Commit Graph

305 Commits

Author SHA1 Message Date
Tobias Grosser 55bc4c0767 Add support for pointer types in expressions
llvm-svn: 225464
2015-01-08 19:26:53 +00:00
Tobias Grosser 3f29619614 Drop all constant scheduling dimensions
Schedule dimensions that have the same constant value accross all statements do
not carry any information, but due to the increased dimensionality of the
schedule cost compile time. To not pay this cost, we remove constant dimensions
if possible.

llvm-svn: 225067
2015-01-01 23:01:11 +00:00
Tobias Grosser 683b8e4462 Remove -polly-codegen-scev option and related code
SCEV based code generation has been the default for two weeks after having
been tested for a long time. We now drop the support the non-scev-based code
generation.

llvm-svn: 222978
2014-11-30 14:33:31 +00:00
Tobias Grosser 7b50beebe4 Assume GetElementPtr offsets to be inbounds
In case a GEP instruction references into a fixed size array e.g., an access
A[i][j] into an array A[100x100], LLVM-IR does not guarantee that the subscripts
always compute values that are within array bounds. We now derive the set of
parameter values for which all accesses are within bounds and add the assumption
that the scop is only every executed with this set of parameter values.

Example:

void foo(float A[][20], long n, long m {
    for (long i = 0; i < n; i++)
      for (long j = 0; j < m; j++)
        A[i][j] = ...

This loop yields out-of-bound accesses if m is at least 20 and at the same time
at least one iteration of the outer loop is executed. Hence, we assume:

  n <= 0 or m <= 20.

Doing so simplifies the dependence analysis problem, allows us to perform
more optimizations and generate better code.

TODO: The location where the GEP instruction is executed is not necessarily the
location where the memory is actually accessed. As a result scanning for GEP[s]
is imprecise. Even though this is not a correctness problem, this imprecision
may result in missed optimizations or non-optimal run-time checks.

In polybench where this mismatch between parametric loop bounds and fixed size
arrays is common, we see with this patch significant reductions in compile time
(up to 50%) and execution time (up to 70%). We see two significant compile time
regressions (fdtd-2d, jacobi-2d-imper), and one execution time regression
(trmm).  Both regressions arise due to additional optimizations that have been
enabled by this patch. They can be addressed in subsequent commits.

http://reviews.llvm.org/D6369

llvm-svn: 222754
2014-11-25 10:51:12 +00:00
Johannes Doerfert 9c14737b9b Simplify the SCEVAffinator [NFC]
llvm-svn: 222360
2014-11-19 15:36:59 +00:00
Tobias Grosser c98a8fccf7 Use nullptr instead of '0' for pointers
llvm-svn: 221982
2014-11-14 11:12:31 +00:00
Johannes Doerfert 80ef110cca [Refactor][NFC] Generalize the creation of ScopArrayInfo objects.
Differential Revision: http://reviews.llvm.org/D6031

llvm-svn: 221512
2014-11-07 08:31:31 +00:00
Johannes Doerfert 21aa3dc014 [FIX] Move the statistic code to fix the build.
llvm-svn: 221038
2014-11-01 01:30:11 +00:00
Johannes Doerfert 5ad8a6a588 Remove the LoopBounds from the TempScop class.
We will use ScalarEvolution in the ScopInfo.cpp to get the loop trip
  count, not cache it in the TempScop object.

Differential Revision: http://reviews.llvm.org/D6070

llvm-svn: 221035
2014-11-01 01:14:56 +00:00
Johannes Doerfert e3da05ac32 Remove the MaxLoopDepth attribute from the TempScop class
Now MaxLoopDepth only lives in Scops not in TempScops anymore.
  This is the first part of a series of changes to make TempScops
  obsolete.

Differential Revision: http://reviews.llvm.org/D6069

llvm-svn: 221026
2014-11-01 00:12:13 +00:00
Johannes Doerfert 75bd66e51d [Refactor][NFC] Remove unused argument.
llvm-svn: 221016
2014-10-31 23:16:02 +00:00
Johannes Doerfert 7c494217f3 [Refactor][NFC] Map basic blocks to SCoP statements.
This will simplify the construction of domains and the modeling of
  PHI's.

llvm-svn: 221015
2014-10-31 23:13:39 +00:00
Tobias Grosser a5605d34b9 Fix typo
llvm-svn: 220870
2014-10-29 19:58:28 +00:00
Johannes Doerfert a99130f042 [Refactor][NfC] Simplify and clean the handling of (new) access relations
This patch does not change the semantic on it's own. However, the
  dependence analysis as well as dce will now use the newest available
  access relation for each memory access, thus if at some point the json
  importer or any other pass will run before those two and set a new
  access relation the behaviour will be different. In general it is
  unclear if the dependence analysis and dce should be run on the old or
  new access functions anyway. If we need to access the original access
  function from the outside later, we can expose the getter again.

Differential Revision: http://reviews.llvm.org/D5707

llvm-svn: 219612
2014-10-13 12:58:03 +00:00
Johannes Doerfert 341a15a64b Use the new access function (if present) to compute the access stride.
Differential Revision: http://reviews.llvm.org/D5661

llvm-svn: 219499
2014-10-10 14:28:46 +00:00
Johannes Doerfert 4c7ce479ea [Refactor] Rename MemoryAccess::Type as it clashes with llvm::Type
llvm-svn: 219294
2014-10-08 10:11:33 +00:00
Johannes Doerfert 74f6869435 [Fix] Ignore forwarding alias sets in the alias set tracker.
llvm-svn: 219275
2014-10-08 02:23:48 +00:00
Johannes Doerfert 219b20e1a3 [Fix] Non i1 typed select condition for weird pw aff functions.
In case the pieceweise affine function used to create an isl_ast_expr
  had empty cases (e.g., with contradicting constraints on the
  parameters), it was possible that the condition of the isl_ast_expr
  select was not a comparison but a constant (thus of type i64).

  This patch does two thing:
   1) Handle the case the condition of a select is not a i1 type like C.
   2) Try to simplify the pieceweise affine functions for the min/max
      access when we generate runtime alias checks. That step can often
      remove empty or redundant cases as well as redundant constrains.

  This fixes bug: http://llvm.org/PR21167

Differential Revision: http://reviews.llvm.org/D5627

llvm-svn: 219208
2014-10-07 14:37:59 +00:00
Johannes Doerfert f1ee2622be [Fix] Dead statements should not confuse the RTC generation
This fixes http://llvm.org/bugs/show_bug.cgi?id=21166 .

Differential Revision: http://reviews.llvm.org/D5623

llvm-svn: 219131
2014-10-06 17:43:00 +00:00
Johannes Doerfert 1a28a8938e Introduce the ScopArrayInfo class.
This class allows to store information about the arrays in the SCoP.
  For each base pointer in the SCoP one object is created storing the
  type and dimension sizes of the array. The objects can be obtained via
  the SCoP, a MemoryAccess or the isl_id associated with the output
  dimension of a MemoryAccess (the description of what is accessed).

  So far we use the information in the IslExprBuilder to create the
  right base type before indexing into the base array. This fixes the
  bug http://llvm.org/bugs/show_bug.cgi?id=21113 (both test cases are
  included). On top of that we can now build runtime alias checks for
  delinearized arrays as the dimension sizes are also part of the
  ScopArrayInfo objects.

Differential Revision: http://reviews.llvm.org/D5613

llvm-svn: 219077
2014-10-05 11:32:18 +00:00
Johannes Doerfert eeab05a084 [RTC] Use the domain to split alias groups.
We use a parametric abstraction of the domain to split alias groups
  if accesses cannot be executed under the same parameter evaluation.

  The two test cases check that we can remove alias groups if the
  pointers which might alias are never accessed under the same parameter
  evaluation and that the minimal/maximal accesses are not global but
  with regards to the parameter evaluation.

Differential Revision: http://reviews.llvm.org/D5436

llvm-svn: 218758
2014-10-01 12:42:37 +00:00
Johannes Doerfert 13771738d3 [RTC] Split alias groups according to read only base addresses
If there are multiple read only base addresses in an alias group
  we can split it into multiple alias groups each with only one
  read only access. This way we might reduce the number of
  comparisons significantly as it grows linear in the number of
  alias groups but exponential in their size.

Differential Revision: http://reviews.llvm.org/D5435

llvm-svn: 218757
2014-10-01 12:40:46 +00:00
Johannes Doerfert 9143d67aba [RTC] Bail if too many parameters are involved in a RTC access.
If too many parameters are involved in accesses used to create RTCs
  we might end up with enormous compile times and RTC expressions.
  The reason is that the lexmin/lexmax is dependent on all these
  parameters and isl might need to create a case for every "ordering"
  of them (e.g., p0 <= p1 <= p2, p1 <= p0 <= p2, ...).

  The exact number of parameters allowed in accesses is defined by the
  command line option -polly-rtc-max-parameters=XXX and set by default
  to 8.

  Differential Revision: http://reviews.llvm.org/D5500

llvm-svn: 218566
2014-09-27 11:02:39 +00:00
Johannes Doerfert b164c795b7 [RTC] Runtime Alias Checks for the ISL backend
This change will build all alias groups (minimal/maximal accesses
  to possible aliasing base pointers) we have to check before
  we can assume an alias free environment. It will also use these
  to create Runtime Alias Checks (RTC) in the ISL code generation
  backend, thus allow us to optimize SCoPs despite possibly aliasing
  pointers when this backend is used.

  This feature will be enabled for the isl code generator, e.g.,
  --polly-code-generator=isl, but disabled for:
    - The cloog code generator (still the default).
    - The case delinearization is enabled.
    - The case non-affine accesses are allowed.

llvm-svn: 218046
2014-09-18 11:17:17 +00:00
Johannes Doerfert 8f7124ceb9 [Polly] Move MAY/MUST write property to IRAccess
Differential Revision: http://reviews.llvm.org/D5223

llvm-svn: 217665
2014-09-12 11:00:49 +00:00
Johannes Doerfert 9e7b17b0d4 Added arcanist linters and cleaned errors and warnings
Arcanist (arc) will now always run linters before uploading any new
  commit to Phabricator. All errors/warnings (or their absence) will be
  shown in the web interface together with a explanation by the commiter
  (arcanist will ask the commiter if the build was not clean).

  The linters include:
    - clang-format
    - spelling check
    - permissions check (aka. chmod)
    - filename check
    - merge conflict marker check
  Note, that their scope is sometimes limited (see .arclint for
  details).

  This commit also fixes all errors and warnings these linters reported,
  namely:
    - spelling mistakes and typos
    - executable permissions for various text files

Differential Revision: http://reviews.llvm.org/D4916

llvm-svn: 215871
2014-08-18 00:40:13 +00:00
Tobias Grosser f4daf34496 Revert "Added support for modulo expressions"
This reverts commit 215684. The intention of the commit is great, but
unfortunately it seems to be the cause of 14 LNT test suite failures:

http://lab.llvm.org:8011/builders/perf-x86_64-penryn-O3-polly/builds/116

To make our buildbots and performance testers green until this issue is solved,
we temporarily revert this commit.

llvm-svn: 215816
2014-08-16 09:08:55 +00:00
Johannes Doerfert 5130c849aa Added support for modulo expressions
The support is limited to signed modulo access and condition
  expressions with a constant right hand side, e.g., A[i % 2] or
  A[i % 9]. Test cases are modified according to this new feature and
  new test cases are added.

Differential Revision: http://reviews.llvm.org/D4843

llvm-svn: 215684
2014-08-15 01:14:11 +00:00
Johannes Doerfert 13c8cf2544 [Refactor] Remove const keyword from access instruction
llvm-svn: 215309
2014-08-10 08:09:38 +00:00
Tobias Grosser f57d63f906 Do allow negative offsets in the outermost array dimension
There is no needed for neither 1-dimensional nor higher dimensional arrays to
require positive offsets in the outermost array dimension.

We originally introduced this assumption with the support for delinearizing
multi-dimensional arrays.

llvm-svn: 214665
2014-08-03 21:07:30 +00:00
Johannes Doerfert 32868bf4c3 Change the printing of reduction types
We use the C operator representation when applicable.

  + Update all the test cases accordingly.

llvm-svn: 214486
2014-08-01 08:13:25 +00:00
Johannes Doerfert 5d83f09cdd Added base array id's to Memory Accesses
These id's allow us to cread id -> array base value mappings.

llvm-svn: 214169
2014-07-29 08:37:55 +00:00
Johannes Doerfert 770ec7a72f [Refactor] Remove unused MemoryAccess constructor
llvm-svn: 214165
2014-07-29 08:34:03 +00:00
Johannes Doerfert 79fc23f26b Make getIslCompatibleName globaly available
llvm-svn: 213907
2014-07-24 23:48:02 +00:00
Matt Arsenault 8ca36815ee Update for RegionInfo changes.
Mostly related to missing includes and renaming of
the pass to RegionInfoPass.

llvm-svn: 213457
2014-07-19 18:40:17 +00:00
Tobias Grosser 780ce0f8e3 DeadCodeElim: Compute correct liveout for non-affine accesses
Thanks to Johannes Doerfert for narrowing down the bug.

Reported-by: Chris Jenneisch <chrisj@codeaurora.org>
llvm-svn: 212796
2014-07-11 07:12:10 +00:00
Tobias Grosser 483a90d1bd clang-format polly to avoid buildbot noise
llvm-svn: 212609
2014-07-09 10:50:10 +00:00
Tobias Grosser 5e6813d184 Derive run-time conditions for delinearization
As our delinearization works optimistically, we need in some cases run-time
checks that verify our optimistic assumptions. A simple example is the
following code:

void foo(long n, long m, long o, double A[n][m][o]) {

  for (long i = 0; i < 100; i++)
    for (long j = 0; j < 150; j++)
      for (long k = 0; k < 200; k++)
        A[i][j][k] = 1.0;
}

After clang linearized the access to A and we delinearized it again to
A[i][j][k] we need to ensure that we do not access the delinearized array
out of bounds (this information is not available in LLVM-IR). Hence, we
need to verify the following constraints at run-time:

CHECK:   Assumed Context:
CHECK:   [o, m] -> {  : m >= 150 and o >= 200 }
llvm-svn: 212198
2014-07-02 17:47:48 +00:00
Johannes Doerfert f618339a37 Introduce reduction types
This change is particularly useful in the code generation as we need
  to know which binary operator/identity element we need to combine/initialize
  the privatization locations.

  + Print the reduction type for each memory access
  + Adjusted the test cases to comply with the new output format and
    to test for the right reduction type

llvm-svn: 212126
2014-07-01 20:52:51 +00:00
Johannes Doerfert 9890a05287 [FIX] Don't consider reductions which are partially outside the SCoP
+ Test case

llvm-svn: 212080
2014-07-01 00:32:29 +00:00
Tobias Grosser 083d3d3cb3 [C++11] Use more range based fors
llvm-svn: 211981
2014-06-28 08:59:45 +00:00
Tobias Grosser bc4ef90802 Remove redundant code and use C++ range fors
llvm-svn: 211980
2014-06-28 08:59:38 +00:00
Johannes Doerfert e58a012094 Allow multiple reductions per statement
Iterate over all store memory accesses and check for valid binary reduction
  candidate loads by following the operands of the stored value.  For each
  candidate pair we check if they have the same base address and there are no
  other accesses which may overlap with them. This ensures that no intermediate
  value can escape into other memory locations or is overwritten at some point.

  + 17 test cases for reduction detection and reduction dependency modeling

llvm-svn: 211957
2014-06-27 20:31:28 +00:00
Johannes Doerfert 6780bc310a [Refactor] Change memory access dump to preserve space
llvm-svn: 211796
2014-06-26 18:47:03 +00:00
Johannes Doerfert ffa73bd44c Reduction like is now a memory access property
- Remove the statement reduction like property
+ Add the reduction like property to memory accesses

llvm-svn: 211372
2014-06-20 16:58:12 +00:00
Johannes Doerfert 0ee1f21478 Subject: [PATCH-v5] Detect and mark reduction like statements
+ Flag to indicate reduction like statements
+ Command line option to (dis)allow multiplicative reduction opcodes
+ Two simple positive test cases, one fp test case w and w/o fast math
+ One "negative" test case (only reduction like but no reduction)

llvm-svn: 211114
2014-06-17 17:31:36 +00:00
Johannes Doerfert f675289d87 [Refactor] C++11 Memory access iterators in SCoP stmts
+ Added const iterator version
+ Changed name to begin/end to allow range loops
+ Changed call sites to range loops
+ Changed typename to (const_)iterator

llvm-svn: 210927
2014-06-13 18:01:45 +00:00
Sebastian Pop 422e33f363 record delinearization result and reuse it in polyhedral translation
Without this patch, the testcase would fail on the delinearization of the second
array:

; void foo(long n, long m, long o, double A[n][m][o]) {
;   for (long i = 0; i < n; i++)
;     for (long j = 0; j < m; j++)
;       for (long k = 0; k < o; k++) {
;         A[i+3][j-4][k+7] = 1.0;
;         A[i][0][k] = 2.0;
;       }
; }

; CHECK: [n, m, o] -> { Stmt_for_body6[i0, i1, i2] -> MemRef_A[3 + i0, -4 + i1, 7 + i2] };
; CHECK: [n, m, o] -> { Stmt_for_body6[i0, i1, i2] -> MemRef_A[i0, 0, i2] };

Here is the output of FileCheck on the testcase without this patch:

; CHECK: [n, m, o] -> { Stmt_for_body6[i0, i1, i2] -> MemRef_A[i0, 0, i2] };
         ^
<stdin>:26:2: note: possible intended match here
 [n, m, o] -> { Stmt_for_body6[i0, i1, i2] -> MemRef_A[o0] };
 ^

It is possible to find a good delinearization for A[i][0][k] only in the context
of the delinearization of both array accesses.

There are two ways to delinearize together all array subscripts touching the
same base address: either duplicate the code from scop detection to first gather
all array references and then run the delinearization; or as implemented in this
patch, use the same delinearization info that we computed during scop detection.

llvm-svn: 210117
2014-06-03 18:16:31 +00:00
Chandler Carruth 95fef9446c [Modules] Fix potential ODR violations by sinking the DEBUG_TYPE
definition below all of the header #include lines, Polly edition.

If you want to know more details about this, you can see the recent
commits to Debug.h in LLVM. This is just the Polly segment of a cleanup
I'm doing globally for this macro.

llvm-svn: 206852
2014-04-22 03:30:19 +00:00
Tobias Grosser 5a56cbf496 [C++11] Use nullptr
llvm-svn: 206361
2014-04-16 07:33:47 +00:00
Tobias Grosser 20532b8e1b Fixed gcc build warnings
+ vim 'fixed' line endings in json_value.cpp

Contributed-by:  Johannes Doerfert <doerfert@cs.uni-saarland.de>
llvm-svn: 206044
2014-04-11 17:56:49 +00:00
Tobias Grosser 79baa21242 ScopInfo: Scalar accesses are zero dimensional
llvm-svn: 205958
2014-04-10 08:38:02 +00:00
Tobias Grosser 6defb5bd6d ScopInfo: Some code cleanup
llvm-svn: 205957
2014-04-10 08:37:44 +00:00
Sebastian Pop 1801668af3 delinearize memory access functions
llvm-svn: 205799
2014-04-08 21:20:44 +00:00
Tobias Grosser 4eb7ddb13a Fix source code formatting
llvm-svn: 204169
2014-03-18 18:51:11 +00:00
Tobias Grosser 483fdd4b7f Print function and region name in scop description
llvm-svn: 204165
2014-03-18 18:05:38 +00:00
Tobias Grosser 34f0613562 Optimizer: Do not accidentally set schedule to NULL
In case the domain of a statement is empty, the schedule optimizer set by
accident the schedule to a NULL pointer. This is incorrect. Instead, we set
it to an empty isl_map with zero schedule dimensions. We already checked for
this in our test cases, but unfortunately the test cases did not fail as
expected. The assert we add in this commit now ensures that the test cases
fail properly in case we regress on this again.

llvm-svn: 201886
2014-02-21 20:51:36 +00:00
Tobias Grosser 37eb422f69 Add polyhedral dead code elimination.
This pass eliminates loop iterations that compute results that are not used
later on. This can help e.g. in D, where the default zero-initialization is
often unnecessary if right after new values are assigned to an array.

Contributed-by: Peter Conn <conn.peter@gmail.com>
llvm-svn: 201817
2014-02-20 21:43:54 +00:00
Tobias Grosser 3fc9154c40 Fix formatting
llvm-svn: 201816
2014-02-20 21:43:45 +00:00
Tobias Grosser d6aafa7c2e Do not track location of scalar dependences in ScopInfo
We do not have a use for this information at the moment. If we need this at some
point, the "instruction -> access" mapping needs to be enhanced as a single
instruction could then possibly perform multiple accesses.

This patch allows us to build the polyhedral information for scops with scalar
dependences.

llvm-svn: 201815
2014-02-20 21:29:09 +00:00
Tobias Grosser f240b487a3 Remove IR/Writer.h
This should fix the buildbots.

llvm-svn: 198859
2014-01-09 10:42:15 +00:00
Chandler Carruth e243b03a8c Update #include paths for r198688 in LLVM that moved headers in the Assembly
directory to their proper homes.

llvm-svn: 198691
2014-01-07 12:59:58 +00:00
Tobias Grosser e86109f508 ScopInfo: Add support for AssumedContext
When constructing a scop sometimes the exact representation of a statement or
condition would be very complex, but there is a common case which is a lot
simpler, but which is only valid under certain assumptions. The assumed context
records the assumptions taken during the construction of this scop and that need
to be code generated as a run-time test.

At the moment, we do not yet model any assumptions, but only added the
AssumedContext as well as the isl-ast generation support. As a next step,
this needs to be hooked up with the isl code generation.

if (1) /* run-time condition */
  {  /* optimized code */ }
else
  {  /* original code */ }

llvm-svn: 193652
2013-10-29 21:05:49 +00:00
Tobias Grosser dd6dc8276f clang-format: No empty line after 'public:'
llvm-svn: 192710
2013-10-15 14:41:02 +00:00
Sebastian Pop 40408760c1 do not compute isl_map_dim in the loop
llvm-svn: 191969
2013-10-04 17:14:53 +00:00
Tobias Grosser 0d17013cb7 clang-format
llvm-svn: 191895
2013-10-03 13:09:19 +00:00
Tobias Grosser 1ff0be799d ScopInfo: Do not include "isl/int.h" any more
We already removed all uses of isl_int in ScopInfo.

Contributed-by:  Michael Kruse  <MichaelKruse@meinersbur.de>
llvm-svn: 191893
2013-10-03 13:09:07 +00:00
Tobias Grosser 0695ee433e Move SCEVAffinator member definitions out of class body
Instead of defining the relevant functions inline, we now just keep the
declarations in the class itself. This makes the class declaration a lot
easier to read as all functions can be seen at once. We also use this
opportunity to privatize all functions not used in the public interface of the
class.

llvm-svn: 190841
2013-09-17 03:30:31 +00:00
Tobias Grosser e42ddb9ad3 ScopInfo: Split start value from SCEVAddRecExpr to enable parameter sharing.
SCoP invariant parameters with the different start value would deter parameter
sharing. For example, when compiling the following C code:

  void foo(float *input) {
    for (long j = 0; j < 8; j++) {
      // SCoP begin
      for (long i = 0; i < 8; i++) {
        float x = input[j * 64 + i + 1];
        input[j * 64 + i] = x * x;
      }
    }
  }

Polly would creat two parameters for these memory accesses:

    p_0: {0,+,256}
    p_2: {4,+,256}
    [j * 64 + i + 1] => MemRef_input[o0] : 4o0 = p_1 + 4i0
    [j * 64 + i]     => MemRef_input[o0] : 4o0 = p_0 + 4i0

These parameters only differ from start value. To enable parameter sharing,
we split the start value from SCEVAddRecExpr, so they would share a single
parameter that always has zero start value:

    p0: {0,+,256}<%for.cond1.preheader>
    [j * 64 + i + 1] => MemRef_input[o0] : 4o0 = 4 + p_1 + 4i0
    [j * 64 + i]     => MemRef_input[o0] : 4o0 = p_0 + 4i0

Such translation can make the polly-dependence much faster.

Contributed-by: Star Tan <tanmx_star@yeah.net>
llvm-svn: 187728
2013-08-05 15:14:15 +00:00
Tobias Grosser ff9bfdfa80 ScopInfo/IndependentBlocks: clang-format
llvm-svn: 187023
2013-07-24 06:10:30 +00:00
Hongbin Zheng c6aa9f5c2a Make sure the each instruction is mapped to one memory access.
llvm-svn: 186417
2013-07-16 15:18:51 +00:00
Tobias Grosser b58f6a4211 ScopInfo: Add getTupe() method to MemoryAccess
We also move the enum to UPPERCASE.

llvm-svn: 186259
2013-07-13 20:41:24 +00:00
Tobias Grosser 04d6ae65b7 Fix my own formatting mistakes
llvm-svn: 184659
2013-06-23 06:04:54 +00:00
Tobias Grosser 4f96749351 ScopInfo: Clarify may-write and must-write accesses
llvm-svn: 184658
2013-06-23 05:21:18 +00:00
Tobias Grosser edab1359a0 Use isl_val instead of isl_int in the core of Polly
isl recently introduced isl_val as an abstract interface to represent arbitrary
precision numbers. This interface superseeds the old isl_int interface. In
contrast to the old interface which implemented arbitrary precision arithmetic
using macros that forward to the gmp library, the new library hides the math
library implementation in isl. This allows us to switch the math library used by
isl without affecting users such as Polly.

llvm-svn: 184529
2013-06-21 06:41:31 +00:00
Tobias Grosser 83628182f7 Sort includes
llvm-svn: 181297
2013-05-07 08:11:54 +00:00
Tobias Grosser e602a07662 Reformat with clang-format
clang-format become way more stable. This time we mainly reformat function
signatures.

llvm-svn: 181294
2013-05-07 07:30:56 +00:00
Tobias Grosser d7e58640a5 Update formatting to latest version of clang-format
llvm-svn: 179160
2013-04-10 06:55:45 +00:00
Tobias Grosser 4d96c8d714 clang-format: Many more files
After this commit, polly is clang-format clean. This can be tested with
'ninja polly-check-format'. Updates to clang-format may change this, but the
differences will hopefully be both small and general improvements to the
formatting.

We currently have some not very nice formatting for a couple of items, DEBUG()
stmts for example. I believe the benefit of being clang-format clean outweights
the not perfect layout of this code.

llvm-svn: 177796
2013-03-23 01:05:07 +00:00
Sebastian Pop 27c10c6b1f ScopInfo: do not call getCanonicalInductionVariable for SCEVCodegen
llvm-svn: 177771
2013-03-22 22:07:43 +00:00
Tobias Grosser 826b2af112 Remove last uses of canoncial induction variable when scev code generating
We now detect scops without a canonical induction variable and can generate a
polyhedral representation for them. There was no modification necessary to
code generate these scops.

llvm-svn: 177643
2013-03-21 16:14:50 +00:00
Sebastian Pop f30d3b2d89 don't store a pointer to the loop in IVS
llvm-svn: 175304
2013-02-15 21:26:48 +00:00
Sebastian Pop 860e021fe6 add NestLoops to remove some uses of IVS
llvm-svn: 175303
2013-02-15 21:26:44 +00:00
Tobias Grosser abfbe637bc ScopInfo: clang-format file
clang-format was able to format the entire file except the final
INITIALIZE_PASS_* macros.

llvm-svn: 174366
2013-02-05 12:09:06 +00:00
Tobias Grosser 74394f0dcf clang-format goodness
llvm-svn: 172486
2013-01-14 22:40:23 +00:00
Tobias Grosser ae2d83ec41 Formatting: Break lines after binary operators such as '&&'
assert(Condition
       && "Text");

->

assert(Condition &&
       "Text);

This aligns Polly with the style used in LLVM.

llvm-svn: 171242
2012-12-29 23:57:18 +00:00
Tobias Grosser 1bb59b0dcf Fix obvious formatting problems.
We fix the following formatting problems found by clang-format:

  - 80 cols violations
  - Obvious problems with missing or too many spaces
  - multiple new lines in a row

clang-format suggests many more changes, most of them falling in the following
two categories:

  1) clang-format does not at all format a piece of code nicely

  2) The style that clang-format suggests does not match the style used in
     Polly/LLVM

I consider differences caused by reason 1) bugs, which should be fixed by
improving clang-format. Differences due to 2) need to be investigated closer
to understand the cause of the difference and the solution that should be taken.

llvm-svn: 171241
2012-12-29 23:47:38 +00:00
Sebastian Pop a00a029115 change interface for isStride
isStride now takes a partial schedule as input.

Patch from Tobias Grosser <tobias@grosser.es>.

llvm-svn: 170419
2012-12-18 07:46:06 +00:00
Sebastian Pop 05c1f00cae remove dead code
llvm-svn: 168621
2012-11-26 22:07:30 +00:00
Tobias Grosser ed29566c4e ScopInfo: Align parameters when using -polly-allow-nonaffine
This ensures that the isl sets/maps we operate on have the same parameter
dimensions. Operations on objects with different parameter dimensions are not
allow and trigger assertions.

llvm-svn: 163618
2012-09-11 13:50:21 +00:00
Tobias Grosser cd95b77330 Pocc: Fix some bugs in the PoCC optimizer pass
This includes:
  - The isl_id of the domain of the scattering must be copied from the original
    domain
  - Remove outdated references to a 'FinalRead' statement
  - Print of the Pocc output, if -debug is provided.
  - Add line breaks to some error messages.

Reported and Debugged by:  Dustin Feld  <d3.feld@gmail.com>

llvm-svn: 162901
2012-08-30 11:49:38 +00:00
Hongbin Zheng cea35f60d5 Add an Instruction member to MemoryAccess Class.
Patched by TangKK <dengjunqi06323011@hotmail.com>.

llvm-svn: 159808
2012-07-06 06:47:03 +00:00
Tobias Grosser 3cc99748b6 Fix some coding convention problems.
llvm-svn: 158081
2012-06-06 16:33:15 +00:00
Tobias Grosser 084d8f7d4c ScopInfo: Store ScopStmt pointer in the domain
Store a pointer to each ScopStmt in the isl_id associated with the space of its
domain. This will later allow us to recover the statement during code
generation with isl.

llvm-svn: 157607
2012-05-29 09:29:44 +00:00
Tobias Grosser 18daacad61 ScopInfo: Add parameter bounds to context
Derive the maximal and minimal values of a parameter from the type it has. Add
this information to the scop context. This information is needed, to derive
optimal types during code generation.

llvm-svn: 157245
2012-05-22 10:47:27 +00:00
Tobias Grosser 3b2cf96bae Replace some asserts with llvm_unreachable
llvm-svn: 157244
2012-05-22 10:47:21 +00:00
Tobias Grosser 42b69e5745 ScopInfo: SCEVUnknowns are always parameters
There is no need for special code to handle SCEVUnknowns. SCEVUnkowns are always
parameters and will be handled by the generic parameter handling code in
visit().

llvm-svn: 157243
2012-05-22 10:47:17 +00:00
Hongbin Zheng 86a37745eb Minor change: Replace convertInt in ScopInfo.cpp by utostr_32, which is
defined in StringExtras.h.

llvm-svn: 155539
2012-04-25 08:01:38 +00:00
Tobias Grosser f9fbbdfd74 Fix typos.
Pointed out by: Sebastian Pop  <sebpop@gmail.com>

llvm-svn: 154337
2012-04-09 19:46:05 +00:00
Tobias Grosser 3cbe5cfff3 Remove FinalRead
The FinalRead statement represented a virtual read that is executed after the
SCoP. It was used when we verified the correctness of a schedule by checking if
it yields the same FLOW dependences as the original code. This is only works, if
we have a final read that reads all memory at the end of the SCoP.
We now switched to just checking if a schedule does not introduce negative
dependences and also consider WAW WAR dependences. This restricts the schedules
a little bit more, but we do not have any optimizer that would calculate a more
complex schedule. Hence, for now final reads are obsolete.

llvm-svn: 152319
2012-03-08 15:21:51 +00:00
Tobias Grosser 5f9a762056 ScopInfo: Add Scop::getDomains()
llvm-svn: 150482
2012-02-14 14:02:40 +00:00
Tobias Grosser c327932cf5 ScopInfo: Simplify some isl code
llvm-svn: 149503
2012-02-01 14:23:36 +00:00
Tobias Grosser 28dd48613e ScopInfo: Add isStrideX to unify stride checking
llvm-svn: 148810
2012-01-24 16:42:16 +00:00
Tobias Grosser 030b0d77ed ScopInfo: Further style improvements
llvm-svn: 148328
2012-01-17 20:39:11 +00:00
Tobias Grosser dea982300e ScopInfo: Simplify some code II
llvm-svn: 148327
2012-01-17 20:34:27 +00:00
Tobias Grosser 78d8a3d505 ScopInfo: Simplify some code
llvm-svn: 148326
2012-01-17 20:34:23 +00:00
Tobias Grosser a187964bac Support non-affine access functions in Polly.
In case we can not analyze an access function, we do not discard the SCoP, but
assume conservatively that all memory accesses that can be derived from our base
pointer may be accessed.

Patch provided by: Marcello Maggioni <hayarms@gmail.com>

llvm-svn: 146972
2011-12-20 10:43:14 +00:00
Tobias Grosser 4a8e356f36 Make isl abort when an error is encountered
llvm-svn: 146027
2011-12-07 07:42:51 +00:00
Tobias Grosser 29ee0b14d5 Do not use getNameStr() anymore.
Instead we switch to the recommended getName(). This fixes compilation with
recent versions of LLVM.

llvm-svn: 144909
2011-11-17 14:52:36 +00:00
Tobias Grosser 7ffe4e8b0b Fix placement of the '*' that marks a pointer
Suggested by Sebastian Pop.

llvm-svn: 144902
2011-11-17 12:56:10 +00:00
Tobias Grosser 8f99c167cd ScopInfo: Use names of simple parameters to name the isl parameter dimensions.
Parameters can be complex SCEV expressions, but they can also be single scalar
values. If a parameters is such a simple scalar value and the value is named,
use this name to name the isl parameter dimensions.

llvm-svn: 144641
2011-11-15 11:38:55 +00:00
Tobias Grosser ff9b54d5a9 JScop: Allow to update the context
llvm-svn: 144639
2011-11-15 11:38:44 +00:00
Tobias Grosser e5e171eadd Reuse the old BaseAddress checking in SCEVValidator to make sure that no base
address is part of the access function. Also remove unused special cases that
were necessery when the base address was still contained in the access function

llvm-svn: 144280
2011-11-10 12:45:03 +00:00
Tobias Grosser 9759f85060 Use getBasePtr in TempScop/ScopInfo
llvm-svn: 144279
2011-11-10 12:44:55 +00:00
Tobias Grosser e4e2f7b16e TempScop: Rename SCEVAffFunc to IRAccess
The SCEVAffFunc is now only used to express memory accesses. Give it a proper
name and rework the class such that this is obvious.

llvm-svn: 144231
2011-11-09 22:35:09 +00:00
Tobias Grosser a601fbd682 Remove SCEVAffFunc from polly::Comparison
llvm-svn: 144226
2011-11-09 22:34:44 +00:00
Tobias Grosser 5683df4a23 Remove more of SCEVAffineFunc
llvm-svn: 144223
2011-11-09 22:34:34 +00:00
Tobias Grosser 63069c60cf ScopInfo: Use getParamsInAffineExpr to get the BaseAddress
llvm-svn: 144222
2011-11-09 22:34:28 +00:00
Tobias Grosser 60b54f19e6 Detect Parameters directly on the SCEV.
Instead of using TempScop to find parameters, we detect them directly
on the SCEV. This allows us to remove the TempScop parameter detection
in a subsequent commit.

This fixes a bug reported by Marcello Maggioni <hayarms@gmail.com>

llvm-svn: 144087
2011-11-08 15:41:28 +00:00
Tobias Grosser 65b0058b56 Remove const
llvm-svn: 144086
2011-11-08 15:41:19 +00:00
Tobias Grosser 6be480c4ab ScopInfo: Don't add common parameters during realignment to the context.
Previously we built a context that contained already all parameter dimensions
from the start. We now build a context without any parameter dimensions and
extend the context as needed. All parameter dimensions are added during final
realignment.

llvm-svn: 144085
2011-11-08 15:41:13 +00:00
Tobias Grosser 8cae72f186 ScopInfo: Realign parameters after the scop is built
llvm-svn: 144084
2011-11-08 15:41:08 +00:00
Tobias Grosser 9a38ab8a04 Use a map to store the dimension of the the parameters
llvm-svn: 144083
2011-11-08 15:41:03 +00:00
Tobias Grosser 76c2e32a8e ScopInfo: Extract function getIdForParam()
llvm-svn: 143961
2011-11-07 12:58:59 +00:00
Tobias Grosser 7b0ee0ec9f ScopInfo: Add two new asserts.
They show a bug recently reported by Marcello Maggioni <hayarms@gmail.com>

llvm-svn: 143691
2011-11-04 10:08:03 +00:00
Tobias Grosser 1179afafca TempScop: Remove SCEVAffFunc from LoopBoundInfo
This is not needed anymore -> Reduce impact of SCEVAffFunc.

llvm-svn: 143575
2011-11-02 21:37:51 +00:00
Tobias Grosser 048c87943c ScopInfo: Remove unneeded code
llvm-svn: 142768
2011-10-23 20:59:20 +00:00
Tobias Grosser 23b3666850 ScopInfo: Fix ctx->ref != 0 problem
Reported by: Yabin Hu <yabin.hwu@gmail.com>

llvm-svn: 142184
2011-10-17 08:32:36 +00:00
Tobias Grosser 0ad4caa263 ScopInfo: Proporly free the local space
llvm-svn: 141459
2011-10-08 00:35:17 +00:00
Tobias Grosser 73600b8edd Initialize the passes early and properly.
llvm-svn: 141455
2011-10-08 00:30:40 +00:00
Tobias Grosser e19661e0ca ScopInfo: Some cleanups
- Use __isl_give and __isl_take
- Convert variables to start with Uppercase letter
- Only assign the 'domain' after it is fully constructed
- Only name it after it is fully constructed

llvm-svn: 141361
2011-10-07 08:46:57 +00:00
Tobias Grosser 9b13d3dbe5 ScopInfo: Simplify the construction of the iteration domain.
llvm-svn: 141325
2011-10-06 22:32:58 +00:00
Tobias Grosser 5d45381294 ScopInfo: Only give away a copy of the access relation.
Also take the chance and rename access functions to access relations. This is
because we do not only allow plain functions to describe an access, but we
can have any access relation that can be described with linear constraints.

llvm-svn: 141257
2011-10-06 00:04:11 +00:00
Tobias Grosser cf3942dfa6 ScopInfo: Only give away a copy of the schedule.
llvm-svn: 141256
2011-10-06 00:04:05 +00:00
Tobias Grosser 4da8d9fc15 ScopInfo: Only give away a copy of the Context
llvm-svn: 141255
2011-10-06 00:03:59 +00:00
Tobias Grosser 3c69fab0e8 ScopInfo: Get the isl_ctx always with getIslCtx()
llvm-svn: 141254
2011-10-06 00:03:54 +00:00
Tobias Grosser 0e27e24751 ScopInfo: Use separate function to build context
llvm-svn: 141253
2011-10-06 00:03:48 +00:00
Tobias Grosser 3748705fff Export the parameter space directly from the SCoP.
Use this to simplify some code.

llvm-svn: 141252
2011-10-06 00:03:42 +00:00
Tobias Grosser f53388034d Adapt to introduction of isl_space
Polly should now be compiled with CLooG 0c252c88946b27b7b61a1a8d8fd7f94d2461dbfd
and isl 56b7d238929980e62218525b4b3be121af386edf. The most convenient way to
update is utils/checkout_cloog.sh.

llvm-svn: 141251
2011-10-06 00:03:35 +00:00
Tobias Grosser 604c981f40 Temporarily remove reduction support and interchange pass
I am planning to eliminate the TempScopInfo pass. To simplify this I remove
some features that may later be added to the ScopInfo pass.

The interchange pass is currently strongly tested and furthermore ment to be
replaced by the general scheduling optimizer. Reductions itself can later
be added easily.

llvm-svn: 138219
2011-08-21 14:57:58 +00:00
Tobias Grosser b76f385334 Free isl_ctx and fix several memory leaks
Because of me not understanding the LLVM pass structure well, I did not find a
good way to allocate isl_ctx and to free it later without getting issues with
reference counting. I now found this place, such that we can free isl_ctx. This
patch also fixes the memory leaks that were ignored beforehand.

llvm-svn: 138204
2011-08-20 11:11:25 +00:00
Tobias Grosser eec4d56e63 Move common code into the function it belongs to
llvm-svn: 138202
2011-08-20 11:11:14 +00:00
Tobias Grosser fa7bc2f1a3 ScopInfo/Dependences: Use parameter ids everywhere
llvm-svn: 138155
2011-08-20 00:03:28 +00:00
Tobias Grosser 7d4cee4b02 ScopInfo: Construct the accessrelations with isl_pw_aff
This simplifies the code quite a bit.

llvm-svn: 138147
2011-08-19 23:34:28 +00:00
Tobias Grosser d2795d0365 ScopInfo: Remove adhoc comparision of affine expressions
Until today, we compared two affine expressions by defining two maps describing
them, creating an union of those maps, adding constraints that do the comparison
and projecting out unneeded dimensions.

This was simplified to using the isl_pw_aff representation of the affine
expressions and using the relevant isl functions to compare them.

llvm-svn: 137932
2011-08-18 07:51:40 +00:00
Tobias Grosser 30b8a092d8 ScopInfo: Create all isl data structures with dimension ids
At the moment, we still remove the ids after all data structures are created,
as later passes do not yet support ids. This limitation will be removed later.

llvm-svn: 137931
2011-08-18 07:51:37 +00:00
Tobias Grosser 33ba62ad11 ScopInfo: Build isl_pw_aff directly from SCEV.
Do not use AffFunc to derive the affine expressions, but use isl_pw_aff to
analyze the original SCEV directly. This will allow several simplifications in
follow up patches, with the final goal of removing AffFunc completely.

llvm-svn: 137930
2011-08-18 06:31:50 +00:00
Tobias Grosser 54a86e6c2a ScopInfo: Simplify code
llvm-svn: 137929
2011-08-18 06:31:46 +00:00
Raghesh Aloor 129e867865 Memaccess: Code generation for constant access function change
Support for generating code for an access function change which is
a constant is added. 

llvm-svn: 137603
2011-08-15 02:33:39 +00:00
Raghesh Aloor 7a04f4f9ba Memaccess: Display Changed Access Relation
The changed access relations imported from JSCOP file is shown
as output of -analyze pass.

llvm-svn: 136774
2011-08-03 13:47:59 +00:00
Tobias Grosser 3b660f8585 ScopInfo: Do not use >"< in statement names.
Signed-off-by: Tobias Grosser <tobias@grosser.es>

llvm-svn: 136750
2011-08-03 00:12:11 +00:00
Raghesh Aloor 3cb6628d7c MemAccess: Reading Change in Access Function
This patch reads the change in access functions from
imported JSCOP file. A test case is also added.

llvm-svn: 134991
2011-07-12 17:14:03 +00:00
Tobias Grosser d5a7bfc51d ScopInfo: Do not return reference to member variable 'domain'.
Instead of returning a pointer to the domain, we return a new copy of it. This
is safer, as we do not give access to internal objects. It is also not
expensive, as isl will just increment a reference counter.

llvm-svn: 131010
2011-05-06 19:52:19 +00:00
Hongbin Zheng 27f3afbc1a ScopInfo: Rememeber the induction variable and its parent loop at the same thime.
llvm-svn: 130586
2011-04-30 03:26:51 +00:00
Tobias Grosser 758053788b Add initial version of Polly
This version is equivalent to commit ba26ebece8f5be84e9bd6315611d412af797147e
in the old git repository.

llvm-svn: 130476
2011-04-29 06:27:02 +00:00