**** BEGIN LOGGING AT Wed Feb 17 02:59:59 2016
Feb 17 09:15:35 <ant_work> bluelightning: I'm encountering all sort of issues lately...
Feb 17 09:16:00 <bluelightning> ant_work: sorry to hear that :(
Feb 17 09:16:02 <bluelightning> what's the latest?
Feb 17 09:16:11 <ant_work> I got a strange failure building 4.5-rc4 because 'fonts not defined'... well, I tried to run menuconfig and
Feb 17 09:16:27 <ant_work> it failed as ususal needing ncurses
Feb 17 09:16:31 <ant_work> ok, that's old
Feb 17 09:17:07 <ant_work> what's new is, instead of menuconfig it did compile the kernel in a separate shell and it was like single-task
Feb 17 09:17:13 <ant_work> it tok lot of time..
Feb 17 09:17:39 <ant_work> so menuconfig task seems now totally b0rked
Feb 17 09:19:02 <ant_work> ah, yes, I've got that funny ld error too...
Feb 17 09:19:43 <ant_work> this evening I'll have to decide where to start debugging :)
Feb 17 09:21:15 <ant_work> atm I see a nasty breakage when building two kernels (i.e. linux-handheld_4.x and linux-kexecboot_3.x)
Feb 17 09:21:27 <ant_work> the problem is the shared workdir
Feb 17 10:08:12 <bluelightning> oh... ouch
Feb 17 10:08:36 <bluelightning> I suspect that usecase wasn't considered... how didn't you see that before just now though? the shared workdir was implemented some time ago
Feb 17 10:30:44 <ant_work> heh
Feb 17 10:31:11 <ant_work> original intention was to debug linux-yocto-tiny-kexecboot
Feb 17 10:31:32 <ant_work> which was showing failure in JaMa's world builds
Feb 17 10:32:49 <ant_work> inspecting that I've added +do_shared_workdir[noexec] = "1"
Feb 17 10:33:29 <ant_work> but still any recipe inheriting kernel.bbclass does unpack there
Feb 17 10:33:53 <ant_work> so there can be an horrible mixture
Feb 17 10:35:05 <ant_work> now, if you build virtual/kernel first then a second one (-kexecboot) all is fine
Feb 17 10:35:39 <ant_work> but they get built in a paralell way so you get a build error
Feb 17 10:38:00 <ant_work> usually the MACHINE add dependency on virtual/kernel so no luck even if you start from scratch and build straight the -kexecboot kernel
Feb 17 10:39:16 <ant_work> this was masked when i.e. for qemux86 both linux-yocto and linux-yocto-tiny-kexecboot where 4.4
Feb 17 10:39:25 <ant_work> err ...4.1
Feb 17 10:39:52 <ant_work> now that linux-yocto is at 4.4 there is mismatch with l-y-t-k 4.1
Feb 17 10:40:03 <ant_work> ..so I moved the latter to 4.4
Feb 17 10:40:43 <ant_work> (there is a missing patch for 4.1 but well, this is another story...)
**** BEGIN LOGGING AT Wed Feb 17 12:00:09 2016
Feb 17 21:29:36 <ant_home> bluelightning, hi
Feb 17 21:29:52 <ant_home> so, do_menuconfig disappeared
Feb 17 21:31:11 <ant_home> http://pastebin.com/qjBEWH33
Feb 17 21:32:49 <ant_home> something is borking it...the log exists
Feb 17 21:33:01 <ant_home> DEBUG: Executing python function do_menuconfig
Feb 17 21:33:01 <ant_home> DEBUG: Attempting to spawn terminal "custom"
Feb 17 21:33:01 <ant_home> DEBUG: No custom terminal (OE_TERMINAL_CUSTOMCMD) set
Feb 17 21:33:01 <ant_home> DEBUG: Attempting to spawn terminal "tmux-running"
Feb 17 21:33:01 <ant_home> DEBUG: Attempting to spawn terminal "tmux-new-window"
Feb 17 21:33:02 <ant_home> DEBUG: Attempting to spawn terminal "gnome"
Feb 17 21:34:48 <ant_home> heh..it is compiling it but there is not such task yet...
Feb 17 21:35:55 <ant_home> this is log.task_orderright now (still building) http://pastebin.com/JbRDhhps
Feb 17 21:38:38 <ant_home> there are two tasks each: http://pastebin.com/irmSXAfT
Feb 17 21:41:44 <bluelightning> hi ant_home
Feb 17 21:41:46 <bluelightning> yuck
Feb 17 21:41:53 <bluelightning> just reproduced that here :(
Feb 17 21:42:41 <ant_home> ah..there is even a bug in the title of spawned terminal... ${PN} Configuration
Feb 17 21:42:47 <ant_home> hm..unexpanded
Feb 17 21:43:46 <ant_home> this is rather recent regression
Feb 17 21:44:50 <bluelightning> yes, I saw that too
Feb 17 21:47:16 <bluelightning> ant_home: bug filed for the first issue: https://bugzilla.yoctoproject.org/show_bug.cgi?id=9134
Feb 17 21:47:23 <ant_home> it is slow...seems only one cpu at time is compiling, maybe two
Feb 17 21:59:00 <ant_home> see, it has just done .oldconfig it seems http://pastebin.com/WUyMqAnx
Feb 17 22:07:06 <ant_home> bluelightning, last but not least
Feb 17 22:07:08 <ant_home> http://pastebin.com/kWCjsBQn
Feb 17 22:07:45 <bluelightning> wtf
Feb 17 22:08:04 <ant_home> oe_runmake failed
Feb 17 22:08:20 <bluelightning> something strange is going on here... if the kernel do_compile didn't work generally we'd surely have seen it on our autobuilders
Feb 17 22:09:08 <ant_home> last weekend all was building, known  issues apart
Feb 17 22:09:43 <ant_home> I've built both linux-yocto_4.4 and its -tiny-kexecboot variant
Feb 17 22:10:02 <ant_home> more than this, I even tested it on qemu :)
Feb 17 22:10:22 <ant_home> it fails because of a check in linux-yocto
Feb 17 22:10:44 <ant_home> we have a patch for this in meta-handheld .bbappends
Feb 17 22:11:02 <ant_home> better put that in the main recipe in meta-initramfs
Feb 17 22:11:32 <ant_home> I'll send a patch for this
Feb 17 22:14:43 <ant_home> bluelightning, don't worry..eventually it builds after a cleansstate ;)
Feb 17 22:15:46 <ant_home> w/out do_menuconfig the autobuilder is unaffected
Feb 17 22:16:21 <ant_home> there is something wrong there...not only mixing 4.4 and 3.2
Feb 17 22:17:00 <ant_home> I'll better stop testing older kernels
Feb 17 22:17:07 <ant_home> for the moment
Feb 17 22:25:09 <ant_home> good night
**** ENDING LOGGING AT Thu Feb 18 02:59:59 2016