**** BEGIN LOGGING AT Wed Nov 20 02:59:58 2013
Nov 20 09:02:43 <ant_work> gm lumag_
Nov 20 10:14:58 <lumag> hello ant_work
Nov 20 10:15:02 <lumag> hello bluelightning
Nov 20 10:17:02 <bluelightning> morning ant_work, lumag
Nov 20 22:08:03 <ant_home> bluelightning: lumag_: so who dares playing with the collie nor flash first?
Nov 20 22:09:29 <ant_home> I'm moving locally to 3.12 hoping something has changed (for the better)
Nov 20 22:14:15 <lumag> ant_home, what should I do?
Nov 20 22:14:35 <ant_home> repeatable failure here is with flashcp
Nov 20 22:15:15 <lumag> What exactly should I do?
Nov 20 22:15:52 <ant_home> you create an image 14680064bytes padded (core-image-minimal)
Nov 20 22:16:17 <ant_home> then you need mtd-utils (core-image-base)
Nov 20 22:16:28 <ant_home> I boot in core-image-base and do
Nov 20 22:17:06 <ant_home> # flashcp ./initrd.bin /dev/mtd2
Nov 20 22:17:36 <lumag> I'm going to update to current HEAD in both oe-core and meta-hh. Is that safe or should I have any other patches?
Nov 20 22:17:49 <lumag> And how to pad core-image-minimal?
Nov 20 22:17:54 <lumag> Or it will be done automatically?
Nov 20 22:18:06 <ant_home> auto
Nov 20 22:18:32 <ant_home> I can send you the image and the binary in a couple of mins
Nov 20 22:18:41 <lumag> n/p, I'll build them.
Nov 20 22:18:44 <lumag> Is HEAD ok ?
Nov 20 22:18:55 <ant_home> I'm building with extra patches in github
Nov 20 22:19:49 <ant_home> linux-yocto_3.10
Nov 20 22:20:13 <ant_home> right now, I see there is a possible error for 3.12 (-dev) because I think I've copied one 3.10 patch for 3.12
Nov 20 22:20:20 <ant_home> 3.10 is ok
Nov 20 22:22:45 <ant_home> bluelightning: pls take on hold last commit for linux-yocto-dev, refreshing it now
Nov 20 22:36:00 <ant_home> bluelightning: fixed. you can pull the top 6 patches from git@github.com:andrea-adami/meta-handheld.git
Nov 20 22:40:38 <bluelightning> ant_home: ok, was just looking over the earlier version
Nov 20 22:41:20 <ant_home> updated one single patch for 3.12
Nov 20 22:41:43 <ant_home> lumag_ collie-fixup-strataflash.patch is probably to remove
Nov 20 22:41:55 <ant_home> my diagnosis was wrong
Nov 20 22:42:18 <lumag> ant_home, still building.
Nov 20 22:42:32 <lumag> Thaks to somebody removing FREESMARTPHONE_GIT :(
Nov 20 22:42:42 <lumag> Should I disable strataflash patch in 3.10?
Nov 20 22:42:44 <ant_home> hm.. 3.12
Nov 20 22:42:46 <ant_home> ++gitAUTOINC+75feecae03_1f274173a8-r0/linux/fs/yaffs2/yaffs_vfs.c:1812:2: error: unknown field 'readdir' specified in initializer
Nov 20 22:42:46 <ant_home> |   .readdir = yaffs_readdir,
Nov 20 22:43:02 <ant_home> those yaffs things are yocto patches
Nov 20 22:43:38 <ant_home> ok, remove with raid and and and...
Nov 20 22:51:21 <ant_home> zeddii is away today, let it be for tomorrow
Nov 20 22:53:59 <ant_home> lumag: that patch is done for a buggy part/serie. Disabling erase on suspend will slow down a bit
Nov 20 22:54:57 <ant_home> havinbg it or not made no difference...flaschpy fails the verification
Nov 20 23:03:32 <lumag> Ah distro name was changed. Thus it takes that long.
Nov 20 23:04:01 <ant_home> yes, to fix the ordering bug
Nov 20 23:04:21 <ant_home> now the decompressed-cpio bug remains
Nov 20 23:05:24 <ant_home> heh, an unfortunate series of patches in oe-core broke our layers left and right
Nov 20 23:09:01 <ant_home> hm..| ERROR: "clk_set_rate" [drivers/usb/phy/phy-generic.ko] undefined!
Nov 20 23:09:19 <ant_home> do we need phy on pxa/sa1100 ?
Nov 20 23:26:23 <lumag> No.
Nov 20 23:26:30 <lumag> At least not yet :)
Nov 20 23:28:02 <ant_home> config USB_PHY  def_bool n
Nov 20 23:28:37 <ant_home> but we get CONFIG_USB_PHY=y
Nov 20 23:29:02 <ant_home> have you discovered th ebeauty of the .meta dir?
Nov 20 23:29:24 <ant_home> andrea@mizar /oe/oe-core/build/tmp-eglibc/work/collie-oe-linux-gnueabi/linux-yoc
Nov 20 23:29:25 <ant_home> to-dev/3.10++gitAUTOINC+75feecae03_1f274173a8-r0/linux/.meta/cfg/standard/collie $
Nov 20 23:29:57 <ant_home> you see how kernel config is processed
Nov 20 23:30:06 <ant_home> and reassembled
Nov 20 23:30:44 <lumag> For me it is quite too complicated a bit.
Nov 20 23:30:53 <lumag> I liked old defconfigs.
Nov 20 23:31:09 <lumag> Not that I liked lack of standardization between those defconfigs
Nov 20 23:31:53 <ant_home> I'm trying to start with mini-defconfig (just hw) and force the common bits on the top (fs, usb, ...)
Nov 20 23:32:18 <ant_home> for kexecboot it makes probably sense to go back to defconfig
Nov 20 23:32:36 <ant_home> I've pushed the defconfigs as well
Nov 20 23:37:40 <ant_home> basically what I'm doing in linux-yocto-dev is to select the patches not yet upstream for the next linux-yocto_3.12
Nov 20 23:38:02 <ant_home> I don't care too much if it builds or not (today)
Nov 20 23:38:33 <ant_home> 3.10 seems working rather well, maybe just the usual battery/power man issues to verify
Nov 20 23:39:28 <ant_home> hm...
Nov 20 23:39:30 <ant_home> include/config/auto.conf:CONFIG_USB_PHY=y
Nov 20 23:39:30 <ant_home> include/generated/autoconf.h:#define CONFIG_USB_PHY 1
Nov 20 23:44:49 <ant_home> lumag: http://patchwork.ozlabs.org/patch/266502/
Nov 20 23:45:10 <ant_home> afais is only for tegra
Nov 20 23:47:25 <lumag> for omap I think.
Nov 20 23:48:25 <lumag> but I might be mistaken
Nov 20 23:48:47 <ant_home> https://lkml.org/lkml/2013/7/30/25
Nov 20 23:48:54 <ant_home> similar, something has changed
Nov 20 23:49:54 <lumag> Ah no, we need it
Nov 20 23:49:57 <lumag> At least for gpio-vbus
Nov 20 23:50:54 <ant_home> I've disabled the modules under usb and now it did build
Nov 20 23:51:16 <ant_home> *all* modules and the last U* thing for tegra
Nov 20 23:51:21 <lumag> I think you have disabled/broken gadget for sme of devices
Nov 20 23:51:25 <ant_home> heh
Nov 20 23:51:58 <ant_home> ok, as it is 3.12 won't compile then maybe 3.12.x will be fixed
Nov 20 23:53:08 <lumag> Yes.
Nov 20 23:53:20 <ant_home> pls just check the right patches are there
Nov 20 23:53:21 <lumag> phy-gpio-vbus-usb.c is necessary at least for tosa.
Nov 20 23:54:01 <lumag> The proper patch would be to add dummy clk_set_rate() to pxa/sa1100.
Nov 20 23:54:15 <lumag> I hope to refactor clocks sometimes in the future.
Nov 20 23:54:24 <lumag> *some time
Nov 20 23:57:07 <lumag> As usual sa1100 is easy (nearly nothing to refactor there).
Nov 20 23:58:20 <lumag> PXA is complicated (I mostly prepared proper clock drivers for pxa25x and 27x, but not 3xx), so still some issues there.
Nov 21 00:02:43 <lumag> OE is still building: Currently 2 running tasks (1835 of 2839)
Nov 21 00:10:00 <ant_home> lumag: those tests: http://pastebin.com/wg63U19n
Nov 21 00:10:40 <lumag> ok, I will run flashcp once it compiles
Nov 21 00:11:17 <lumag> Should I erase it, or flashcp will do that?
Nov 21 00:12:32 <ant_home> http://pastebin.com/FRJuWYnB
Nov 21 00:12:42 <ant_home> does that
Nov 21 00:13:07 <lumag> ack
Nov 21 00:13:27 <ant_home> works with small images (i.e. sum.jffs2, unpadded to the end)
Nov 21 00:18:45 <ant_home> the C+D method (internal flasher) allows you to flash anything
Nov 21 00:19:18 <ant_home> it would be handy doing it from console
Nov 21 00:20:22 <ant_home> ok, I'll read tomorrow the resultzZZzzzz
Nov 21 00:20:24 <ant_home> gn
Nov 21 00:26:40 <lumag> gn
Nov 21 02:13:09 <lumag> ant_home, I still got kernel hang with original cramfs being flashed :(
**** ENDING LOGGING AT Thu Nov 21 02:59:59 2013