**** BEGIN LOGGING AT Tue Jan 24 03:00:02 2017
Jan 24 08:12:58 <VartiWork> hi ant
Jan 24 08:13:05 <ant_work> hello
Jan 24 08:14:04 <ant_work> ka6sox, we are fine with nslu2-log. No need for chanserv right now. thanks!
Jan 24 21:14:12 <ant_home> Varti, heh, didn't know Lineo has joint Yocto project
Jan 24 21:15:42 <ant_home> greguu, still waiting for 4.10 final
Jan 24 21:15:55 <bluelightning> ant_home: hey
Jan 24 21:15:59 <bluelightning> ant_home: btw:
Jan 24 21:16:05 <bluelightning> ant: <ka6sox> bluelightning, ant asked me to get chanserv back in here...I'm not able to at this time...and we aren't on the same schedule so I keep missing him.
Jan 24 21:16:14 <bluelightning> ant_home: <ka6sox> can you tell him we would all need to get out and then come back in?
Jan 24 21:16:35 <bluelightning> that sounds like it's going to be difficult to coordinate
Jan 24 21:16:44 <ant_home> ah, thanks, I've read it and answered that we are fine with just nslu-log services
Jan 24 21:16:51 <bluelightning> ah ok, great
Jan 24 21:16:54 <ant_home> no need for chanserv afaik
Jan 24 21:16:58 <ant_home> or?
Jan 24 21:17:47 <bluelightning> I'm not sure what chanserv does either to be honest
Jan 24 21:17:58 <ant_home> ka6sox, I can ping jay7 and ask him to kick us out
Jan 24 21:18:05 <bluelightning> probably something to do with ensuring there's an op round
Jan 24 21:18:09 <bluelightning> s/round/around/
Jan 24 21:18:48 <ant_home> I think we mostly need logs
Jan 24 21:21:04 <ant_home> bluelightning, ot
Jan 24 21:21:18 <ant_home> it seems kernel images are now misnamed/mangled
Jan 24 21:21:57 <ant_home> zImage--4.4-r0-akita-20170124210818.bin
Jan 24 21:31:27 <ant_home> argh..and
Jan 24 21:31:30 <ant_home> zImage-zImage-kexecboot-4.4-akita.bin
Jan 24 21:32:12 <ant_home> from ERNEL_IMAGE_BASE_NAME = "${KERNEL_IMAGETYPE}-kexecboot-${PV}-${MACHINE}"
Jan 24 21:32:22 <ant_home> hmm
Jan 24 21:35:52 <ant_home> run.do_deploy:		base_name=${type}-zImage-kexecboot-4.4-akita
Jan 24 21:35:52 <ant_home> run.do_deploy:		symlink_name=${type}-zImage-kexecboot-akita
Jan 24 21:38:08 <ant_home> bluelightning, well, one of this patches must be the culprit
Jan 24 21:38:10 <ant_home> http://cgit.openembedded.org/openembedded-core/log/?qt=grep&q=IMAGETYPE
Jan 24 21:38:18 <ant_home> s/this/these
Jan 24 21:41:22 <bluelightning> ant_home: relates to this I guess: http://cgit.openembedded.org/openembedded-core/commit/?id=849b67b2e4820564b5e5c9bd4bb293c44351c5f3
Jan 24 21:43:01 <ant_home> yes, must be that...but I'm slow reading python
Jan 24 21:44:19 <ant_home> I can revert the changes to kernel.bbclass and test
Jan 24 21:45:03 <ant_home> but must be a very obvious bug, it prepends i.e. zImage unconditionally it seems
Jan 24 21:45:14 <ant_home> zImage- even
Jan 24 21:49:14 <ant_home> bah..
Jan 24 21:49:15 <ant_home> KERNEL_IMAGETYPES[doc] = "The list of types of kernel to build for a device, usually set by the machine configuration files and defaults to KERNEL_IMAGETYPE."
Jan 24 21:49:26 <ant_home> where is it defined in nodistro?
Jan 24 21:50:23 <bluelightning> I don't think it has to be defined
Jan 24 21:52:50 <ant_home> right, my python foo is low but still
Jan 24 21:52:51 <ant_home> type = d.getVar('KERNEL_IMAGETYPE', True) or ""
Jan 24 21:54:05 <ant_home> so 'zImage'
Jan 24 21:54:40 <ant_home> types = d.getVar('KERNEL_IMAGETYPES', True) or ""
Jan 24 21:54:43 <ant_home> so ""
Jan 24 21:56:35 <ant_home> it's the following two lines problematic
Jan 24 21:56:38 <ant_home> +    if type not in types.split():
Jan 24 21:56:38 <ant_home> +        types = (type + ' ' + types).strip()
Jan 24 21:59:08 <bluelightning> hang on
Jan 24 21:59:46 <bluelightning> ant_home: if you look back at that patch, the default KERNEL_IMAGE_BASE_NAME has been changed to not include ${KERNEL_IMAGETYPE}
Jan 24 21:59:56 <bluelightning> ant_home: you need to do the same for the kexecboot value
Jan 24 22:04:40 <ant_home> bluelightning, I see now, testing
Jan 24 22:04:53 <ant_home> still ,two dashes in linux-handheld
Jan 24 22:05:04 <ant_home> one by one...
Jan 24 22:05:15 <bluelightning> that will be as a result of a blank variable - which one is it?
Jan 24 22:05:42 <ant_home> linux-handheld doesn't declare anything special
Jan 24 22:06:12 <ant_home> we just have KERNEL_IMAGETYPE in zaurus.inc
Jan 24 22:06:30 <bluelightning> yes but can we figure out where the blank value is coming from?
Jan 24 22:06:52 <ant_home> mom
Jan 24 22:06:53 <ant_home> │ zImage--kexecboot-4.4-akita.bin
Jan 24 22:07:04 <ant_home> almost there, just the two dashes now
Jan 24 22:08:00 <ant_home> gah, my bad here...
Jan 24 22:08:09 <ant_home> KERNEL_IMAGE_BASE_NAME = "-kexecboot-${PV}-${MACHINE}"
Jan 24 22:08:09 <ant_home> KERNEL_IMAGE_SYMLINK_NAME = "-kexecboot-${MACHINE}"
Jan 24 22:08:13 <bluelightning> ah right
Jan 24 22:08:18 <ant_home> so I remove dash as well
Jan 24 22:08:26 <bluelightning> yes
Jan 24 22:08:31 <ant_home> ..still linux-handheld...
Jan 24 22:08:45 <bluelightning> let's look at that then
Jan 24 22:09:56 <ant_home> heh, thanks for helping me sorting out this. Sysroot fixes will be more of a  nightmare
Jan 24 22:11:06 <ant_home> bluelightning, zImage-kexecboot-4.4-akita.bin      bingo
Jan 24 22:11:15 <ant_home> next on now
Jan 24 22:28:42 <ant_home> bluelightning, kid in bed now...
Jan 24 22:28:46 <ant_home> soo
Jan 24 22:28:56 <ant_home> run.do_deploy
Jan 24 22:28:59 <ant_home> kernel_do_deploy() {
Jan 24 22:28:59 <ant_home> <------>for type in zImage ; do
Jan 24 22:28:59 <ant_home> <------><------>base_name=${type}--4.4-r0-akita-20170124221217
Jan 24 22:29:18 <ant_home> double slash is already there
Jan 24 22:29:27 <ant_home> going backwards now
Jan 24 22:32:39 <ant_home> so in kernel.bbclass
Jan 24 22:32:41 <ant_home> kernel_do_deploy() {
Jan 24 22:32:41 <ant_home> <------>for type in ${KERNEL_IMAGETYPES} ; do
Jan 24 22:32:41 <ant_home> <------><------>base_name=${type}-${KERNEL_IMAGE_BASE_NAME}
Jan 24 22:34:13 <ant_home> a couple of lines earlier:
Jan 24 22:34:31 <ant_home> KERNEL_IMAGE_BASE_NAME ?= "${PKGE}-${PKGV}-${PKGR}-${MACHINE}-${DATETIME}"
Jan 24 22:36:40 <ant_home> so the empty PKGE ?
Jan 24 22:37:53 <ant_home> held/4.4-r0/temp$ grep -R PKGE
Jan 24 22:37:53 <ant_home> run.emit_pkgdata.4043:        write_if_exists(sf, pkg, 'PKGE')
Jan 24 22:37:53 <ant_home> run.do_package.4043:        write_if_exists(sf, pkg, 'PKGE')
Jan 24 22:37:53 <ant_home> run.do_package:        write_if_exists(sf, pkg, 'PKGE')
Jan 24 22:40:00 <ant_home> bluelightning, http://pastebin.com/WQEKqjbX
Jan 24 22:40:53 <bluelightning> ant_home: judging by what I see in my tmp/deploy/images, this is now expected
Jan 24 22:41:05 <bluelightning> so linux-handheld isn't special in this regard
Jan 24 22:41:10 <ant_home> ok, this is for oe-core
Jan 24 22:41:24 <ant_home> I'll send the fix for the kernels in meta-handheld
Jan 24 22:41:34 <ant_home> one patch for all I think
Jan 24 22:42:25 <ant_home> however, we could define KERNEL_IMAGE_BASENAME w/out PKGE in the recipes and be free
Jan 24 22:42:43 <ant_home> but if it is expected...
Jan 24 22:47:29 <bluelightning> well it's in the default value, so all kernels are getting this
Jan 24 22:47:38 <bluelightning> the question is is it a problem?
Jan 24 22:47:55 <ant_home> bluelightning, my bad I've noticed the problem in meta-handheld only now
Jan 24 22:48:19 <ant_home> you'll need to include it once you tag morty
Jan 24 22:48:39 <ant_home> wrt two dashes..well, just an eyesore, not a problem :)
Jan 24 22:48:53 <bluelightning> right yes this was a morty change AFAIK
Jan 24 22:48:57 <bluelightning> right, ok :)
Jan 24 22:49:17 <ant_home> what I mostly hate are the double slashes
Jan 24 22:49:21 <ant_home> ;)
Jan 24 22:50:29 <ant_home> ok, I'll redo a build from start and test the patches
Jan 24 22:50:34 <ant_home> good night
Jan 24 22:50:48 <bluelightning> we could probably fix it in core somehow
Jan 24 22:50:52 <bluelightning> 'night
Jan 24 23:16:06 <ka6sox> I keep missing him!
Jan 24 23:16:32 <ka6sox> okay lets find out when Jay7 is around
Jan 24 23:16:47 <ka6sox> if he is an Op then we can have him invite Chanserv back in.
**** ENDING LOGGING AT Wed Jan 25 03:00:01 2017