| 11:11 | Ke> | is anyone still alive here |
| 14:38 | Ke> | linux-4.19-rc5 hangs on boot on my mcbin |
| 15:06 | suihkulokki> | Ke: tried with clk_ignore_unused ? |
| 15:06 | Ke> | nope |
| 15:06 | Ke> | would it be a regression after 4.18? |
| 15:12 | suihkulokki> | could be |
| 15:17 | Ke> | well, why not |
| 15:17 | Ke> | I guess that's command line parameter |
| 15:24 | Ke> | yeah, seems to work, is it like Keanu Reeves in Matrix? |
| 15:24 | Ke> | I have heard someone else pointing to that particular option also |
| 15:35 | Ke> | suihkulokki: let's assume it was the clk_ignore_unused, what would it tell? |
| 15:38 | suihkulokki> | Ke: it tells that the kernel and device-tree have different idea on what clocks are really neccessary |
| 15:39 | Ke> | on some of the boots I still had my old 4.16 device tree in use |
| 15:40 | Ke> | I don't need it anymore though, since I have new EFI |
| 15:40 | suihkulokki> | As discussed last week with Grant Likely and Rob Herring, we need to start considering kernel's failing to boot with in-firmware device trees as regressions |
| 15:41 | suihkulokki> | in case of McBin they can still cop out with "we didn't ship it in firmware, you build the firmware yourself..." |
| 15:42 | Ke> | if you run unmodified dts in it, it's not really different |
| 15:44 | Ke> | though it's not like my build script could not handle updating the dts |
| 15:44 | Ke> | I guess you noticed that mcbin efi is mostly upstream now |
| 15:45 | Ke> | I guess it's actually linaro maintaining the edk2-open-platform |
| 15:46 | Ke> | https://github.com/Linaro/OpenPlatformPkg or this |
| 15:47 | Ke> | other than that, there is mv-ddr-marvell and ble-marvell that did not yet fit into arm-trusted-firmware |
| 15:50 | Ke> | there is rumours of 4.19 being frozen for debian buster so I can stop worrying |