IRC Logs for #crux-devel Thursday, 2012-12-13

*** acrux has quit IRC02:03
*** acrux has joined #crux-devel02:05
*** acrux has joined #crux-devel02:05
*** joe9 has joined #crux-devel06:44
*** horrorStruck has quit IRC09:27
*** horrorStruck has joined #crux-devel09:28
*** c0x has quit IRC10:45
teK_jue10:51
jueteK_: heyo10:53
teK_hello10:53
teK_I saw you updated opt/valgrind10:53
teK_not by accident I suppose10:53
teK_did you sync it with the 64bit version?10:53
jueyep, I saw you did it for 64bit but not for 32, so I thought it might be helpful todo the update ;)10:55
teK_just got the conflict before commiting /pulling10:55
teK_good to knowm thanks10:55
jueand yes, merged it into 3.010:55
teK_I should be much less busy after christmas or so..10:55
teK_at least I hope so10:55
juewell, not at problem at all, the main reason I jumped in was the already closed ticket :)10:57
teK_maintaining packages halfway sucks which was the original reason I jumped in ;)10:57
jueone thing to note: I got a different footprint in our 3.0 branch, compared to the x86_64 branch10:59
jueprobably related to the multilib environment?11:00
teK_diffing right now11:00
teK_hm11:01
teK_s/amd64/x86/g11:01
teK_well.. would you mind running valgrind on 3.0 on a 64bit binary?11:01
juesure, if you tell me what I have to do11:02
juenever used it ;)11:02
teK_for a quick check: valgrind $RANDOM_64BIT_BINARY11:03
teK_tbh I did not check out 3.0 to any extent..11:04
teK_feels kinda stupid having to ask: are there any pure 64bit binaries on 3.0 (I suppose so)11:04
juewell, everything is pure 64bit if you install nothing from compat-3211:05
jueseems to work11:08
jueteK_: looks like we have 2 versions of the vgpreload_* modules/libs11:10
juee.g. vgpreload_core-amd64-linux.so and vgpreload_core-x86-linux.so11:11
juesame for others, helgrind-* etc.11:13
horrorStruckfrinnst: i seems that core-x86_64/util-linux could be dropped. with 2.22.2: configure: WARNING: unrecognized options: --disable-runuser11:35
frinnstwhat about setarch?11:43
teK_jue: meh so why does the current footprint only show the 32bit and not only 64bit11:43
frinnstusr/bin/x86_64 -> setarch11:43
jueteK_: sorry, don't get it, I see x86 libs on 2.8 and both on 3.0?11:46
jueboth means amd64 and x8611:46
teK_woot? a second.. :)11:47
horrorStruckfrinnst: i may have missed something but the diff between core/u-l and core-x86_64/u-l seems to only be this option, which seems not to exist anymore11:49
frinnstno, its the footprint too11:50
horrorStruckah crap, of course11:50
horrorStrucksorry11:50
frinnstnp :)11:50
frinnstbtw, are there any issues left for 3.0? the wiki seems mostly (done) :)11:51
frinnstoh right, flyspray maybe11:51
frinnstone critical bug left :)11:51
jaegerAnyone done a 2.8 x86 -> 3.0 upgrade yet? It's on my list to do but I haven't had time yet11:51
frinnstjaeger: yeah I run 3.0 here11:51
frinnstbut not with the ISO11:51
frinnstcould do that tomorrow at work - its still on 2.8 multilib11:52
jaegerI mean specifically a 2.8 x86 to 3.0 upgrade, though11:52
juejaeger: no, did a fresh install11:52
jaegerok11:52
frinnstah11:52
teK_so I must be blind (and grep, too) jue11:52
teK_wtf11:52
jaegerI still plan to test that before 3.011:52
teK_curl -s http://crux.nu/gitweb/\?p\=ports/opt.git\;a\=blob_plain\;f\=valgrind/.footprint\;h\=7469065d78c734164b68ac038109c55e58df9859 | grep vgpreload11:53
teK_this if for the 3.0 branch and only yields x8611:53
jueteK_: http://crux.nu/gitweb/?p=ports/opt.git;a=blob;f=valgrind/.footprint;h=8b1e5cbd6d69a6a7e18e67c58823379e71f2a980;hb=0d17ef34eacaf5f729b4d4fcd5b1ecf6f83f020a11:54
teK_yeah seems right11:55
teK_what have I been looking at.. oh my...:)11:55
teK_so we sould be fine. god balgrind.11:55
teK_+o11:55
frinnsthm, i get a footprint missmatch on xfsprogs11:57
frinnstasdf, md5sum11:57
frinnstnot footprint11:57
jue.oO11:58
juefrinnst: http://thread.gmane.org/gmane.comp.file-systems.xfs.general/4942011:59
frinnstah11:59
jueseems I was a bit too fast12:00
frinnst:)12:04
frinnstjaeger: running an upgrade of a 32bit 2.8 to 3.012:04
jaegercool12:04
juesorry for the hassle12:04
frinnstit has X and some other stuff, so should be a good test :)12:04
jaegerexcellent12:05
juejaeger: FYI, I've committed two small things to iso.git12:05
jaegerok12:06
jaegerI saw the harfbuzz and init version commits if those are the ones you mean12:06
jueyeah, that was it12:06
jueat all 3.0 works brilliant, thanks jaeger12:08
jaegerGlad to hear that. :) I've been using it as well on 2 machines but haven't done upgrades yet12:08
frinnstyeah same here12:10
frinnstwell, it booted up once i had run the 3.0 media a second time to rebuild the kernel :)12:15
horrorStruck3.0 branch of xorg repo needs the 2.8 libpixman update :)12:16
jaegerdidn't want to rebuild it on the first boot?12:16
frinnstyeah it probably hasnt been rebased for a while12:16
frinnstheh, nah. i forgot :)12:16
jaegerAh, ok12:17
jaegerJust wanted to make sure it wasn't a technical issue12:17
frinnstX wouldnt start - wrong elfclass for modesetting_drv.so12:17
frinnstits not included on the cd? xorg-xf86-video-modesetting12:17
jueno, it isn't, we should probably do that?12:18
frinnstits probably not a big deal. other stuff will still be broken :)12:20
jueguess a lot :)12:21
frinnstyeah - everything :)12:22
frinnsthmm, easy way of locating 32bit binaries?12:24
frinnstcheck all executeables with "file" and grep for "32" ?12:31
frinnstand libraries..12:31
jaegerthat's what I usually do12:39
horrorStruckcustom :P root:~# crux12:49
horrorStruckCRUX version 3.0 x86_6412:49
teK_jue jaeger so almost every commit to 2.8 is supposed to go to 3.0, right12:53
jaegerany time that it makes sense to do so, I'd say12:54
*** jue_ has joined #crux-devel14:40
*** jue has quit IRC14:47
*** pitillo has quit IRC14:47
*** jue_ is now known as jue14:48
*** pitillo has joined #crux-devel14:52
*** deus_ex has quit IRC18:31
*** deus_ex has joined #crux-devel18:32
*** acrux has quit IRC19:02
*** __mavrick61 has quit IRC21:28
*** __mavrick61 has joined #crux-devel21:29

Generated by irclog2html.py 2.11.0 by Marius Gedminas - find it at mg.pov.lt!