| 09:11 | vpeter | jnettlet: I finally got response from the user regarding Cubox-i with old SOM which is mistakenly identified as v1.5. GPIO pin values from me and him: https://pastebin.com/hDzz4W7N |
| 10:05 | rabeeh | vpeter: is that with bcm4329? |
| 10:22 | jnettlet | Yes, most likely a bcm4329 version. He said from 2013 |
| 10:24 | jnettlet | I think I am above ground for a while. |
| 10:25 | jnettlet | vpeter, if they set the env variable som to nothing then the can override the autodetect |
| 10:25 | jnettlet | Or just set fdtfile |
| 10:27 | vpeter | Or rename dtb file :) |
| 10:27 | vpeter | I'm just passing this info just to be known. |
| 10:30 | jnettlet | Probably better to set the u-boot env variable and save it. That will survive upgrades |
| 10:45 | vpeter | And why exactly this detection doesn't work correctly? Not enough gpio pins? :) |
| 11:12 | topi` | rabeeh: what is the meaning of R47 on the HB Gate? when I applied solder to R48, it also contacted the already-installed resistor on R47, but the board seems to boot OK nevertheless and mmc2 (emmc) gets detected. However mmc1 does not |
| 11:13 | topi` | if I boot the 3.14.79-fslc on that board, it detects mmc1 but not the mmc2(emmc) |
| 11:14 | topi` | so I haven't (yet) been able to run a kernel which would recognize both at the same time |
| 11:25 | jnettlet | topi, which device-tree file is it selecting for 3.14.79? |
| 11:49 | topi` | jnettlet: I symlinked the imx6q-hummingboard2.dtb to -v15-som.dtb |
| 11:49 | topi` | so, that one. |
| 11:50 | topi` | is it a good idea to clone rmk's linux-4.11 tree? to get the newest HB patches? |
| 11:51 | topi` | maybe russell's for-next branch |
| 11:54 | jnettlet | topi`, the only patches are the last 4 I pointed you to. other than that there is nothing different about that branch |
| 11:55 | topi` | OK |
| 11:56 | topi` | every time I try to clone linus's tree without "--depth 1", I run out of disk space ;) |
| 11:56 | jnettlet | yeah, the kernel tree is huge as this point. |
| 11:56 | topi` | my systems are too weak :( |
| 11:57 | topi` | since I have an allergy against cross compilation, I have an 8-core Rockchip 3368 board for building |
| 11:57 | topi` | and a NVIDIA Tegra K1 board as well, but it has less RAM |
| 12:01 | vpeter | topi`: You hate yourself? :) |
| 12:03 | topi` | yeah, I am a generic hater :) |
| 12:04 | topi` | actually, kernel compiles with a cross-compiler is easy and not error-prone at all. It's the userspace stuff that gets problematic |
| 12:04 | jnettlet | there is a major problem looming at the end of this year. Everyone in the EU will be able to roam to other countries but use their existing mobile plan. Except that all existing mobile phones suck at this. |
| 12:05 | jnettlet | there is no way to tell our stupid phones, yes I know I am roaming but it is fine. Please don't go crazy constantly scanning to get me back to my home network. I would like some battery life. |
| 12:23 | bencoh | jnettlet: I guess they'll just add some kludge to check phone location as well (?) |
| 15:07 | topi` | jnettlet: do you intend to push the -v15-som.dts into the mainline kernel as well? and does it even exist in the 3.14.79+ fslc kernel? |
| 15:08 | topi` | in the meantime, I'll continue symlinking it to the corresponding hummingboard2.dtb :) |
| 17:31 | jnettlet | topi`, they do exist in 3.14.79, mainline is tricky. The ti drivers don't support device-tree in mainline. So they won't accept patches until that is fixed |
| 17:32 | jnettlet | And adding proper device tree support requires a full driver rewrite for their kim protocol |
| 17:33 | jnettlet | I may try and push wifi support, and then bt can be initialized from userspace. |
| 18:05 | topi` | hm, maybe I need to git pull my 3.14.79 tree since I did not see the -v15-som.dtb there |
| 18:06 | topi` | hm, I thought mainline has a driver for Wilink8?? |
| 18:11 | jnettlet | It does but device tree support for the non wifi side was rejected |
| 18:14 | topi` | https://e2e.ti.com/blogs_/b/connecting_wirelessly/archive/2015/06/09/wilink-8-drivers-now-available-in-mainline-linux |
| 18:14 | topi` | oh, ok. |
| 18:14 | topi` | but do we get working wifi nevertheless? |
| 18:40 | topi` | I'll try it out. |
| 18:40 | jnettlet | Yes wifi should work although I haven't tested it. |
| 18:40 | topi` | I'll do some extensive testing during the following days |
| 18:41 | topi` | my biggest gripe now is the emmc / sdcard detection :) |
| 18:43 | topi` | first, I have to clear another, unrelated project to be able to proceed with this :/ |
| 18:43 | topi` | next thing they'll tell us is that our vacations will be cancelled ;) |