Hi,
i have exactly the same problem since the last anjuta update in karmic.
Running under valgrind:

==5514== Memcheck, a memory error detector.
==5514== Copyright (C) 2002-2008, and GNU GPL'd, by Julian Seward et al.
==5514== Using LibVEX rev 1884, a library for dynamic binary translation.
==5514== Copyright (C) 2004-2008, and GNU GPL'd, by OpenWorks LLP.
==5514== Using valgrind-3.4.1-Debian, a dynamic binary instrumentation framework.
==5514== Copyright (C) 2000-2008, and GNU GPL'd, by Julian Seward et al.
==5514== For more details, rerun with: -v
==5514== 

(anjuta:5514): libanjuta-WARNING **: Invalid gconf type for key: anjuta.toolbar.visible
--5514-- WARNING: unhandled syscall: 332
--5514-- You may be able to write your own handler.
--5514-- Read the file README_MISSING_SYSCALL_OR_IOCTL.
--5514-- Nevertheless we consider this a bug.  Please report
--5514-- it at http://valgrind.org/support/bug_reports.html.

(anjuta:5514): libanjuta-WARNING **: Invalid gconf type for key: margin.linenumber.visible

(anjuta:5514): libanjuta-WARNING **: Invalid gconf type for key: margin.marker.visible

(anjuta:5514): libanjuta-WARNING **: Invalid gconf type for key: margin.fold.visible

(anjuta:5514): libanjuta-WARNING **: Invalid gconf type for key: view.indentation.guides

(anjuta:5514): libanjuta-WARNING **: Invalid gconf type for key: view.whitespace
==5523== 
==5523== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 437 from 2)
==5523== malloc/free: in use at exit: 8,047,184 bytes in 54,886 blocks.
==5523== malloc/free: 890,424 allocs, 835,538 frees, 87,408,019 bytes allocated.
==5523== For counts of detected errors, rerun with: -v
==5523== searching for pointers to 54,886 not-freed blocks.
==5523== checked 8,050,768 bytes.
==5523== 
==5523== LEAK SUMMARY:
==5523==    definitely lost: 23,032 bytes in 320 blocks.
==5523==      possibly lost: 4,580,916 bytes in 3,548 blocks.
==5523==    still reachable: 3,443,236 bytes in 51,018 blocks.
==5523==         suppressed: 0 bytes in 0 blocks.
==5523== Rerun with --leak-check=full to see details of leaked memory.
==5514== Invalid read of size 4
==5514==    at 0x7AD6E09: symbol_db_engine_file_exists (in /usr/lib/anjuta/libanjuta-symbol-db.so)
==5514==    by 0x7AD75BB: symbol_db_engine_update_files_symbols (in /usr/lib/anjuta/libanjuta-symbol-db.so)
==5514==    by 0x7AD7CE9: symbol_db_engine_update_project_symbols (in /usr/lib/anjuta/libanjuta-symbol-db.so)
==5514==    by 0x7ABBD0D: (within /usr/lib/anjuta/libanjuta-symbol-db.so)
==5514==    by 0x4AA3773: (within /usr/lib/libanjuta.so.0.0.0)
==5514==    by 0x4A98A6E: anjuta_cclosure_marshal_VOID__STRING_BOXED (in /usr/lib/libanjuta.so.0.0.0)
==5514==    by 0x497A091: g_closure_invoke (in /usr/lib/libgobject-2.0.so.0.2200.0)
==5514==    by 0x498F6F7: (within /usr/lib/libgobject-2.0.so.0.2200.0)
==5514==    by 0x4990A7C: g_signal_emit_valist (in /usr/lib/libgobject-2.0.so.0.2200.0)
==5514==    by 0x4990D91: g_signal_emit_by_name (in /usr/lib/libgobject-2.0.so.0.2200.0)
==5514==    by 0x4A99E9D: anjuta_shell_add_value (in /usr/lib/libanjuta.so.0.0.0)
==5514==    by 0x70A0EC4: (within /usr/lib/anjuta/libanjuta-project-manager.so)
==5514==  Address 0x0 is not stack'd, malloc'd or (recently) free'd
==5514== 
==5514== Process terminating with default action of signal 11 (SIGSEGV)
==5514==  Access not within mapped region at address 0x0
==5514==    at 0x7AD6E09: symbol_db_engine_file_exists (in /usr/lib/anjuta/libanjuta-symbol-db.so)
==5514==    by 0x7AD75BB: symbol_db_engine_update_files_symbols (in /usr/lib/anjuta/libanjuta-symbol-db.so)
==5514==    by 0x7AD7CE9: symbol_db_engine_update_project_symbols (in /usr/lib/anjuta/libanjuta-symbol-db.so)
==5514==    by 0x7ABBD0D: (within /usr/lib/anjuta/libanjuta-symbol-db.so)
==5514==    by 0x4AA3773: (within /usr/lib/libanjuta.so.0.0.0)
==5514==    by 0x4A98A6E: anjuta_cclosure_marshal_VOID__STRING_BOXED (in /usr/lib/libanjuta.so.0.0.0)
==5514==    by 0x497A091: g_closure_invoke (in /usr/lib/libgobject-2.0.so.0.2200.0)
==5514==    by 0x498F6F7: (within /usr/lib/libgobject-2.0.so.0.2200.0)
==5514==    by 0x4990A7C: g_signal_emit_valist (in /usr/lib/libgobject-2.0.so.0.2200.0)
==5514==    by 0x4990D91: g_signal_emit_by_name (in /usr/lib/libgobject-2.0.so.0.2200.0)
==5514==    by 0x4A99E9D: anjuta_shell_add_value (in /usr/lib/libanjuta.so.0.0.0)
==5514==    by 0x70A0EC4: (within /usr/lib/anjuta/libanjuta-project-manager.so)
==5514==  If you believe this happened as a result of a stack overflow in your
==5514==  program's main thread (unlikely but possible), you can try to increase
==5514==  the size of the main thread stack using the --main-stacksize= flag.
==5514==  The main thread stack size used in this run was 8388608.
==5514== 
==5514== ERROR SUMMARY: 1 errors from 1 contexts (suppressed: 437 from 2)
==5514== malloc/free: in use at exit: 8,541,953 bytes in 59,090 blocks.
==5514== malloc/free: 936,189 allocs, 877,099 frees, 91,044,656 bytes allocated.
==5514== For counts of detected errors, rerun with: -v
==5514== searching for pointers to 59,090 not-freed blocks.
==5514== checked 16,835,644 bytes.
==5514== 
==5514== LEAK SUMMARY:
==5514==    definitely lost: 28,606 bytes in 618 blocks.
==5514==      possibly lost: 4,919,126 bytes in 4,184 blocks.
==5514==    still reachable: 3,594,221 bytes in 54,288 blocks.
==5514==         suppressed: 0 bytes in 0 blocks.
==5514== Rerun with --leak-check=full to see details of leaked memory.

So it seems there is a access attempt at address 0x00 (a NULL pointer maybe).

Best regards,
Paulo