**** BEGIN LOGGING AT Wed Jan 20 02:59:59 2016
Jan 20 08:13:52 <daal> Hi ant_work, long time no see
Jan 20 08:14:06 <ant_work> hello
Jan 20 08:14:16 <daal> how are you ? all good ?
Jan 20 08:14:28 <daal> just want to catch up about 4.4 kernels
Jan 20 08:14:45 <ant_work> I'm slowly following on OESF
Jan 20 08:15:00 <ant_work> in the meantime I've created recipes for 4.4 in OE
Jan 20 08:15:19 <daal> I have made some progress but there are some odd issues
Jan 20 08:15:27 <ant_work> I see there are issues with SD card?
Jan 20 08:15:32 <ant_work> on pxa270?
Jan 20 08:15:33 <daal> yes
Jan 20 08:15:42 <ant_work> hm..
Jan 20 08:15:59 <daal> well, when I build the 4.4 with gcc 5.3 there was no SD at all
Jan 20 08:16:10 <daal> I did rebuild with gcc 4.9 and had SD but read only
Jan 20 08:16:53 <daal> then I "backported" the drivers/mmc/host/pxamci.c from 4.2.3
Jan 20 08:17:00 <daal> SD started working again, read write
Jan 20 08:17:20 <daal> however I am not sure this is save, as there could be corruption happening ? no issues so far
Jan 20 08:17:42 <daal> s/save/safe
Jan 20 08:17:44 <ant_work> I have only tested -rc7 iirc
Jan 20 08:18:00 <ant_work> I remember having added PXA_DMA for this 4.4 serie
Jan 20 08:18:05 <daal> also since the 4.x series the first console "blank" goes white
Jan 20 08:18:41 <ant_work> ah.. imagine we have just fixed similar issue for poodle and collie (locomo chip)
Jan 20 08:19:09 <ant_work> if you have time please diff the defconfig used here
Jan 20 08:19:10 <daal> another thing I noticed, sound will work only when compiled in kernel, not as module
Jan 20 08:19:11 <ant_work> https://github.com/LinuxPDA/meta-handheld
Jan 20 08:19:23 <ant_work> hm...also bad
Jan 20 08:19:48 <daal> you say you found a fix for the white blank ?
Jan 20 08:20:11 <ant_work> yes but for different devices
Jan 20 08:20:30 <daal> ok, I have a look, this may be portable ?
Jan 20 08:20:38 <ant_work> https://github.com/LinuxPDA/meta-handheld/commit/526de214f4f88b41c41adc74da30637757acbee7
Jan 20 08:21:21 <daal> old story.. but what is the status about online charging and charging LED amber ?
Jan 20 08:22:08 <daal> this seems to be broken for ages
Jan 20 08:25:14 <ant_work> heh
Jan 20 08:25:32 <ant_work> Dmitry added recently battery driver for poodle and fixed the one for collie
Jan 20 08:25:43 <ant_work> I have to ask him to look at corgi/spitz
Jan 20 08:26:30 <ant_work> you see, we spent the last couple of years fixing collie and poodle
Jan 20 08:26:31 <daal> well, that is good news
Jan 20 08:26:46 <ant_work> spitz and corgi were almost OK...now maybe lag behind :/
Jan 20 08:27:01 <ant_work> I have a C860 and C3200 so will test
Jan 20 08:27:06 <daal> ha, wonder why you focus on the older platform, but hey, at least some progress
Jan 20 08:27:13 <ant_work> he he
Jan 20 08:27:41 <ant_work> we aim at least at having old OPIE image running on those old devices
Jan 20 08:27:44 <daal> good to know you guys are still active too
Jan 20 08:27:45 <ant_work> nothing more
Jan 20 08:28:09 <daal> you know I focus on the Arch port but in terms of kernel, we should work together ?
Jan 20 08:28:33 <ant_work> sure, I'm regularly looking for patches to send upstream
Jan 20 08:28:48 <ant_work> now the actual issue is pxa platform is being moved to devicetree
Jan 20 08:29:03 <ant_work> so there have been many changes in the last 2-3 kernels
Jan 20 08:29:27 <ant_work> the kernel maintainer for PXa is very friendly
Jan 20 08:30:09 <ant_work> (today is Robert Jarzmik robert.jarzmik@free)
Jan 20 08:30:36 <daal> does pxa270 get much attention ?
Jan 20 08:30:51 <ant_work> well, how to say...more than pxa25x for sure :)
Jan 20 08:30:53 <daal> I try to follow github commits on the platform
Jan 20 08:31:32 * daal brb
Jan 20 08:40:31 <daal> I noticed someone was asking for a kexec to boot zgrom
Jan 20 08:40:56 <daal> I have tried with my arch-kexecboot and it does find the kernel and boots
Jan 20 08:41:17 <daal> however there is framebuffer distortion and ext3 errors just before the login prompt
Jan 20 08:41:23 <daal> maybe a corrupt sd card
Jan 20 08:42:11 <daal> the boot.cfg config in zgrom seems to be the "updated" one
Jan 20 08:44:23 <ant_work> yes, I think dromede used the 'last version' of boot cfg
Jan 20 08:45:17 <ant_work> I was looking yesterday at the commit history and I think I can easily rebuild a serie of 3.2 kernels
Jan 20 08:45:41 <ant_work> I'll add the last version of kexecboot of course
Jan 20 08:46:11 <daal> I was looking for it too, I must have it somewhere, just can not find it
Jan 20 08:46:51 <ant_work> the actual 0.6 version on kexecboot github is fine
Jan 20 08:47:13 <VartiWork> hi
Jan 20 08:47:18 <daal> my kexecboot is based on 3.5 and has ext4 support , nothing else different as far as I know, just the logos
Jan 20 08:47:55 <daal> Hi VartiWork
Jan 20 08:48:00 * daal brb
Jan 20 08:48:13 <VartiWork> daal: me and ant were wondering on what version is based your kexecboot, yesterday :)
Jan 20 08:48:30 <daal> It is mentioned on the github as far as I know :)
Jan 20 08:48:41 <ant_work> hi VartiWork
Jan 20 08:49:19 <ant_work> daal: there are only minor patches to add to kexecboot 0.6
Jan 20 08:49:23 <VartiWork> ah, missed that then, sorry
Jan 20 08:49:28 <VartiWork> hi ant
Jan 20 08:49:38 <VartiWork> anzi, direi ciao ant ;)
Jan 20 08:49:49 <daal> ah, yes, also kexecboot 0.6 is used in arch-kexecboot
Jan 20 08:50:08 <ant_work> https://github.com/andrea-adami/kexecboot/commits/master
Jan 20 08:50:37 <ant_work> we'll merge this in a 0.7 release soon (when Jay7 will be free...)
Jan 20 08:51:34 <daal> what kernel you intend to use for kexecboot ? 4.x series ? they tend to get big, even with xz
Jan 20 08:51:50 <ant_work> correct
Jan 20 08:51:58 <ant_work> I have yet to test how big...
Jan 20 08:52:34 <daal> not tried yet, I sticked with 3.5 as I did not see any reason to upgrade so far
Jan 20 08:54:19 <ant_work> yes, it's only for collie that we are stuck at 2.x
Jan 20 08:54:33 <daal> ant_work : if you find the time, can you look into the SD card issue I found ? Just to see if it was only my build chain..
Jan 20 08:54:34 <ant_work> the 3.x serie lacks SD card driver
Jan 20 08:54:43 <ant_work> yes, I'll try
Jan 20 08:54:59 <daal> 3.x lacks SD ? for spitz ?
Jan 20 08:55:24 <ant_work> no, collie. Note that at first I didn't add the new DMA config to pxa machines so I had nasty issues
Jan 20 08:55:30 <daal> I see
Jan 20 08:56:28 <ant_work> btw  http://lkml.iu.edu/hypermail/linux/kernel/1601.0/00292.html
Jan 20 08:56:36 <daal> VartiWork : I could boot zgrom of arch-kexecboot but had some issues before login screen ? how far did you get ?
Jan 20 08:57:53 <daal> ant_work : thanks I will use this as a base for next recompile
Jan 20 08:59:41 <ant_work> I plan to diff with this new upstream config as well
Jan 20 09:00:22 <daal> You remember the 4.x series would not boot when compiled with gcc5+
Jan 20 09:00:32 <ant_work> it's ok now
Jan 20 09:01:05 <daal> yeah, but odd with the SD card issue I faced, and 4.9 would bring it back, still had to backport to get read/write
Jan 20 09:01:12 <ant_work> pls test https://github.com/LinuxPDA/builds_2015-12-22
Jan 20 09:01:29 <ant_work> these were built with gcc5
Jan 20 09:02:14 <daal> ok, I will test SD card with these
Jan 20 09:02:39 <VartiWork> daal: not tried yet, yesterday I spent some time in trying to find out what version of kexecboot to use
Jan 20 09:03:31 <VartiWork> also I need first to install Cacko on NAND and move it to SD, before I can proceed with kexec, alarm and zgrom
Jan 20 09:04:00 <daal> sound like you have a complex setup :)
Jan 20 09:04:33 <VartiWork> heh, I'm not sure I will manage to fit it within the 7 partitions limit :)
Jan 20 09:05:27 <VartiWork> I guess someday I'll just ditch zgrom and compile/use the emus I need on alarm
Jan 20 09:06:01 <daal> VartiWork : well danboid and me were working on that
Jan 20 09:06:21 <VartiWork> just curious, are you guys using a crosscompiler targetting ARM, or you compile all natively on the Z?
Jan 20 09:06:32 <daal> toolchain
Jan 20 09:06:34 <daal> for kernel
Jan 20 09:06:55 <ant_work> I use OpenEmbedded
Jan 20 09:06:56 <daal> I use the ArchLinux ARM cross-ng or what it is called
Jan 20 09:07:15 <VartiWork> ok
Jan 20 09:10:36 <daal> VartiWork : I hope to find a way to get DirectFB to support the pxa270 overlay and add iWMMXt rotation and SRAM etc like dromede used
Jan 20 09:10:51 <daal> may never happen but I will have a look
Jan 20 09:11:29 <ant_work> sounds good
Jan 20 09:12:06 <VartiWork> SRAM should give it a nice speedup
Jan 20 09:12:17 <VartiWork> overlay too I guees
Jan 20 09:12:19 <daal> wonder why dromede never released src
Jan 20 09:12:26 <VartiWork> good luck with that :)
Jan 20 09:12:38 <VartiWork> I wonder that too
Jan 20 09:12:46 <daal> overlay seems to work alread with DirectFB , 240x320 rotated
Jan 20 09:12:58 <VartiWork> I hope that he'll eventually find them
Jan 20 09:13:01 <daal> luck I will need
Jan 20 09:13:20 <daal> DirectFB also beeing abandoned since Aug last year...
Jan 20 09:13:45 <VartiWork> on the latest official Linux kernel?
Jan 20 09:14:20 <daal> DirectFB is not part of the kernel, its a seperate thing, however I added the linux-fusion module in the latest Arch kernel
Jan 20 09:14:28 <VartiWork> ah ok
Jan 20 09:14:56 <VartiWork> for multiple framebuffers I read
Jan 20 09:15:23 <daal> well, multiple DirectFB applications running at the same time, with out the module, only one can run
Jan 20 09:15:40 <daal> there will be only one FB
Jan 20 09:16:30 <VartiWork> I see, so that you can run multiple programs using e.g. "screen"
Jan 20 09:17:20 <daal> well, sort of, more like an X replacement. You can run GTK+ apps, SDL apps etc without X, using Framebuffer
Jan 20 09:17:39 <VartiWork> (unrelated, I found a nice Akita for a good price on EBay, unfortunately I have been outbidded :( http://offer.ebay.com/ws/eBayISAPI.dll?ViewBidderProfile&mode=1&item=221990664238&aid=e***u&eu=&bidtid=1681353398012&view=BUYER&ssPageName=PageBidderProfileViewBids_Buyer_ViewLink )
Jan 20 09:18:02 <VartiWork> didn't know you could do that with GTK as well, nice
Jan 20 09:18:27 <daal> if my C3100 ever dies, I will need to find one but probably again only a C3100
Jan 20 09:18:46 <VartiWork> weren't you trying to port MATE too? How did that project go?
Jan 20 09:18:58 <daal> borzoi seems to be the best model Sharp made, please correct me if wrong
Jan 20 09:19:10 <VartiWork> heh, I'd like a spare Zaurus as well
Jan 20 09:19:30 <daal> MATE ? no way
Jan 20 09:19:32 <daal> not me
Jan 20 09:19:56 <daal> I ported E17 back in 2005 on pdaxrom, worked quite well
Jan 20 09:20:14 <daal> now I do not focus on X at all
Jan 20 09:20:14 <VartiWork> http://forums.mate-desktop.org/viewtopic.php?t=3609
Jan 20 09:20:20 <VartiWork> it wasn't you then
Jan 20 09:20:44 <daal> nah, that was Archi
Jan 20 09:22:06 <VartiWork> k.. I wonder if it was usable at all
Jan 20 09:22:42 <VartiWork> I'd like to have an X desktop too
Jan 20 09:23:16 <daal> well, use pdaxrom, back then it was usable with tinyx fbdev
Jan 20 09:23:17 <VartiWork> pdaxrom... used pdaxii for a while, it was a nice distro
Jan 20 09:23:44 <daal> full blown Xorg, no way
Jan 20 09:23:53 <VartiWork> unfortunately AFAIK no one ported the driver for SDHC cards for it
Jan 20 09:24:36 <daal> not sure, I can read a 16GB class 2 card fine, at least I could, should try again
Jan 20 09:24:53 <daal> an arch
Jan 20 09:25:29 <VartiWork> I seem to recall that I tried it on pdaxii and failed
Jan 20 09:26:05 <ant_work> daal: there is another patch needed for C3000/3100/3200
Jan 20 09:26:10 <daal> well back then I did not have a card like that. I meant now, with Arch or recent kernels
Jan 20 09:26:19 <ant_work> it's for mtd detection
Jan 20 09:26:23 <daal> ant_work : ?
Jan 20 09:26:34 <ant_work> each model has a different nand partitioning
Jan 20 09:26:42 <daal> mtd , well ok , I do not use, Arch does not fit in there
Jan 20 09:26:43 <ant_work> we read that in userspace in kexecboot
Jan 20 09:26:54 <daal> but would be good to have
Jan 20 09:26:57 <ant_work> need to create a kernel-parser
Jan 20 09:27:06 <daal> for ubifs ?
Jan 20 09:27:42 <ant_work> not for a specific fs, just for the correct mtdpart
Jan 20 09:28:13 <daal> ah ok, I read somewhere that the mtd partions are different from model to model
Jan 20 09:28:21 <ant_work> silly Sharp...
Jan 20 09:28:42 <daal> ah, they produces such a nice gadget, no need to blame them :)
Jan 20 09:29:11 <ant_work> we can live with it being done in kexecboot but ideally it should be detected by the kexecboot kernel
Jan 20 09:29:28 <ant_work> then it would be one kernel for all 3 models
Jan 20 09:29:43 <ant_work> (one kexecboot kernel I mean)
Jan 20 09:30:29 <daal> agree, but the mtd would be more for legacy distribtions that fit in the nand (OE ?)
Jan 20 09:30:54 <ant_work> you can store in flash pretty beefy console-images
Jan 20 09:31:02 <ant_work> i.e. rescue
Jan 20 09:31:16 <daal> yes, true, or a busybox image ,etc
Jan 20 09:31:41 <ant_work> even on spitz 3000 with 5MiB flash
Jan 20 09:35:01 * daal brb
Jan 20 21:40:32 <ant_home> bluelightning, hi
Jan 20 21:40:41 <bluelightning> hi ant_home
Jan 20 21:41:38 <ant_home> I'm struggling a bit rewinding to danny
Jan 20 21:41:55 <ant_home> all the layesr can be easily rest
Jan 20 21:41:59 <bluelightning> danny? are you trying to see how long ago something worked?
Jan 20 21:42:00 <ant_home> *reset
Jan 20 21:42:05 <ant_home> apart bitbake
Jan 20 21:42:16 <ant_home> I can't apparently match bitbake
Jan 20 21:42:36 <ant_home> tryin to rebuild kexecboot 3.2 kernels)
Jan 20 21:42:56 <bluelightning> https://wiki.yoctoproject.org/wiki/Releases - danny would match up with the bitbake 1.16 branch
Jan 20 21:43:12 <ant_home> I'm about downloading the yocto release in fact :)
Jan 20 21:44:22 <ant_home> I get immediately failure with pseudo-native on ubuntu 14.04 lts
Jan 20 21:44:35 <ant_home> with 1.12, 1.18
Jan 20 21:44:58 <ant_home> I'll test Yocto sources now
Jan 20 21:46:20 <ant_home> hm.. 1.16 it says...
Jan 20 21:50:47 <ant_home> same
Jan 20 21:50:51 <ant_home> bluelightning, http://pastebin.com/DUDn0fNH
Jan 20 21:55:00 <bluelightning> don't expect to be able to build this on a modern distro I'm afraid...
Jan 20 21:55:17 <bluelightning> you could backport the fix for that issue of course, I think it was just one patch
Jan 20 21:55:26 <bluelightning> but you may hit other things
Jan 20 22:03:07 <ant_home> well, here it is ubuntu's tar, as found googling
Jan 20 22:03:25 <ant_home> and I can't build tar-native, heh
Jan 20 22:04:05 <ant_home> https://communities.intel.com/thread/53094?start=0&tstart=0
Jan 20 22:05:18 <ant_home> maybe this  http://git.yoctoproject.org/cgit/cgit.cgi/poky/commit/?id=d3846f06d1e3faedcbc3c28e22c427bb0088683d
Jan 20 22:05:55 <bluelightning> yep, that's the one
Jan 20 22:26:50 <ant_home> I went up to that commit in oe-core then it asked for bitbake 1.19.1
Jan 20 22:26:52 <ant_home> then
Jan 20 22:26:56 <ant_home> http://pastebin.com/kkveUmf0
Jan 20 22:27:13 <ant_home> harder than I thought
**** ENDING LOGGING AT Thu Jan 21 02:59:58 2016