| 06:46 | jnettlet[m]> | Ke: yeah that is one of the things that Intel still has over ARM based Chromebooks. Even though Intel is just getting into industrial supported chips with the new Apollo Lake (15 years lifetime). They just have better mainline support so it is much easier for Google to support them. |
| 06:47 | jnettlet[m]> | A chromebook based on the new Snapdragon 820 industrial would be an excellent ARM choice. |
| 08:09 | Ke> | perhaps, though the main reason to want arm goes away, if you have the baseband there, or locked bootloader |
| 08:10 | bencoh> | indeed |
| 09:18 | paulk-gagarine> | jnettlet[m], Snapdragon? |
| 09:18 | paulk-gagarine> | jnettlet[m], do you know any that can run free boot software? |
| 09:18 | paulk-gagarine> | Ke, depends what you're looking for, but it sure is a nice machine, yeah |
| 09:18 | paulk-gagarine> | Ke, the huge screen resolution is kind of a problem to me though |
| 09:22 | jnettlet[m]> | paulk-gagarine: the dragonboard from 96boards can boot from OSS UEFI |
| 09:22 | paulk-gagarine> | jnettlet[m], do you know details about that or should I look them up? |
| 09:23 | paulk-gagarine> | especially, whether this is just a chainloaded free UEFI boot software or whether it runs just after the bootro, |
| 09:23 | paulk-gagarine> | bootrom* |
| 09:27 | jnettlet[m]> | paulk-gagarine: well the primary bootloader and DDR initialization are blobs |
| 09:27 | paulk-gagarine> | jnettlet[m], ah right, that's what I remembered :) |
| 09:27 | paulk-gagarine> | so I hope they never choose these for chromebooks |
| 09:28 | jnettlet[m]> | but that is not much different than the Marvell a38x. their boot init is just hard coded into the chip. |
| 09:29 | paulk-gagarine> | well at least if it's hardcoded in the chip then it's not a problem from the free software perspective |
| 09:29 | bencoh> | "not a problem" from the fsf* perspective |
| 09:29 | paulk-gagarine> | nope |
| 09:29 | paulk-gagarine> | instructions hardcoded in silicon are not software |
| 09:30 | bencoh> | this code still runs on main cpu |
| 09:30 | paulk-gagarine> | they're part of the hardware design |
| 09:30 | paulk-gagarine> | and it doesn't need to be distributed, which makes it possible to have fully free boot software |
| 09:31 | paulk-gagarine> | it doesn't mean there's no freedom issue though |
| 09:31 | paulk-gagarine> | it means it's a hardware problem, not a software problem |
| 09:31 | bencoh> | well, yeah |
| 09:31 | jnettlet[m]> | all of this amuses me considering Linux is built on top of x86 which runs a BIOS/UEFI that is completely software, and not free at all. |
| 09:32 | jnettlet[m]> | I love how everyone loves to overlook that point |
| 09:32 | paulk-gagarine> | jnettlet[m], well freedom in x86 boot is definitely a problem :p |
| 09:32 | paulk-gagarine> | and the few x86 platform that can boot without blobs are old |
| 09:33 | paulk-gagarine> | well not that much actually, I think we have it up to ivybridge |
| 09:33 | Ke> | paulk-gagarine: how is libre parity with Asus C201 |
| 09:33 | paulk-gagarine> | Ke, libre parity? I don't understand that. |
| 09:34 | Ke> | is libreboot still plausible, does mainline kernel work equally well |
| 09:34 | paulk-gagarine> | mainline for rk3288 is still work in progress |
| 09:34 | Ke> | sure, though it works well enough for me |
| 09:34 | jnettlet[m]> | we have a working coreboot for our Braswell boards, however it still requires Intel binary blobs |
| 09:34 | paulk-gagarine> | rockchip is contributing bits and pieces, but there's work left to do |
| 09:34 | paulk-gagarine> | jnettlet[m], yeah, at this point you need the FSP |
| 09:35 | jnettlet[m]> | yes |
| 09:35 | paulk-gagarine> | jnettlet[m], but for up to ivy bridge, we have native ram and memory init in coreboot now |
| 09:35 | paulk-gagarine> | ram and video* |
| 09:35 | Ke> | paulk-gagarine: but I mean rk3399 chromebook plus if the baseline is C201 |
| 09:35 | paulk-gagarine> | Ke, probably a lot worse in terms of upstream linux support |
| 09:35 | paulk-gagarine> | but I haven't booted it already |
| 09:36 | paulk-gagarine> | people in #linux-rockchip may have a more accurate idea |
| 09:36 | Ke> | and libreboot? |
| 09:37 | paulk-gagarine> | Ke, I have it supported in my build system, but it's not integrated in libreboot |
| 09:37 | Ke> | it builds and is able to boot chromebook kernel or something? |
| 09:37 | paulk-gagarine> | yeah sure |
| 09:37 | Ke> | nice |
| 09:37 | Ke> | thanks |
| 09:37 | paulk-gagarine> | same as c201 |
| 09:37 | paulk-gagarine> | I also added EC support some weeks ago |
| 09:37 | paulk-gagarine> | so you can rebuild and reflash the EC |
| 09:38 | Ke> | are you running mmind kernel or not running at all? |
| 09:38 | Ke> | or chromebook kernel |
| 09:38 | paulk-gagarine> | running the chromeos kernel |
| 09:39 | Ke> | I actually have firefly-rk3399 so I could check SoC support myself |
| 09:39 | paulk-gagarine> | indeed |
| 09:42 | paulk-gagarine> | but I should definitely start caring about upstream for rk3399, as well as mt8173 |
| 09:43 | Ke> | but as mentioned, long term availability chromebook with libreboot would be awesome |
| 09:43 | paulk-gagarine> | Ke, well honestly, I'm not too worried about that, they are mass-produced |
| 09:44 | Ke> | but as noted Chromebook plus has disappeared from amazon and only has some availability over different sources |
| 09:44 | Ke> | C201 also |
| 09:44 | paulk-gagarine> | I'm sure you can find lots on ebay |
| 09:44 | Ke> | yes, I can |
| 09:46 | paulk-gagarine> | Ke, and I mean, amazon is often temporarily out of stock, it doesn't mean they have stopped producing it ;) |
| 09:46 | Ke> | ok, let's hope |
| 13:51 | topi`> | jnettlet[m]: do you know if Ilya is back from vacations? I just sent him email about HB availability |
| 13:51 | jnettlet[m]> | topi`: just got back today |
| 13:58 | topi`> | there will be a hefty vacation backlog of emails :) |
| 14:32 | topi`> | jnettlet[m]: do you ship any i.mx6 SOMs with ready-flashed eMMC? I'd like to have UBOOT readily installed, as was the case with the macchiatoBIN |
| 14:33 | jnettlet[m]> | topi`: it can be requested |
| 14:35 | topi`> | so if you have any i.mx6 SOMs in stock (already out of factory, so no lead-time), they most often are with blank eMMCs? |
| 14:35 | topi`> | and to have something custom, you'll need to wait 12 weeks |
| 14:36 | jnettlet[m]> | correct |
| 14:37 | jnettlet[m]> | topi`: however new boards will have rev 1.4 boards so the boot jumpers are functional |
| 14:37 | jnettlet[m]> | so you can easily recover eMMC |
| 14:38 | jnettlet[m]> | I also have DFU support working in mainline u-boot if you would prefer to use that for flashing |