The fixup value comes in as the whole 32-bit value, so for the lo16 fixup,
the upper bits need to be masked off. Previously we assumed the masking had
already been done and asserted.
rdar://9635991
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@133818 91177308-0d34-0410-b5e6-96231b3b80d8
instructions can be used to match combinations of multiply/divide and VCVT
(between floating-point and integer, Advanced SIMD). Basically the VCVT
immediate operand that specifies the number of fraction bits corresponds to a
floating-point multiply or divide by the corresponding power of 2.
For example, VCVT (floating-point to fixed-point, Advanced SIMD) can replace a
combination of VMUL and VCVT (floating-point to integer) as follows:
Example (assume d17 = <float 8.000000e+00, float 8.000000e+00>):
vmul.f32 d16, d17, d16
vcvt.s32.f32 d16, d16
becomes:
vcvt.s32.f32 d16, d16, #3
Similarly, VCVT (fixed-point to floating-point, Advanced SIMD) can replace a
combinations of VCVT (integer to floating-point) and VDIV as follows:
Example (assume d17 = <float 8.000000e+00, float 8.000000e+00>):
vcvt.f32.s32 d16, d16
vdiv.f32 d16, d17, d16
becomes:
vcvt.f32.s32 d16, d16, #3
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@133813 91177308-0d34-0410-b5e6-96231b3b80d8
target machine from those that are only needed by codegen. The goal is to
sink the essential target description into MC layer so we can start building
MC based tools without needing to link in the entire codegen.
First step is to refactor TargetRegisterInfo. This patch added a base class
MCRegisterInfo which TargetRegisterInfo is derived from. Changed TableGen to
separate register description from the rest of the stuff.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@133782 91177308-0d34-0410-b5e6-96231b3b80d8
1. (((x) & 0xFF00) >> 8) | (((x) & 0x00FF) << 8)
=> (bswap x) >> 16
2. ((x&0xff)<<8)|((x&0xff00)>>8)|((x&0xff000000)>>8)|((x&0x00ff0000)<<8))
=> (rotl (bswap x) 16)
This allows us to eliminate most of the def : Pat patterns for ARM rev16
revsh instructions. It catches many more cases for ARM and x86.
rdar://9609108
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@133503 91177308-0d34-0410-b5e6-96231b3b80d8
Also switch the return type to ArrayRef<unsigned> which works out nicely
for ARM's implementation of this function because of the clever ArrayRef
constructors.
The name change indicates that the returned allocation order may contain
reserved registers as has been the case for a while.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@133216 91177308-0d34-0410-b5e6-96231b3b80d8
This is intended to support using REG_SEQUENCE SDNode's with type MVT::untyped, and is part of the long road to eliminating some of the hacks we currently use to support register pairs and other strange constraints, particularly on ARM NEON.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@133178 91177308-0d34-0410-b5e6-96231b3b80d8
accumulator forwarding. Specifically (from SVN log entry):
Distribute (A + B) * C to (A * C) + (B * C) to make use of NEON multiplier
accumulator forwarding:
vadd d3, d0, d1
vmul d3, d3, d2
=>
vmul d3, d0, d2
vmla d3, d1, d2
Make sure it catches cases where operand 1 is add/fadd/sub/fsub, which was
intended in the original revision.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@133127 91177308-0d34-0410-b5e6-96231b3b80d8
This simplifies many of the target description files since it is common
for register classes to be related or contain sequences of numbered
registers.
I have verified that this doesn't change the files generated by TableGen
for ARM and X86. It alters the allocation order of MBlaze GPR and Mips
FGR32 registers, but I believe the change is benign.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@133105 91177308-0d34-0410-b5e6-96231b3b80d8
the bits being cleared by the AND are not demanded by the BFI.
The previous BFI dag combine rule was actually incorrect (or used to be
correct until BFI representation changed).
rdar://9609030
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@133034 91177308-0d34-0410-b5e6-96231b3b80d8
The logic for reserving R4 for use as a scratch needs to match that for
actually using it. Also, it's not necessary for immediate <=508, so adjust
the value checked.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132934 91177308-0d34-0410-b5e6-96231b3b80d8
causing an assertion failure downstream. This fixes <rdar://problem/9562908>.
This really seems like it should always be set at CCState creation time, so mistakes like
this can never happen. I'll take a look at doing that.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132811 91177308-0d34-0410-b5e6-96231b3b80d8
addressing mode problem mentioned in r132559.
Backend part of rdar://9037836 and part of rdar://9119939
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132561 91177308-0d34-0410-b5e6-96231b3b80d8
must be encoded decremented by one. Only add encoding tests for ssat16
because ssat can't be parsed yet.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132324 91177308-0d34-0410-b5e6-96231b3b80d8
This is important for the correct lowering of unwind instructions
(which doesn't matter at all) and llvm.eh.resume calls (which does).
Take 2, now with more basic competence.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132295 91177308-0d34-0410-b5e6-96231b3b80d8
This is important for the correct lowering of unwind instructions
(which doesn't matter at all) and llvm.eh.resume calls (which does).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132291 91177308-0d34-0410-b5e6-96231b3b80d8
to load/store i64 values. Since there's no current support to explicitly
declare such restrictions, implement it by using specific hardcoded register
pairs during isel.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132248 91177308-0d34-0410-b5e6-96231b3b80d8
register allocation dependent and will occasionally break. WIP in the
register allocator to model paired/etc registers.
rdar://9119939
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132242 91177308-0d34-0410-b5e6-96231b3b80d8
mode (only the "mov.w" variant). Now, when parsing "mov" in thumb mode,
default to the Thumb 1 versions/encodings.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132233 91177308-0d34-0410-b5e6-96231b3b80d8
The practical effects here are that x86-64 fast-isel can now handle trunc from i8 to i1, and ARM fast-isel can handle many more constructs involving integers narrower than 32 bits (including loads, stores, and many integer casts).
rdar://9437928 .
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132099 91177308-0d34-0410-b5e6-96231b3b80d8
Modified the patch to .td file supplied by Jyun-Yan You. Add a test case and
modified ARMDisassemblerCore.cpp a little bit.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@131859 91177308-0d34-0410-b5e6-96231b3b80d8
text section.
Assume the following bit of annotated assembly:
.section .data.rel.ro,"aw",%progbits
.align 2
.LAlpha:
.long startval(GOTOFF)
.text
.align 2
.type main,%function
.align 4
main: ;;; assume "main" starts at offset 0x20
0x0 push {r11, lr}
0x4 movw r0, :lower16:(.LAlpha-(.LBeta+8))
;;; ==> (.AddrOf(.LAlpha) - ((.AddrOf(.LBeta) - .AddrOf(".")) + 8)
;;; ==> (??? - ((16-4) + 8) = -20
0x8 movt r0, :upper16:(.LAlpha-(.LBeta+8))
;;; ==> (.AddrOf(.LAlpha) - ((.AddrOf(.LBeta) - .AddrOf(".")) + 8)
;;; ==> (??? - ((16-8) + 8) = -16
0xc ... blah
.LBeta:
0x10 add r0, pc, r0
0x14 ... blah
.LGamma:
0x18 add r1, pc, r1
Above snippet results in the following relocs in the .o file for the
first pair of movw/movt instructions
00000024 R_ARM_MOVW_PREL_NC .LAlpha
00000028 R_ARM_MOVT_PREL .LAlpha
And the encoded instructions in the .o file for main: must be
00000020 <main>:
20: e92d4800 push {fp, lr}
24: e30f0fec movw r0, #65516 ; 0xffec i.e. -20
28: e34f0ff0 movt r0, #65520 ; 0xfff0 i.e. -16
However, llc (prior to this commit) generates the following sequence
00000020 <main>:
20: e92d4800 push {fp, lr}
24: e30f0fec movw r0, #65516 ; 0xffec - i.e. -20
28: e34f0fff movt r0, #65535 ; 0xffff - i.e. -1
What has to happen in the ArmAsmBackend is that if the relocation is PC
relative, the 16 bits encoded as part of movw and movt must be both addends,
not addresses. It makes sense to encode addresses by right shifting the value
by 16, but the result is incorrect for PIC.
i.e., the right shift by 16 for movt is ONLY valid for the NON-PCRel case.
This change agrees with what GNU as does, and makes the PIC code run.
MC/ARM/elf-movt.s covers this case.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@131674 91177308-0d34-0410-b5e6-96231b3b80d8
("T is 1 if the target symbol S has type STT_FUNC and the
symbol addresses a Thumb instruction ;it is 0 otherwise."
from "ELF for the ARM Architecture" 4.7.1.2)
Patch by Koan-Sin Tan!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@131406 91177308-0d34-0410-b5e6-96231b3b80d8
intrinsic call. This prevents it from being reordered so that it appears
*before* the setjmp intrinsic (thus making it completely useless).
<rdar://problem/9409683>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@131174 91177308-0d34-0410-b5e6-96231b3b80d8
DWARF stuff also gets fixed up by ELFARMAsmBackend::ApplyFixup(),
but the offset is not guaranteed to be mod 4 == 0 as in text/data.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@131137 91177308-0d34-0410-b5e6-96231b3b80d8
LiveVariables doesn't understand that clobbering D0 and D1 completely overwrites
Q0, so if Q0 is live-in to a function, its live range will extend beyond a
function call that only clobbers D0 and D1. This shows up in the
ARM/2009-11-01-NeonMoves test case.
LiveVariables should probably implement the much stricter rules for physreg
liveness that RAFast imposes - a physreg is killed by the first use of any
alias.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@130801 91177308-0d34-0410-b5e6-96231b3b80d8
model constants which can be added to base registers via add-immediate
instructions which don't require an additional register to materialize
the immediate.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@130743 91177308-0d34-0410-b5e6-96231b3b80d8
Fix a rather obscure crash caused by ARM fast-isel generating code which redefines a register.
rdar://problem/9338332 .
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@130539 91177308-0d34-0410-b5e6-96231b3b80d8
The hook will be used by the register allocator when recomputing register
classes after removing constraints.
Thumb1 code doesn't allow anything larger than tGPR, and x86 needs to ensure
that the spill size doesn't change.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@130228 91177308-0d34-0410-b5e6-96231b3b80d8
Fixes Thumb2 ADCS and SBCS lowering: <rdar://problem/9275821>.
t2ADCS/t2SBCS are now pseudo instructions, consistent with ARM, so the
assembly printer correctly prints the 's' suffix.
Fixes Thumb2 adde -> SBC matching to check for live/dead carry flags.
Fixes the internal ARM machine opcode mnemonic for ADCS/SBCS.
Fixes ARM SBC lowering to check for live carry (potential bug).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@130048 91177308-0d34-0410-b5e6-96231b3b80d8
add <rd>, sp, #<imm8>
ldr <rd>, [sp, #<imm8>]
When the offset from sp is multiple of 4 and in range of 0-1020.
This saves code size by utilizing 16-bit instructions.
rdar://9321541
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@129971 91177308-0d34-0410-b5e6-96231b3b80d8
On the x86-64 and thumb2 targets, some registers are more expensive to encode
than others in the same register class.
Add a CostPerUse field to the TableGen register description, and make it
available from TRI->getCostPerUse. This represents the cost of a REX prefix or a
32-bit instruction encoding required by choosing a high register.
Teach the greedy register allocator to prefer cheap registers for busy live
ranges (as indicated by spill weight).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@129864 91177308-0d34-0410-b5e6-96231b3b80d8
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
Add a avoidWriteAfterWrite() target hook to identify register classes that
suffer from write-after-write hazards. For those register classes, try to avoid
writing the same register in two consecutive instructions.
This is currently disabled by default. We should not spill to avoid hazards!
The command line flag -avoid-waw-hazard can be used to enable waw avoidance.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@129772 91177308-0d34-0410-b5e6-96231b3b80d8
is, it assumes addresses are 64-bit aligned (which should be the more common
case). If the alignment is found not to be aligned, then getOperandLatency()
would adjust the operand latency computation by one to compensate for it.
rdar://9294833
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@129742 91177308-0d34-0410-b5e6-96231b3b80d8
The ARMARM specifies these instructions as unpredictable when storing the
writeback register. This shouldn't affect code generation much since storing a
pointer to itself is quite rare.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@129409 91177308-0d34-0410-b5e6-96231b3b80d8
In addition, the base register is not rGPR, but GPR with th exception that:
if n == 15 then UNPREDICTABLE
rdar://problem/9273836
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@129391 91177308-0d34-0410-b5e6-96231b3b80d8
disassembler API. Hooked this up to the ARM target so such tools as Darwin's
otool(1) can now print things like branch targets for example this:
blx _puts
instead of this:
blx #-36
And even print the expression encoded in the Mach-O relocation entried for
things like this:
movt r0, :upper16:((_foo-_bar)+1234)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@129284 91177308-0d34-0410-b5e6-96231b3b80d8