[04:53] <twb> lilstevie: quick question -- is XV supposed to work on the TF101, using the oooold oneiric armel images you built (and the android 2.6.36 kernel)?
[05:01] <lilstevie> probably not :p
[05:05] <twb> Good-o, because it isn't for me
[07:35] <janimo> marvin24, so the 3.1 ubuntu kernel's CONFIG_PROFILING should be off?
[07:35] <janimo> btw do you use for-next with defconfig? that did not boot for me on Friday
[07:35] <janimo> I see you made an update since
[07:51] <marvin24> janimo: it boots
[07:51] <marvin24> did you concatenate the compiled device tree to it?
[07:51] <marvin24> paz00_defconfig should be fine
[07:51] <marvin24> for the 3.1 kernel ...
[07:52] <marvin24> I tried it on my local branch (which is source identical to the package tree)
[07:52] <marvin24> the strange thing is, disabling PROFILING helps with my own tree
[07:52] <marvin24> but after building the package, it still doesn't work (same config of course)
[07:54] <marvin24> also the binaries are different
[07:54] <marvin24> and yes, I double checked that the source and config is identical
[07:55] <marvin24> mmh, thing about it, the package used the hf compiler
[07:58] <janimo> yes I use the hf compiler only lately, 4.6
[07:58] <janimo> marvin24, concatenate DT to it? I never built DT kernel before so I did not know something extra is needed that make zImage does not do
[08:09] <janimo> marvin24, what I see when googling is make zimage, cat boar.dt >>zImage;make uImage
[08:09] <janimo> but that is for uboot
[08:09] <janimo> or wait, the current ac100 bootloader cannot boot DT kernels?
[08:24] <marvin24> janimo: it can, but the device tree needs to be added to the kernel
[08:24] <marvin24> for u-boot, device tree can come from the bootloader
[08:25] <marvin24> you need to do a make zImage dtbs
[08:25] <marvin24> and then cat arch/arm/boot/tegra20-paz00.dtb >> arch/arm/boot/zImage
[08:26] <janimo> marvin24, so you make a generic zimage and cat as many board DT files as you want it to be able to support?
[08:26] <marvin24> no, only one
[08:27] <marvin24> ideally, the bootloader supplied the device tree
[08:27] <marvin24> but ours can't  right now
[08:27] <marvin24> but this will change with uboot
[08:27] <janimo> ah, so the same kernel booting does not mean an exact same blob but a same zImage + custom DTs
[08:27] <marvin24> so you can directly boot the same kernel on different boards
[08:27] <marvin24> yes
[08:27] <janimo> ok, by same kernel initially I though one binary. This works too though
[08:28] <janimo> will try it out once the build finishes
[08:28] <marvin24> I think it must be similar on omap
[08:28] <janimo> marvin24, also isn't nvec a generic driver for all tegra ECs?
[08:28] <janimo> I see only paz00 uses nvec
[08:28] <janimo> do the others use other EC implementations?
[08:28] <marvin24> the ec is not inside the tegra
[08:28] <marvin24> but somewhere else
[08:28] <marvin24> so it is board specific
[08:29] <janimo> so only the communication protocol is common?
[08:29] <marvin24> there are other "first gen" board using nvec
[08:29] <janimo> I just grepped the dt files for nvec
[08:29] <marvin24> we (and trimslice) are the only non nvidia board supported by mainline kernel
[08:30] <marvin24> so there is no wonder why we are the only user ;-)
[08:30] <janimo> so other ECs do not necessarily use the same SMBus protocol as described in the nvdia docs
[08:30] <marvin24> yes
[08:31] <janimo> I thought  tegra somehow implied nvec
[08:31] <marvin24> toshiba folio uses it for power supply readings
[08:31] <marvin24> same for advent vega
[08:32] <marvin24> but unfornately, they have their own implementations
[08:33] <marvin24> janimo: if you feel lucky today, you can also try https://gitorious.org/uboot-ac100 with this kernel ;-)
[08:35] <janimo> marvin24, nice. Do I just build it and flash it to the bootloader partition as an android bootimage?
[08:35] <janimo> cloning it now
[08:36] <marvin24> I use tegrarcm to load it to memory
[08:36] <marvin24> I think that's the best way for now, because flashing it to emmc is a bit more complicated
[08:37] <marvin24> http://nv-tegra.nvidia.com/gitweb/?p=tools/tegrarcm.git;a=summary
[08:37] <marvin24> but you need the bct first ...
[08:39] <marvin24> (the first 4080 byte of the bct partition #2 I think)
[09:05] <janimo> marvin24, mainline boots showing two penguins qwith blue beaks and feet, then cannot find root, but impressive still :)
[09:05] <marvin24> janimo: needs to boot from sd card
[09:05] <ogra_> whats impressive ? the penguins ? :P
[09:05] <janimo> the fact it boots :)
[09:05] <marvin24> ogra_: maybe the blue feets
[09:06] <janimo> marvin24, why sdcard? no mmc suport implemented?
[09:06] <marvin24> janimo: there is some bug in the drm driver which causes color switch
[09:06] <marvin24> janimo: no nvidia partition support
[09:06] <ogra_> marvin24, well, penguins live on ice :)
[09:06] <marvin24> I could add it though
[09:06] <ogra_> you mean we could use mainline with just one patch added ?
[09:07] <ogra_> or is there more that doesnt work
[09:07] <janimo> that's what I was hoping but marvin24 said no way yet :)
[09:07] <ogra_> what else is missing ? nvec ?
[09:07] <marvin24> ogra_: we need the drm support (one patch) and optional partition support
[09:08] <marvin24> everything else is there
[09:08] <ogra_> hmm, so three then
[09:08] <janimo> power management?
[09:08] <marvin24> not yet
[09:08] <ogra_> aha
[09:08] <ogra_> k, that somewhat rules it out for now then
[09:08] <marvin24> ogra_: but it is boot
[09:08] <marvin24> *booting
[09:08] <ogra_> (as  a default)
[09:26] <marvin24> ogra_: don't you have a nvec-leds entry in sysfs?
[09:26] <ogra_> i didnt look :P
[12:15] <ogra_> http://ograblog.wordpress.com/2012/08/06/the-bamboo-feeder-automating-continuous-arm-image-tests/
[12:32] <janimo> ogra_, too much hardware in the blogpost for my taste but still good :)
[12:35] <ogra_> i should have scattered more cute panda pics all over the place i guess :)
[13:33] <lilstevie> ogra_, that is damn cool
[13:33] <ogra_> :)
[13:34] <rsalveti> ogra_: you need to enable facebook/g+ sharing at your wordpress blog, can't share directly from your post :-)
[13:34] <rsalveti> but it's shared at g+ now :-)
[13:34] <ogra_> heh, thx
[13:35] <ogra_> i dont have a fecaebook account and dont plan to get one
[13:36] <ogra_> (do i ned one to enable the sharing) ?
[13:37] <ogra_> G+ enabled now
[13:37] <rsalveti> ogra_: nops
[13:38] <rsalveti> ogra_: you just enable the sharing plugin at wordpress for it
[13:38] <ogra_> ok
[13:39] <ogra_> added too now
[15:04] <FunkyPenguin> anyone know if u-boot in precise supports extload?
[15:05] <ogra_> should, yep
[15:05] <ogra_> i think all ubuntu versions since lucid do ...
[15:07] <FunkyPenguin> thanks, i'll give it a go and see
[15:10] <FunkyPenguin> i tried Linaro's 12.05 image on the snowball's emmc but that didnt seem to want to see the sdcard and boot from that
[15:10] <ogra_> oh, snowball
[15:10] <FunkyPenguin> just building the latest 12.07 release and will try that
[15:10] <ogra_> note that ubuntu doesnt have any support for snowball beyond what linaro puts in the archive
[15:11] <ogra_> so if linaros u-boot doesnt have extload for the snowball enabled, you wont have it
[15:11] <ogra_> i can only speak for all u-boots for the supported ubuntu arches
[15:19] <FunkyPenguin> which is omap?
[15:20] <ogra_> omap, omap4, highbank and armadaxp currently
[15:20] <ogra_> and tegra2 ac100 netbooks but there is no u-boot
[15:21] <ogra_> (android based netbook which uses a fastboot based loader)
[15:22] <ogra_> we used to have mx5 images too until precise
[15:28] <janimo> ogra in the initial ac100 tarball, where are the hooks that create the new initramfs for the installer reboot?
[15:28] <ogra_> you mean ac100-tarball-installer ?
[15:29] <ogra_> there are no specific initrd hooks, it just runs a chrooted flash-kernel and puts a default initrd in place
[15:30] <ogra_> (it sets MODULES=dep and diverts plymouth and cryptsetup hooks though)
[15:30] <ogra_> (before doing that)
[19:31] <zma> I've got a silly question. I'm trying to build kernel I fetched from git://github.com/RobertCNelson/stable-kernel.git , build_kernel script fails in the beginning to complain there is no directory "KERNEL". Am I missing something?
[19:36] <zma> ahh, sorry that, I think it's for mainline kernel source