| 03:37.14 | *** join/#brlcad packrat (~packrator@c-98-209-146-133.hsd1.mi.comcast.net) |
| 03:56.11 | *** join/#brlcad packrat (~packrator@c-98-209-146-133.hsd1.mi.comcast.net) |
| 04:02.38 | *** join/#brlcad hackrat (~packrator@c-98-209-146-133.hsd1.mi.comcast.net) |
| 05:49.01 | *** join/#brlcad Beetle (~beetle@h70.20.131.174.dynamic.ip.windstream.net) |
| 05:49.27 | *** part/#brlcad Beetle (~beetle@h70.20.131.174.dynamic.ip.windstream.net) |
| 07:35.28 | *** join/#brlcad jordisayol (~jordisayo@unaffiliated/jordisayol) |
| 13:02.39 | *** join/#brlcad dtidrow (~dtidrow@c-68-84-167-135.hsd1.mi.comcast.net) |
| 17:11.03 | Maloeran | Hey. I'm trying to investigate a bug in VSL, which crashes somewhere deep in BRL-CAD while opening a .g file, in bu_vls_trimspace(). |
| 17:11.17 | Maloeran | The (simplified) backtrace is bu_vls_trimspace(), bu_str_true(), db_attr_add(), db5_standardize_avs(), rt_db_external5_to_internal5(), rt_db_get_internal5(), rt_db_get_internal(), db_update_nref() |
| 17:11.40 | Maloeran | Does anyone have a pointer or two? |
| 17:13.01 | Maloeran | And at first glance, I don't really trust that while(isspace(...)) loop |
| 18:02.46 | Maloeran | Actually... (gdb) p *vp $2 = {vls_magic = 2301836219, vls_str = 0x7fffcc002050 "R", vls_offset = 4294967296, vls_len = 140733193388072, vls_max = 140737028006102} |
| 21:24.44 | brlcad | at a glance, it looks like an uninitialized vls being used |
| 21:25.18 | brlcad | compiled without optimization, gcc automatically zero-initializes |
| 21:25.52 | brlcad | with optimization, it's random data so you can run into that situation rather easily |
| 21:27.32 | brlcad | that said, the vls in question inside bu_str_true() is initialized, so don't know what the issue is there |
| 21:49.59 | Maloeran | Valgrind does say the memory was unitialized |
| 21:51.15 | Maloeran | I would be more inclined to blame VSL than BRL-CAD, but I didn't notice anything clearly wrong with VSL's calls... |
| 22:14.18 | *** join/#brlcad pawleeq (~pawleeq@212-96-188-229.cust.selfnet.cz) |
| 22:53.36 | *** join/#brlcad pawleeq (~pawleeq@212-96-188-229.cust.selfnet.cz) |
| 23:14.44 | starseeker | growl... the tree highlighting in Archer isn't happy |
| 23:17.57 | starseeker | makes a note to ask Bob tomorrow... |