[redland-dev] Error when removing triples

Mon Feb 23 08:53:42 EST 2015

All right, here is the full trace and heap state. I have also 
uncommented DEBUG in rdf_storage_virtuoso.c

Looks like the origin is in Virtodbc.

$ valgrind --tool=memcheck --num-callers=50 --leak-check=full rdfproc -t 
"dsn='Virtuoso',user=*,password=*" -s virtuoso 
http://www.3top.com/graph/fscharf remove 
http://dbpedia.org/resource/Apple 
http://www.3top.com/entity/ranked_in_category 
http://www.3top.com/category/530/asian-fruits
==4710== Memcheck, a memory error detector
==4710== Copyright (C) 2002-2013, and GNU GPL'd, by Julian Seward et al.
==4710== Using Valgrind-3.10.0.SVN and LibVEX; rerun with -h for 
copyright info
==4710== Command: rdfproc -t dsn='Virtuoso',user=*,password=* -s 
virtuoso http://www.3top.com/graph/fscharf remove 
http://dbpedia.org/resource/Apple 
http://www.3top.com/entity/ranked_in_category 
http://www.3top.com/category/530/asian-fruits
==4710==
librdf_storage_virtuoso_init
librdf_storage_virtuoso_init_connections
librdf_storage_virtuoso_open
librdf_storage_virtuoso_context_remove_statement
librdf_storage_virtuoso_connection
SQL: >>sparql define output:format '_JAVA_' delete from graph iri(??) 
{`iri(??)` `iri(??)` `bif:__rdf_long_from_batch_params(??,??,??)`}<<
librdf_storage_virtuoso_release_handle 0x6eac3b0
rdfproc: removed triple from the graph
librdf_storage_virtuoso_close
librdf_storage_virtuoso_transaction_rollback
librdf_storage_virtuoso_sync
librdf_storage_virtuoso_terminate
librdf_storage_virtuoso_finish_connections
==4710== Invalid free() / delete / delete[] / realloc()
==4710==    at 0x4C2BDEC: free (in 
/usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==4710==    by 0x80A9F18: librdf_storage_virtuoso_terminate 
(rdf_storage_virtuoso.c:1199)
==4710==    by 0x50B21DA: librdf_free_storage (rdf_storage.c:723)
==4710==    by 0x403CBB: main (rdfproc.c:1597)
==4710==  Address 0x6eab860 is 0 bytes inside a block of size 34 free'd
==4710==    at 0x4C2BDEC: free (in 
/usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==4710==    by 0x80AC9ED: 
librdf_storage_virtuoso_context_remove_statement 
(rdf_storage_virtuoso.c:2089)
==4710==    by 0x4056A8: main (rdfproc.c:1280)
==4710==
==4710==
==4710== HEAP SUMMARY:
==4710==     in use at exit: 148,416 bytes in 3,255 blocks
==4710==   total heap usage: 5,487 allocs, 2,233 frees, 787,600 bytes 
allocated
==4710==
==4710== 24 bytes in 1 blocks are possibly lost in loss record 598 of 749
==4710==    at 0x4C2AB80: malloc (in 
/usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==4710==    by 0x9A01ADC: dk_alloc_reserve_malloc (Dkernel.c:5430)
==4710==    by 0x99ECE2C: dk_alloc_box (Dkbox.c:191)
==4710==    by 0x99EDB82: box_dv_short_string (Dkbox.c:1427)
==4710==    by 0x9A019B8: PrpcConnect1 (Dkernel.c:4458)
==4710==    by 0x99CCEEC: internal_sql_connect (CLIsql1.c:900)
==4710==    by 0x99EA5ED: virtodbc__SQLDriverConnect.isra.1 (CLIsql3.c:767)
==4710==    by 0x82C49C5: ???
==4710==    by 0x80AA285: ???
==4710==    by 0x80AC840: ???
==4710==    by 0x4056A8: main (rdfproc.c:1280)
==4710==
==4710== 224 (40 direct, 184 indirect) bytes in 1 blocks are definitely 
lost in loss record 681 of 749
==4710==    at 0x4C2CC70: calloc (in 
/usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==4710==    by 0x50AC045: librdf_new_hash_from_hash (rdf_hash.c:482)
==4710==    by 0x50B24D7: librdf_new_storage_with_options 
(rdf_storage.c:579)
==4710==    by 0x4035EC: main (rdfproc.c:683)
==4710==
==4710== 1,698 (72 direct, 1,626 indirect) bytes in 1 blocks are 
definitely lost in loss record 737 of 749
==4710==    at 0x4C2CC70: calloc (in 
/usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==4710==    by 0x82FE02E: ???
==4710==    by 0x82FE30E: ???
==4710==    by 0x82BE8BC: ???
==4710==    by 0x82C496D: ???
==4710==    by 0x80AA285: ???
==4710==    by 0x80AC840: ???
==4710==    by 0x4056A8: main (rdfproc.c:1280)
==4710==
==4710== 4,232 bytes in 1 blocks are definitely lost in loss record 745 
of 749
==4710==    at 0x4C2CC70: calloc (in 
/usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==4710==    by 0x80AB945: ???
==4710==    by 0x50B2346: librdf_new_storage_from_factory 
(rdf_storage.c:698)
==4710==    by 0x4035EC: main (rdfproc.c:683)
==4710==
==4710== LEAK SUMMARY:
==4710==    definitely lost: 4,344 bytes in 3 blocks
==4710==    indirectly lost: 1,810 bytes in 101 blocks
==4710==      possibly lost: 24 bytes in 1 blocks
==4710==    still reachable: 142,238 bytes in 3,150 blocks
==4710==         suppressed: 0 bytes in 0 blocks
==4710== Reachable blocks (those to which a pointer was found) are not 
shown.
==4710== To see them, rerun with: --leak-check=full --show-leak-kinds=all
==4710==
==4710== For counts of detected and suppressed errors, rerun with: -v
==4710== ERROR SUMMARY: 5 errors from 5 contexts (suppressed: 1 from 1)

Thanks for your help

Francois

Le 2/20/15 7:36 AM, Michael Stahl a écrit :
> On 13.02.2015 19:49, François Scharffe wrote:
>> I get errors when trying to remove triples, both through the python
>> bindings or using rdfproc.
>>
>> rdfproc gives the following output:
>>
>> rdfproc: removed triple from the graph
>> *** Error in `rdfproc': double free or corruption (!prev):
>> 0x00000000016d5130 ***
>> Aborted (core dumped)
>>
>>
>> The triple is actually removed.
>>
>> I use Redland with Virtuoso through unixODBC.
>
> this is definitely a bug somewhere, and you can use valgrind to track
> down where exactly.  you need to install debuginfo packages, or rebuild
> things with debug enabled until valgrind produces useful output (with
> line numbers).
>
>   valgrind --tool=memcheck --num-callers=50  rdfproc ...
>
> _______________________________________________
> redland-dev mailing list
> redland-dev at lists.librdf.org
> http://lists.librdf.org/mailman/listinfo/redland-dev
>