| 20:37 | heap_ | hi |
| 20:39 | heap_ | what is the most stable 4.x kernel for the cuboxi? |
| 20:41 | Humpelstilzchen | 3.14? |
| 20:43 | heap_ | 4,x |
| 20:51 | vpeter | heap_: Are there more than one? |
| 20:52 | heap_ | linux-image-4.3.0-0.bpo.1-armmp - Linux 4.3 for ARMv7 multiplatform compatible SoCs\ |
| 20:59 | heap_ | the heck |
| 20:59 | heap_ | why that stupid debian is completely broken from solidrun? |
| 20:59 | heap_ | i cant even install kernel from backports |
| 20:59 | heap_ | Processing triggers for systemd (215-17+deb8u2) ... |
| 20:59 | heap_ | Errors were encountered while processing: /var/cache/apt/archives/irqbalance_1.0.6-3_armhf.deb |
| 21:10 | heap_ | ok looks like really customized debian from solidrun is crap |
| 21:26 | heap_ | why solid run uses its own uboot not the one from debian? |
| 21:26 | heap_ | crap |
| 21:40 | heap_ | why you have your own uboot - debian have its own one |
| 21:40 | heap_ | why do you have your own kernel patches on prehistoric kernel... debian backports has 4x |
| 21:40 | heap_ | etc |
| 21:41 | heap_ | the intention is to confuse user as much as possible? |
| 21:42 | dexter93 | heap_, welcome to arm lol |
| 21:42 | heap_ | that is stupid. |
| 22:01 | heap_ | eveyone is working on its sand and results are shits |
| 22:40 | jnettlet | heap_, if you want to use upstream packages then go and download the deb bootstrap and knock yourself out. We provide a customized distribution that works for our customers |
| 22:41 | jnettlet | you want to use something a feature that we do not support and have never claimed to support |
| 22:41 | heap_ | jnettlet: i just tried |
| 22:41 | heap_ | even official uboot is broken |
| 22:41 | heap_ | it doenst support boot from sata drive |
| 22:41 | heap_ | ;//// |
| 22:41 | jnettlet | I know, which is why we provide a customized working u-boot |
| 22:41 | heap_ | so i just completely messed it up |
| 22:41 | heap_ | jnettlet: ;-9 |
| 22:41 | heap_ | ;-( |
| 22:41 | heap_ | jnettlet: they told me that ppl not from debian like vendors doesnt understand those debian things.. |
| 22:42 | heap_ | jnettlet: where is that last debian image pls? |
| 22:42 | jnettlet | https://images.solid-build.xyz/IMX6/Debian/ |
| 22:42 | heap_ | this is like breakdown |
| 22:42 | jnettlet | We discussed having proper support in debian and debian said they were not interested. Thus we have provided our own |
| 22:43 | jnettlet | The only proper distribution that is interested in supporting upstream hardware as best as possible is Fedora |
| 22:43 | jnettlet | although they will not support any binary drivers |
| 22:43 | heap_ | jnettlet: whey they were not interested? |
| 22:44 | jnettlet | because debian is a terribly maintained distribution. They are still using libpng 1.2 which doesn't even have neon support |
| 22:44 | heap_ | ah shit |
| 22:45 | heap_ | i break my system down |
| 22:45 | heap_ | jnettlet: pls and how can i flash that image from your site |
| 22:45 | heap_ | to sd |
| 22:46 | jnettlet | if you are so interested in btrfs then I would suggest back porting whatever version you want to a stable kernel |
| 22:46 | jnettlet | from linux? |
| 22:46 | heap_ | yes |
| 22:46 | heap_ | what do you mean by backporting? |
| 22:46 | heap_ | like i have to do it? |
| 22:47 | jnettlet | merge the upstream btrfs code back into the 3.14 kernel |
| 22:47 | heap_ | hm i dont have skill for that |
| 22:47 | heap_ | ;/ |
| 22:50 | heap_ | jnettlet: and whats that command pls |
| 22:50 | jnettlet | xzcat | sudo dd of=/dev/sdX bs=1M; sync |
| 22:50 | heap_ | thanks! |
| 22:51 | heap_ | is it hard to backport that to 3x? |
| 22:51 | heap_ | kernel |
| 22:51 | jnettlet | I have no idea. I have not tracked the btrfs development work at all |
| 22:52 | vagrantc | jnettlet: who did you talk to about cubox support in debian? |
| 22:53 | jnettlet | I can't remember it was so long ago |
| 22:53 | vagrantc | jnettlet: there are several debian developers who've been working on getting support for cubox/hummingboard in debian properly |
| 22:54 | vagrantc | cubox-i, at least, not sure about the old cubox boards |
| 22:54 | heap_ | vagrantc: looks like you guys should discuss this |
| 22:54 | jnettlet | really at this point we don't have much interest. Fedora is being much more open about things. |
| 22:54 | vagrantc | main thing was u-boot support wasn't in mainline u-boot, although that's changed now |
| 22:55 | jnettlet | except that mainline u-boot is broken and causes all sorts of instability because they don't use our stress tested DDR settings |
| 22:55 | jnettlet | and they are missing other various patches |
| 22:55 | vagrantc | jnettlet: i've been the u-boot maintainer in debian for over two years now, and only random users ever asked about getting support |
| 22:55 | vagrantc | jnettlet: and patches have been submitted upstream? |
| 22:55 | jnettlet | I can dig up the email thread but really I don't have the time or energy at this point |
| 22:56 | vagrantc | ok. |
| 22:56 | jnettlet | vagrantc, no. I don't have the time to revisit work done a year and a half ago, again when u-boot wasn't interested in merging SPL support for iMX6 |
| 22:56 | vagrantc | well, i've been working on improving cubox-i support in debian ever since i got my cubox-i4pro, as have several other developers, so comments about debian not being interested seem really odd. |
| 22:56 | jnettlet | with no direction for patches to be fixed for inclusion. The community got, no that is not how we want it done. we are working on something |
| 22:57 | jnettlet | which was then produced 5 months later |
| 22:57 | vagrantc | sure, mainline support usually takes time |
| 22:57 | jnettlet | yes, but we were shipping product |
| 22:57 | jnettlet | we can't go back and redo all our dev work because upstream finally has time |
| 22:57 | jnettlet | our bootloader works, and works better than upstream |
| 22:58 | vagrantc | oh well. |
| 22:58 | jnettlet | and now people are using upstream and complaining to us that their boards are crashing etc. |
| 22:58 | heap_ | yes |
| 22:58 | heap_ | im example |
| 22:58 | heap_ | ;-) |
| 22:59 | vagrantc | i've had a cubox-i4pro and hummingboard-i2ex running constant builds for months using an upstreamed u-boot and upstreamed kernel. |
| 22:59 | heap_ | and we just proved that upstream uboot is borkend and cant even handle external drive |
| 22:59 | jnettlet | u-boot gave us a big long speech on how they only merge stable stuff, and then they merge our board "support" and let it be busted |
| 22:59 | jnettlet | because you happen to have the two soms that are kind of correct ddr settings |
| 22:59 | jnettlet | allthough even they will have problems on warm reset |
| 23:00 | vagrantc | heap_: well, actually the u-boot you tested is based on solid-run's patches ... |
| 23:01 | vagrantc | jnettlet: guess i'm just lucky with the 2 hummingboard and cubox-i systems i've been running for 1-2 years... |
| 23:01 | dexter93 | https://www.youtube.com/watch?v=Vhh_GeBPOhs |
| 23:01 | jnettlet | vagrantc, you haven't been running them on upstream u-boot for 1-2 years |
| 23:01 | heap_ | vagrantc: tthen why it doesnt work |
| 23:01 | vagrantc | jnettlet: true, only been running on upstream u-boot for about 8 months |
| 23:02 | vagrantc | heap_: that's a complicated issue that we haven't even begun to delve into |
| 23:03 | vagrantc | haven't encountered stability issues, but missing sata support is unfortunate |
| 23:03 | jnettlet | and if you want to know how "stable" upstream u-boot is you will find that ethernet was broken for 3 months before I tested and found it |
| 23:04 | vagrantc | well, thanks for that |
| 23:04 | heap_ | jnettlet: looks like noone will backport btrfs to 3. |
| 23:04 | heap_ | even dev ppl from btrfs said its hard\ |
| 23:05 | jnettlet | heap_, sorry about that |
| 23:05 | vagrantc | still don't understand why heap_ 's kernel isn't detecting sata. |
| 23:05 | vagrantc | though i admittedly didn't test 4.3, just 3.16 with that |
| 23:06 | jnettlet | 4.x has been in various on and off broken states throughout the series |
| 23:06 | jnettlet | rmk keeps fixing things and then freescale developers break them again |
| 23:06 | jnettlet | ethernet still produces MDIO read timeout errors |
| 23:06 | vagrantc | the msata on 4.x works fine on the hummingboard |
| 23:06 | vagrantc | (though not from u-boot, if i recall) |
| 23:07 | vagrantc | jnettlet: well, thanks for your work on these platforms, they're quite nice to work with! |
| 23:07 | heap_ | jnettlet: do you mean rmk or mk? where is those rmk's sources? |
| 23:08 | jnettlet | everything is pushed upstream |
| 23:08 | vagrantc | unfortunately upstreaming often takes too long |
| 23:08 | jnettlet | rmk is the linux arm maintainer |
| 23:08 | heap_ | but can i download rmk's 4.x kernel directly? |
| 23:09 | jnettlet | it is the mainline linux kernel |
| 23:09 | vagrantc | heap_: could try pulling the kernel from debian experimental, which is 4.4.x ... |
| 23:09 | heap_ | ahh then i need the solid-run kernel patched |
| 23:10 | vagrantc | heap_: or just get you back to whatever you had before |
| 23:10 | heap_ | vagrantc: but hm i can try to pull 4.4 but as packae? |
| 23:10 | heap_ | will it work with solidrun uboot? |
| 23:12 | vagrantc | heap_: it may take some manual configuration |
| 23:15 | heap_ | hmhm |
| 23:15 | heap_ | jnettlet: how stable is one from xbian? |
| 23:16 | jnettlet | I don't use it. I either work on 3.14 or mainline |
| 23:16 | heap_ | ah |
| 23:20 | vagrantc | so, with the cubox-i4x4, have the ram timings changed even more? |
| 23:21 | vagrantc | or is that likely to work at least as "badly" as before? |
| 23:22 | jnettlet | I don't know if upstream supports 4gb configs. They don't support the HB2 |
| 23:22 | vagrantc | HB2 being... ? |
| 23:22 | vagrantc | the -edge series, or what? |
| 23:23 | vagrantc | i've got a hummingboard-i2ex working with mainline u-boot and kernel, and a cubox-i4pro, and a hummingboard-i1 |
| 23:25 | vagrantc | jnettlet: well, sounds like you're worn out on some of this, but i've had some luck getting patches into u-boot mainline, and would love to improve support for these boards in Debian. |
| 23:28 | vagrantc | jnettlet: so if there's a way i could help, i'd be interested |
| 23:30 | jnettlet | Not really. I am trying to work with fedora at this point. Fabio is trying to fix uboot but again I have no time to revisit work that I did a year and a half ago |
| 23:30 | jnettlet | Thanks for the offer. |
| 23:32 | vagrantc | alright, if it's moving upstream, debian will get it eventually |