| 00:00.00 | *** mode/#brlcad [+o tofu] by ChanServ | |
| 00:07.19 | starseeker | brlcad: I got better at testing for java/jni, but apparently not better enough |
| 00:09.29 | starseeker | that zlib incompatibility is Really Bad News |
| 00:09.42 | starseeker | why the *bleep* did they have to to mess it up? |
| 00:12.45 | CIA-128 | BRL-CAD: 03starseeker * r49502 10/brlcad/trunk/CMakeLists.txt: |
| 00:12.45 | CIA-128 | BRL-CAD: Default RTSERVER to OFF until we figure out what to do about the zlib issue - if |
| 00:12.45 | CIA-128 | BRL-CAD: someone needs it on and the need to use a java compiled against an older zlib, |
| 00:12.45 | CIA-128 | BRL-CAD: they'll need to set -DBRLCAD_ZLIB=OFF and -DBRLCAD_ENABLE_RTSERVER=ON. |
| 00:15.35 | starseeker | has half a notion to downgrade our zlib until they straighten out the mess... |
| 00:16.26 | starseeker | ah, wait - 1.2.6 got released last month |
| 00:16.29 | starseeker | didn't see that |
| 00:16.48 | starseeker | claims "Fixed a problem with large file support macros" |
| 00:18.57 | starseeker | won't help with the libgio issue though |
| 00:19.19 | starseeker | ZLIB_1.2.2 ??? |
| 00:25.52 | starseeker | is torn... to upgrade or not to upgrade, that is the question... |
| 00:26.41 | brlcad | on the system where I'm hitting the problem, I can't upgrade |
| 00:27.32 | brlcad | so I'll need to turn it off |
| 00:27.46 | starseeker | brlcad: it's off by default now |
| 00:27.47 | brlcad | cmake should detect the situation and autodisable, just not sure how other than maybe trying to link something against jnilib |
| 00:27.57 | brlcad | ahh, that might help :) |
| 00:28.12 | *** join/#brlcad CIA-128 (~CIA@cia.atheme.org) | |
| 00:28.12 | brlcad | at least short term |
| 00:28.23 | starseeker | only one known situation where it's needed, and we can turn it on with the right options for that |
| 00:28.32 | brlcad | the coverity case study is looking pretty sweet |
| 00:29.26 | starseeker | what the *bleep* are the zlib devs doing breaking compatibility with point releases like this? |
| 00:30.41 | starseeker | is trying to decide if it's worth it to put zlib 1.2.6 in src/other - if they've really addressed the large file macro issue that would be a step back towards vanilla, but we'd still have to fix the zconf.h stuff |
| 00:31.05 | starseeker | brlcad: oh, is the study out yet? |
| 00:31.41 | *** join/#brlcad User531 (~User@189.71.122.246) | |
| 00:33.00 | brlcad | it's not the first time zlib guys break compatibility .. actually you can break compatibility yourself if you set the right configure options |
| 00:33.14 | starseeker | arrgh |
| 00:33.19 | *** join/#brlcad anaozin (~gvifs@189.71.122.246) | |
| 00:33.30 | brlcad | the study isn't out, they sent me a draft for approval |
| 00:33.38 | starseeker | oh, sweet! |
| 00:33.50 | brlcad | I think it's finalized tomorrow |
| 00:33.59 | brlcad | don't know when it posts, but I'll repost when it does |
| 00:34.13 | starseeker | awesome - nice website news item :-) |
| 00:34.16 | anaozin | scan on? |
| 00:34.35 | brlcad | anaozin: hm? |
| 00:35.00 | anaozin | 1[br02,00lc12,00ad1,00]01: RFI scan on? |
| 00:35.01 | brlcad | starseeker: upgrading to 1.2.6 won't fix the rtserver issue |
| 00:35.07 | starseeker | I know |
| 00:35.45 | starseeker | wasn't proposing to upgrade to fix rtserver issue, was looking more at the claim they've addressed that large file support macro problem |
| 00:35.54 | brlcad | anaozin: coverity scan |
| 00:36.08 | brlcad | brl-cad participates in the coverity scan initiative |
| 00:36.32 | brlcad | it's static code analysis, looking for defects across the source using one of the best commercial scanners available |
| 00:36.48 | brlcad | starseeker: ah |
| 00:36.53 | brlcad | probably not worth it for just that |
| 00:37.13 | anaozin | im looking for servers with rfi scan |
| 00:37.13 | brlcad | given our version works, it's only potential for it to NOT work :) |
| 00:37.17 | anaozin | you know? |
| 00:37.27 | brlcad | anaozin: okay, now you're really confusing me |
| 00:37.33 | brlcad | rfi == request for information |
| 00:37.34 | starseeker | yeah - will have to check the changelog, but now that we've finally got it behaving I'm loath to break it |
| 00:38.00 | anaozin | rfi scan = scan vulnerability |
| 00:38.19 | brlcad | what does that mean? |
| 00:38.51 | brlcad | starseeker: I'd wait until there's some other compelling reason (like a CVE or some improvement etc) |
| 00:39.31 | brlcad | anaozin: oooh, I get it .. sorry, being a bit slow |
| 00:39.47 | brlcad | anaozin: why the hell would you think to ask for that here?? |
| 00:40.31 | anaozin | because i found this channel in google |
| 00:40.31 | anaozin | lol |
| 00:41.12 | *** kick/#brlcad [anaozin!~Adium@c-68-34-98-23.hsd1.md.comcast.net] by tofu (lame) | |
| 00:41.13 | *** join/#brlcad anaozin (~gvifs@189.71.122.246) | |
| 00:41.24 | anaozin | loll |
| 00:41.27 | anaozin | 12,00:01D |
| 00:41.29 | tofu | forewarned, ask for malware elsewhere |
| 00:45.51 | brlcad | not really funny, just sad |
| 00:46.09 | brlcad | so much more interesting constructive things one could do with their time |
| 00:46.15 | starseeker | no kidding |
| 00:52.02 | starseeker | shelves zlib 1.2.6 |
| 00:53.02 | brlcad | idiot |
| 02:00.51 | *** join/#brlcad tofu1 (~Adium@c-68-34-98-23.hsd1.md.comcast.net) | |
| 04:27.15 | *** join/#brlcad kaushik_ (~kaushik@122.164.44.178) | |
| 05:20.36 | *** join/#brlcad kaushik_ (~kaushik@122.164.44.178) | |
| 05:24.52 | CIA-128 | BRL-CAD: 03brlcad * r49503 10/brlcad/trunk/doc/docbook/articles/en/ (CMakeLists.txt about.xml): add the original about page from the website as a docbook article. we've gotten a lot of mileage out of that text since I pulled it all together 8+ years ago, long overdue some basic revision control. |
| 06:14.05 | *** join/#brlcad kaushik_ (~kaushik@122.164.44.178) | |
| 07:02.56 | CIA-128 | BRL-CAD: 03brlcad * r49505 10/brlcad/trunk/ (6 files in 3 dirs): include an even older translation of the about page to russian by Ilya in 2008. don't have many details on this contributor but again the translation seems sane. |
| 08:59.20 | *** join/#brlcad jordisayol (~jordisayo@unaffiliated/jordisayol) | |
| 10:46.32 | *** join/#brlcad louipc (~louipc@archlinux/fellow/louipc) | |
| 14:00.50 | *** join/#brlcad tofu (~Adium@c-68-34-98-23.hsd1.md.comcast.net) | |
| 14:00.50 | *** mode/#brlcad [+o tofu] by ChanServ | |
| 16:39.51 | brlcad | implements a CheckCInline for icc |
| 16:54.03 | CIA-128 | BRL-CAD: 03brlcad * r49506 10/brlcad/trunk/sh/footer.sh: use cmake-mode instead of sh-mode now that it works, even though users have to install it manually |
| 16:57.44 | ``Erik | libcheck for unit testing? I'm using it in that side C lib project, so I could copy over the cmake bits and do up my tesselators test suite with it |
| 17:01.53 | ``Erik | also; ffi friendly shootray? { struct application a; struct partition *p; load_geom(&a, filename, object); p=rt_shootray_simple(&a, /*point_t*/origin, /*vect_t*/dir); do{ ...} while(p=p->next); } ? (to dump librtserver in favor of jna down the road) |
| 17:02.44 | *** join/#brlcad umashankar (IceChat77@117.204.161.208) | |
| 17:33.19 | *** join/#brlcad KimK (~Kim__@209.248.147.2.nw.nuvox.net) | |
| 17:37.53 | CIA-128 | BRL-CAD: 03brlcad * r49507 10/brlcad/trunk/src/proc-db/menger.c: capture the mask values in an int before assigning to the mask to appease icc |
| 17:41.17 | *** join/#brlcad umashankar (IceChat77@117.204.161.208) | |
| 17:41.29 | *** part/#brlcad umashankar (IceChat77@117.204.161.208) | |
| 17:44.39 | CIA-128 | BRL-CAD: 03brlcad * r49508 10/brlcad/trunk/src/libged/ (edit.c tables.c): icc portability, undo over-constification decls |
| 17:51.05 | CIA-128 | BRL-CAD: 03brlcad * r49509 10/brlcad/trunk/src/libged/ (exists.c typein.c): icc does not like code that implicitly pretends enums are ints. cast accordingly to make the intent explicit. |
| 18:26.27 | CIA-128 | BRL-CAD: 03brlcad * r49510 10/brlcad/trunk/src/libged/attr.c: name is already the right type for bu_strcmp() |
| 18:31.48 | CIA-128 | BRL-CAD: 03brlcad * r49511 10/brlcad/trunk/src/librt/search.c: init to one of the enum types, not an int |
| 18:32.51 | CIA-128 | BRL-CAD: 03brlcad * r49512 10/brlcad/trunk/src/util/fix_polysolids.c: init at least the first union field to zero. |
| 18:34.45 | CIA-128 | BRL-CAD: 03brlcad * r49513 10/brlcad/trunk/src/mged/mged.c: mged_notify() is used as an void(*)(int) pointer, so give it an unused int param |
| 18:35.40 | CIA-128 | BRL-CAD: 03brlcad * r49514 10/brlcad/trunk/src/librt/primitives/pnts/pnts.c: cast int to enum to make icc happy |
| 18:37.26 | CIA-128 | BRL-CAD: 03brlcad * r49515 10/brlcad/trunk/src/vdeck/vdeck.c: init op to obuf so it's not potentially used before init. |
| 18:42.55 | CIA-128 | BRL-CAD: 03brlcad * r49516 10/brlcad/trunk/src/librt/primitives/revolve/revolve.c: bonefide bug detected by icc, the sketch verts array are 2d entities so we should be calling V2MOVE() instead of VMOVE() and V2JOIN1() instead of VJOIN1(). |
| 18:43.13 | CIA-128 | BRL-CAD: 03brlcad * r49517 10/brlcad/trunk/src/liboptical/sh_text.c: fill out the callback functions that do nothing just to appease pedantic compilation. uncovered the need for a second mlib_void() due to the different signature. |
| 18:50.25 | CIA-128 | BRL-CAD: 03brlcad * r49518 10/brlcad/trunk/src/ (conv/dem-g.c libtclcad/tclcad_obj.c): cast int to enum type to appease icc |
| 18:53.08 | CIA-128 | BRL-CAD: 03brlcad * r49519 10/brlcad/trunk/src/liboptical/photonmap.c: avoid division by zero |
| 19:01.04 | starseeker | brlcad: any idea how xpush might be managing to corrupt dbip->dbiHead? |
| 19:02.19 | starseeker | I can see *when* it's happening - *headp = dp; in db_lookup.c - but I'm not sure how it's managing to get the right pointer badness to do it |
| 19:23.58 | *** join/#brlcad andymanak (~andy@sjs-mh-wifi-1-0-lc-int.sjsu.edu) | |
| 19:30.23 | starseeker | no, that might not be it... when I add print statements, something else happens?? |
| 19:32.36 | brlcad | sure, a bug ;) |
| 19:32.52 | brlcad | otherwise, not much use without a debugger and loads of time to investigate |
| 19:33.29 | brlcad | }hello andymanak |
| 19:37.48 | andymanak | hello |
| 19:55.32 | starseeker | is beyond his depth here... |
| 20:03.30 | brlcad | but you've been debugging less than a day! :) |
| 20:08.39 | brlcad | starseeker: given the potential corruption lead, you might have better luck running through valgrind |
| 20:08.57 | brlcad | then it might detect a direct memory overrun for you |
| 20:10.57 | CIA-128 | BRL-CAD: 03starseeker * r49520 10/brlcad/trunk/ (NEWS src/tclscripts/rtwizard/lib/MGEDpage.itk): Apply alphabetical sorting to the toplevel objects list in rtwizard. |
| 20:12.34 | starseeker | brlcad: it looks like a dp->d_use_hd somehow is initialized and has what looks like a valid pointer to the if test, but when the free routine goes after it... kablooey |
| 20:12.54 | starseeker | probably several ways that could happen... |
| 20:13.03 | starseeker | maybe valgrind is a good idea |
| 20:15.39 | CIA-128 | BRL-CAD: 03starseeker * r49521 10/brlcad/branches/STABLE/src/tclscripts/rtwizard/lib/MGEDpage.itk: put rtwizard tweak in stable too. |
| 20:39.50 | CIA-128 | BRL-CAD: 03starseeker * r49522 10/brlcad/branches/STABLE/src/librt/primitives/revolve/revolve.c: merge bugfix from trunk r49516 |
| 21:24.48 | CIA-128 | BRL-CAD: 03starseeker * r49523 10/brlcad/trunk/src/libdm/CMakeLists.txt: clear previous definitions in case something (like opengl settings) changed. |
| 21:27.50 | CIA-128 | BRL-CAD: 03starseeker * r49524 10/brlcad/trunk/src/librt/db_alloc.c: (log message trimmed) |
| 21:27.51 | CIA-128 | BRL-CAD: Doubtful this gets to the heart of the matter, but when investigating an xpush |
| 21:27.51 | CIA-128 | BRL-CAD: crash valgrind reports memory being called uninitialized in the xpush free |
| 21:27.51 | CIA-128 | BRL-CAD: routines at xpush.c:90 is initially allocated by db_alloc_directory_block. |
| 21:27.51 | CIA-128 | BRL-CAD: Somehow the dp->d_use_hd entry in question has a 'used' pointer pointing to |
| 21:27.51 | CIA-128 | BRL-CAD: uninitialized memory. Should have it in the first place, but if we use |
| 21:27.52 | CIA-128 | BRL-CAD: bu_calloc in db_alloc_directory_block instead of bu_malloc we at least don't |
| 21:44.44 | CIA-128 | BRL-CAD: 03starseeker * r49525 10/brlcad/trunk/src/ (adrt/CMakeLists.txt libtclcad/CMakeLists.txt): Couple more tweaks for turning opengl on/off |
| 21:52.22 | *** part/#brlcad andymanak (~andy@sjs-mh-wifi-1-0-lc-int.sjsu.edu) | |
| 22:19.59 | CIA-128 | BRL-CAD: 03starseeker * r49526 10/brlcad/trunk/src/ (librt/CMakeLists.txt libwdb/CMakeLists.txt): few more initializations |
| 22:29.12 | *** join/#brlcad Yoshi47 (~jan@d72-39-60-53.home1.cgocable.net) | |
| 22:29.49 | CIA-128 | BRL-CAD: 03n_reed * r49527 10/brlcad/trunk/src/libged/CMakeLists.txt: initialize GED defines |