**** BEGIN LOGGING AT Fri May 27 02:59:58 2016 May 27 06:20:24 Hi, can someone explain me the difference between meta-ti layer and meta-bbb ? May 27 06:59:50 Gwak: meta-ti is general BSP for multiple TI platforms, while meta-bbb is focused on BeagleBone black/green May 27 07:01:10 the main difference is kernel, meta-ti supplies kernel from TI tree, meta-bbb uses linux-yocto with some patches on top. May 27 07:05:42 maciejjo: Thanks, i'm looking to create a custom distro for BBB and green, I need a bare image and to change some kernel conf (PREEMPT), what are the pros/cons for meta-ti vs meta-bbb ? May 27 07:10:48 Gwak: AFAIK meta-bbb is unnecessary because this machine is already in the Yocto/Poky main layer May 27 07:27:08 good morning May 27 07:35:38 Gwak_: yes meta-bbb really does not provide much, if you want basic kernel for bbb you can just use bare poky/meta-yocto-bsp May 27 07:56:55 does meta-ti still make sense for BBB? IIRC linux-ti was more stable and came with more BBB specific stuff than linux-yocto May 27 08:21:19 so according to what you are saying it does make sense May 27 08:23:52 but there is yet another bsp layer for bbb, meta-beagleboard, that uses linux from beagleboard github May 27 08:27:08 maciejjo: so it may be worthwhile to figure out what is the best and what it offers every possible solution May 27 08:31:50 I am sticking with meta-ti, as it is up-to-date with openembedded development and provides support for PRU May 27 10:34:28 Hello May 27 10:34:36 have anyone seen this problem here on jethro? May 27 10:34:38 cfns.gperf:101:1: error: 'const char* libc_name_p(const char*, unsigned int)' redeclared inline with 'gnu_inline' attribute May 27 10:34:56 when building gcc-crosssdk_5.2 May 27 10:35:32 I've seen this patch for gcc 4.9 that seems familiar, but it's not applied to 5.2: http://patchwork.openembedded.org/patch/122469/ May 27 11:26:49 OK, that patch worked. May 27 11:26:57 With minor ajustment. May 27 11:27:23 That's because I am on Arch and I have gcc6 already here. Other distros tested for Jethro probably still uses gcc 5 May 27 12:54:17 hi all May 27 12:55:39 we have an issue that when we use uppercase characters in machine name creating of sdk fails (we use also PACKAGE_CLASSES = "package_ipk" but when we drop it it works) May 27 12:56:29 any ideas? May 27 13:34:38 I'm on ArchLinux but I still really really would like to use yocto on it raw instead of in a virtual machine because of speed. I installed gcc-5 and made it my default one by softlinking it to gcc, g++ etc. and I get quite long but not the whole way because I get this error https://gist.github.com/rpannek/e8e29c70ef869b36bb7b68549e4ec581 May 27 13:34:45 anybody seen it? May 27 13:35:04 and is it a gcc error? May 27 13:58:09 for variables that are made package-specific with a _${PN} suffix, is there an easy way to make them apply to multilib variants? May 27 13:59:04 or do I have to have duplicate entries like LICENSE_FLAGS_WHITELIST = "commercial_libmad commercial_lib32-mad" ? May 27 14:05:10 it works with LICENSE_FLAGS_WHITELIST = "commercial_${MLPREFIX}mad" May 27 14:05:32 thank you May 27 14:34:23 Have anyone had a problem debugging remotelly? Program received signal SIGSEGV, Segmentation fault. May 27 14:34:26 _dl_debug_initialize (ldbase=4294967292, ns=195728) at dl-debug.c:55 May 27 14:34:35 Any ideas? May 27 15:24:49 Has anyone else tried omitting mountpoint from part lines in wic wks files? The documentation indicates doing so will create the partition but will not modify fstab. I'm guessing it's just broken and I need to file a bug report. May 27 17:50:46 billr, Maintenance is done. Whatever changes were made on the others need to be done to centos7. May 27 17:51:40 halstead: thanks. I'll login and update it. **** ENDING LOGGING AT Sat May 28 02:59:58 2016