**** BEGIN LOGGING AT Mon Feb 29 02:59:59 2016 Feb 29 08:03:44 good morning Feb 29 13:54:00 I'm getting the dreaded "Package version for package X went backwards..." QA error when I change MACHINE Feb 29 13:54:27 If a dependency of X has PACKAGE_ARCH = "${MACHINE_ARCH}" set, should X itself also have that set? Feb 29 14:28:28 Hi. Does anybody know why this channel logs stop at end of 2015? https://www.yoctoproject.org/irc/ Feb 29 14:29:19 diego_r: I've raised that before, would be good to get it fixed! Feb 29 14:31:34 CTtpollard: I see, thanks Feb 29 14:31:35 diego_r: You can use http://logs.nslu2-linux.org/livelogs/ Feb 29 14:34:35 paulbarker: thanks Feb 29 14:36:58 Hi rburton. I'm confused about 'meta package' and package group, with respect to what we discussed here: http://logs.nslu2-linux.org/livelogs/yocto/yocto.20160212.txt (search for 'not a packagegroup, but maybe a metapackage') Feb 29 14:38:12 in another reply you say that 'a meta packages whose sole purpose is to bring in dependencies is a packagegroup' https://lists.yoctoproject.org/pipermail/yocto/2014-March/018610.html Feb 29 15:01:04 diego_r: a packagegroup is a specific type of meta package Feb 29 15:01:21 a meta package is a generic name for a package that has no content itself, just dependencies Feb 29 15:01:40 a packagegroup is something that is produced by a recipe inheriting packagegroup Feb 29 15:02:10 whereas gstreamer-plugins-good is a metapackage that depends on all the individual gstreaer plugin packages but contains nothing itself Feb 29 15:03:27 rburton: thanks, should be clearer now Feb 29 15:50:02 zeddii_home: do you understand the perf build system? Feb 29 15:50:05 Hi Guys, Feb 29 15:50:14 i am newbie yocto Feb 29 15:50:24 i use yocto fido Feb 29 15:50:33 rburton: I do. and then it changes. I was hunting a bug about two weeks ago and was seriously confused/pissed off at it. Feb 29 15:50:53 i have variscite kit whis is VAR_STK_6_UL based on imx6 Feb 29 15:51:29 zeddii_home: http://errors.yoctoproject.org/Errors/Details/38137/ Feb 29 15:51:32 anybody know usefull package manager ? Anybody can answer me ? Feb 29 15:51:41 rburton: looking. Feb 29 15:52:22 zeddii_home: presumably a race as it appears to be compiling that file twice Feb 29 15:52:38 zeddii_home: thought you might be able to fix it quicker than i can :) Feb 29 15:52:46 (failed twice on the ab over the weekend) Feb 29 15:53:02 Hi Guys, i am newbie yocto, i use yocto fido, i have variscite kit whis is VAR_STK_6_UL based on imx6 anybody know usefull package manager ? Anybody can answer me ?, Feb 29 15:53:36 rburton: yah. there is something racing in that area. I swear I’ve seen that as well. Looking at the Makefiles to remember where I was. Feb 29 15:55:21 rburton: the truncated file is intereseting. something copying to work-shared while the build starts ? hmm. Feb 29 15:55:25 zeddii_home: also its doing a lot of compilation in do_install. is that rebuilding or just not doing any building in compile? Feb 29 15:55:47 zeddii_home: my guess would be one gcc has truncated the file to write to it whilst another is linking with it Feb 29 15:55:51 there are two CC lines for that plugin Feb 29 15:56:36 if we serialize the build for perf, I wonder … Feb 29 15:57:05 that's horrible! Feb 29 15:58:17 but yah, the building in the install phase should be minimal. but de-tangling the perf makefiles (with all the shared tools junk) is the trick. Feb 29 15:58:19 zeddii_home: so it doesn't get lost i just filed #9182 Feb 29 16:00:40 Hi Guys, i am newbie yocto, i use yocto fido, i have variscite kit whis is VAR_STK_6_UL based on imx6 anybody know usefull package manager ? Anybody can answer me ?, Feb 29 16:01:09 how can i install gcc on this board ? Feb 29 16:01:20 kageja: depending on how they're using yocto you could have smart or opkg installed Feb 29 16:03:25 No opkg was not installed Feb 29 16:04:24 rburton: you know another package manager ? Feb 29 16:04:29 as i said, smart Feb 29 16:04:46 but you're asking for help on something that i've never used Feb 29 16:05:32 variscite has a support forum, you might get better answers there Feb 29 16:05:49 ohoo sorry i mean smart is smart technology which is anything, ok i try immediatley Feb 29 16:17:29 hmm, is there a way to force the creation of startup.nsh also for non-iso, or will i need to override that function from the class? Feb 29 16:55:14 kergoth: for minnow i just tweaked the mkefidisk.sh script to do it Feb 29 17:05:57 tlwoerner: yeah, that's what we've been doing, just wanted to use wic instead, and ran into problems with it :) Feb 29 17:10:37 Hi folks, I used openembedded alot in the past and now going to switch to poky/yocto. I have question regarding "coding style". Is it legit to use generig %.bbappend files in an overlay layer instead of fixed version.bbappend files? Feb 29 17:11:41 cpo_: it is common practice to wildcard Feb 29 17:11:56 if you know you want to append something against a moving recipe Feb 29 17:11:59 ok, so a submitted patch would be welcome? Feb 29 17:41:54 hi everyone, i am trieng install apache on yocto fido, and i take an error which is configure: error: no acceptable C compiler found in $PATH Feb 29 17:41:54 , How can i fix this? how can i install gcc, meanwhile my board is var_stk_6_ul based on imx 6 ultralite, its varissciye product Feb 29 17:42:51 kageja: you're trying to install apache on a built system? Feb 29 17:43:22 kageja: also your system might have gcc, but not set in path Feb 29 17:45:55 CTtpollard: yes i think as you, how can i added gcc to $PATH Feb 29 17:46:45 kageja: PATH=$PATH:./usr/bin/gcc ? Feb 29 17:48:03 it would be odd though Feb 29 17:52:50 CTtpollard: that isn't work, ls -la /usr/bin | grep gcc return nothiing, and i tried your code after then ./configure that isnt work Feb 29 17:53:25 kageja: that would suggest gcc isn't in that location, which would usually mean it's not on your system Feb 29 17:55:13 CTtpollard: okey, what must i do for installing gcc on my system Feb 29 17:55:56 kageja: can you do "gcc --version" or "which gcc" ? Feb 29 17:56:09 its return nothing Feb 29 17:56:17 CTtpollard: its return nothing Feb 29 17:56:55 kageja: I would suggest rebuilding your system then to include gcc Feb 29 17:57:42 its so hard, there must be another way, as like installing gcc Feb 29 17:57:48 CTtpollard: its so hard, there must be another way, as like installing gcc Feb 29 17:58:48 kageja: http://www.linuxfromscratch.org/blfs/view/svn/general/gcc.html ? :) Feb 29 17:59:51 kageja: I would suggest it is easier to rebuild your roots with gcc in it Feb 29 17:59:57 *rootfs Feb 29 18:07:51 CTtpollard: so many thanks for your helping, but i think its so hard, there must be another wat like as smart install gcc Feb 29 19:45:51 hey anyone around a C/python module expert? Feb 29 19:46:06 RPM between 5.4.14 and 5.4.16 updated to a newer version of the upstream rpm-python module.. Feb 29 19:46:18 this appears to have broken the file listing capabilities, but I don't see the problem.. Feb 29 19:46:34 but I'm in no way an expert on how python/C interface works.. Feb 29 19:46:54 the sympom is that python is getting a list (which is correct), in the old way it has many items.. in the new version it only has 1 item in the index Feb 29 19:47:03 (the first item) Feb 29 19:52:52 morning all Feb 29 20:17:21 Hey, RP, any updates on 9112? I have thoughts about further refactoring of pseudo's startup to improve diagnostics, but I'm also wondering whether we're getting anything else useful from existing logs, etc. Feb 29 20:19:55 seebs: The only thing I have to add since we last talked is I bumped retries to 250 and we've not seen a 134 exit code since Feb 29 20:20:11 Huh. Interesting. Feb 29 20:20:21 seebs: didn't know what to test so that seemed like my only option. Nothing further ever came from the logs Feb 29 20:20:25 So there's a typo in the code, which is that there's an X+3 that's supposed to be (X*3). Feb 29 20:20:45 * RP did wonder about that Feb 29 20:20:53 I've got a patch that adds some event logging, and in particular, I also added shutdown stats on database close time. Feb 29 20:21:42 I was going to suggest a thing, though, which is (1) set retries somewhat lower, (2) add --disable-memory-db. Feb 29 20:21:56 Because I *suspect* that what's biting us is the time for the database copy on shutdown. Feb 29 20:22:32 I also suspect that the memory database is not buying as much performance-wise as I'd originally hoped, because my other pseudo optimization work suggests that I was massively overestimating the actual database access costs. Feb 29 20:22:53 seebs: I suspect you're probably right, however we just hit M3 feature freeze so my ability to run test builds to debug this just got a lot harder Feb 29 20:23:08 This is why I've been trying to hard to fix this for the past few weeks :/ Feb 29 20:23:23 But what it looks like to me is that the problems we're seeing seem to only happen if the server spends an abnormally long time shutting down, and there's virtually nothing that happens after the socket is closed... except the pdb_backup in the close. Feb 29 20:23:53 Yeah. I had come up with this theory on the 24th, but I don't know whether you get pings on updates on the ticket. Feb 29 20:24:11 I'm still trying to find a real-world reproducer I can actually hit on my hardware. Feb 29 20:25:19 Anyway, now I have an event logger that, in addition to its other features, has fancy-looking timestamps so I can see how long components of a thing are taking. Mar 01 01:38:27 Hi all, I'm trying to introduce myself to Yocto. I am on an OSX machine and running Yocto on a headless Ubuntu 15.10. I've been using `runqemu qemux86 publicvnc`, but whenever I try to connect from my mac (realvnc client), the connection closes immediately after initializing (CConnection: close: [EndOfStream] The connection closed unexpectedly.). Any thoughts or suggestions? Mar 01 01:39:32 I've been following http://www.yoctoproject.org/docs/2.0/mega-manual/mega-manual.html#qs-building-images Mar 01 02:07:04 solved the issue, using a different vnc client worked (in this case TigerVNC) **** ENDING LOGGING AT Tue Mar 01 02:59:58 2016