
may determine that they cannot be used uninitialized. But that might be a bit too much for the compiler to determine. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120916 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-04Support/PathV2: Remove redundant calls to make_error_code.Michael J. Spencer
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120913 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-04APInt: microoptimize a few methods.Benjamin Kramer
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120912 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-04Remove unneeded zero arrays.Benjamin Kramer
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120910 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-04Apparently APFloat::getZero doesn't like PPCDoubleDoubles.Benjamin Kramer
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120909 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-04Simplify code. No functionality change.Benjamin Kramer
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120907 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-04The Thumb tADDrSPi instruction is not valid when the destination is SP.Bob Wilson
Check for that and try narrowing it to tADDspi instead. Radar 8724703. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120892 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-04There are two reasons why we might want to useRafael Espindola
foo = a - b .long foo instead of just .long a - b First, on darwin9 64 bits the assembler produces the wrong result. Second, if "a" is the end of the section all darwin assemblers (9, 10 and mc) will not consider a - b to be a constant but will if the dummy foo is created. Split how we handle these cases. The first one is something MC should take care of. The second one has to be handled by the caller. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120889 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-04Support/FileSystem: Add status implementation.Michael J. Spencer
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120870 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-04Support/Windows/FileSystem: Fix MinGW warnings.Michael J. Spencer
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120868 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-04Support/FileSystem: Add file_size implementation.Michael J. Spencer
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120867 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-04Next step: Only pad debug_line when the target is darwin. Add a FIXME to avoidRafael Espindola
doing that if the target is darwin10 or newer. This fixes *) Direct object emission was producing objects without the workaround on darwin9. *) Assembly printing was producing objects with the workaround on linux. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120866 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-04Encode condition code for Thumb1 conditional branch instruction.Jim Grosbach
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120865 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-03Correctly size-reduce the t2CMPzrr instruction to tCMPzr when possible.Jim Grosbach
tCMPzhir has undefined behavior when both source registers are low registers. rdar://8728577 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120858 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-03Use correct variable names to match the patterns.Bill Wendling
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120857 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-03Also inore '()' while creating mdnode name from ObjC symbol name.Jakob Stoklund Olesen
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120856 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-03First step in fixing MC. Make it clear that we are avoiding a bug in theRafael Espindola
darwin9 linker, what is needed to avoid it and where to get more information. Also make the workaround simpler. Just the regular end_sequence we normally create is more than 4 bytes. Tested by building cctools and ld64 from darwin9 on a darwin10 system and using those. I checked that I was able to reproduce the bootstrap failure when the the workaround was disabled. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120854 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-03Ignore '+' while creating mdnode name from ObjC symbol name.Devang Patel
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120853 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-03Match pattern operand names to expected encoding field names. This corrects theJim Grosbach
operand encoding ordering of the instruction. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120852 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-03Remove incorrect BL target encoding (it's similar to, but not the same as theJim Grosbach
ARM instruction). Add encoding of bits 13 and 11. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120849 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-03Encode the 32-bit wide Thumb (and Thumb2) instructions with the high orderJim Grosbach
halfword being emitted to the stream first. rdar://8728174 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120848 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-03Revert this change since it breaks a couple of the AVX tests.Nate Begeman
I'm unclear if the tests are actually correct or not, but reverting for now. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120847 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-03Rename virtRegMap to avoid confusion with the VirtRegMap that it isn't.Jakob Stoklund Olesen
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120846 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-03Coalesce debug locations when possible, causing less DBG_VALUE instructions toJakob Stoklund Olesen
be emitted. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120845 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-03Scalar f32/f64 are also subregs of ymm regsNate Begeman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120844 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-03Remove SSE1-4 disable when AVX is enabled. While this may be useful for ↵Nate Begeman
development, it completely breaks scalar fp in xmm regs when AVX is enabled. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120843 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-03Emit DBG_VALUE instructions from LiveDebugVariables.Jakob Stoklund Olesen
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120842 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-03Also update virtRegMap when renaming virtual registers.Jakob Stoklund Olesen
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120841 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-03When using the 'push' mnemonic for Thumb2 stmdb, be explicit when it's theJim Grosbach
32-bit wide version by adding the .w suffix. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120838 91177308-0d34-0410-b5e6-96231b3b80d8
generated by cgit v1.2.3-18-g5258 (git 2.32.0) at 2025-08-05 07:38:49 +0000