Summary:
This information is necessary for PPCG to perform correct life range reordering.
With these changes applied we can live-range reorder some of the important
kernels in COSMO.
We also update and rename one test case, which previously could not be optimized
and now is optimized thanks to live-range reordering. To preserve test coverage
we add a new test case scalar-writes-in-scop-requires-abort.ll, which exercises
our automatic abort in case of scalar writes in the kernel.
Reviewers: Meinersbur, bollu, singam-sanjay
Subscribers: nemanjai, pollydev, llvm-commits, kbarton
Tags: #polly
Differential Revision: https://reviews.llvm.org/D36929
llvm-svn: 311259
Kernel argument sizes now only get appended to the kernel launch parameter list if the OpenCL runtime is selected, not if CUDA runtime is chosen.
Differential revision: D36925
llvm-svn: 311248
When using -polly-ignore-integer-wrapping and -polly-acc-codegen-managed-memory
we add parameter dimensions lazily to the domains, which results in PPCG not
including parameter dimensions that are only used in memory accesses in the
kernel space. To make sure these parameters are still passed to the kernel, we
collect these parameter dimensions and align the kernel's parameter space
before code-generating it.
llvm-svn: 311239
Summary:
When trying to expand memory accesses, the current version of Polly uses statement Level dependences. The actual implementation is not working in case of multiple dependences per statement. For example in the following source code :
```
void mse(double A[Ni], double B[Nj], double C[Nj], double D[Nj]) {
int i,j;
for (j = 0; j < Ni; j++) {
for (int i = 0; i<Nj; i++)
S: B[i] = i;
for (int i = 0; i<Nj; i++)
T: D[i] = i;
U: A[j] = B[j];
C[j] = D[j];
}
}
```
The statement U has two dependences with S and T. The current version of polly fails during expansion.
This patch aims to fix this bug. For that, we use Reference Level dependences to be able to filter dependences according to statement and memory ref. The principle of expansion remains the same as before.
We also noticed that we need to bail out if load come after store (at the same position) in same statement. So a check was added to isExpandable.
Contributed by: Nicholas Bonfante <nicolas.bonfante@insa-lyon.fr>
Reviewers: Meinersbur, simbuerg, bollu
Reviewed By: Meinersbur, simbuerg
Subscribers: pollydev, llvm-commits
Differential Revision: https://reviews.llvm.org/D36791
llvm-svn: 311165
Summary:
Drop unused parameter dimensions to reduce the size of the sets we are working
with. Especially the computed dependences tend to accumulate a lot of parameters
that are present in the input memory accesses, but often not necessary to
express the actual dependences. As isl represents maps and sets with dense
matrices, reducing the dimensionality of isl sets commonly reduces code
generation performance.
This reduces compile time from 17 to 11 seconds for our test case. While this is
not impressive, this patch helped me to identify the previous two performance
improvements and additionally also increases readability of the isl data
structures we use.
Reviewers: Meinersbur, bollu, singam-sanjay
Reviewed By: bollu
Subscribers: nemanjai, pollydev, llvm-commits, kbarton
Tags: #polly
Differential Revision: https://reviews.llvm.org/D36869
llvm-svn: 311161
Summary:
They are not used and consequently do not even need to be computed. This reduces
the overall compile time for our kernel from 1m33s to 17s.
Reviewers: Meinersbur, bollu, singam-sanjay
Reviewed By: bollu
Subscribers: nemanjai, pollydev, llvm-commits, kbarton
Tags: #polly
Differential Revision: https://reviews.llvm.org/D36868
llvm-svn: 311157
Summary:
This change reduces the overall number of synchronize calls for kernels with
a lot of output data at the cost of additional synchronize calls for kernels
launched in sequence without any device to host transfers in between. As the
latter pattern is a lot less frequent, this seems a better tradeoff.
Even though the above motivation would be motivation enough, this is just
a step towards enabling ppcg to not compute to and from device copy calls
at all, which would be incorrect in case we still relied on these calls to
place our synchronization statements.
Reviewers: Meinersbur, bollu, singam-sanjay
Reviewed By: bollu
Subscribers: nemanjai, kbarton, pollydev, llvm-commits
Tags: #polly
Differential Revision: https://reviews.llvm.org/D36867
llvm-svn: 311155
This avoid the construction of very large sets and in many cases also keeps the
number of parameters low. As a result, we see a compile time reduction from 5
minutes to only slightly above 1 minute for one of our larger test cases.
llvm-svn: 311127
We add a ScopInliner pass which inlines functions based on a simple heuristic:
Let `g` call `f`.
If we can model all of `f` as a Scop, we inline `f` into `g`.
This requires `-polly-detect-full-function` to be enabled. So, the pass
asserts that `-polly-detect-full-function` is enabled.
Differential Revision: https://reviews.llvm.org/D36832
llvm-svn: 311126
Reuse the machinery built for replacing global arrays to replace malloc/free as
well. Example replacement that was missed earlier:
```
call void \
bitcast (void (i8*)* @free to void (%custom_type*)*) (%custom_type* %13)
```
- Since the `bitcast` is a `ConstantExpr`, `replaceAllUsesWith` would miss
this. We don't miss this anymore.
Differential Revision: https://reviews.llvm.org/D36825
llvm-svn: 311121
In release builds LLVM may not pass along LLVM names consistently. We make the
test cases independent of the LLVM-IR names to avoid spurious test case
failures.
llvm-svn: 311118
- If we have global arrays, we would like to rewrite them to global
pointers which are allocated using `cudaMallocManaged`.
- If we have allocas in a function, we would like to rewrite them to
heap-allocations with `cudaMallocManaged` and `cudaFree`.
- With these rewrite mechanisms, we can offload _any_ function to the
GPU with no code rewrite whatsover.
Differential Revision: https://reviews.llvm.org/D36516
llvm-svn: 311080
Summary:
This pass detangles induction variables from functions, which take variables by
reference. Most fortran functions compiled with gfortran pass variables by
reference. Unfortunately a common pattern, printf calls of induction variables,
prevent in this situation the promotion of the induction variable to a register,
which again inhibits any kind of loop analysis. To work around this issue
we developed a specialized pass which introduces separate alloca slots for
known-read-only references, which indicate the mem2reg pass that the induction
variables can be promoted to registers and consquently enable SCEV to work.
We currently hardcode the information that a function
_gfortran_transfer_integer_write does not read its second parameter, as
dragonegg does not add the right annotations and we cannot change old dragonegg
releases. Hopefully flang will produce the right annotations.
Reviewers: Meinersbur, bollu, singam-sanjay
Reviewed By: bollu
Subscribers: mgorny, pollydev, llvm-commits
Tags: #polly
Differential Revision: https://reviews.llvm.org/D36800
llvm-svn: 311066
ReportLoopHasNoExit started failing after r310940 that added
infinite loops to postdominators. The change made regions not
contain infinite loops anymore.
This patch unbreaks the polly tree by XFAILING the
ReportLoopHasNoExit test. Full fix is under review in D36776.
llvm-svn: 310980
Requesting size 0 allocations from `cuMalloc` / `cuMallocManaged` fails.
If there is a size 0 allocation that can be statically proved, the we
fail at PPCGCodeGeneration. This is because if size 0 allocation could
take place, we should not generate code that tries to use this array.
However, there are cases where we cannot statically prove this, and at
runtime we get a request for 0 bytes of memory. We choose to allocate
size 1 to allow the program to continue running.
Differential Revision: https://reviews.llvm.org/D36751
llvm-svn: 310941
Summary:
Before, if we fail to parse a jscop file, this will be reported as an
error and importing is aborted. However, this isn't actually strong
enough, since although the import is aborted, the scop has already been
modified and is very likely broken. Instead, make this a hard failure
and throw an LLVM error. This new behaviour requires small changes to
the tests for the legacy pass, namely using `not` to verify the error.
Further, fixed the jscop file for the
base_pointer_load_is_inst_inside_invariant_1 testcase.
Reviewed By: Meinersbur
Split out of D36578.
llvm-svn: 310599
Summary:
I pulled out all functionality into static functions, and use those both
in the legacy passes and in the new ones.
Reviewers: grosser, Meinersbur, bollu
Reviewed By: Meinersbur
Subscribers: llvm-commits, pollydev
Differential Revision: https://reviews.llvm.org/D36578
llvm-svn: 310597
Summary:
During code generation for a Scop we modify the IR of a function.
While this shouldn't affect a Scop in the formal sense, the implementation
caches various information about the IR such as SCEV expressions for bounds or
parameters. This cached information needs to be updated or invalidated. To this
end, SPMUpdater allows passes to report when they've invalidated a Scop to the
PassManager, which will then flush and recompute all Scops. This in turn
invalidates all iterators, so references to Scops shouldn't be held.
Reviewers: grosser, Meinersbur, bollu
Reviewed By: grosser
Subscribers: llvm-commits, pollydev
Differential Revision: https://reviews.llvm.org/D36524
llvm-svn: 310551
We are working towards removing uses of Scop::getStmtFor(BB). In this
patch, we remove dependency of Scop::getStmtFor(Inst) on getStmtFor(BB).
To do so, we introduce a map of instructions to their corresponding scop
statements and use it to get the instructions' statement.
Contributed-by: Nandini Singhal <cs15mtech01004@iith.ac.in>
Differential Revision: https://reviews.llvm.org/D35663
llvm-svn: 310494
This pass is useful to automatically convert a codebase that uses malloc/free
to use their managed memory counterparts.
Currently, rewrite malloc and free to the `polly_{malloc,free}Managed` variants.
A future patch will teach ManagedMemoryRewrite to rewrite global arrays
as pointers to globally allocated managed memory.
Differential Revision: https://reviews.llvm.org/D36513
llvm-svn: 310471
Codegen with -polly-parallel queried the unmapped MemoryAccess, but only
the MemoryKind after mapping is relevant for codegen.
This should fix various fails of the
perf-x86_64-penryn-O3-polly-parallel-fast buildbot.
llvm-svn: 310466
The previous value of "polly-delicm" was forgotten to to be changed when
ForwardOpTree was split from DeLICM.
Thanks to Tobias for noticing!
llvm-svn: 310465
distributeDomain() and filterKnownValInst() are used in a scop
of ForwardOpTree that limits the number of isl operations.
Therefore some isl functions may return null after any operation.
Remove assertion that assume non-null results and handle
isl_*_foreach returning isl::stat::error.
I hope this fixes the crash of the asop buildbot at ihevc_recon.c.
llvm-svn: 310461
Previously, we used to compute this with `elementSizeInBits / 8`. This
would yield an element size of 0 when the array had element size < 8 in
bits.
To fix this, ask data layout what the size in bytes should be.
Differential Revision: https://reviews.llvm.org/D36459
llvm-svn: 310448
We introduce another level of alias metadata to distinguish the individual
non-aliasing accesses that have inter iteration alias-free base pointers
marked with "Inter iteration alias-free" mark nodes. To distinguish two
accesses, the comparison of raw pointers representing base pointers is used.
In case of, for example, ublas's prod function that implements GEMM, and
DeLiCM we can get accesses to same location represented by different raw
pointers. Consequently, we create different alias sets that can prevent
accesses from, for example, being sinked or hoisted.
To avoid the issue, we compare the corresponding SCEV information instead
of the corresponding raw pointers.
Reviewed-by: Tobias Grosser <tobias@grosser.es>
Differential Revision: https://reviews.llvm.org/D35761
llvm-svn: 310380
Currently, only convex isolation sets can be efficiently processed by isl.
Consequently, as a temporary solution, we use a different algorithm for partial
tile isolation that helps to build convex isolation sets in some cases.
Reviewed-by: Tobias Grosser <tobias@grosser.es>
Differential Revision: https://reviews.llvm.org/D36278
llvm-svn: 310374
This allows us to get rid of stores that are overwritten within the very same
basic block, without ever being read beforehand. This simplification is
necessary for delicm to run on pb4's correlation.
llvm-svn: 310369
To do this, we replicate what `CodeGeneration` does. We expose
`markNodeUnreachable` from `CodeGeneration` to `PPCGCodeGeneration`.
Differential Revision: https://reviews.llvm.org/D36457
llvm-svn: 310350
It is possible that partial writes are empty (write is never executed).
In this case, when in PHINode's incoming edge is never taken such that
the incoming write becomes an empty partial write, if enabled. The
issue is that when converting the union_map to an map, it's space
cannot be derived from the union_map itself. Rather, we need to
determine its space independently.
This fixes test-suite's MultiSource/Benchmarks/ASC_Sequoia/CrystalMk.
llvm-svn: 310348
- It's useful to know the amount of memory asked for since, for example,
asking for `0` bytes of memory is illegal.
- Line number is helpful since we print the same message in the function
at different points.
llvm-svn: 310340
In certain cases delicm might decide to not leave the original array write in
the loop body, but to remove it and instead leave a transformed phi node as
write access. This commit teached the matmul pattern detection to order the
memory accesses according to when the access actually happens and use this
information to detect the new pattern. This makes pattern based matmul
optimization work for 2mm and 3mm in polybench 4 after
polly-position=before-vectorizer has been enabled.
llvm-svn: 310338
Polly has traditionally always been executed at the beginning of the pass
pipeline as LLVM's inliner and DeLICM passes introduced plenty of scalar
dependences which prevented any kind of useful high-level loop optimizations
later in the pass pipeline. With DeLICM now being available, Polly can also
run optimizations when folded into the pass pipeline. This has the benefit
that Polly should now be more effective on C++ code and as an additional bonus,
no additional early canonicalization phase must be run. As a result, Polly
touches the code only if it applies a transformation. Code that does not
benefit from Polly is not touched and consequently will have the very same
execution time as without Polly enabled. Random performance changes, as could
sometimes be observed with polly-position=early are consequently not possible
any more. If performance is changed, this is due to Polly is choosing to
perform a transformation. If this choice is wrong, it can be fixed directly
in Polly.
http://polly.llvm.org/docs/Architecture.html#polly-in-the-llvm-pass-pipeline
llvm-svn: 310319
This allows us to remove more scalar dependences. While this feature is still
rather experimental, we want to give it sufficient test coverage.
llvm-svn: 310314
Two write statements which write into the very same array slot generally are
conflicting. However, in case the value that is written is identical, this
does not cause any problem. Hence, allow such write pairs in this specific
situation.
llvm-svn: 310311
This commit implements the initial version of fully-indexed static
expansion.
```
for(int i = 0; i<Ni; i++)
for(int j = 0; j<Ni; j++)
S: B[j] = j;
T: A[i] = B[i]
```
After the pass, we want this :
```
for(int i = 0; i<Ni; i++)
for(int j = 0; j<Ni; j++)
S: B[i][j] = j;
T: A[i] = B[i][i]
```
For now we bail (fail) in the following cases:
- Scalar access
- Multiple writes per SAI
- MayWrite Access
- Expansion that leads to an access to the original array
Furthermore: We still miss checks for escaping references to the array
base pointers. A future commit will add the missing escape-checks to
stay correct in those cases. The expansion is still locked behind a
CLI-Option and should not yet be used.
Patch contributed by: Nicholas Bonfante <bonfante.nicolas@gmail.com>
Reviewers: simbuerg, Meinersbur, bollu
Reviewed By: Meinersbur
Subscribers: mgorny, llvm-commits, pollydev
Differential Revision: https://reviews.llvm.org/D34982
llvm-svn: 310304
This is an addition to the -polly-optree pass that reuses the array
content analysis from DeLICM to find array elements that contain the
same value as the value loaded when the target statement instance
is executed.
The analysis is now enabled by default.
The known content analysis could also be used to rematerialize any
llvm::Value that was written to some array element, but currently
only loads are forwarded.
Differential Revision: https://reviews.llvm.org/D36380
llvm-svn: 310279
This update is mostly a maintenance update, but also exposes a couple of new
functions that will be needed for the next version of the isl++ bindings.
llvm-svn: 310205
Summary:
Small patch to fix the JSON exporter.
Currently, using "opt -polly-export-jscop" does not generate jscop files, but gives an error:
*** Error in `opt': corrupted double-linked list: 0x0000000000bc4bb0 ***
Updated the function getAccessRelationStr() to work with the current version of getAccessRelation(), fixing the JSON exporter
Reviewers: bollu, grosser
Reviewed By: grosser
Subscribers: grosser, llvm-commits, pollydev
Tags: #polly
Differential Revision: https://reviews.llvm.org/D36370
llvm-svn: 310199
Summary:
This resolves some "instruction does not dominate use" errors, as we used to
prepare the arrays at the location of the first kernel, which not necessarily
dominated all other kernel calls.
Reviewers: Meinersbur, bollu, singam-sanjay
Subscribers: nemanjai, pollydev, llvm-commits, kbarton
Differential Revision: https://reviews.llvm.org/D36372
llvm-svn: 310196
A Scop with a loop outside it is not handled currently by
PPCGCodeGeneration. The test case is such that the Scop has only one inner loop
that is detected. This currently breaks codegen.
The fix is to reuse the existing mechanism in `IslNodeBuilder` within
`GPUNodeBuilder.
Differential Revision: https://reviews.llvm.org/D36290
llvm-svn: 310193
This logic is duplicated, so we refactor it into a separate function.
This will be used in a later patch to teach PPCGCodeGen code generation
for loops that are outside the scop.
Differential Revision: https://reviews.llvm.org/D36310
llvm-svn: 310192
In https://reviews.llvm.org/D36278 it was pointed out that the behavior of
getPartialTilePrefixes is not very well understood. To allow for a better
understanding, we first provide some basic unittests.
llvm-svn: 310175
The method forwardSpeculatable forwards speculatively executable
instructions and is currently the only way to forward an
instruction.
In the future we intend to add more methods.
llvm-svn: 310056
Summary:
Testing the new-pm passes becomes much easier once they behave more like the
old passes in terms of the order in which Scops are processed and printed. This
requires three changes:
- ScopInfo: Use an ordered map to store scops
- ScopInfo: Iterate and print Scops in reverse order to match legacy PM behaviour
- ScopDetection: print function name in ScopAnalysisPrinter
Reviewers: grosser, Meinersbur, bollu
Reviewed By: grosser
Subscribers: pollydev, llvm-commits
Differential Revision: https://reviews.llvm.org/D36303
llvm-svn: 310052
The complication of bspatch.cc of the AOSP buildbot currently fails
presumably because the occurance of a MetadataAsValue in an operand.
This kind of value can occur as operands of intrinsics, the typical
example being the debug intrinsics.
Polly currently ignores the debug intrinsics and it is not yet clear
which other intrinic might occur. For such cases, and to unbreak the
AOSP buildbot, treat a MetadataAsValue as a constant because it can be
referenced without modification in generated code.
llvm-svn: 309992
With this patch, we get rid of the last use of getStmtFor(BB). Here
this is done by getting the last statement of the incoming block in
case the user is a phi node; otherwise just fetching the statement
comprising the instruction for which the virtual use is being created.
Differential Revision: https://reviews.llvm.org/D36268
llvm-svn: 309947
Summary:
In case the option -polly-ignore-parameter-bounds is set, not all parameters
will be added to context and domains. This is useful to keep the size of the
sets and maps we work with small. Unfortunately, for AST generation it is
necessary to ensure all parameters are part of the schedule tree. Hence,
we modify the GPGPU code generation to make sure this is the case.
To obtain the necessary information we expose a new function
Scop::getFullParamSpace(). We also make a couple of functions const to be
able to make SCoP::getFullParamSpace() const.
Reviewers: Meinersbur, bollu, gareevroman, efriedma, huihuiz, sebpop, simbuerg
Subscribers: nemanjai, kbarton, pollydev, llvm-commits
Tags: #polly
Differential Revision: https://reviews.llvm.org/D36243
llvm-svn: 309939
When we have `-polly-ignore-parameter-bounds`, `Scop::Context` does not contain
all the paramters present in the program.
The construction of the `isl_multi_pw_aff` requires all the indivisual `pw_aff`
to have the same parameter dimensions. To achieve this, we used to realign
every `pw_aff` with `Scop::Context`. However, in conjunction with
`-polly-ignore-parameter-bounds`, this is now incorrect, since `Scop::Context`
does not contain all parameters.
We set this up correctly by creating a space that has all the parameters
used by all the `isl_pw_aff`. Then, we realign all `isl_pw_aff` to this space.
llvm-svn: 309934
These passes have been tested over the last month and should generally help
to remove scalar data dependences in Polly. We enable them to give them even
wider test coverage. Large performance regressions and any kind of correctness
regressions are not expected.
llvm-svn: 309878
When compiling with clang, explicit instantiation of the
OwningScopAnalysisManagerFunctionProxy needs to happen within the polly
namespace. Same goes with the specialization of its run method.
llvm-svn: 309835
Summary:
This patch is a first attempt at registering Polly passes with the LLVM tools. Tool plugins are still unsupported, but this registration is usable from the tools if Polly is linked into them (albeit requiring minimal patches to those tools). Registration requires a small amount of machinery (the owning analysis proxies), necessary for injecting ScopAnalysisManager objects into the calling tools.
This patch is marked WIP because the registration is incomplete. Parsing manual pipelines is fully supported, but default pass injection into the O3 pipeline is lacking, mostly because there is opportunity for some redesign here, I believe. The first point of order would be insertion points. I think it makes sense to run before the vectorizers. Running Polly Early, however, is weird. Mostly because it actually is the default (which to me is unexpected), and because Polly runs it's own O1 pipeline. Why not instead insert it at an appropriate place somewhere after simplification happend? Running after the loop optimizers seems intuitive, but it also seems wasteful, since multiple consecutive loops might well be a single scop, and we don't need to run for all of them.
My second request for comments would be regarding all those smallish helper passes we have, like PollyViewer, PollyPrinter, PollyImportJScop. Right now these are controlled by command line options, deciding whether they should be part of the Polly pipeline. What is your opinion on treating them like real passes, and have the user write an appropriate pipeline if they want to use any of them?
Reviewers: grosser, Meinersbur, bollu
Reviewed By: grosser
Subscribers: llvm-commits, pollydev
Tags: #polly
Differential Revision: https://reviews.llvm.org/D35458
llvm-svn: 309826
Summary:
**Remove debug metadata from instruction to be copied to prevent the source file's debug metadata being copied into GPUModule and eventually failing Module verification and ASM string codegeneration.**
When copying the instruction onto the Module meant for the GPU, debug metadata attached to an instruction causes all related metadata to be pulled into the Module, including the DICompileUnit, which is not listed in llvm.dbg.cu of the Module. This fails the verification of the Module and generation of the ASM string.
The only debug metadata of the instruction, the DebugLoc, is unset by this patch.
This patch reattempts https://reviews.llvm.org/D35630 by targeting only those instructions that are to end up in a Module meant for the GPU.
Reviewers: grosser, bollu
Reviewed By: grosser
Subscribers: pollydev
Tags: #polly
Differential Revision: https://reviews.llvm.org/D36161
llvm-svn: 309822
Summary: I made a mistake in handling transitive invalidation of analysis results. I've updated the list of preserved analyses as well as the correct result dependences.
The Invalidator passed through the invalidate() path can be used to
transitively invalidate analyses. It frequently happens that analysis
results depend on other analyses, and thus store references to their
results. When the dependee now gets invalidated, the depender needs to
be invalidated as well. This is the purpose of the Invalidator object,
which can be used to check whether some dependee analysis is in the
process of being invalidated. I originally was checking the wrong
dependee analyses, which is an actual error, you can only check analysis
results that are in the cache (which they are if you've captured their
reference). The invalidation I'm handling inside the proxy deals with
the standard analyses the proxy passes into the Scop pipeline, since I'm
capturing their reference.
This checking allows us to actually preserve a couple of results outside
of the proxy, since the Scop pipeline shouldn't break those, or
otherwise should update them accordingly.
Reviewers: grosser, Meinersbur, bollu
Reviewed By: grosser
Subscribers: pollydev, llvm-commits
Differential Revision: https://reviews.llvm.org/D36216
llvm-svn: 309811
We introduce `polly_mallocManaged` and `polly_freeManaged` as
proxies for `cudaMallocManaged` / `cudaFree`. This is currently not
used by Polly. It is auxiliary code that is used in `COSMO`.
This is useful because `polly_mallocManaged` matches the signature of `malloc`,
while `cudaMallocManaged` does not. We introduce `polly_freeManaged` for
symmetry.
We use this in COSMO to use the unified memory feature of the newer
CUDA APIs (>= 6).
Differential Revision: https://reviews.llvm.org/D35991
llvm-svn: 309808
On mixing the driver and runtime APIs, it is quite possible that a
context already exists due to runtime API usage. In this case, Polly should
try to use the same context.
This patch teaches GPUJIT to detect that a context exists and how to
pick up this context.
Without this, calling `cudaMallocManaged`, for example, before a
polly-generated kernel launch causes P100 to *hang*.
This is a part of (https://reviews.llvm.org/D35991) that was extracted
out.
Differential Revision: https://reviews.llvm.org/D36162
llvm-svn: 309802
Commit r309730 moved the call to canForwardTree into an assert(), even
though this function has side-effects if its DoIt parameter is true. To
avoid a warning in release builds, do an (void)Execution of its result
instead.
To avoid such confusion in the future, rename
canForwardTree() to forwardTree().
llvm-svn: 309753
The previous algorithm was to search a writes and the sours of its value
operand, and see whether the write just stores the same read value back,
which includes a search whether there is another write access between
them. This is O(n^2) in the max number of accesses in a statement
(+ the complexity of isl comparing the access functions).
The new algorithm is more similar to the one used for searching for
overwrites and coalescable writes. It scans over all accesses in order
of execution while tracking which array elements still have the same
value since it was read. This is O(n), not counting the complexity
within isl. It should be more reliable than trying to catch all
non-conforming cases in the previous approach. It is also less code.
We now also support if the write is a partial write of the read's
domain, and to some extent non-affine subregions.
Differential Revision: https://reviews.llvm.org/D36137
llvm-svn: 309734
With a lot of reads and writes to the same array in a statement,
some isl sets that capture the state between access can become
complex such that isl takes more considerable time and memory
for operations on them.
The problems identified were:
- is_subset() takes considerable time with many disjoints in the
arguments. We limit the number of disjoints to 4, any additional
information is thrown away.
- subtract() can lead to many disjoints. We instead assume that any
array element is possibly accessed, which removes all disjoints.
- subtract_domain() may lead to considerable processing, even if all
elements are are to be removed. Instead, we remove determine and
remove the affected spaces manually. No behaviour is changed.
llvm-svn: 309728
This is necessary to get accurate traces from `nvprof` / `nvcc`.
Otherwise, we lose some profiling information.
Differential Revision: https://reviews.llvm.org/D35940
llvm-svn: 309682
It is possible that the `HostPtr` that coresponds to an array could be
invariant load hoisted. Make sure we use the invariant load hoisted
value by using `IslNodeBuilder::getLatestValue`.
Differential Revision: https://reviews.llvm.org/D36001
llvm-svn: 309681
We populate `IslNodeBuilder::ValueMap` which contains replacements for
`llvm::Value`s. There was no simple method to pick up a replacement if
it exists, otherwise fall back to the original.
Create a method `IslNodeBuilder::getLatestValue` which provides this
functionality.
This will be used in a later patch to fix bugs in `PPCGCodeGeneration`
where the latest value is not being used.
Differential Revision: https://reviews.llvm.org/D36000
llvm-svn: 309674
This allows -polly-optree to move instructions that depend on
synthesizable values.
The difficulty for synthesizable values is that their value depends on
the location. When it is moved over a loop header, and the SCEV
expression depends on the loop induction variable (SCEVAddRecExpr), it
would use the current induction variable instead of the last one.
At the moment we cannot forward PHI nodes such that crossing the header
of loops referenced by SCEVAddRecExpr is not possible (assuming the loop
header has at least two incoming blocks: for entering the loop and the
backedge, such any instruction to be forwarded must have a phi between
use and definition).
A remaining issue is when the forwarded value is used after the loop,
but is only synthesizable inside the loop. This happens e.g. if
ScalarEvolution is unable to determine the number of loop iterations or
the initial loop value. We do not forward in this situation.
Differential Revision: https://reviews.llvm.org/D36102
llvm-svn: 309609
In addition to array and PHI writes, also allow scalar value writes.
The only kind of write not allowed are writes by functions
(including memcpy/memmove/memset).
llvm-svn: 309582
Summary:
This allows us to map functions such as exp, expf, expl, for which no
LLVM intrinsics exist. Instead, we link to NVIDIA's libdevice which provides
high-performance implementations of a wide range of (math) functions. We
currently link only a small subset, the exp, cos and copysign functions. Other
functions will be enabled as needed.
Reviewers: bollu, singam-sanjay
Reviewed By: bollu
Subscribers: tstellar, tra, nemanjai, pollydev, mgorny, llvm-commits, kbarton
Tags: #polly
Differential Revision: https://reviews.llvm.org/D35703
llvm-svn: 309560
This reverts commit r309490 as it triggers on our AOSP buildbut error messages
of the form:
inlinable function call in a function with debug info must have a !dbg location
llvm-svn: 309556
Summary:
**Remove debug metadata from instruction to be copied to prevent the source file's debug metadata being copied into GPUModule and eventually failing Module verification and ASM string codegeneration.**
When copying the instruction onto the Module meant for the GPU, debug metadata attached to an instruction causes all related metadata to be pulled into the Module, including the DICompileUnit, which is not listed in llvm.dbg.cu of the Module. This fails the verification of the Module and generation of the ASM string.
The only debug metadata of the instruction, the DebugLoc, is unset by this patch.
Reviewers: grosser, bollu, Meinersbur
Reviewed By: grosser, bollu
Subscribers: pollydev
Tags: #polly
Differential Revision: https://reviews.llvm.org/D35630
llvm-svn: 309490
Write coalescing combines write accesses that
- Write the same llvm::Value.
- Write to the same array.
- Unless they do not write anything in a statement instance (partial
writes), write to the same element.
- There is no other access between them that accesses the same element.
This is particularly useful after DeLICM, which leaves partial writes to
disjoint domains.
Differential Revision: https://reviews.llvm.org/D36010
llvm-svn: 309489
ScopStmt::contains is currently implemented on the basis of BasicBlock
and does not take the instruction list into account. Therefore any
instruction copied by -polly-optree into another statement currently
triggers that assertion.
Remove that assertion for now. We might re-enable it when the
implementation of ScopStmt::contains changes.
llvm-svn: 309421
This fixes a bug in isl_flow where triggering the compute out could result in
undefined or unexpected behavior. This fixes some recent regressions we saw
in the android buildbots. Thanks Eli Friedman for reducing the corresponding
test cases.
llvm-svn: 309274
ScopBuilder and Simplify (through VirtualInstruction.cpp) previously
used this functionality in their own implementation. Refactor them
both into a common one into the Scop class.
BlockGenerator also makes use of a similiar functionality, but also
records outside users and takes place after region simplification.
Merging it as well would be more complicated.
llvm-svn: 309273
After region exit simplification, the incoming block of a phi node in
the SCoP region's exit block lands outside of the region. Since we
treat SCoPs as if this already happened, we need to account for that
when looking for outside uses of scalars (i.e. escaping scalars).
llvm-svn: 309271
A region statement's instruction list is always empty and ignored by the code
generator. Don't give the impression that it means anything.
llvm-svn: 309197
A PHI node's incoming block is the user of its operand, not the PHI's parent.
Assuming the PHINode's parent being the user lead to the removal of a
MemoryAccesses because its use was assumed to be inside of the SCoP.
llvm-svn: 309164
Translate the ScheduleOptimizer to use the new isl C++ bindings.
Reviewed-by: Michael Kruse <llvm@meinersbur.de>
Differential Revision: https://reviews.llvm.org/D35845
llvm-svn: 309119
Since there will be no more a 1:1 correspondence between statements and
basic blocks, we would like to get rid of the method getStmtFor(BB)
and its uses. Here we remove one of its uses in ScopInfo by fetching
the statement in which the call instruction lies.
Contributed-by: Nandini Singhal <cs15mtech01004@iith.ac.in>
Differential Revision: https://reviews.llvm.org/D35691
llvm-svn: 309110
In the following loop:
int i;
for (i = 0; i < func(); i+=1)
;
SCoP:
for (int j = 0; j<n; j+=1)
S(i, j)
The value i is synthesizable in the SCoP that includes only the j-loop.
This is because i is fixed within the SCoP, it is irrelevant whether
it originates from another loop.
This fixes a strange case where a PHI was synthesiable in a SCoP,
but not its incoming value, triggering an assertion.
This should fix MultiSource/Applications/sgefa/sgefa of the
perf-x86_64-penryn-O3-polly-before-vectorizer-unprofitable buildbot.
llvm-svn: 309109
It seems I still had some incomplete changes in the tree when committing.
In general, we only import changes from isl upstream. In this case, the
changes were especially unfortunate, as they broke the error management
in isl_flow.c and consequently caused regressions.
Thanks to Michael Kruse for spotting this mistake.
llvm-svn: 309039
In future, there will be no more a 1:1 correspondence between statements
and basic blocks, the name `contains` does not correctly capture their
relationship. A BB may infact comprise of multiple statements; hence we
describe a statement 'representing' a basic block.
Differential Revision: https://reviews.llvm.org/D35838
llvm-svn: 308982
Summary:
This consists instances of two changes:
- Accept any order of checks for a specific loop form, that appear in different order in the new vs legacy-PM.
- Remove checks for specific regions.
Reviewers: grosser
Reviewed By: grosser
Subscribers: pollydev, llvm-commits
Tags: #polly
Differential Revision: https://reviews.llvm.org/D35837
llvm-svn: 308976
Invariant load hoisted scalars, and arrays whose size we can statically compute
to be 0 do not need to be allocated as arrays.
Invariant load hoisted scalars are sent to the kernel directly as parameters.
Earlier, we used to allocate `0` bytes of memory for these because our
computation of size from `PPCGCodeGeneration::getArraySize` would result in `0`.
Now, since we don't invariant loads as arrays in PPCGCodeGeneration, this
problem does not occur anymore.
Differential Revision: https://reviews.llvm.org/D35795
llvm-svn: 308971
Read-only values (values defined before the SCoP) require special
handing with -polly-analyze-read-only-scalars=true (which is the
default). If active, each use of a value requires a read access.
When a copied value uses a read-only value, we must also ensure that
such a MemoryAccess is available or is created.
Differential Revision: https://reviews.llvm.org/D35764
llvm-svn: 308876
Summary:
- We were using `.count` in `StringRef`, which matches substrings.
- We may want to use this for equality as well.
- Generalise this, so allow regexes as a parameter to `polly-only-func`.
Differential Revision: https://reviews.llvm.org/D35728
llvm-svn: 308875
If the access relation's domain is empty, the access will never be
executed. We can just remove it.
We only remove write accesses. Partial read accesses are not yet
supported and instructions in the statement might require the
llvm::Value holding the read's result to be defined.
llvm-svn: 308830
Change the indention of the last brace to align with the opening line.
Before:
Instructions {
%val = fadd double %arg, 2.100000e+01
store double %val, double* %A
}
After:
Instructions {
%val = fadd double %arg, 2.100000e+01
store double %val, double* %A
}
llvm-svn: 308828
Hoisted loads can be trivially supported because there are no
MemoryAccess to be modified, the loaded value is just available
at code generation.
llvm-svn: 308826
This pass 'forwards' operand trees into statements that use them in
order to avoid scalar dependencies.
This minimal implementation handles only the case of speculatable
instructions. We will successively add support for:
- Hoisted loads
- Read-only values
- Synthesizable values
- Loads
- PHIs
- Forwarding only parts of the tree
Differential Revision: https://reviews.llvm.org/D35754
llvm-svn: 308825
Summary:
For the ScopInfo lit testsuite, this patch removes some dependences on output behaviour of the legacy PM.
In most cases, these tests checked the tool output for labels created by the pass printer in the legacy PM. This doesn't work for the new PM anymore. Untangling the testcases is the first step to porting the testsuite for the new PM infrastructure.
Reviewers: grosser, Meinersbur, bollu
Reviewed By: grosser
Subscribers: llvm-commits, pollydev
Tags: #polly
Differential Revision: https://reviews.llvm.org/D35727
llvm-svn: 308754
Summary:
Added SPIR Code Generation to the PPCG Code Generator. This can be invoked using
the polly-gpu-arch flag value 'spir32' or 'spir64' for 32 and 64 bit code respectively.
In addition to that, runtime support has been added to execute said SPIR code on Intel
GPU's, where the system is equipped with Intel's open source driver Beignet (development
version). This requires the cmake flag 'USE_INTEL_OCL' to be turned on, and the polly-gpu-runtime
flag value to be 'libopencl'.
The transformation of LLVM IR to SPIR is currently quite a hack, consisting in part of regex
string transformations.
Has been tested (working) with Polybench 3.2 on an Intel i7-5500U (integrated graphics chip).
Reviewers: bollu, grosser, Meinersbur, singam-sanjay
Reviewed By: grosser, singam-sanjay
Subscribers: pollydev, nemanjai, mgorny, Anastasia, kbarton
Tags: #polly
Differential Revision: https://reviews.llvm.org/D35185
llvm-svn: 308751
Print a statement's instruction on dump() regardless of
-polly-print-instructions. dump() is supposed to be used in the debugger
only and never in regression tests. While debugging, get all the
information we have and we are not bound to break anything. For non-dump
purposes of print, forward the setting of -polly-print-instructions as
parameters.
Some calls to print() had to be changed because the
PollyPrintInstructions setting is only available in ScopInfo.cpp.
In ScheduleOptimizer.cpp, dump() was used in regression tests.
That's not what dump() is for.
The print parameter "PrintInstructions" will also be useful for an
explicit print SCoP pass in a future patch.
llvm-svn: 308746
When performing invariant load hoisting we check that invariant load expressions
are not too complex. Up to this commit, we performed this check by counting the
sum of dimensions in the access range as a very simple heuristic. This heuristic
is a little too conservative, as it prevents hoisting for any scops with a
very large number of parameters. Hence, we update the heuristic to only count
existentially quantified dimensions and set dimensions. We expect this to still
detect the problematic expressions in h264 because of which this check was
originally introduced.
For some unknown reason, this complexity check was originally committed in
IslNodeBuilder. It really belongs in ScopInfo, as there is no point in
optimizing a program which we could have known earlier cannot be code generated.
The benefit of running the check early is that we can avoid to even hoist checks
that are expensive to code generate as invariant loads. This can be seen in
the changed tests, where we now indeed detect the scop, but just not invariant
load hoist the complicated access.
We also improve the formatting of the code, document it, and use isl++ to
simplify expressions.
llvm-svn: 308659
Visual Studio, even the 2017 version, does not support C99 VLAs.
For VLA paramters, the length of the outermost dimension is not
required anyway, so remove it.
llvm-svn: 308643
When constructing a schedule true and there are multiple statements for
a basic block, create a sequence node for these statements.
Contributed-by: Nandini Singhal <cs15mtech01004@iith.ac.in>
Differential Revision: https://reviews.llvm.org/D35679
llvm-svn: 308635
We are working towards removing uses of Scop::getStmtFor(BB). In this
patch, we remove dependency of Scop::getLastStmtFor(BB) on
getStmtFor(BB). To do so, we get the list of all statements
corresponding to the BB and then fetch the last one.
Contributed-by: Nandini Singhal <cs15mtech01004@iith.ac.in>
Differential Revision: https://reviews.llvm.org/D35665
llvm-svn: 308633
Introduce previously missing PHIReads analogous the the already existing
PHIWrites/ValueWrites/ValueReads maps. PHIReads was initially not
required and the later introduced lookupPHIReadOf() used a linear
search instead.
With PHIReads, lookupPHIReadOf() can now also do a map lookup and remove
any surprising performance/behaviour differences to lookupPHIWriteOf(),
lookupValueWriteOf() and lookupValueReadOf().
llvm-svn: 308630
Use a mark-and-sweep algorithm to find and remove unused instructions
and MemoryAccesses. This is useful in particular to remove scalar
writes that are never used anywhere. A scalar write in a loop induces
a write-after-write dependency that stops the loop iterations to be
rescheduled. Such writes can be a result of previous transformations
such as DeLICM and operand tree forwarding.
It adds a new class VirtualInstruction that represents an instruction in
a particular statement. At the moment an instruction can only belong to
the statement that represents a BasicBlock. In the future, instructions
can be in one of multiple statements representing a BasicBlock
(Nandini's work), in different statements than its BasicBlock would
indicate, and even multiple statements at once (by forwarding operand
trees). It also integrates nicely with the VirtualUse class.
ScopStmt::contains(Instruction*) currently uses the instruction's parent
BasicBlock to check whether it contains the instruction. It will need to
check the actual statement list when one of the aforementioned features
become possible.
Differential Revision: https://reviews.llvm.org/D35656
llvm-svn: 308626
This commit *WILL COMPILE*.
1. `PPCG` now uses `isl_multi_pw_aff` instead of an array of `pw_aff`.
This needs us to adjust how we index array bounds and how we construct
array bounds.
2. `PPCG` introduces two new kinds of nodes: `init_device` and `clear_device`.
We should investigate what the correct way to handle these are.
3. `PPCG` has gotten smarter with its use of live range reordering, so some of
the tests have a qualitative improvement.
4. `PPCG` changed its output style, so many test cases need to be updated to
fit the new style for `polly-acc-dump-code` checks.
Differential Revision: https://reviews.llvm.org/D35677
llvm-svn: 308625
- This commit *WILL NOT COMPILE*. `PPCGCodeGeneration` requires changes
since some of PPCG's internal data structures have been modified.
- Has polly-speific changes to PPCG. Polly exports certain functionality that
is private to PPCG. It also creates stubs for large parts of the pet API as
well as other functions in `ppcg/external.c` to keep the linker happy.
- This commit includes changes to CMakeLists.txt.
Differential Revision: https://reviews.llvm.org/D35676
llvm-svn: 308624
- This commit *WILL NOT COMPILE*, as it checks in vanilla PPCG 0.07
- We choose to introduce this commit into the history to cleanly display
the Polly-specific changes made to PPCG.
Differential Revision: https://reviews.llvm.org/D35675
llvm-svn: 308623
Since there will be no more a 1:1 correspondence between statements
and basic blocks, we would like to get rid of the method getStmtFor(BB)
and its uses. Here we remove one of its uses in ScopBuilder by fetching
the statement in which the instruction lies.
Contributed-by: Nandini Singhal <cs15mtech01004@iith.ac.in>
Differential Revision: https://reviews.llvm.org/D35610
llvm-svn: 308610
This is one possible solution to implement wrap-arounds for integers in
unsigned icmp operations. For example,
store i32 -1, i32* %A_addr
%0 = load i32, i32* %A_addr
%1 = icmp ult i32 %0, 0
%1 should hold false, because under the assumption of unsigned integers,
-1 should wrap around to 2^32-1. However, previously. it was assumed
that the MSB (Most Significant Bit - aka the Sign bit) was never set for
integers in unsigned operations.
This patch modifies the buildConditionSets function in ScopInfo.cpp to
give better information about the integers in these unsigned
comparisons.
Contributed-by: Annanay Agarwal <cs14btech11001@iith.ac.in>
Differential Revision: https://reviews.llvm.org/D35464
llvm-svn: 308608
Before this patch, ScalarDefUseChain was a tool used by DeLICM to find
all reads and writes of scalar accesses. It iterated once over all
accesses and stores the accesses into maps.
By integrating it into the Scop class, we can keep the maps up-to-date
without the need for recomputing them. It will be needed for more than
DeLICM in the future, such as SCoP simplification, code movement between
virtual statements, and array expansion (GSoC project).
Compared to ScalarUseDefChain, we save two maps by finding the ScopStmt
a Def/PHIRead must reside in, and use its already existing lookup
function to find the MemoryAccess.
Differential Revision: https://reviews.llvm.org/D35631
llvm-svn: 308495
Some optimizations (e.g., DeLICM) can modify memory accesses (e.g., change
their MemoryKind). Consequently, the pattern matching should take it into
the account.
Reviewed-by: Tobias Grosser <tobias@grosser.es>,
Michael Kruse <llvm@meinersbur.de>
Differential Revision: https://reviews.llvm.org/D33138
llvm-svn: 308494
Summary:
This makes code more readable and allows to reuse this functionality in
the future at other places.
Suggested-by Michael Kruse in post-commit review of r307660.
Reviewers: Meinersbur, bollu, gareevroman, efriedma, huihuiz, sebpop, simbuerg
Reviewed By: Meinersbur
Subscribers: pollydev, llvm-commits
Tags: #polly
Differential Revision: https://reviews.llvm.org/D35585
llvm-svn: 308435
Once statements are split, a BasicBlock will comprise of multiple
statements. To prepare for this change in future, we introduce a list
of statements in the statement map.
Contributed-by: Nandini Singhal <cs15mtech01004@iith.ac.in>
Differential Revision: https://reviews.llvm.org/D35301
llvm-svn: 308318
find_library(lib) stores the result in the variable "lib", which is also
cached in CMakeCache.txt. This could theoretically conflict if jsoncpp
required two libraries, which each would get cached as "lib". Use a
more descriptive and disambiguative "jsoncpp_${libname}" for that.
llvm-svn: 308289
Use ${libname} instead of ${lib}. By a coincidence, this worked
because ${lib} also the variable used for finding the libjsoncpp.so
full path.
llvm-svn: 308288
pkg_search_module(JSONCPP) should set
JSONCPP_LIBDIR/JSONCPP_LIBRARY_DIRS to where the libjsoncpp.so can be
found. However, on Ubuntu 14.04 LTS (Trusty Tahr) it returns /usr/lib
while the libjsoncpp library can be found at
/usr/lib/x86_64-linux-gnu/libjsoncpp.so. Thus, while searching for
the full path of the jsoncpp library, it is not found.
JSONCPP_LIBDIR is correctly set to /usr/lib/x86_64-linux-gnu on e.g.,
Ubuntu 16.04 LTS (Xenial Xerus )
Fix by removing the NO_DEFAULT_PATH flag, in order to search the system
default paths even if the library is not found in
JSONCPP_LIBDIR/JSONCPP_LIBRARY_DIRS.
This fixes bug llvm.org/PR33798.
llvm-svn: 308287
We extended kills in Polly to handle both `phi` nodes and scalars that
are not used within the Scop. Update the comments and choice of
variable names to reflect this.
llvm-svn: 308279
Utilizing newer LLVM diagnostic remark API in order to enable use of
opt-viewer tool. Polly Diagnostic Remarks also now appear in YAML
remark file.
In this patch, I've added the OptimizationRemarkEmitter into certain
classes where remarks are being emitted and update the remark emit calls
itself. I also provide each remark a BasicBlock or Instruction from where
it is being called, in order to compute the hotness of the remark.
Patch by Tarun Rajendran!
Differential Revision: https://reviews.llvm.org/D35399
llvm-svn: 308233
Summary: Since there will be no more a 1-1 correspondence between statements and basic block, we would like to get rid of the method `getStmtFor(BB)` and its uses. Here we remove one of its uses in PolyhedralInfo, as suggested by Michael Sir.
Reviewers: grosser, Meinersbur, bollu
Reviewed By: grosser
Subscribers: pollydev
Tags: #polly
Differential Revision: https://reviews.llvm.org/D35300
llvm-svn: 308220
Summary:
We do not keep domain constraints on access functions when building the
scop. Hence, for consistency reasons, it makes also sense to not include
them when storing a new access function. This change results in simpler
access functions that make output easier to read.
This patch also helps to make DeLICMed memory accesses to be understood by
our matrix multiplication pattern matching pass. Further changes to the
matrix multiplication pattern matching are needed for this to work, so the
corresponding test case will be added in a future commit.
Reviewers: Meinersbur, bollu, gareevroman, efriedma, huihuiz, sebpop, simbuerg
Subscribers: pollydev, llvm-commits
Tags: #polly
Differential Revision: https://reviews.llvm.org/D35237
llvm-svn: 308215
- We should call `preloadInvariantLoads` to make sure that code is
generated for invariant loads in the kernel.
Differential Revision: https://reviews.llvm.org/D35410
llvm-svn: 308187
This patch makes sure that in case a loop is not fully contained within a region
that later forms a SCoP, none of the loop backedges are allowed to be part of
the region. We currently do not support the situation where only some of a loops
backedges are part of a scop. Today, this can break both scop modeling and code
generation. One such breaking test case is for example
test/ScopDetectionDiagnostics/loop_partially_in_scop-2.ll, where we totally
forgot to code generate some of the backedges. Fortunately, it is commonly not
necessary to support these partial loops, it is way more common that either
no backedge is included in a region or all loop backedge are included.
This fixes a recent miscompile in
MultiSource/Benchmarks/MiBench/consumer-typeset which was exposed after
r306477.
llvm-svn: 308113
Tobias Grosser