| 3a32363 | Pere Mato | 01 September 2016, 12:35:07 UTC | Update ROOT version files to v6.06/08. | 01 September 2016, 12:35:07 UTC |
| e2a3709 | Pere Mato | 01 September 2016, 12:28:39 UTC | More release notes | 01 September 2016, 12:28:39 UTC |
| 3bae07b | Pere Mato | 01 September 2016, 12:17:06 UTC | Update release notes | 01 September 2016, 12:22:36 UTC |
| 3077a57 | Danilo Piparo | 26 August 2016, 07:31:49 UTC | Fix algorithm to handle paths with spaces. | 29 August 2016, 06:38:24 UTC |
| 0114a89 | Danilo Piparo | 25 August 2016, 16:00:06 UTC | Do not try to link library names with a space at the end | 29 August 2016, 06:38:24 UTC |
| 92ce55f | Danilo Piparo | 22 August 2016, 13:54:07 UTC | Improve heuristics to identify paths with spaces in aclic. | 29 August 2016, 06:38:23 UTC |
| 81ad91d | Danilo Piparo | 22 August 2016, 10:24:35 UTC | Fix ROOT-8332: support linked libraries paths with spaces in aclic | 29 August 2016, 06:38:23 UTC |
| 6fd7b78 | Lorenzo Moneta | 26 August 2016, 16:23:05 UTC | Fix test35, by defining the correct number of parameters for the functions to test | 26 August 2016, 16:23:05 UTC |
| 63af0e7 | Lorenzo Moneta | 14 March 2016, 10:55:38 UTC | Copy from master commit which implement bigaus as a new TFormula pre-defined function | 18 August 2016, 08:50:17 UTC |
| b63f514 | Philippe Canal | 17 August 2016, 15:43:18 UTC | Publish ListOfBases only when it is complete | 17 August 2016, 18:02:23 UTC |
| d6aa2c7 | Philippe Canal | 11 August 2016, 04:08:58 UTC | Fix ROOT-8307: Crash when leaving ROOT from GUI menu item. Opening any menu item after clicking 'Quit ROOT' but before the windows is actually closed (there is a non-negligeable time window where you can), will lead to a crash (or at least the reuse of a delete object). This is because open the menu item was registering the pop menu with the TGClient but the canvas destructor was deleting the C++ object representing the pop menu without unregistering it from the TGClient. This is fixed by forcing the unregistration from the TGPopupMenu destructor. | 11 August 2016, 16:32:37 UTC |
| 4823030 | Philippe Canal | 09 August 2016, 16:26:50 UTC | Use ROOT error logger. In addition having #include <iostream> in header files is 'somewhat' costly as it induces static variables in each compilation unit containing it. | 09 August 2016, 16:31:28 UTC |
| 2e70925 | Lorenzo Moneta | 18 April 2016, 13:30:01 UTC | Fix for ROOT-8105 and similar bugs caused by parsing for pre-defined functions. Add new tests showing cases which were failing before (test34,35,36) | 09 August 2016, 16:28:45 UTC |
| f87d854 | Philippe Canal | 08 August 2016, 20:12:12 UTC | Update release notes (ROOT-8105) | 08 August 2016, 20:12:12 UTC |
| e10f496 | Lorenzo Moneta | 18 April 2016, 13:30:01 UTC | Fix for ROOT-8105 and similar bugs caused by parsing for pre-defined functions. Add new tests showing cases which were failing before (test34,35,36) | 08 August 2016, 19:19:40 UTC |
| b87a4aa | Manuel Tobias Schiller | 07 August 2016, 12:00:37 UTC | BidirMMapPipe: undo inadvertent change in unit test code | 07 August 2016, 12:05:22 UTC |
| abc427f | Manuel Tobias Schiller | 05 August 2016, 12:04:48 UTC | BidirMMapPipe: silence message about mmapping strategy by default Since this class has been used for quite some time now without major trouble, it doesn't seem imprudent to hide the status messages about the mmapping strategy by default. Since I have had requests to that end, this is what I do with this commit. BidirMMapPipe gets two static methods which allow toggling the message back on if the user so wishes. | 05 August 2016, 12:16:48 UTC |
| 34c331d | Manuel Tobias Schiller | 01 August 2016, 11:21:37 UTC | RooMappedCategory: speed up evaluation This patch introduces a cache for index codes into RooMappedCategory. The previous implementation had to do a regex match on the different category labels, leading to slow evaluation. We now keep a cache indexed by the input category's index code, leading to a significant speedup. The persistent data format of the class remains unchanged, therefore the class version number was not changed. Backports to the various ROOT 6 and ROOT 5 bugfix branches may be in order. | 04 August 2016, 15:56:55 UTC |
| 57a5402 | Manuel Tobias Schiller | 01 August 2016, 11:17:33 UTC | RooDataSet: avoid crash with ImportFromFile("path/to/file.root","treename") In the RooDataSet constructor, the wrong TTree pointer was picked up when the code is supposed to read from an external ROOT file, causing a crash. This code fixes the null pointer deference by picking up the right TTree pointer. Backports to the ROOT 6 and ROOT 5 bugfix branches may be in order. | 04 August 2016, 15:52:33 UTC |
| 5d055aa | Philippe Canal | 27 July 2016, 20:05:37 UTC | Don't internally rely on global state changes to prevent TF1 add to global list. Doing so (without putting a lock around the whole section that update back and forth the global state or using thread local global state) will result in randomly wrong behavior if the code is used from multiple thread. Instead, add a direct parameter to the TF1 constructor (similar to the one in TFormula) that explicit the wanted behavior. | 27 July 2016, 20:08:44 UTC |
| be067b5 | Pere Mato | 26 July 2016, 06:38:28 UTC | Fix for ROOT-8293 - CMake does not find Castor libraries | 26 July 2016, 06:38:50 UTC |
| 6832343 | Pere Mato | 25 July 2016, 21:10:03 UTC | Fix for ROOT-8239 - "python: free(): invalid next size (fast)" in PyROOT 6.06.02 and patches | 25 July 2016, 21:12:30 UTC |
| f5bc345 | Lorenzo Moneta | 21 April 2016, 14:25:29 UTC | Make the TFormula::kIsNotGlobal bit public. Fix method TF1::AddToGlobalList and add also a default method TF1::DefaultAddToGlobalList. Use TF1::DefaultAddToGlobalList in TLinearMinimizer::SetFunction This fixes ROOT-8089 | 25 July 2016, 20:06:35 UTC |
| bc56c41 | Lorenzo Moneta | 19 April 2016, 10:27:28 UTC | Fix for memory leaks in TLInearMinimizer-TVirtualFitter. When calling SetBasisFunction the function list must be owner, while in case of SetFormula the list should not be owner | 25 July 2016, 20:04:59 UTC |
| 01425d6 | Danilo Piparo | 19 April 2016, 06:12:44 UTC | Remove leak in function fitting See https://root.cern.ch/phpBB3/viewtopic.php?f=3&t=21466 | 25 July 2016, 20:04:44 UTC |
| 726d939 | Philippe Canal | 13 May 2016, 19:04:25 UTC | Match malloc with free (rather than delete) | 25 July 2016, 20:04:31 UTC |
| 3e3caaf | Paul Russo | 15 July 2016, 18:55:19 UTC | Fix roottest/root/multicore/exectsinclude.C failure. | 15 July 2016, 18:55:19 UTC |
| 93e1f65 | Dan Riley | 13 July 2016, 14:12:15 UTC | Adjust dictionary size downward for small baskets LZMA by default creates very large hash tables for its dictionaries, e.g., at compression level 4, the hash table is 4Mi 4 byte entries, 16 MiB total. The hash table has to be zeroed before use so it is allocated via calloc(), which means all the pages have to be allocated, mapped and written. ROOT baskets are often much smaller than the default LZMA dictionaries; for small baskets, the large dictionary has very little compression benefit, while zeroing the hash table can be more expensive than the actual compression operation. Since R__zipLZMA() is actually being used to compress a buffer of known size, not a stream, we can use the size of the buffer to estimate an appropriate size for the dictionary. This PR uses a slightly more advanced part of the LZMA API to set the dictionary size to 1/4 the size of the input buffer, if that is smaller than the default size from the selected preset compression level. In tests with CMS data, this results in less than 1% increase in the output size and (in one test job) a 25% reduction in job total run time, with LZMA compression time reduced by 80% (all of that time that was being spent in memset() zeroing the hash table). I also tested this with the "Event" test program with Brian's changes from #59. With the same test parameters as Brian ("./Event 4000 6 99 1 1000 2"), I get ZLIB level-6: 14.4 MB/s Original LZMA level-6: 2.3 MB/s Modified LZMA level-6: 3.0 MB/s With 100 tracks per event (and hence smaller baskets) the improvement is from 2.2 MB/s to 3.9 MB/s. This change should be fully transparent and backwards compatible. | 15 July 2016, 14:52:12 UTC |
| 19083e5 | Pere Mato | 15 July 2016, 06:30:58 UTC | Fix for ROOT-8270 - C++14 | 15 July 2016, 06:39:45 UTC |
| 2bb2f80 | Pere Mato | 07 July 2016, 15:07:05 UTC | Further fixes for ROOT-8226 - CMake 3.5.2: circular dependency libMathMore | 07 July 2016, 15:07:51 UTC |
| 6b7c425 | Pere Mato | 07 July 2016, 08:25:34 UTC | Better formatting of release notes | 07 July 2016, 08:25:34 UTC |
| 48e867a | Pere Mato | 06 July 2016, 17:08:44 UTC | Update ROOT version files to v6.06/07. | 06 July 2016, 17:08:44 UTC |
| f7d9d49 | Pere Mato | 06 July 2016, 16:36:49 UTC | Update ROOT version files to v6.06/06. | 06 July 2016, 16:36:49 UTC |
| b9c1d82 | Pere Mato | 06 July 2016, 16:21:03 UTC | Updated the release notes. | 06 July 2016, 16:21:03 UTC |
| eef786f | Paul Russo | 02 July 2016, 21:08:39 UTC | Update release notes. | 02 July 2016, 21:08:39 UTC |
| 51e8158 | Paul Russo | 02 July 2016, 20:42:39 UTC | Avoid autoparse looking for default constructor. | 02 July 2016, 20:42:39 UTC |
| bdd08ad | Paul Russo | 02 July 2016, 20:41:47 UTC | Do not insert macro definition of __ROOTCLING__ into the pch. | 02 July 2016, 20:41:47 UTC |
| 8e198a1 | Paul Russo | 02 July 2016, 20:40:18 UTC | Check for new-style empty pcm with key named "EMPTY" created since commit 90047b0cba6fd295f5c5722749a0d043fbc11ea5. | 02 July 2016, 20:40:18 UTC |
| 35f6544 | Paul Russo | 02 July 2016, 20:38:51 UTC | Avoid autoparse on IsForeign() if possible. | 02 July 2016, 20:38:51 UTC |
| d26502f | Paul Russo | 02 July 2016, 20:37:23 UTC | Do not autoparse headers for classes in the pch. | 02 July 2016, 20:37:23 UTC |
| b26812b | Pere Mato | 26 June 2016, 02:13:46 UTC | Fix for ROOT-8226 - CMake 3.5.2: circular dependency libMathMore. Including libraries is an idiom for CINT only. | 26 June 2016, 02:14:11 UTC |
| 0f2650e | Omar Zapata | 02 January 2016, 07:41:00 UTC | PyMVA: added support for python3 | 22 June 2016, 16:12:09 UTC |
| 017b62b | Omar Zapata | 02 January 2016, 05:16:49 UTC | PyMVA: fixed warning from scikit-learn in data shape validation.py:386: DeprecationWarning: Passing 1d arrays as data is deprecated in 0.17 and willraise ValueError in 0.19. Reshape your data either using X.reshape(-1, 1) if your data has a single feature or X.reshape(1, -1) if it contains a single sample. | 22 June 2016, 16:12:09 UTC |
| 5fb3771 | Omar Zapata | 02 January 2016, 04:05:27 UTC | PyMVA:working in serialization to support python3 | 22 June 2016, 16:12:09 UTC |
| 545e5a7 | Omar Zapata | 02 January 2016, 01:58:40 UTC | PyMVA: working in serialization | 22 June 2016, 16:12:09 UTC |
| f0b4af2 | Philippe Canal | 25 April 2016, 18:05:08 UTC | Fix fatal typo | 21 June 2016, 20:47:38 UTC |
| 165fc69 | Philippe Canal | 25 April 2016, 20:01:43 UTC | Suppress warning about misleading indentation | 21 June 2016, 20:35:00 UTC |
| cc28308 | Philippe Canal | 25 April 2016, 19:53:36 UTC | Suppress warning about misleading indentation | 21 June 2016, 20:34:56 UTC |
| 2c179cc | Philippe Canal | 25 April 2016, 19:45:23 UTC | Suppress warning about misleading indentation (and gain a tiny bit of run-time) | 21 June 2016, 20:34:51 UTC |
| e6abd05 | Philippe Canal | 25 April 2016, 19:19:15 UTC | Fix wrong 'fixing' of misleading indentation | 21 June 2016, 20:34:47 UTC |
| b2bec8b | Philippe Canal | 25 April 2016, 17:23:17 UTC | Suppress warning about misleading indentation | 21 June 2016, 20:34:42 UTC |
| 89d6441 | Philippe Canal | 25 April 2016, 17:20:08 UTC | Suppress warning about misleading indentation | 21 June 2016, 20:34:36 UTC |
| 4867ccb | Philippe Canal | 25 April 2016, 17:18:51 UTC | Suppress warning about misleading indentation | 21 June 2016, 20:34:04 UTC |
| 28c2edc | Philippe Canal | 25 April 2016, 16:57:47 UTC | Suppress warning about misleading indentation | 21 June 2016, 20:33:46 UTC |
| de9c5bf | Philippe Canal | 25 April 2016, 16:55:31 UTC | Suppress warning about misleading indentation | 21 June 2016, 20:33:35 UTC |
| 24d32e1 | Philippe Canal | 25 April 2016, 16:23:26 UTC | Suppress warning about misleading indentation | 21 June 2016, 20:33:31 UTC |
| d389662 | Philippe Canal | 25 April 2016, 16:14:40 UTC | Suppress warning about misleading indentation | 21 June 2016, 20:33:17 UTC |
| d536eb8 | Philippe Canal | 09 April 2016, 23:09:21 UTC | Suppress warning about misleading indentation (and gain a tiny bit of run-time) | 21 June 2016, 20:33:12 UTC |
| 1fae2eb | Philippe Canal | 09 April 2016, 22:51:25 UTC | Avoid confusing/misleading code layout (were triggering misleading indentation warning) | 21 June 2016, 20:33:07 UTC |
| 4d4952b | Philippe Canal | 09 April 2016, 22:48:30 UTC | Fix incorrect indentation (revealed by 'misleading indentation warning') | 21 June 2016, 20:33:02 UTC |
| c7bb1a6 | Philippe Canal | 09 April 2016, 22:32:37 UTC | Suppress warning about misleading indentation (and gain a tiny bit of run-time) | 21 June 2016, 20:32:56 UTC |
| 09f0e6b | Philippe Canal | 09 April 2016, 21:52:40 UTC | Fix misleading indentation. This was discovered by gcc 6's warning: .../root/math/matrix/src/TMatrixTBase.cxx: In member function ‘void TMatrixTBase<Element>::Print(Option_t*) const’: .../root/math/matrix/src/TMatrixTBase.cxx:877:10: error: this ‘for’ clause does not guard... [-Werror=misleading-indentation] for (Int_t j = sheet_counter; j < sheet_counter+cols_per_sheet && j <= ncols; j++) ^~~ .../root/math/matrix/src/TMatrixTBase.cxx:879:13: note: ...this statement, but the latter is misleadingly indented as if it is guarded by the ‘for’ printf("\n"); ^~~~~~ | 21 June 2016, 20:32:35 UTC |
| 69bcba3 | Philippe Canal | 25 April 2016, 17:15:39 UTC | Suppress warning about misleading indentation | 21 June 2016, 20:28:10 UTC |
| 6e54d92 | Pere Mato | 21 June 2016, 09:46:10 UTC | Use further PATH_SUFFIX to clean module | 21 June 2016, 10:11:31 UTC |
| ca8d080 | Oliver Freyermuth | 12 April 2016, 21:51:16 UTC | FindPostgreSQL.cmake: Add PATH_SUFFIX for searching. This slightly changes lookup order: Namely, the system-folder "/usr/include/postgresql" is now preferred over "/usr/include/" even if both container libpq-fe.h. This finds the correct path on a standard Gentoo installation, where "/usr/include/postgresql" is a symlink to the folder with all include-files for the user-selected postgresql-version, while "/usr/include/libpq-fe.h" is a single symlink provided for backwards-compatibility. Since ROOT uses also e.g. "pg_config.h", selecting "/usr/include/" over "/usr/include/postgresql" leads to failure on Gentoo. | 21 June 2016, 10:11:23 UTC |
| c370d48 | Mattias Ellert | 01 March 2016, 06:03:40 UTC | Remove executable permissons | 21 June 2016, 08:55:48 UTC |
| 269e4df | Enric Tejedor Saavedra | 21 June 2016, 07:45:36 UTC | Fix ROOT-8232 - Use kMAXPATHLEN when getting unix home directory | 21 June 2016, 07:57:01 UTC |
| 90b9cb3 | Pere Mato | 19 June 2016, 20:14:35 UTC | Fix ROOT-8221 - Enum value kKey_Shift not accessible from python with ROOT 6. Thanks to Mattias Ellert for the hint. | 19 June 2016, 20:15:19 UTC |
| 0b7e7f5 | Pere Mato | 17 June 2016, 16:45:16 UTC | Fix for ROOT-8231 - _GLIBCXX_USE_CXX11_ABI=0 should be propagated into root-config --cflags | 17 June 2016, 16:46:07 UTC |
| c61a1b2 | Pere Mato | 17 June 2016, 14:38:51 UTC | Applied changes suggested by Pepe Le Pew and Mattias Ellert | 17 June 2016, 14:42:46 UTC |
| 84ca24c | Pere Mato | 17 June 2016, 13:57:25 UTC | Use pkg-config tp get glib-2.0 include dirs for GFAL | 17 June 2016, 13:57:25 UTC |
| 3752251 | Pere Mato | 17 June 2016, 12:01:46 UTC | Fixes to deal with gfal version 1 and 2 | 17 June 2016, 12:02:32 UTC |
| 1ce2f14 | Pere Mato | 16 June 2016, 16:16:57 UTC | Fixes for ROOT v6-06-04 on Ubuntu 16.04 (gfal, odbc). See https://root.cern.ch/phpBB3/viewtopic.php?p=95425 | 16 June 2016, 16:18:29 UTC |
| dbc9815 | Pere Mato | 16 June 2016, 14:21:58 UTC | Remove all c++ flags from ROOT_DEFINITIONS set by ROOTConfig.cmake | 16 June 2016, 16:18:04 UTC |
| cc23e9d | Pere Mato | 13 June 2016, 13:39:05 UTC | Check new GCC ABI in configure step and eventually fail configuration | 13 June 2016, 13:39:44 UTC |
| cc6d740 | Helge Voss | 13 June 2016, 11:54:41 UTC | fix error in communicating with the event setting window | 13 June 2016, 12:52:56 UTC |
| 5c8d0ae | Philippe Canal | 10 June 2016, 17:00:04 UTC | Update release notes | 10 June 2016, 17:00:04 UTC |
| 3a1da50 | Enric Tejedor Saavedra | 08 June 2016, 14:33:01 UTC | Safely remove lock guard in TSystem::ExpandFileName. In order to make TSystem::ExpandFileName thread safe without the lock guard, the calls to HomeDirectory and WorkingDirectory have been replaced by their corresponding thread safe versions: GetHomeDirectory and GetWorkingDirectory. These new functions do not return static local variables nor modify the internal state of the TSystem object. | 10 June 2016, 16:08:34 UTC |
| fae0f2a | Enric Tejedor Saavedra | 07 June 2016, 14:11:26 UTC | Make TSystem::ExpandFileName(const char*) thread safe. A TLS variable is used for the path to expand and the method is kept for backward compatibility. In addition, a new method with signature TSystem::ExpandFileName(TString&) is defined, so that it can be invoked internally from TUnixSystem and TWinNTSystem without paying the price of the TLS. | 10 June 2016, 16:08:34 UTC |
| 78b19cd | Lorenzo Moneta | 14 March 2016, 15:46:49 UTC | Fix memory leaks when reading RooFitResult and HypoTestInverterResult from disk. Patch from Geert-Jan Besjes & Ingrid Deigaard. Fix for ROOT-7986. | 10 June 2016, 13:31:58 UTC |
| e1157c3 | Pere Mato | 08 June 2016, 09:09:41 UTC | Do not randomize the test results | 08 June 2016, 09:09:41 UTC |
| b7403be | Pere Mato | 08 June 2016, 08:08:17 UTC | Added additional dependency to avoid several tutorials running hsimple.C at the same time | 08 June 2016, 08:09:03 UTC |
| 8a073c1 | Philippe Canal | 07 June 2016, 15:00:32 UTC | Avoid race condition on gMinuit. Reduce reliance on gMinuit on constructor. | 07 June 2016, 15:09:28 UTC |
| 738502c | Philippe Canal | 07 June 2016, 14:59:41 UTC | Avoid useless lock (left over from da33f0a) | 07 June 2016, 15:09:28 UTC |
| e2a482a | Philippe Canal | 06 June 2016, 15:52:29 UTC | Update release notes | 06 June 2016, 15:52:29 UTC |
| ec2e4cc | Philippe Canal | 04 June 2016, 13:03:27 UTC | Fix detection of errors that appears in nested TTreeFormula. See https://root.cern.ch/phpBB3/viewtopic.php?f=3&t=21715 and roottest/root/treeformula/array/execNestedFail.C. | 06 June 2016, 15:51:07 UTC |
| 3a49159 | Pere Mato | 27 May 2016, 17:26:26 UTC | Fix for ROOT-8207 - tmva/tmva/CMakeLists.txt unnecessarily includes TreeViewer | 27 May 2016, 17:26:53 UTC |
| be55575 | Pere Mato | 24 May 2016, 16:59:15 UTC | Fix for ROOT-8197 - rootcp not preserving kSingleKey | 24 May 2016, 17:01:37 UTC |
| aaed06f | Philippe Canal | 22 May 2016, 18:28:38 UTC | Update releases notes | 22 May 2016, 18:28:38 UTC |
| c87978c | Pere Mato | 18 May 2016, 14:49:46 UTC | Applied fix for ROOT-8038 [patch] "for e in getSTLLikeCollection()" crashes pyroot | 18 May 2016, 14:52:24 UTC |
| 050307d | Philippe Canal | 17 May 2016, 12:36:01 UTC | Update relase notes | 17 May 2016, 12:36:01 UTC |
| 0b9db52 | Philippe Canal | 08 April 2016, 20:35:24 UTC | Fix ROOT-8096 Add proper support for template parameter pack in cling::ForwardDeclPrinter::VisitTemplateArgument | 17 May 2016, 12:34:50 UTC |
| 2878e3f | Philippe Canal | 13 May 2016, 16:06:48 UTC | Update release notes | 13 May 2016, 16:06:48 UTC |
| 5592f3a | Philippe Canal | 03 March 2016, 18:07:28 UTC | In AutoSynthesizer handle the case of try-catch function [ROOT-8034]. Example of try-catch function: void function_with_try_block() try { // try block body } catch (...) { // catch block body } | 13 May 2016, 16:05:42 UTC |
| e4b3d70 | Philippe Canal | 11 May 2016, 19:47:17 UTC | Update release notes | 11 May 2016, 19:47:17 UTC |
| f108389 | Philippe Canal | 11 May 2016, 19:39:21 UTC | Fix ROOT-7258: MinIf and MaxIf TTreeFormula functions ignore last element. From: Will Buttinger | 11 May 2016, 19:46:35 UTC |
| 9168c00 | Pere Mato | 09 May 2016, 06:41:09 UTC | Disable vc for Mac latest Xcode | 09 May 2016, 06:41:09 UTC |
| 18f8977 | Pere Mato | 16 February 2016, 17:31:51 UTC | Update the traceback message to guide users to use the forum instead of JIRA. Update INSTALL instructions. | 04 May 2016, 12:05:31 UTC |
| 2c7d10d | Axel Naumann | 03 May 2016, 18:45:26 UTC | Update ROOT version files to v6.06/05. | 03 May 2016, 18:45:26 UTC |
| 9297ecf | Axel Naumann | 03 May 2016, 18:22:18 UTC | Update ROOT version files to v6.06/04. | 03 May 2016, 18:23:11 UTC |
