We now always print the reason why the code did not pass the LLVM verifier and
we also allow to disable verfication with -polly-codegen-verify=false. Before
this change the first assertion had generally no information why or what might
have gone wrong and it was also impossible to -view-cfg without recompile. This
change makes debugging bugs that result in incorrect IR a lot easier.
llvm-svn: 261320
We also disable this feature by default, as there are still some issues in
combination with invariant load hoisting that slipped through my initial
testing.
llvm-svn: 260025
Re-run canonicalization passes after Polly's code generation.
The set of passes currently added here are nearly all the passes between
--polly-position=early and --polly-position=before-vectorizer, i.e. all
passes that would usually run after Polly.
In order to run these only if Polly actually modified the code, we add a
function attribute "polly-optimzed" to a function that contains
generated code. The cleanup pass is skipped if the function does not
have this attribute.
There is no support by the (legacy) PassManager to run passes only under
some conditions. One could have wrapped all transformation passes to run
only when CodeGeneration changed the code, but the analyses would run
anyway. This patch creates an independent pass manager. The
disadvantages are that all analyses have to re-run even if preserved and
it does not honor compiler switches like the PassManagerBuilder does.
Differential Revision: http://reviews.llvm.org/D14333
llvm-svn: 254150
When we bail out early we make the partially build new code path
practically dead, though it was not unreachable. To remove dominance
problems we now make it not only dead but also prevent the control
flow to join with the original code path, thus allow to use original
values after the SCoP without any PHI nodes.
This fixes bug 25447.
llvm-svn: 252420
The bail out in r252412 left the code generation without verifying the (so
far) generated IR. This will change now and ensure we always run the
verifier.
Suggested-by: Tobias Grosser <tobias@grosser.es>
llvm-svn: 252419
While the program cannot cause a dependence cycle between invariant
loads, additional constraints (e.g., to ensure finite loops) can
introduce them. It is hard to detect them in the SCoP description,
thus we will only check for them at code generation time. If such a
recursion is detected we will bail out the code generation and place a
"false" runtime check to guarantee the original code is used.
This fixes bug 25443.
llvm-svn: 252412
Remove all the implicit ilist iterator conversions from polly, in
preparation for making them illegal in ADT. There was one oddity I came
across: at line 95 of lib/CodeGen/LoopGenerators.cpp, there was a
post-increment `Builder.GetInsertPoint()++`.
Since it was a no-op, I removed it, but I admit I wonder if it might be
a bug (both before and after this change)? Perhaps it should be a
pre-increment?
llvm-svn: 252357
Polly can now be used as a analysis only tool as long as the code
generation is disabled. However, we do not have an alternative to the
independent blocks pass in place yet, though in the relevant cases
this does not seem to impact the performance much. Nevertheless, a
virtual alternative that allows the same transformations without
changing the input region will follow shortly.
llvm-svn: 250652
This patch allows invariant loads to be used in the SCoP description,
e.g., as loop bounds, conditions or in memory access functions.
First we collect "required invariant loads" during SCoP detection that
would otherwise make an expression we care about non-affine. To this
end a new level of abstraction was introduced before
SCEVValidator::isAffineExpr() namely ScopDetection::isAffine() and
ScopDetection::onlyValidRequiredInvariantLoads(). Here we can decide
if we want a load inside the region to be optimistically assumed
invariant or not. If we do, it will be marked as required and in the
SCoP generation we bail if it is actually not invariant. If we don't
it will be a non-affine expression as before. At the moment we
optimistically assume all "hoistable" (namely non-loop-carried) loads
to be invariant. This causes us to expand some SCoPs and dismiss them
later but it also allows us to detect a lot we would dismiss directly
if we would ask e.g., AliasAnalysis::canBasicBlockModify(). We also
allow potential aliases between optimistically assumed invariant loads
and other pointers as our runtime alias checks are sound in case the
loads are actually invariant. Together with the invariant checks this
combination allows to handle a lot more than LICM can.
The code generation of the invariant loads had to be extended as we
can now have dependences between parameters and invariant (hoisted)
loads as well as the other way around, e.g.,
test/Isl/CodeGen/invariant_load_parameters_cyclic_dependence.ll
First, it is important to note that we cannot have real cycles but
only dependences from a hoisted load to a parameter and from another
parameter to that hoisted load (and so on). To handle such cases we
materialize llvm::Values for parameters that are referred by a hoisted
load on demand and then materialize the remaining parameters. Second,
there are new kinds of dependences between hoisted loads caused by the
constraints on their execution. If a hoisted load is conditionally
executed it might depend on the value of another hoisted load. To deal
with such situations we sort them already in the ScopInfo such that
they can be generated in the order they are listed in the
Scop::InvariantAccesses list (see compareInvariantAccesses). The
dependences between hoisted loads caused by indirect accesses are
handled the same way as before.
llvm-svn: 249607
When the ScopAnnotator was a class member variable some of the maps it contains
have not been properly cleared. As a result we had dangling pointers to
llvm::Value(s) which got detected by the AssertingVH we recently added.
No test case as this issue is hard to reproduce reliably as subsequent
optimizations need to delete some of the llvm::Values we still keep in our
lists.
llvm-svn: 249269
As a first step in the direction of assumed invariant loads (loads
that are not written in some context) we now detect and hoist
definitively invariant loads. These invariant loads will be preloaded
in the code generation and used in the optimized version of the SCoP.
If the load is only conditionally executed the preloaded version will
also only be executed under the same condition, hence we will never
access memory that wouldn't have been accessed otherwise. This is also
the most distinguishing feature to licm.
As hoisting can make statements empty we will simplify the SCoP and
remove empty statements that would otherwise cause artifacts in the
code generation.
Differential Revision: http://reviews.llvm.org/D13194
llvm-svn: 248861
The changes affect methods that are part of the Pass interface and
include:
- Comments that describe the methods purpose.
- A consistent use of the keywords override and virtual.
Additionally, the printScop method is now optional and removed from
SCoP passes that do not implement it.
llvm-svn: 248685
The TempScopInfo (-polly-analyze-ir) pass is removed and its work taken
over by ScopInfo (-polly-scops). Several tests depend on
-polly-analyze-ir and use -polly-scops instead which for the moment
prints the output of both passes. This again is not expected by some
other tests, especially those with negative searches, which have been
adapted.
Differential Version: http://reviews.llvm.org/D12694
llvm-svn: 247288
This prepares for a series of patches that merges TempScopInfo into ScopInfo to
reduce Polly's code complexity. Only ScopInfo.{cpp|h} will be left thereafter.
Moving the code of TempScopInfo in one commit makes the mains diffs simpler to
understand.
In detail, merging the following classes is planned:
TempScopInfo into ScopInfo
TempScop into Scop
IRAccess into MemoryAccess
Only moving code, no functional changes intended.
Differential Version: http://reviews.llvm.org/D12693
llvm-svn: 247274
This modifies the order in which Polly passes are executed.
Assuming a function has two scops (A and B), the order before was:
FunctionPassManager
ScopDetection
IndependentBlocks
TempScopInfo for A and B
RegionPassManager
ScopInfo for A
DependenceInfo for A
IslScheduleOptimizer for A
IslAstInfo for A
CodeGeneration for A
ScopInfo for B
DependenceInfo for B
IslScheduleOptimizer for B
IslAstInfo for B
CodeGeneration for B
After this patch:
FunctionPassManager
ScopDetection
IndependentBlocks
RegionPassManager
TempScopInfo for A
ScopInfo for A
DependenceInfo for A
IslScheduleOptimizer for A
IslAstInfo for A
CodeGeneration for A
TempScopInfo for B
ScopInfo for B
DependenceInfo for B
IslScheduleOptimizer for B
IslAstInfo for B
CodeGeneration for B
TempScopInfo for B might store information and references to the IR
that CodeGeneration for A might modify. Changing the order ensures that
the IR is not modified from the analysis of a region until code
generation.
Reviewers: grosser
Differential Revision: http://reviews.llvm.org/D12014
llvm-svn: 245091
In order to have a valid region analysis, we assign all newly created blocks to the parent of the scop's region. This is correct for any pre-existing regions (including the scop's region and its parent), but does not discover any region inside the generated code. For Polly this is not necessary because we do not want to re-run Polly on its own generated code anyway.
Reviewers: grosser
Part of Differential Revision: http://reviews.llvm.org/D11867
llvm-svn: 244608
The previous code had several problems:
For newly created BasicBlocks it did not (always) call RegionInfo::setRegionFor in order to update its analysis. At the moment RegionInfo does not verify its BBMap, but will in the future. This is fixed by determining the region new BBs belong to and set it accordingly. The new executeScopConditionally() requires accurate getRegionFor information.
Which block is created by SplitEdge depends on the incoming and outgoing edges of the blocks it connects, which makes handling its output more difficult than it needs to be. Especially for finding which block has been created an to assign a region to it for the setRegionFor problem above. This patch uses an implementation for splitEdge that always creates a block between the predecessor and successor. simplifyRegion has also been simplified by using SplitBlockPredecessors instead of SplitEdge. Isolating the entries and exits have been refectored into individual functions.
Previously simplifyRegion did more than just ensuring that there is only one entering and one exiting edge. It ensured that the entering block had no other outgoing edge which was necessary for executeScopConditionally(). Now the latter uses the alternative splitEdge implementation which can handle this situation so simplifyRegion really only needs to simplify the region.
Also, executeScopConditionally assumed that there can be no PHI nodes in blocks with one incoming edge. This is wrong and LCSSA deliberately produces such edges. However, previous passes ensured that there can be no such PHIs in exit nodes, but which will no longer hold in the future.
The new code that the property that it preserves the identity of region block (the property that the memory address of the BasicBlock containing the instructions remains the same; new blocks only contain PHI nodes and a terminator), especially the entry block. As a result, there is no need to update the reference to the BasicBlock of ScopStmt that contain its instructions because they have been moved to other basic blocks.
Reviewers: grosser
Part of Differential Revision: http://reviews.llvm.org/D11867
llvm-svn: 244606
To reduce compile time and to allow more and better quality SCoPs in
the long run we introduced scalar dependences and PHI-modeling. This
patch will now allow us to generate code if one or both of those
options are set. While the principle of demoting scalars as well as
PHIs to memory in order to communicate their value stays the same,
this allows to delay the demotion till the very end (the actual code
generation). Consequently:
- We __almost__ do not modify the code if we do not generate code
for an optimized SCoP in the end. Thus, the early exit as well as
the unprofitable option will now actually preven us from
introducing regressions in case we will probably not get better
code.
- Polly can be used as a "pure" analyzer tool as long as the code
generator is set to none.
- The original SCoP is almost not touched when the optimized version
is placed next to it. Runtime regressions if the runtime checks
chooses the original are not to be expected and later
optimizations do not need to revert the demotion for that part.
- We will generate direct accesses to the demoted values, thus there
are no "trivial GEPs" that select the first element of a scalar we
demoted and treated as an array.
Differential Revision: http://reviews.llvm.org/D7513
llvm-svn: 238070
Besides class, function and file names, we also change the command line option
from -polly-codegen-isl to just -polly-codegen. The isl postfix is a leftover
from the times when we still had the CLooG based -polly-codegen. Today it is
just redundant and we drop it.
llvm-svn: 237099
This commit drops the Cloog support for Polly. The scripts and
documentation are changed to only use isl as prerequisity. In the code
all Cloog specific parts have been removed and all relevant tests have
been ported to the isl backend when it was created.
llvm-svn: 223141
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
By adding braces into the DEBUG statement we can make clang-format format code
such as:
DEBUG(stmt1(); stmt2())
as multi-line code:
DEBUG({
stmt1();
stmt2();
});
This makes control-flow in debug statements easier to read.
llvm-svn: 220441
+ Generalized function names and comments
+ Removed OpenMP (omp) from the names and comments
+ Use common names (non OpenMP specific) for runtime library call creation
methodes
+ Commented the parallel code generator and all its member functions
+ Refactored some values and methodes
Differential Revision: http://reviews.llvm.org/D4990
llvm-svn: 219003
Summary:
+ Refactor the runtime check (RTC) build function
+ Added helper function to create an PollyIRBuilder
+ Change the simplify region function to create not
only unique entry and exit edges but also enfore that
the entry edge is unconditional
+ Cleaned the IslCodeGeneration runOnScop function:
- less post-creation changes of the created IR
+ Adjusted and added test cases
Reviewers: grosser, sebpop, simbuerg, dpeixott
Subscribers: llvm-commits, #polly
Differential Revision: http://reviews.llvm.org/D5076
llvm-svn: 217508
This previous code added in r216842 most likely created unnecessary copies.
Reported-by: Duncan P. N. Exon Smith <dexonsmith@apple.com>
llvm-svn: 217507
This reverts commit 215466 (and 215528, a trivial formatting fix).
The intention of these commits is a good one, but unfortunately they broke
our LNT buildbot:
http://lab.llvm.org:8011/builders/perf-x86_64-penryn-O3-polly-codegen-isl
Several of the cleanup changes that have been combined in this 'fixup' are
trivial and could probably be committed as obvious changes without risking to
break the build. The remaining changes are little and it should be easy to
figure out what went wrong.
llvm-svn: 215817
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
PollyIRBuilder is currently just a typedef to IRBuilder<>. Consequently, this
change should not affect behavior. In subsequent patches we will extend its
functionality to emit loop.parallel metadata.
llvm-svn: 202853
This removes the last isl_int dependency in the default build. There are
still some in OpenScop and Scoplib. For those isl-0.12.2 still needs to be used.
llvm-svn: 199585
We now use __isl_take to annotate the uses of the isl_set where we got the
memory management wrong.
Thanks to Rafael! His pipefail work hardened our test environment and exposed
this bug nicely.
llvm-svn: 187338
When the Polly code generation was written we did not correctly update the
LoopInfo data, but still claimed that the loop information is correct. This
does not only lead to missed optimizations, but it can also cause
miscompilations in case passes such as LoopSimplify are run after Polly.
Reported-by: Sergei Larin <slarin@codeaurora.org>
llvm-svn: 181987
BeforeBB
|
v
GuardBB
/ \
__ PreHeaderBB \
/ \ / |
latch HeaderBB |
\ / \ /
< \ /
\ /
ExitBB
This does not only remove the need for an explicit loop rotate pass, but it also
gives us the possibility to skip the construction of the guard condition in case
the loop is known to be executed at least once. We do not yet exploit this, but
by implementing this analysis in the isl code generator we should be able to
remove more guards than the generic loop rotate pass can. Another point is that
loop rotation can introduce additional PHI nodes, which may hide that a loop can
be executed in parallel. This change avoids this complication and will make it
easier to move the openmp code generation into a separate pass.
llvm-svn: 181986
Use the new cl::OptionCategory support to move the Polly options into a separate
option category. The aim is to hide most options and show by default only the
options a user needs to influence '-O3 -polly'. The available options probably
need some care, but here is the current status:
Polly Options:
Configure the polly loop optimizer
-enable-polly-openmp - Generate OpenMP parallel code
-polly - Enable the polly optimizer (only at -O3)
-polly-no-tiling - Disable tiling in the scheduler
-polly-only-func=<function-name> - Only run on a single function
-polly-report - Print information about the activities
of Polly
-polly-vectorizer - Select the vectorization strategy
=none - No Vectorization
=polly - Polly internal vectorizer
=unroll-only - Only grouped unroll the vectorize
candidate loops
=bb - The Basic Block vectorizer driven by
Polly
llvm-svn: 181295
Regions that have multiple entry edges are very common. A simple if condition
yields e.g. such a region:
if
/ \
then else
\ /
for_region
This for_region contains two entry edges 'then' -> 'for_region' and 'else' -> 'for_region'.
Previously we scheduled the RegionSimplify pass to translate such regions into
simple regions. With this patch, we now support them natively when the region is
in -loop-simplify form, which means the entry block should not be a loop header.
Contributed by: Star Tan <tanmx_star@yeah.net>
llvm-svn: 179586
Regions that have multiple exit edges are very common. A simple if condition
yields e.g. such a region:
if
/ \
then else
\ /
after
Region: if -> after
This regions contains the bbs 'if', 'then', 'else', but not 'after'. It has
two exit edges 'then' -> 'after' and 'else' -> 'after'.
Previously we scheduled the RegionSimplify pass to translate such regions into
simple regions. With this patch, we now support them natively.
Contributed-by: Star Tan <tanmx_star@yeah.net>
llvm-svn: 179159
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
When using the scev based code generation, we now do not rely on the presence
of a canonical induction variable any more. This commit prepares the path to
(conditionally) disable the induction variable canonicalization pass.
llvm-svn: 177548
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
generation.
We don't use the exact same way to build loop body for GPGPU codegen as openmp
codegen and other transformations do currently, in which cases 'createLoop'
function is called recursively. GPGPU codegen may fail due to improper restore
of ValueMap and ClastVars .
Contributed by: Yabin Hu <yabin.hwu@gmail.com>
llvm-svn: 168966
The detection of values that need to be copied in to the generated OpenMP
subfunction also detects the array base addresses needed in the SCoP. Hence, it
is not necessary to unconditionally copy all the base addresses to the generated
function.
Test cases are modified to reflect this change. Arrays which are global
variables do not occur in the struct passed to the subfunction anymore. A test
case for base address copy-in is added in copy_in_array.{c,ll}.
Committed with slight modifications
Contributed by: Armin Groesslinger <armin.groesslinger@uni-passau.de>
llvm-svn: 167215
In addition to the arrays and clast variables a SCoP statement may also refer to
values defined before the SCoP or to function arguments. Detect these values and
add them to the set of values passed to the function generated for OpenMP
parallel execution of a clast.
Committed with additional test cases and some refactoring.
Contributed by: Armin Groesslinger <armin.groesslinger@uni-passau.de>
llvm-svn: 167214
When generating OpenMP or GPGPU code the original ValueMap and ClastVars must be
kept. We already recovered the original ClastVars by reverting the changes, but
we did not keep the content of the ValueMap. This patch keeps now an explicit
copy of both maps and restores them after generating OpenMP or GPGPU code.
This is an adapted version of a patch contributed by:
Armin Groesslinger <armin.groesslinger@uni-passau.de>
llvm-svn: 167213
Previously isl always generated '<=' or '>='. However, in many cases '<' or '>'
leads to simpler code. This commit updates isl and adds the relevant code
generation support to Polly.
llvm-svn: 166020
Translate the selected parallel loop body into a ptx string and run it with the
cuda driver API. We limit this preliminary implementation to target the
following special test cases:
- Support only 2-dimensional parallel loops with or without only one innermost
non-parallel loop.
- Support write memory access to only one array in a SCoP.
The patch was committed with smaller changes to the build system:
There is now a flag to enable gpu code generation explictly. This was required
as we need the llvm.codegen() patch applied on the llvm sources, to compile this
feature correctly. Also, enabling gpu code generation does not require cuda.
This requirement was removed to allow 'make polly-test' runs, even without an
installed cuda runtime.
Contributed by: Yabin Hu <yabin.hwu@gmail.com>
llvm-svn: 161239
I did not take into account, that this patch fails to compile without the
llvm.codegen patch applied. This breaks buildbots.
I revert this until we found a solution to commit this without buildbots
complaining.
This reverts commit cb43ab80e94434e780a66be3b9a6ad466822fe33.
llvm-svn: 160165
Translate the selected parallel loop body into a ptx string and run it
with cuda driver API. We limit this preliminary implementation to
target the following special test cases:
- Support only 2-dimensional parallel loops with or without only one
innermost non-parallel loop.
- Support write memory access to only one array in a SCoP.
Contributed by: Yabin Hu <yabin.hwu@gmail.com>
llvm-svn: 160164
Hongbin Zheng