Hi, yes, we made with Borek the TestIndexSearcher test working. In order to do this we made following changes:
1) removing the assert - the problem was that the method WaitForSingleObject was used only within the assert and so in release version not called! This must be definitely a bug! Very strange, this could never work in release! So the call must be outside of the assert, but you are right we still can use the assert to check its result. (I will change the code in order to do this) 2) RemoveThreadLocal method added. This method is called from the destructor of _ThreadLocal and removes all occurrences of the deleted _ThreadLocal object from all ThreadLocals stored in the static threadData multimap. In order to make the RemoveThreadLocal methos more efficient it would be great if ThreadLocals inherit from a set not a vector - I will also trie to change it in this way. Jiri -----Original Message----- From: Itamar Syn-Hershko [mailto:ita...@code972.com] Sent: Tuesday, September 14, 2010 8:25 PM To: clucene-developers@lists.sourceforge.net Subject: Re: [CLucene-dev] AddIndexesNoOptimize test added to intensive_testing branch, some problems in core detected Jiri, I checked your branch. On my main machine, Win XP and compiling with VS2008, all tests that are expected to pass, do pass. I'm not sure what you did in commit b0156de tho - why did you remove the assert? Itamar. On 14/9/2010 10:51 AM, Šplíchal Jiří wrote: > Hello Itamar, > > I merged the current master into my branch. There were only a few conflicts, > most of them could be easily solved. There are a lot of memory leaks in the > test and > there are two tests that do not work correctly on my computer: > > 1) IndexModifierTest::testIMinsertDelete - this test crashes in release and > also in debug (but not when traced) > with R6025 - pure virtual function call. This happens when the FSIndexOuput > is being created for the file > "C:\Users\splichal\AppData\Local\Temp\test.search\segments.gen". Then > something goes wrong, the file can not > be created (it is not in the directory) and the contructor fires an exception > - and so the destructor of the base > class is called which leades to the call of a pure virtual method > flushBuffer() = crash. > > 2) testAddIndexesNoOptimize::testHangOnClose - crashes only in release and > only if I run the tests without -v option. > I tried to find the last call by inserting prints and the call > writer = _CLNEW IndexWriter4Test(dir2, false,&an, true); > never returns. > > Do those test work for you? Could you check my branch? > > Jiri > > > > -----Original Message----- > From: Itamar Syn-Hershko [mailto:ita...@code972.com] > Sent: Monday, September 13, 2010 7:15 PM > To: clucene-developers@lists.sourceforge.net > Subject: Re: [CLucene-dev] AddIndexesNoOptimize test added to > intensive_testing branch, some problems in core detected > > Borek, is this fix tested to make sure it is still valid also in extreme > cases? > > On 2/9/2010 4:12 PM, Kostka Bořivoj wrote: > >> The problem was caused by redundant NULL test in >> SegmentMerger::mergeVectors(). >> termVectorsWriter->addAllDocVectors(tmp) contains code for case tmp is NULL, >> so it should be always called >> (see fix in intensive_testing branch) >> >> Borek >> >> >> >>> -----Original Message----- >>> From: Ben van Klinken [mailto:bvanklin...@gmail.com] >>> Sent: Thursday, September 02, 2010 11:48 AM >>> To: clucene-developers@lists.sourceforge.net >>> Subject: Re: [CLucene-dev] AddIndexesNoOptimize test added to >>> intensive_testing >>> branch, some problems in core detected >>> >>> Try running the index test program in Luke. It may show up more >>> >>> ben >>> >>> On Wednesday, September 1, 2010, Veit Jahns<nuncupa...@googlemail.com> >>> wrote: >>> >>> >>>> 2010/8/23 Kostka Bořivoj<kos...@tovek.cz>: >>>> >>>> >>>>> The last test in suite (testHangOnClose) fails on CND_PRECONDITION in >>>>> >>>>> >>> SegmentMerger on line 415. >>> >>> >>>>> Is anybody familiar enough with sement merging code to easily detect why? >>>>> >>>>> >>>> I'm not familiar with the code, but it seems to me, that the problem >>>> is not the merging code. Here it is tried to read term vector of some >>>> documents, but these documents have no fields with term vectors to be >>>> merged. But the index seems fine. I changed to the test, so the index >>>> was written to the filesystem and opened it with Luke without any >>>> problems. >>>> >>>> Veit >>>> >>>> ------------------------------------------------------------------------------ >>>> This SF.net Dev2Dev email is sponsored by: >>>> >>>> Show off your parallel programming skills. >>>> Enter the Intel(R) Threading Challenge 2010. >>>> http://p.sf.net/sfu/intel-thread-sfd >>>> _______________________________________________ >>>> CLucene-developers mailing list >>>> CLucene-developers@lists.sourceforge.net >>>> https://lists.sourceforge.net/lists/listinfo/clucene-developers >>>> >>>> >>>> >>> ------------------------------------------------------------------------------ >>> This SF.net Dev2Dev email is sponsored by: >>> >>> Show off your parallel programming skills. >>> Enter the Intel(R) Threading Challenge 2010. >>> http://p.sf.net/sfu/intel-thread-sfd >>> _______________________________________________ >>> CLucene-developers mailing list >>> CLucene-developers@lists.sourceforge.net >>> https://lists.sourceforge.net/lists/listinfo/clucene-developers >>> >>> >> ------------------------------------------------------------------------------ >> This SF.net Dev2Dev email is sponsored by: >> >> Show off your parallel programming skills. >> Enter the Intel(R) Threading Challenge 2010. >> http://p.sf.net/sfu/intel-thread-sfd >> _______________________________________________ >> CLucene-developers mailing list >> CLucene-developers@lists.sourceforge.net >> https://lists.sourceforge.net/lists/listinfo/clucene-developers >> >> > ------------------------------------------------------------------------------ > Start uncovering the many advantages of virtual appliances > and start using them to simplify application deployment and > accelerate your shift to cloud computing > http://p.sf.net/sfu/novell-sfdev2dev > _______________________________________________ > CLucene-developers mailing list > CLucene-developers@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/clucene-developers > ------------------------------------------------------------------------------ > Start uncovering the many advantages of virtual appliances > and start using them to simplify application deployment and > accelerate your shift to cloud computing. > http://p.sf.net/sfu/novell-sfdev2dev > _______________________________________________ > CLucene-developers mailing list > CLucene-developers@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/clucene-developers > ------------------------------------------------------------------------------ Start uncovering the many advantages of virtual appliances and start using them to simplify application deployment and accelerate your shift to cloud computing. http://p.sf.net/sfu/novell-sfdev2dev _______________________________________________ CLucene-developers mailing list CLucene-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/clucene-developers ------------------------------------------------------------------------------ Start uncovering the many advantages of virtual appliances and start using them to simplify application deployment and accelerate your shift to cloud computing. http://p.sf.net/sfu/novell-sfdev2dev _______________________________________________ CLucene-developers mailing list CLucene-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/clucene-developers
