reference, declarationdefinition
definition → references, declarations, derived classes, virtual overrides
reference to multiple definitions → definitions
unreferenced
    1
    2
    3
    4
    5
    6
    7
    8
    9
   10
   11
   12
   13
   14
   15
   16
   17
   18
   19
   20
   21
   22
   23
   24
   25
   26
   27
   28
   29
   30
   31
   32
   33
   34
   35
   36
   37
   38
   39
   40
   41
   42
   43
   44
   45
   46
   47
   48
   49
   50
   51
   52
   53
   54
   55
   56
   57
   58
   59
   60
   61
   62
   63
   64
   65
   66
   67
   68
   69
   70
   71
   72
   73
   74
   75
   76
   77
   78
   79
   80
   81
   82
   83
   84
   85
   86
   87
   88
   89
   90
   91
   92
   93
   94
   95
   96
   97
   98
   99
  100
  101
  102
  103
  104
  105
  106
  107
  108
  109
  110
  111
  112
  113
  114
  115
  116
  117
  118
  119
  120
  121
  122
  123
  124
  125
  126
  127
  128
  129
  130
  131
  132
  133
  134
  135
  136
  137
  138
  139
  140
  141
  142
  143
  144
  145
  146
  147
  148
  149
  150
  151
  152
  153
  154
  155
  156
  157
  158
  159
  160
  161
  162
  163
  164
  165
  166
  167
  168
  169
  170
  171
  172
  173
  174
  175
  176
  177
  178
  179
  180
  181
  182
  183
  184
  185
  186
  187
  188
  189
  190
  191
  192
  193
  194
  195
  196
  197
  198
  199
  200
  201
  202
  203
  204
  205
  206
  207
  208
  209
  210
  211
  212
  213
  214
  215
  216
  217
  218
  219
  220
  221
  222
  223
  224
  225
  226
  227
  228
  229
  230
  231
  232
  233
  234
  235
  236
  237
  238
  239
  240
  241
  242
  243
  244
  245
  246
  247
  248
  249
  250
  251
  252
  253
  254
  255
  256
  257
  258
  259
  260
  261
  262
  263
  264
  265
  266
  267
  268
  269
  270
  271
  272
  273
  274
  275
  276
  277
  278
  279
  280
  281
  282
  283
  284
  285
  286
  287
  288
  289
  290
  291
  292
  293
  294
  295
  296
  297
  298
  299
  300
  301
  302
  303
  304
  305
  306
  307
  308
  309
  310
  311
  312
  313
  314
  315
  316
  317
  318
  319
  320
  321
  322
  323
  324
  325
  326
  327
  328
  329
  330
  331
  332
  333
  334
  335
  336
  337
  338
  339
  340
  341
  342
  343
  344
  345
  346
  347
  348
  349
  350
  351
  352
  353
  354
  355
  356
  357
  358
  359
  360
  361
  362
  363
  364
  365
  366
  367
  368
  369
  370
  371
  372
  373
  374
  375
  376
  377
  378
  379
  380
  381
  382
  383
  384
  385
  386
  387
  388
  389
  390
  391
  392
  393
  394
  395
  396
  397
  398
  399
  400
  401
  402
  403
  404
  405
  406
  407
  408
  409
  410
  411
  412
  413
  414
  415
  416
  417
  418
  419
  420
  421
  422
  423
  424
  425
  426
  427
  428
  429
  430
  431
  432
  433
  434
  435
  436
  437
  438
  439
==========================
Auto-Vectorization in LLVM
==========================

.. contents::
   :local:

LLVM has two vectorizers: The :ref:`Loop Vectorizer <loop-vectorizer>`,
which operates on Loops, and the :ref:`SLP Vectorizer
<slp-vectorizer>`. These vectorizers
focus on different optimization opportunities and use different techniques.
The SLP vectorizer merges multiple scalars that are found in the code into
vectors while the Loop Vectorizer widens instructions in loops
to operate on multiple consecutive iterations.

Both the Loop Vectorizer and the SLP Vectorizer are enabled by default.

.. _loop-vectorizer:

The Loop Vectorizer
===================

Usage
-----

The Loop Vectorizer is enabled by default, but it can be disabled
through clang using the command line flag:

.. code-block:: console

   $ clang ... -fno-vectorize  file.c

Command line flags
^^^^^^^^^^^^^^^^^^

The loop vectorizer uses a cost model to decide on the optimal vectorization factor
and unroll factor. However, users of the vectorizer can force the vectorizer to use
specific values. Both 'clang' and 'opt' support the flags below.

Users can control the vectorization SIMD width using the command line flag "-force-vector-width".

.. code-block:: console

  $ clang  -mllvm -force-vector-width=8 ...
  $ opt -loop-vectorize -force-vector-width=8 ...

Users can control the unroll factor using the command line flag "-force-vector-interleave"

.. code-block:: console

  $ clang  -mllvm -force-vector-interleave=2 ...
  $ opt -loop-vectorize -force-vector-interleave=2 ...

Pragma loop hint directives
^^^^^^^^^^^^^^^^^^^^^^^^^^^

The ``#pragma clang loop`` directive allows loop vectorization hints to be
specified for the subsequent for, while, do-while, or c++11 range-based for
loop. The directive allows vectorization and interleaving to be enabled or
disabled. Vector width as well as interleave count can also be manually
specified. The following example explicitly enables vectorization and
interleaving:

.. code-block:: c++

  #pragma clang loop vectorize(enable) interleave(enable)
  while(...) {
    ...
  }

The following example implicitly enables vectorization and interleaving by
specifying a vector width and interleaving count:

.. code-block:: c++

  #pragma clang loop vectorize_width(2) interleave_count(2)
  for(...) {
    ...
  }

See the Clang
`language extensions
<http://clang.llvm.org/docs/LanguageExtensions.html#extensions-for-loop-hint-optimizations>`_
for details.

Diagnostics
-----------

Many loops cannot be vectorized including loops with complicated control flow,
unvectorizable types, and unvectorizable calls. The loop vectorizer generates
optimization remarks which can be queried using command line options to identify
and diagnose loops that are skipped by the loop-vectorizer.

Optimization remarks are enabled using:

``-Rpass=loop-vectorize`` identifies loops that were successfully vectorized.

``-Rpass-missed=loop-vectorize`` identifies loops that failed vectorization and
indicates if vectorization was specified.

``-Rpass-analysis=loop-vectorize`` identifies the statements that caused
vectorization to fail. If in addition ``-fsave-optimization-record`` is
provided, multiple causes of vectorization failure may be listed (this behavior
might change in the future).

Consider the following loop:

.. code-block:: c++

  #pragma clang loop vectorize(enable)
  for (int i = 0; i < Length; i++) {
    switch(A[i]) {
    case 0: A[i] = i*2; break;
    case 1: A[i] = i;   break;
    default: A[i] = 0;
    }
  }

The command line ``-Rpass-missed=loop-vectorized`` prints the remark:

.. code-block:: console

  no_switch.cpp:4:5: remark: loop not vectorized: vectorization is explicitly enabled [-Rpass-missed=loop-vectorize]

And the command line ``-Rpass-analysis=loop-vectorize`` indicates that the
switch statement cannot be vectorized.

.. code-block:: console

  no_switch.cpp:4:5: remark: loop not vectorized: loop contains a switch statement [-Rpass-analysis=loop-vectorize]
    switch(A[i]) {
    ^

To ensure line and column numbers are produced include the command line options
``-gline-tables-only`` and ``-gcolumn-info``. See the Clang `user manual
<http://clang.llvm.org/docs/UsersManual.html#options-to-emit-optimization-reports>`_
for details

Features
--------

The LLVM Loop Vectorizer has a number of features that allow it to vectorize
complex loops.

Loops with unknown trip count
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

The Loop Vectorizer supports loops with an unknown trip count.
In the loop below, the iteration ``start`` and ``finish`` points are unknown,
and the Loop Vectorizer has a mechanism to vectorize loops that do not start
at zero. In this example, 'n' may not be a multiple of the vector width, and
the vectorizer has to execute the last few iterations as scalar code. Keeping
a scalar copy of the loop increases the code size.

.. code-block:: c++

  void bar(float *A, float* B, float K, int start, int end) {
    for (int i = start; i < end; ++i)
      A[i] *= B[i] + K;
  }

Runtime Checks of Pointers
^^^^^^^^^^^^^^^^^^^^^^^^^^

In the example below, if the pointers A and B point to consecutive addresses,
then it is illegal to vectorize the code because some elements of A will be
written before they are read from array B.

Some programmers use the 'restrict' keyword to notify the compiler that the
pointers are disjointed, but in our example, the Loop Vectorizer has no way of
knowing that the pointers A and B are unique. The Loop Vectorizer handles this
loop by placing code that checks, at runtime, if the arrays A and B point to
disjointed memory locations. If arrays A and B overlap, then the scalar version
of the loop is executed.

.. code-block:: c++

  void bar(float *A, float* B, float K, int n) {
    for (int i = 0; i < n; ++i)
      A[i] *= B[i] + K;
  }


Reductions
^^^^^^^^^^

In this example the ``sum`` variable is used by consecutive iterations of
the loop. Normally, this would prevent vectorization, but the vectorizer can
detect that 'sum' is a reduction variable. The variable 'sum' becomes a vector
of integers, and at the end of the loop the elements of the array are added
together to create the correct result. We support a number of different
reduction operations, such as addition, multiplication, XOR, AND and OR.

.. code-block:: c++

  int foo(int *A, int *B, int n) {
    unsigned sum = 0;
    for (int i = 0; i < n; ++i)
      sum += A[i] + 5;
    return sum;
  }

We support floating point reduction operations when `-ffast-math` is used.

Inductions
^^^^^^^^^^

In this example the value of the induction variable ``i`` is saved into an
array. The Loop Vectorizer knows to vectorize induction variables.

.. code-block:: c++

  void bar(float *A, float* B, float K, int n) {
    for (int i = 0; i < n; ++i)
      A[i] = i;
  }

If Conversion
^^^^^^^^^^^^^

The Loop Vectorizer is able to "flatten" the IF statement in the code and
generate a single stream of instructions. The Loop Vectorizer supports any
control flow in the innermost loop. The innermost loop may contain complex
nesting of IFs, ELSEs and even GOTOs.

.. code-block:: c++

  int foo(int *A, int *B, int n) {
    unsigned sum = 0;
    for (int i = 0; i < n; ++i)
      if (A[i] > B[i])
        sum += A[i] + 5;
    return sum;
  }

Pointer Induction Variables
^^^^^^^^^^^^^^^^^^^^^^^^^^^

This example uses the "accumulate" function of the standard c++ library. This
loop uses C++ iterators, which are pointers, and not integer indices.
The Loop Vectorizer detects pointer induction variables and can vectorize
this loop. This feature is important because many C++ programs use iterators.

.. code-block:: c++

  int baz(int *A, int n) {
    return std::accumulate(A, A + n, 0);
  }

Reverse Iterators
^^^^^^^^^^^^^^^^^

The Loop Vectorizer can vectorize loops that count backwards.

.. code-block:: c++

  int foo(int *A, int *B, int n) {
    for (int i = n; i > 0; --i)
      A[i] +=1;
  }

Scatter / Gather
^^^^^^^^^^^^^^^^

The Loop Vectorizer can vectorize code that becomes a sequence of scalar instructions 
that scatter/gathers memory.

.. code-block:: c++

  int foo(int * A, int * B, int n) {
    for (intptr_t i = 0; i < n; ++i)
        A[i] += B[i * 4];
  }

In many situations the cost model will inform LLVM that this is not beneficial
and LLVM will only vectorize such code if forced with "-mllvm -force-vector-width=#".

Vectorization of Mixed Types
^^^^^^^^^^^^^^^^^^^^^^^^^^^^

The Loop Vectorizer can vectorize programs with mixed types. The Vectorizer
cost model can estimate the cost of the type conversion and decide if
vectorization is profitable.

.. code-block:: c++

  int foo(int *A, char *B, int n, int k) {
    for (int i = 0; i < n; ++i)
      A[i] += 4 * B[i];
  }

Global Structures Alias Analysis
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

Access to global structures can also be vectorized, with alias analysis being
used to make sure accesses don't alias. Run-time checks can also be added on
pointer access to structure members.

Many variations are supported, but some that rely on undefined behaviour being
ignored (as other compilers do) are still being left un-vectorized.

.. code-block:: c++

  struct { int A[100], K, B[100]; } Foo;

  int foo() {
    for (int i = 0; i < 100; ++i)
      Foo.A[i] = Foo.B[i] + 100;
  }

Vectorization of function calls
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

The Loop Vectorize can vectorize intrinsic math functions.
See the table below for a list of these functions.

+-----+-----+---------+
| pow | exp |  exp2   |
+-----+-----+---------+
| sin | cos |  sqrt   |
+-----+-----+---------+
| log |log2 |  log10  |
+-----+-----+---------+
|fabs |floor|  ceil   |
+-----+-----+---------+
|fma  |trunc|nearbyint|
+-----+-----+---------+
|     |     | fmuladd |
+-----+-----+---------+

The loop vectorizer knows about special instructions on the target and will
vectorize a loop containing a function call that maps to the instructions. For
example, the loop below will be vectorized on Intel x86 if the SSE4.1 roundps
instruction is available.

.. code-block:: c++

  void foo(float *f) {
    for (int i = 0; i != 1024; ++i)
      f[i] = floorf(f[i]);
  }

Partial unrolling during vectorization
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

Modern processors feature multiple execution units, and only programs that contain a
high degree of parallelism can fully utilize the entire width of the machine. 
The Loop Vectorizer increases the instruction level parallelism (ILP) by 
performing partial-unrolling of loops.

In the example below the entire array is accumulated into the variable 'sum'.
This is inefficient because only a single execution port can be used by the processor.
By unrolling the code the Loop Vectorizer allows two or more execution ports
to be used simultaneously.

.. code-block:: c++

  int foo(int *A, int *B, int n) {
    unsigned sum = 0;
    for (int i = 0; i < n; ++i)
        sum += A[i];
    return sum;
  }

The Loop Vectorizer uses a cost model to decide when it is profitable to unroll loops.
The decision to unroll the loop depends on the register pressure and the generated code size. 

Performance
-----------

This section shows the execution time of Clang on a simple benchmark:
`gcc-loops <http://llvm.org/viewvc/llvm-project/test-suite/trunk/SingleSource/UnitTests/Vectorizer/>`_.
This benchmarks is a collection of loops from the GCC autovectorization
`page <http://gcc.gnu.org/projects/tree-ssa/vectorization.html>`_ by Dorit Nuzman.

The chart below compares GCC-4.7, ICC-13, and Clang-SVN with and without loop vectorization at -O3, tuned for "corei7-avx", running on a Sandybridge iMac.
The Y-axis shows the time in msec. Lower is better. The last column shows the geomean of all the kernels.

.. image:: gcc-loops.png

And Linpack-pc with the same configuration. Result is Mflops, higher is better.

.. image:: linpack-pc.png

Ongoing Development Directions
------------------------------

.. toctree::
   :hidden:

   Proposals/VectorizationPlan

:doc:`Proposals/VectorizationPlan`
   Modeling the process and upgrading the infrastructure of LLVM's Loop Vectorizer.

.. _slp-vectorizer:

The SLP Vectorizer
==================

Details
-------

The goal of SLP vectorization (a.k.a. superword-level parallelism) is
to combine similar independent instructions
into vector instructions. Memory accesses, arithmetic operations, comparison
operations, PHI-nodes, can all be vectorized using this technique.

For example, the following function performs very similar operations on its
inputs (a1, b1) and (a2, b2). The basic-block vectorizer may combine these
into vector operations.

.. code-block:: c++

  void foo(int a1, int a2, int b1, int b2, int *A) {
    A[0] = a1*(a1 + b1)/b1 + 50*b1/a1;
    A[1] = a2*(a2 + b2)/b2 + 50*b2/a2;
  }

The SLP-vectorizer processes the code bottom-up, across basic blocks, in search of scalars to combine.

Usage
------

The SLP Vectorizer is enabled by default, but it can be disabled
through clang using the command line flag:

.. code-block:: console

   $ clang -fno-slp-vectorize file.c

LLVM has a second basic block vectorization phase
which is more compile-time intensive (The BB vectorizer). This optimization
can be enabled through clang using the command line flag:

.. code-block:: console

   $ clang -fslp-vectorize-aggressive file.c