summaryrefslogtreecommitdiffstats
path: root/polly
Commit message (Collapse)AuthorAgeFilesLines
* polly: update for LLVM API changeSaleem Abdulrasool2014-05-192-5/+6
| | | | | | | SVN r209103 removed the OwningPtr variant of the MemoryBuffer APIs. Switch to the equivalent std::unique_ptr versions. This should clear up the build bots. llvm-svn: 209104
* Do not run GPGPU test cases without nvptx targetTobias Grosser2014-05-144-1/+16
| | | | | | | | Tag the GPGPU codegen test cases as unsupported if the nvptx target is not included in the current llvm build. Contributed-by: Yabin Hu <yabin.hwu@gmail.com> llvm-svn: 208779
* fix clang formatSebastian Pop2014-05-121-2/+1
| | | | llvm-svn: 208626
* do not build llc and lli for polly testSebastian Pop2014-05-122-6/+6
| | | | llvm-svn: 208619
* correct the delinearization failing caseSebastian Pop2014-05-123-9/+68
| | | | | | collect terms from affine and non affine memory accesses llvm-svn: 208616
* delinearize together all accesses to the same arraySebastian Pop2014-05-092-11/+52
| | | | llvm-svn: 208457
* Adapt to DomTree changes in r199104Tobias Grosser2014-05-081-1/+1
| | | | | | | This fixes the compilation of our GPGPU code generation experiment in Polly. Contributed-by: Yabin Hu <yabin.hwu@gmail.com> llvm-svn: 208287
* unxfail passing testcaseSebastian Pop2014-05-071-2/+0
| | | | llvm-svn: 208233
* Remove superflous semicolonTobias Grosser2014-05-071-19/+17
| | | | llvm-svn: 208208
* Remove unnecessary struct in return typeTobias Grosser2014-05-071-2/+2
| | | | llvm-svn: 208191
* Catch all functions containing a certain stringTobias Grosser2014-05-071-3/+4
| | | | llvm-svn: 208190
* Replace llvm::IRBuilder with PollyIRBuilderTobias Grosser2014-05-071-3/+3
| | | | | Contributed-by: Yabin Hu <yabin.hwu@gmail.com> llvm-svn: 208182
* Remove unnecessary explicit typing in std::make_pairTobias Grosser2014-05-072-4/+3
| | | | | Contributed-by: Yabin Hu <yabin.hwu@gmail.com> llvm-svn: 208181
* Add the missing 'LoopToScev' parameterTobias Grosser2014-05-071-1/+1
| | | | | Contributed-by: Yabin Hu <yabin.hwu@gmail.com> llvm-svn: 208180
* Add delinearization testcase for ivs that do not follow the loop orderTobias Grosser2014-04-301-0/+71
| | | | | | | This is a test case that is currently failing, but that should start working with an upcoming version of our delinearization pass. llvm-svn: 207678
* Add missing include.Benjamin Kramer2014-04-301-0/+1
| | | | llvm-svn: 207618
* www: Reference phone call notesTobias Grosser2014-04-231-0/+3
| | | | llvm-svn: 207006
* Fix formattingTobias Grosser2014-04-221-2/+2
| | | | llvm-svn: 206903
* We missed two files in the last commit.Tobias Grosser2014-04-222-2/+2
| | | | | Contributed-by: Johannes Doerfert <doerfert@cs.uni-saarland.de> llvm-svn: 206901
* Fixed missing cloog test with automake/configure build setupTobias Grosser2014-04-222-1/+3
| | | | | Contributed-by: Johannes Doerfert <doerfert@cs.uni-saarland.de> llvm-svn: 206900
* IslCodeGeneration: Document the IslExprBuilderTobias Grosser2014-04-221-3/+86
| | | | llvm-svn: 206898
* [Modules] Fix potential ODR violations by sinking the DEBUG_TYPEChandler Carruth2014-04-2217-37/+40
| | | | | | | | | | 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
* Really fix the load case.Tobias Grosser2014-04-182-2/+11
| | | | | | | | Commit r206510 falsely advertised to fix the load cases, even though it only fixed the store case. This commit adds the same fix for the load case including the missing test coverage. llvm-svn: 206577
* Ensure a scalar pointer when issuing a vector loadTobias Grosser2014-04-172-0/+36
| | | | | | | | | | | Even tough we may want to generate a vector load, the address from which to load still is a scalar. Make sure even if previous address computations may have been vectorized, that the addresses are also available as scalars. This fixes http://llvm.org/PR19469 Reported-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> llvm-svn: 206510
* Link to notes of our last phone callTobias Grosser2014-04-171-0/+3
| | | | llvm-svn: 206488
* [C++11] Use nullptrTobias Grosser2014-04-1610-59/+59
| | | | llvm-svn: 206361
* Fix formattingTobias Grosser2014-04-151-1/+0
| | | | llvm-svn: 206333
* Fix for vector codegen in OpenMP subfunctionsTobias Grosser2014-04-152-0/+55
| | | | | Contributed-by: Johannes Doerfert <doerfert@cs.uni-saarland.de> llvm-svn: 206332
* Avoid -Wunused-const-variable warningTobias Grosser2014-04-151-0/+2
| | | | llvm-svn: 206329
* Fix more build errors in Polly after r206310. David caught one of theseChandler Carruth2014-04-152-6/+4
| | | | | | | in r206312, but others don't seem to show up on build bots? Unsure of why, they showed up for me. llvm-svn: 206326
* Dependences: Do not fail in case a schedule eliminates all dependencesTobias Grosser2014-04-153-0/+63
| | | | | | | | | | | | | | | | | | | The following example shows a non-parallel loop void f(int a[]) { int i; for (i = 0; i < 10; ++i) A[i] = A[i+5]; } which, in case we import a schedule that limits the iteration domain to 0 <= i < 5, becomes parallel. Previously we crashed in such cases, now we just recognize it as parallel. This fixes http://llvm.org/PR19435 Reported-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> llvm-svn: 206318
* Blind fix Polly for r206310David Blaikie2014-04-151-9/+9
| | | | llvm-svn: 206312
* Codegeneration: Free memory correctly when using -polly-vectorizer=pollyTobias Grosser2014-04-142-0/+22
| | | | | | | This fixes PR19421. Reported-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> llvm-svn: 206156
* Update isl to fix memory bugsTobias Grosser2014-04-131-1/+1
| | | | llvm-svn: 206134
* Dependences: Refine the compute out facilityTobias Grosser2014-04-122-2/+2
| | | | | | | | | | | | We update to a newer version of isl, which includes changes to the compute out facility that make it a lot more predicable. With our new value, we can reliably bail out for all reported bugs, while still being able to compute dependences for all but two test cases in the LLVM test suite. For the remaining two test cases, the dependence problem we construct is unnecessarily complex, so there is hope we can improve on this. However, to avoid any future issues, having a reliable compute out facility in place is important. llvm-svn: 206106
* Fixed gcc build warningsTobias Grosser2014-04-113-85/+93
| | | | | | | + vim 'fixed' line endings in json_value.cpp Contributed-by: Johannes Doerfert <doerfert@cs.uni-saarland.de> llvm-svn: 206044
* Remove OpenScopTobias Grosser2014-04-1118-1229/+296
| | | | | | | | | We only supported a very old version of OpenScop that was entirely different to what OpenScop is today. To not confuse people, we remove this old and unusable support. If anyone is interested to add OpenScop support back in, the relevant patches are available in version control. llvm-svn: 206026
* todo: Update todo listTobias Grosser2014-04-102-29/+64
| | | | llvm-svn: 205985
* www: Fix typoTobias Grosser2014-04-101-1/+1
| | | | llvm-svn: 205984
* todo: Building against an installed LLVM worksTobias Grosser2014-04-101-1/+1
| | | | | | We only care about cmake and the buildbots test that it works. llvm-svn: 205983
* www: We have no interest in FORTRAN support in test suiteTobias Grosser2014-04-101-4/+1
| | | | llvm-svn: 205982
* Add link to meeting notesTobias Grosser2014-04-101-1/+4
| | | | llvm-svn: 205980
* only delinearize when the access function is not affineSebastian Pop2014-04-104-31/+41
| | | | llvm-svn: 205971
* ScopInfo: Scalar accesses are zero dimensionalTobias Grosser2014-04-105-24/+21
| | | | llvm-svn: 205958
* ScopInfo: Some code cleanupTobias Grosser2014-04-102-11/+7
| | | | llvm-svn: 205957
* clang format last patchSebastian Pop2014-04-081-7/+7
| | | | llvm-svn: 205804
* delinearize memory access functionsSebastian Pop2014-04-0815-176/+202
| | | | llvm-svn: 205799
* Add another hint for fixing check-polly errors to get_started.htmlTobias Grosser2014-04-031-1/+4
| | | | | | | | Reversed the order in which LD_LIBRARY_PATH is defined in order to make sure the ${CLOOG_INSTALL} prefix is found first. Contributed-by: Christian Bielert <cib123@googlemail.com> llvm-svn: 205556
* Note of last Polly phone callTobias Grosser2014-04-031-0/+7
| | | | llvm-svn: 205553
* Add a paragraph describing how to configure the python interpreter.Tobias Grosser2014-04-031-0/+8
| | | | | Contributed-by: cib123@googlemail.com llvm-svn: 205549
OpenPOWER on IntegriCloud