| Commit message (Collapse) | Author | Age | Files | Lines |
... | |
|
|
|
| |
llvm-svn: 129096
|
|
|
|
|
|
| |
And two test cases.
llvm-svn: 129090
|
|
|
|
|
|
| |
match.
llvm-svn: 129078
|
|
|
|
|
|
| |
vector type (vectors of size 3). Also included test cases.
llvm-svn: 129074
|
|
|
|
|
|
| |
rdar://problem/9246844
llvm-svn: 129050
|
|
|
|
|
|
|
|
|
|
|
| |
is equivalent to any other relevant value; it isn't true in general.
If it is equivalent, the LoopPromoter will tell the AST the equivalence.
Also, delete the PreheaderLoad if it is unused.
Chris, since you were the last one to make major changes here, can you check
that this is sane?
llvm-svn: 129049
|
|
|
|
|
|
|
|
|
|
| |
checking for register values
for USAD8 and USADA8.
rdar://problem/9247060
llvm-svn: 129047
|
|
|
|
| |
llvm-svn: 129045
|
|
|
|
|
|
| |
rdar://problem/9246650
llvm-svn: 129042
|
|
|
|
|
|
| |
folded comparisons, just like ADD and SUB.
llvm-svn: 129038
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The ARM disassembler should reject invalid (type, align) encodings as invalid instructions.
So, instead of:
Opcode=1641 Name=VST2b32_UPD Format=ARM_FORMAT_NLdSt(30)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
-------------------------------------------------------------------------------------------------
| 1: 1: 1: 1| 0: 1: 0: 0| 0: 0: 0: 0| 0: 0: 1: 1| 0: 0: 0: 0| 1: 0: 0: 1| 1: 0: 1: 1| 0: 0: 1: 1|
-------------------------------------------------------------------------------------------------
vst2.32 {d0, d2}, [r3, :256], r3
we now have:
Opcode=1641 Name=VST2b32_UPD Format=ARM_FORMAT_NLdSt(30)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
-------------------------------------------------------------------------------------------------
| 1: 1: 1: 1| 0: 1: 0: 0| 0: 0: 0: 0| 0: 0: 1: 1| 0: 0: 0: 0| 1: 0: 0: 1| 1: 0: 1: 1| 0: 0: 1: 1|
-------------------------------------------------------------------------------------------------
mc-input.txt:1:1: warning: invalid instruction encoding
0xb3 0x9 0x3 0xf4
^
llvm-svn: 129033
|
|
|
|
|
|
|
|
|
|
| |
Since these "Advanced SIMD and VFP" instructions have more specfic encoding bits
specified, if coproc == 10 or 11, we should reject the insn as invalid.
rdar://problem/9239922
rdar://problem/9239596
llvm-svn: 129027
|
|
|
|
|
|
|
|
|
| |
Also set the encoding bits (for A8.6.303, A8.6.328, A8.6.329) Inst{3-0} = 0b0000,
in class NVLaneOp.
rdar://problem/9240648
llvm-svn: 129015
|
|
|
|
|
|
| |
test fail (without the fix). Thanks Dan.
llvm-svn: 128999
|
|
|
|
|
|
|
|
|
|
| |
Add more complete sanity check for LdStFrm instructions where if IBit (Inst{25})
is 1, Inst{4} should be 0. Otherwise, we should reject the insn as invalid.
rdar://problem/9239347
rdar://problem/9239467
llvm-svn: 128977
|
|
|
|
|
|
|
|
|
|
| |
illegal register
encodings for DisassembleArithMiscFrm().
rdar://problem/9238659
llvm-svn: 128958
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Qd -> bit[12] == 0
Qn -> bit[16] == 0
Qm -> bit[0] == 0
If one of these bits is 1, the instruction is UNDEFINED.
rdar://problem/9238399
rdar://problem/9238445
llvm-svn: 128949
|
|
|
|
|
|
|
|
| |
Added checks for regs which should not be 15.
rdar://problem/9237734
llvm-svn: 128945
|
|
|
|
|
|
| |
still used by RegionInfo :(
llvm-svn: 128943
|
|
|
|
|
|
|
|
|
| |
For register-controlled shifts, we should check that the encoding constraint
Inst{7} = 0 and Inst{4} = 1 is satisfied.
rdar://problem/9237693
llvm-svn: 128941
|
|
|
|
|
|
| |
coalescing is now universal.
llvm-svn: 128936
|
|
|
|
| |
llvm-svn: 128935
|
|
|
|
|
|
|
|
| |
Modify DisassembleCoprocessor() of ARMDisassemblerCore.cpp to react to the change.
rdar://problem/9236873
llvm-svn: 128922
|
|
|
|
| |
llvm-svn: 128919
|
|
|
|
|
|
|
|
| |
def into the remaining use.
Rematerialization can leave single-use loads behind that we might as well fold whenever possible.
llvm-svn: 128918
|
|
|
|
| |
llvm-svn: 128913
|
|
|
|
| |
llvm-svn: 128909
|
|
|
|
| |
llvm-svn: 128906
|
|
|
|
|
|
|
|
|
|
| |
An alternative syntax is available for a modified immediate constant that permits the programmer to specify
the encoding directly. In this syntax, #<const> is instead written as #<byte>,#<rot>, where:
<byte> is the numeric value of abcdefgh, in the range 0-255
<rot> is twice the numeric value of rotation, an even number in the range 0-30.
llvm-svn: 128897
|
|
|
|
|
|
|
|
|
| |
if dLo == 15 || dHi == 15 || n == 15 || m == 15 then UNPREDICTABLE;
if dHi == dLo then UNPREDICTABLE;
rdar://problem/9230202
llvm-svn: 128895
|
|
|
|
| |
llvm-svn: 128891
|
|
|
|
|
|
|
|
|
|
|
|
| |
addRegisterDead().
There can be multiple defs for a single virtual register when they are defining
sub-registers.
The missing <dead> flag was stopping the inline spiller from eliminating dead
code after rematerialization.
llvm-svn: 128888
|
|
|
|
| |
llvm-svn: 128887
|
|
|
|
|
|
|
| |
space info. We crash with an assert in this case. This change checks that the
address space of the bitcasted pointer is the same as the gep ptr.
llvm-svn: 128884
|
|
|
|
| |
llvm-svn: 128870
|
|
|
|
| |
llvm-svn: 128867
|
|
|
|
|
|
|
|
|
| |
rdar://problem/9230801 ARM disassembler discrepancy: erroneously accepting SRS
Plus add invalid-RFEorLDMIA-arm.txt test which should have been checked in with
http://llvm.org/viewvc/llvm-project?view=rev&revision=128859.
llvm-svn: 128864
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
When a virtual register has a single value that is defined as a copy of a
reserved register, permit that copy to be joined. These virtual register are
usually copies of the stack pointer:
%vreg75<def> = COPY %ESP; GR32:%vreg75
MOV32mr %vreg75, 1, %noreg, 0, %noreg, %vreg74<kill>
MOV32mi %vreg75, 1, %noreg, 8, %noreg, 0
MOV32mi %vreg75<kill>, 1, %noreg, 4, %noreg, 0
CALLpcrel32 ...
Coalescing these virtual registers early decreases register pressure.
Previously, they were coalesced by RALinScan::attemptTrivialCoalescing after
register allocation was completed.
The lower register pressure causes the mcinst-lowering-cmp0.ll test case to fail
because it depends on linear scan spilling a particular register.
I am deleting 2008-08-05-SpillerBug.ll because it is counting the number of
instructions emitted, and its revision history shows the 'correct' count being
edited many times.
llvm-svn: 128845
|
|
|
|
|
|
| |
rdar://problem/9225433
llvm-svn: 128841
|
|
|
|
|
|
|
|
| |
The code inserted by PPCTargetLowering::EmitInstrWithCustomInserter for ppc64 is
wrong, and I don't know how to fix it. It seems to be using the correct register
classes for pointers, but it inserts all 32-bit instructions.
llvm-svn: 128835
|
|
|
|
|
|
|
|
|
|
| |
also fix the encoding of the later.
- Add a new encoding bit to describe the index mode used in AM3.
- Teach printAddrMode3Operand to check by the addressing mode which
index mode to print.
- Testcases.
llvm-svn: 128832
|
|
|
|
| |
llvm-svn: 128827
|
|
|
|
| |
llvm-svn: 128826
|
|
|
|
|
|
|
| |
returning a scalar value in a function whose return type is a single-
element structure or array.
llvm-svn: 128810
|
|
|
|
|
|
|
|
|
|
|
| |
after the given instruction; make sure to handle that case correctly.
(It's difficult to trigger; the included testcase involves a dead
block, but I don't think that's a requirement.)
While I'm here, get rid of the unnecessary warning about
SimplifyInstructionsInBlock, since it should work correctly as far as I know.
llvm-svn: 128782
|
|
|
|
| |
llvm-svn: 128767
|
|
|
|
|
|
|
|
| |
registers that arise from argument shuffling with the soft float ABI. These
instructions are particularly slow on Cortex A8. This fixes one half of
<rdar://problem/8674845>.
llvm-svn: 128759
|
|
|
|
|
|
|
|
|
|
| |
operand in am2offset;
instead of the second operand in addrmode_imm12.
rdar://problem/9225289
llvm-svn: 128757
|
|
|
|
|
|
| |
rdar://problem/9224276
llvm-svn: 128749
|
|
|
|
|
|
|
|
| |
UNPREDICTABLE.
rdar://problem/9224120
llvm-svn: 128748
|