|Summary:||System bus crashes on Mac OS X 10.4|
|Product:||dbus||Reporter:||Randall Wood <rhwood>|
|Component:||core||Assignee:||Havoc Pennington <hp>|
|Status:||RESOLVED WONTFIX||QA Contact:||John (J5) Palmieri <johnp>|
|OS:||Mac OS X (All)|
|Whiteboard:||unreproducible on 10.6|
|i915 platform:||i915 features:|
0.94 Build log with --enable-tests
0.94 make check after building with --enable-tests
make check log for 0.94 with bus/bus.c patched as in comments
download/configure/build/test log from CVS HEAD
Description Randall Wood 2006-09-20 00:41:57 UTC
Comment 1 Randall Wood 2006-09-20 00:44:38 UTC
Created attachment 7092 [details] Installation log The attached log gives the build history and build environment for the crashing dbus. Lines starting DEBUG: are from port, the build automation tool from macports.org.
Comment 2 John (J5) Palmieri 2006-10-11 10:32:35 UTC
Please compile witch tests and checks. The --enable-tests bug should be fixed in CVS.
Comment 3 Randall Wood 2006-10-16 02:03:19 UTC
Created attachment 7421 [details] 0.94 Build log with --enable-tests Log of configure & build of version 0.94 Log was generated using the MacPorts port command and so includes output from that command that wraps the configure and make commands.
Comment 4 Randall Wood 2006-10-16 02:05:18 UTC
Created attachment 7422 [details] 0.94 make check after building with --enable-tests If I knew how to use gdb or to compile with a backtrace, I would be able to provide that information as well.
Comment 5 Randall Wood 2006-10-16 02:08:09 UTC
Crash when running the command "sudo dbus-daemon --system": ********** Host Name: Getafix Date/Time: 2006-10-15 21:18:28.325 -0400 OS Version: 10.4.8 (Build 8L127) Report Version: 4 Command: dbus-daemon Path: /opt/local/bin/dbus-daemon Parent: launchd  Version: ??? (???) PID: 10950 Thread: 0 Exception: EXC_BAD_ACCESS (0x0001) Codes: KERN_PROTECTION_FAILURE (0x0002) at 0x00000004 Thread 0 Crashed: 0 dbus-daemon 0x000330ac _dbus_list_prepend + 72 1 dbus-daemon 0x00033014 _dbus_list_append + 40 2 dbus-daemon 0x00034ef8 add_callback + 48 3 dbus-daemon 0x000351f8 _dbus_loop_add_watch + 108 4 dbus-daemon 0x00013f7c bus_watch_directory + 340 5 dbus-daemon 0x00033bcc _dbus_list_foreach + 136 6 dbus-daemon 0x00006a1c process_config_postinit + 168 7 dbus-daemon 0x00007104 bus_context_new + 1732 8 dbus-daemon 0x000278f8 main + 1972 9 dbus-daemon 0x0000201c _start + 760 10 dbus-daemon 0x00001d20 start + 48 Thread 0 crashed with PPC Thread State 64: srr0: 0x00000000000330ac srr1: 0x000000000000d030 vrsave: 0x0000000000000000 cr: 0x24000244 xer: 0x0000000000000004 lr: 0x0000000000033088 ctr: 0x0000000090003730 r0: 0x0000000000300904 r1: 0x00000000bffff3f0 r2: 0x0000000000000004 r3: 0x0000000000300904 r4: 0x00000000003008d0 r5: 0x0000000000000000 r6: 0x0000000000000006 r7: 0x0000000000000007 r8: 0x0000000000402015 r9: 0x000000000000000c r10: 0x0000000000400011 r11: 0x00000000003008f0 r12: 0x0000000090003138 r13: 0x0000000000000000 r14: 0x0000000000000000 r15: 0x0000000000000000 r16: 0x0000000000000000 r17: 0x0000000000000000 r18: 0x0000000000000000 r19: 0x0000000000000000 r20: 0x0000000000000000 r21: 0x0000000000000000 r22: 0x0000000000000000 r23: 0x0000000000000000 r24: 0x0000000000000000 r25: 0x0000000000000002 r26: 0x00000000bffff980 r27: 0x00000000bffff920 r28: 0x00000000bffff914 r29: 0x000000000007955c r30: 0x00000000bffff3f0 r31: 0x0000000000013e40 Binary Images Description: 0x1000 - 0x85fff dbus-daemon /opt/local/bin/dbus-daemon 0x9c000 - 0xbafff libexpat.1.dylib /opt/local/lib/libexpat.1.dylib 0x8fe00000 - 0x8fe51fff dyld 45.3 /usr/lib/dyld 0x90000000 - 0x901bcfff libSystem.B.dylib /usr/lib/libSystem.B.dylib 0x90214000 - 0x90219fff libmathCommon.A.dylib /usr/lib/system/libmathCommon.A.dylib
Comment 6 John (J5) Palmieri 2006-10-16 09:13:49 UTC
looks like check_get_connection_unix_process_id isn't getting a pid in the tests. Hmm, perhaps I should get that macbook since I need to return this laptop to the Desktop group ;-) In any case there are two errors here. The unix_process_id error and array out of bounds for the file watch code which is different on *BSD and Linux. My guess is the patch I asked the FreeBSD guys to hack up, which hasn't been submitted yet, is the problem. Can you do me a favor and add this bit: in bus/bus.c function process_config_postinit there is a call to _dbus_list_foreach where the last parameter is NULL. Change that NULL to bus_context_get_loop (context) and try compiling and running the bus. Looking at the FreeBSD patch (Bug #8298) I think it will also fix up the check_get_connection_unix_process_id bug also. If the above fixes the first part I'm going to mark this bug as a duplicate.
Comment 7 Randall Wood 2006-10-31 03:11:59 UTC
Created attachment 7605 [details] make check log for 0.94 with bus/bus.c patched as in comments Fails on the same procedure but in a different test. I tried the full set of patches from the BSD patches (changing #ifdef __FreeBSD__ to #ifdef __APPLE__) and the test fails the same.
Comment 8 John (J5) Palmieri 2006-10-31 08:33:57 UTC
Can you try the source from CVS? Thanks
Comment 9 Randall Wood 2006-12-03 02:29:21 UTC
Created attachment 7946 [details] download/configure/build/test log from CVS HEAD This log is from a download/configure/build/test cycle that was automated using the MacPorts.org port command.
Comment 10 Benjamin Close 2008-01-11 02:36:48 UTC
Bugzilla Upgrade Mass Bug Change NEEDSINFO state was removed in Bugzilla 3.x, reopening any bugs previously listed as NEEDSINFO. - benjsc fd.o Wrangler
Comment 11 Mike McQuaid 2010-12-30 03:19:00 UTC
I can't reproduce this on 10.6. Can you try with the latest master? Otherwise this can be closed.
Comment 12 Randall Wood 2011-01-27 15:59:18 UTC
I no longer have access to a Mac OS X 10.4 system. I suggest that this bug be resolved as "won't fix" and reopened if a Mac OS X 10.4 user experiences this issue and is available to assist in addressing it.
Comment 13 Simon McVittie 2011-04-07 10:40:04 UTC
(In reply to comment #12) > I no longer have access to a Mac OS X 10.4 system. I suggest that this bug be > resolved as "won't fix" and reopened if a Mac OS X 10.4 user experiences this > issue and is available to assist in addressing it. Sounds good to me.