Commit Graph

116 Commits

Author SHA1 Message Date
Jim Grosbach
81fb6bfb85 Tidy up. Trailing whitespace.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@205583 91177308-0d34-0410-b5e6-96231b3b80d8
2014-04-03 23:43:18 +00:00
Alp Toker
ae43cab6ba Fix known typos
Sweep the codebase for common typos. Includes some changes to visible function
names that were misspelt.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@200018 91177308-0d34-0410-b5e6-96231b3b80d8
2014-01-24 17:20:08 +00:00
Artyom Skrobov
ef572e31e2 [ARM] Add support for MVFR2 which is new in ARMv8
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@194416 91177308-0d34-0410-b5e6-96231b3b80d8
2013-11-11 19:56:13 +00:00
Tilmann Scheller
8f3d54d057 Reverting 190043 for now.
Solution is not sufficient to prevent 'mov pc, lr' being emitted for jump table code.
Test case doesn't trigger the added functionality.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@190047 91177308-0d34-0410-b5e6-96231b3b80d8
2013-09-05 11:59:43 +00:00
Tilmann Scheller
10b5086e6e ARM: Add GPR register class excluding LR for use with the ADR instruction.
This improves code generation for jump tables by avoiding the emission of "mov pc, lr" which could fool the processor into believing this is a return from a function causing mispredicts. The code generation logic for jump tables uses ADR to materialize the address of the jump target.

Patch by Daniel Stewart!
   


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@190043 91177308-0d34-0410-b5e6-96231b3b80d8
2013-09-05 11:10:31 +00:00
Jim Grosbach
e2a48fbd9d ARM: R9 is not safe to use for tcGPR.
Indirect tail-calls shouldn't use R9 for the branch destination, as
it's not reliably a call-clobbered register.

rdar://14793425

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@188967 91177308-0d34-0410-b5e6-96231b3b80d8
2013-08-22 00:14:24 +00:00
Mihai Popa
4a378b95aa Support APSR_nzcv as operand for Thumb2 mrc. Deprecate pre-UAL syntax (pc instead of apsr_nzcv)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@187803 91177308-0d34-0410-b5e6-96231b3b80d8
2013-08-06 15:52:36 +00:00
Mihai Popa
16ad92ad3c This patch adds support for FPINST/FPINST2 as operands to vmsr/vmrs. These are optional registers that may be supported some ARM implementations to aid with resolution of floating point exceptions. The manual pages for vmsr and vmrs do not detail their use. Encodings and other information can be found in ARM Architecture Reference Manual section F, chapter 6, paragraph 3.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@183733 91177308-0d34-0410-b5e6-96231b3b80d8
2013-06-11 09:39:51 +00:00
Ahmed Bougacha
bed2308186 Add a way to define the bit range covered by a SubRegIndex.
NOTE: If this broke your out-of-tree backend, in *RegisterInfo.td, change
the instances of SubRegIndex that have a comps template arg to use the
ComposedSubRegIndex class instead.

In TableGen land, this adds Size and Offset attributes to SubRegIndex,
and the ComposedSubRegIndex class, for which the Size and Offset are
computed by TableGen. This also adds an accessor in MCRegisterInfo, and
Size/Offsets for the X86 and ARM subreg indices.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@183020 91177308-0d34-0410-b5e6-96231b3b80d8
2013-05-31 17:08:36 +00:00
Mihai Popa
f86e436fb9 The purpose of the patch is to fix the syntax of ARM mrc and mrc2 instructions when they are used to write to the APSR. In this case, the destination operand should be APSR_nzcv, and the encoding of the target should be 0b1111 (same as for PC). In pre-UAL syntax, this form used the PC register as a textual target. This is still allowed for backward compatibility.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@181705 91177308-0d34-0410-b5e6-96231b3b80d8
2013-05-13 14:10:04 +00:00
Jakob Stoklund Olesen
cd275f5687 Add GPRPair Register class to ARM.
Some instructions in ARM require 2 even-odd paired GPRs. This
patch adds support for such register class.

Patch by Weiming Zhao!

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@166816 91177308-0d34-0410-b5e6-96231b3b80d8
2012-10-26 21:29:15 +00:00
Bob Wilson
eb1641d54a Add LLVM support for Swift.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164899 91177308-0d34-0410-b5e6-96231b3b80d8
2012-09-29 21:43:49 +00:00
Eric Christopher
529f314050 This field isn't used anymore, use it with HWEncoding instead.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@161564 91177308-0d34-0410-b5e6-96231b3b80d8
2012-08-09 01:39:32 +00:00
Jakob Stoklund Olesen
338607ae0d Remove the SubRegClasses field from RegisterClass descriptions.
This information in now computed by TableGen.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@156152 91177308-0d34-0410-b5e6-96231b3b80d8
2012-05-04 03:30:34 +00:00
Jakob Stoklund Olesen
ccca22e9e2 Prefer even-odd D-register pairs.
We are sometimes allocatinog from the DPair register class which
contains odd-even pairs in addition to the Q registers.

Place the Q registers first in the DPair allocation order as they can be
copied with a single instruction. The odd-even pairs should only be
allocated as a last resort.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153699 91177308-0d34-0410-b5e6-96231b3b80d8
2012-03-29 22:54:32 +00:00
Jim Grosbach
9426ac7b57 ARM vmrs system registers mvfr0 and mvfr1 handling.
rdar://11058464

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@152881 91177308-0d34-0410-b5e6-96231b3b80d8
2012-03-16 00:27:18 +00:00
Jakob Stoklund Olesen
14f87e01ca Allow the same types in DPair as in QPR.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@152129 91177308-0d34-0410-b5e6-96231b3b80d8
2012-03-06 18:44:11 +00:00
Lang Hames
4f92b5e616 Split fpscr into two registers: FPSCR and FPSCR_NZCV.
The fpscr register contains both flags (set by FP operations/comparisons) and
control bits. The control bits (FPSCR) should be reserved, since they're always
available and needn't be defined before use. The flag bits (FPSCR_NZCV) should
like to be unreserved so they can be hoisted by MachineCSE. This fixes PR12165.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@152076 91177308-0d34-0410-b5e6-96231b3b80d8
2012-03-06 00:19:55 +00:00
Jim Grosbach
28f08c93e7 ARM refactor away a bunch of VLD/VST pseudo instructions.
With the new composite physical registers to represent arbitrary pairs
of DPR registers, we don't need the pseudo-registers anymore. Get rid of
a bunch of them that use DPR register pairs and just use the real
instructions directly instead.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@152045 91177308-0d34-0410-b5e6-96231b3b80d8
2012-03-05 19:33:30 +00:00
Jia Liu
31d157ae1a Emacs-tag and some comment fix for all ARM, CellSPU, Hexagon, MBlaze, MSP430, PPC, PTX, Sparc, X86, XCore.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@150878 91177308-0d34-0410-b5e6-96231b3b80d8
2012-02-18 12:03:15 +00:00
Jakob Stoklund Olesen
42865588d7 Add pseudo-registers for pairs, triples, and quads of D registers.
NEON loads and stores accept single and double spaced pairs, triples,
and quads of D registers.  This patch adds new register classes to
accurately model those constraints:

  Dn, Dn+1    Dn, Dn+2
  ----------------------
  DPair       DPairSpc
  DTriple     DTripleSpc
  DQuad       DQuadSpc

Also extend the existing QQ and QQQQ register classes to contains all Q
pairs and quads instead of just the aligned ones.

These new register classes will make it possible to accurately model
constraints on NEON loads and stores, and we can get rid of all the NEON
pseudo-instructions.  The late scheduler will be able to accurately
model instruction dependencies from the explicit operands.

This more than doubles the number of ARM registers, but the backend
passes are quite good at handling this. The llc -O0 compile time only
regresses by 1.5%.  Future work on register mask operands will recover
this regression.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@149640 91177308-0d34-0410-b5e6-96231b3b80d8
2012-02-02 22:45:32 +00:00
Jakob Stoklund Olesen
521804a1f7 Move ARM subreg index compositions to the SubRegIndex itself.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@149557 91177308-0d34-0410-b5e6-96231b3b80d8
2012-02-01 23:16:43 +00:00
Jakob Stoklund Olesen
31867660cb Add a CoveredBySubRegs property to Register descriptions.
When set, this bit indicates that a register is completely defined by
the value of its sub-registers.

Use the CoveredBySubRegs property to infer which super-registers are
call-preserved given a list of callee-saved registers.  For example, the
ARM registers D8-D15 are callee-saved.  This now automatically implies
that Q4-Q7 are call-preserved.

Conversely, Win64 callees save XMM6-XMM15, but the corresponding
YMM6-YMM15 registers are not call-preserved because they are not fully
defined by their sub-registers.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@148363 91177308-0d34-0410-b5e6-96231b3b80d8
2012-01-18 00:16:39 +00:00
Jakob Stoklund Olesen
ece8b73eb2 Use RegisterTuples to generate pseudo-registers.
The QQ and QQQQ registers are not 'real', they are pseudo-registers used
to model some vld and vst instructions.

This makes the call clobber lists longer, but I intend to get rid of
those soon.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@148151 91177308-0d34-0410-b5e6-96231b3b80d8
2012-01-13 22:55:42 +00:00
Jakob Stoklund Olesen
9897c622e0 Remove a register class that can just as well be synthesized.
Add the new TableGen register class synthesizer feature to the release
notes.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@146875 91177308-0d34-0410-b5e6-96231b3b80d8
2011-12-19 16:53:40 +00:00
Evan Cheng
342e3161d9 Change ARM / Thumb2 addc / adde and subc / sube modeling to use physical
register dependency (rather than glue them together). This is general
goodness as it gives scheduler more freedom. However it is motivated by
a nasty bug in isel.

When a i64 sub is expanded to subc + sube.
  libcall #1
     \
      \        subc 
       \       /  \
        \     /    \
         \   /    libcall #2
          sube

If the libcalls are not serialized (i.e. both have chains which are dag
entry), legalizer can serialize them in arbitrary orders. If it's
unlucky, it can force libcall #2 before libcall #1 in the above case.

  subc
   |
  libcall #2
   |
  libcall #1
   |
  sube

However since subc and sube are "glued" together, this ends up being a
cycle when the scheduler combine subc and sube as a single scheduling
unit.

The right solution is to fix LegalizeType too chains the libcalls together.
However, LegalizeType is not processing nodes in order so that's harder than
it should be. For now, the move to physical register dependency will do.

rdar://10019576


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@138791 91177308-0d34-0410-b5e6-96231b3b80d8
2011-08-30 01:34:54 +00:00
Jim Grosbach
5b81584f74 Thumb1 ADD/SUB SP instructions are predicable in Thumb2 mode.
Add the predicate operand to the instructions. Update the back end
accordingly where the instructions are used. Restrict the SP operands
to actually only be SP, as otherwise these break assembly parsing for the
normal instruction variants.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@138445 91177308-0d34-0410-b5e6-96231b3b80d8
2011-08-24 17:46:13 +00:00
Owen Anderson
51c9805c4b Create a new register class for the set of all GPRs except the PC. Use it to tighten our decoding of BFI.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@137168 91177308-0d34-0410-b5e6-96231b3b80d8
2011-08-09 22:48:45 +00:00
Jim Grosbach
80d01dd3d1 ARM assembly parsing of MRS instruction.
Teach the parser to recognize the APSR and SPSR system register names. Add
and update tests accordingly.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@135527 91177308-0d34-0410-b5e6-96231b3b80d8
2011-07-19 21:59:29 +00:00
Eric Christopher
73744df0c4 Add support for the 'h' constraint.
Part of rdar://9119939


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134203 91177308-0d34-0410-b5e6-96231b3b80d8
2011-06-30 23:23:01 +00:00
Jakob Stoklund Olesen
0a074ed3ef Switch ARM to using AltOrders instead of MethodBodies.
This slightly changes the GPR allocation order on Darwin where R9 is not
a callee-saved register:

Before: %R0 %R1 %R2 %R3 %R12 %R9 %LR %R4 %R5 %R6 %R8 %R10 %R11
After:  %R0 %R1 %R2 %R3 %R9 %R12 %LR %R4 %R5 %R6 %R8 %R10 %R11

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@133326 91177308-0d34-0410-b5e6-96231b3b80d8
2011-06-18 01:14:46 +00:00
Jakob Stoklund Olesen
f28987b76e Use set operations instead of plain lists to enumerate register classes.
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
2011-06-15 23:28:14 +00:00
Jakob Stoklund Olesen
4f3fb6d08b Flag unallocatable register classes instead of giving them empty
allocation orders.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132509 91177308-0d34-0410-b5e6-96231b3b80d8
2011-06-02 23:07:24 +00:00
Jakob Stoklund Olesen
ca66226e7e Eliminate the ARM sub-register indexes that are not needed by the sources.
Tablegen will invent its own names for these indexes, and the register file is a
bit simpler.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@131059 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-07 21:22:42 +00:00
Devang Patel
64d3d12f37 As per ARM docs, register Dx is described as DW_OP_regx(256+x) in DWARF.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@129922 91177308-0d34-0410-b5e6-96231b3b80d8
2011-04-21 17:51:06 +00:00
Jakob Stoklund Olesen
6bfba2e5af Prefer cheap registers for busy live ranges.
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
2011-04-20 18:19:48 +00:00
Evan Cheng
9fe2009956 Sorry, several patches in one.
TargetInstrInfo:
Change produceSameValue() to take MachineRegisterInfo as an optional argument.
When in SSA form, targets can use it to make more aggressive equality analysis.

Machine LICM:
1. Eliminate isLoadFromConstantMemory, use MI.isInvariantLoad instead.
2. Fix a bug which prevent CSE of instructions which are not re-materializable.
3. Use improved form of produceSameValue.

ARM:
1. Teach ARM produceSameValue to look pass some PIC labels.
2. Look for operands from different loads of different constant pool entries
   which have same values.
3. Re-implement PIC GA materialization using movw + movt. Combine the pair with
   a "add pc" or "ldr [pc]" to form pseudo instructions. This makes it possible
   to re-materialize the instruction, allow machine LICM to hoist the set of
   instructions out of the loop and make it possible to CSE them. It's a bit
   hacky, but it significantly improve code quality.
4. Some minor bug fixes as well.

With the fixes, using movw + movt to materialize GAs significantly outperform the
load from constantpool method. 186.crafty and 255.vortex improved > 20%, 254.gap
and 176.gcc ~10%.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@123905 91177308-0d34-0410-b5e6-96231b3b80d8
2011-01-20 08:34:58 +00:00
Bruno Cardoso Lopes
61505907f5 Create two new generic classes to represent the following VMRS/VMSR variations:
vmrs  reg, fpexc
vmrs  reg, fpsid
vmsr  fpexc, reg
vmsr  fpsid, reg



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@123783 91177308-0d34-0410-b5e6-96231b3b80d8
2011-01-18 21:58:20 +00:00
Bob Wilson
77f42b5278 PR8359: The ARM backend may end up allocating registers D16 to D31 when
"-mattr=+vfp3" is specified. However, this will not work for hardware that
only supports 16 registers.  Add a new flag to support -"mattr=+vfp3,+d16".
Patch by Jan Voung!


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@116310 91177308-0d34-0410-b5e6-96231b3b80d8
2010-10-12 16:22:47 +00:00
Bob Wilson
7d24705f65 Change register allocation order for ARM VFP and NEON registers to put the
callee-saved registers at the end of the lists.  Also prefer to avoid using
the low registers that are in register subclasses required by certain
instructions, so that those registers will more likely be available when needed.
This change makes a huge improvement in spilling in some cases.  Thanks to
Jakob for helping me realize the problem.

Most of this patch is fixing the testsuite.  There are quite a few places
where we're checking for specific registers.  I changed those to wildcards
in places where that doesn't weaken the tests.  The spill-q.ll and
thumb2-spill-q.ll tests stopped spilling with this change, so I added a bunch
of live values to force spills on those tests.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@116055 91177308-0d34-0410-b5e6-96231b3b80d8
2010-10-08 06:15:13 +00:00
Jim Grosbach
e7c1416263 Now that register allocation properly considers reserved regs, simplify the
ARM register class allocation order functions to take advantage of that.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@112841 91177308-0d34-0410-b5e6-96231b3b80d8
2010-09-02 18:14:29 +00:00
Jim Grosbach
7af3a345a9 trivial cleanup
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@112779 91177308-0d34-0410-b5e6-96231b3b80d8
2010-09-02 00:02:26 +00:00
Jim Grosbach
352f23529c Simplify the tGPR register class now that the register allocators know not
to try to allocate reserved registers.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@112774 91177308-0d34-0410-b5e6-96231b3b80d8
2010-09-01 23:50:23 +00:00
Chris Lattner
23e70ebf35 fix emacs language spec's, patch by Edmund Grimley-Evans!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@111241 91177308-0d34-0410-b5e6-96231b3b80d8
2010-08-17 16:20:04 +00:00
Evan Cheng
ac096808a3 Re-apply r110655 with fixes. Epilogue must restore sp from fp if the function stack frame has a var-sized object.
Also added a test case to check for the added benefit of this patch: it's optimizing away the unnecessary restore of sp from fp for some non-leaf functions.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@110707 91177308-0d34-0410-b5e6-96231b3b80d8
2010-08-10 19:30:19 +00:00
Daniel Dunbar
4bd828f781 Revert r110655, "Fix ARM hasFP() semantics. It should return true whenever FP
register is", it breaks a couple test-suite tests.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@110701 91177308-0d34-0410-b5e6-96231b3b80d8
2010-08-10 18:32:02 +00:00
Evan Cheng
c9aed19747 Fix ARM hasFP() semantics. It should return true whenever FP register is
reserved, not available for general allocation. This eliminates all the
extra checks for Darwin.

This change also fixes the use of FP to access frame indices in leaf
functions and cleaned up some confusing code in epilogue emission.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@110655 91177308-0d34-0410-b5e6-96231b3b80d8
2010-08-10 06:26:49 +00:00
Jim Grosbach
6ccfc507dc Many Thumb2 instructions can reference the full ARM register set (i.e.,
have 4 bits per register in the operand encoding), but have undefined
behavior when the operand value is 13 or 15 (SP and PC, respectively).
The trivial coalescer in linear scan sometimes will merge a copy from
SP into a subsequent instruction which uses the copy, and if that
instruction cannot legally reference SP, we get bad code such as:
  mls r0,r9,r0,sp
instead of:
  mov r2, sp
  mls r0, r9, r0, r2

This patch adds a new register class for use by Thumb2 that excludes
the problematic registers (SP and PC) and is used instead of GPR
for those operands which cannot legally reference PC or SP. The
trivial coalescer explicitly requires that the register class
of the destination for the COPY instruction contain the source
register for the COPY to be considered for coalescing. This prevents
errant instructions like that above.

PR7499




git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@109842 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-30 02:41:01 +00:00
Bob Wilson
8af4c54af1 Clean up a comment.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@107882 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-08 16:54:45 +00:00
Dale Johannesen
b0ccb757b3 Fix PR 7433. Silly typo in non-Darwin ARM tail call
handling, plus correct R9 handling in that mode.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106434 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-21 18:21:49 +00:00