| 05:29 | jnettlet[m]> | Ke: upstream u-boot for which? |
| 06:48 | Ke> | mcbin |
| 07:45 | Ke> | obviously there is no regression anymore, when I can start the bisect |
| 07:45 | Ke> | 4.15.10 boots just fine |
| 07:47 | Ke> | -# Linux/arm64 4.15.10 Kernel Configuration |
| 07:47 | Ke> | +# Linux/arm64 4.15.10-rc1 Kernel Configuration |
| 07:47 | Ke> | diff in configs is that much |
| 07:51 | suihkulokki> | Ke: 4.16-rc1 ? |
| 07:51 | Ke> | rc5 was the only 4.16 tried from debian experimental |
| 07:51 | suihkulokki> | that's broken |
| 07:51 | suihkulokki> | forget and wait for rc6 |
| 07:52 | Ke> | yes, I can confirm it was broken, as I already tried |
| 07:52 | Ke> | anyway, as 4.15.10 works, I am just going to drop this for now |
| 07:52 | Ke> | it also has the scsi patch |
| 07:53 | suihkulokki> | pretty much all of arm was broken in -rc5 by someone who decided to muck with core kernel page allocation code to fix and issue with a handful of x86 machines |
| 07:54 | Ke> | so it was one of the x86 commits I disregarded |
| 07:55 | Ke> | perhaps that patch was also dropped between 4.15.10-rc1 and 4.15.10 |
| 07:58 | suihkulokki> | It was CC:d stable, I don't know if it managed to land to 4.15.x before people started screaming |
| 07:59 | Ke> | at least 4.15.10 runs |
| 08:37 | Ke> | wonder if these build guys are starting to strip away relevant make deps again, last time I had to move make install to a separate command, now I guess I'll need to have make -j$(nproc) all && make install && make -j$(nproc) modules_install |
| 08:37 | Ke> | cp: cannot stat './modules.order': No such file or directory |
| 08:50 | suihkulokki> | building kernel as root.. naughty boy! |
| 08:51 | Ke> | =o( |
| 08:51 | Ke> | in gentoo all kernels are owned by the root |
| 08:52 | Ke> | I guess with O= I could be able to build as user with readonly sources |
| 08:53 | suihkulokki | 08:53 * suihkulokki knows nothing about gentoo except the memes people make about it |
| 08:54 | Ke> | gentoo allows you to install debug and dev packages and build deps for all the packages you install |
| 08:54 | Ke> | also had more functional perf and adb than debian |
| 08:55 | Ke> | and some other packages a bit better maintained, like anything in rust/haskell projects |
| 08:56 | Ke> | the maintenance load is though considerably higher, and since I found equivs having packages installed in debian is considerably less painful |
| 08:58 | Ke> | though newest version of equivs no longer allows you to install files somehow |
| 09:10 | Ke> | all debug packages btw. means 50GiB of space |
| 09:10 | Ke> | or 30, who's counting |
| 09:20 | Ke> | actually just removed 7GiB of old kernel debug symbols |
| 16:00 | vpeter> | jnettlet[m]: Do you know what are the changes in latest IBx Bios 0.31-2 ? |
| 16:01 | jnettlet[m]> | vpeter: we have a newer bios that applies all my changes to tweak the thermal management. |
| 16:01 | jnettlet[m]> | do you want to test it? |
| 16:02 | vpeter> | Can it brick? :) |
| 16:03 | jnettlet[m]> | nope |
| 16:03 | jnettlet[m]> | https://drive.google.com/open?id=10pP_YneOjqq4KDHY-wpuKKCWw74sd7Y6 |
| 16:06 | vpeter> | You need to approve it - check mail. |
| 16:40 | jnettlet[m]> | vpeter: sorry try this. https://drive.google.com/file/d/1V_U_FNI9aF1--rvC58BT5ASR1khg6LRC/view?usp=sharing |
| 18:24 | vpeter> | jnettlet[m]: How to create fat43 usb stick? :) |
| 18:24 | vpeter> | "Fat34 USB storage device with at least 11MB of free space" |
| 18:24 | jnettlet[m]> | type. should be fat32 |
| 18:25 | jnettlet[m]> | typo |
| 18:25 | vpeter> | I know :) |
| 18:26 | jnettlet[m]> | it is beta |
| 18:27 | vpeter> | Well, if even text is buggy what can I expect from binary bios then :-) |
| 18:37 | vpeter> | jnettlet[m]: when I type fs0 or fs1 I got "not recognized command". What is wrong? But it list 2 disks on efi enter. |
| 18:37 | jnettlet[m]> | fs0: |
| 18:38 | jnettlet[m]> | I think |
| 18:38 | jnettlet[m]> | just like dos |
| 18:38 | vpeter> | I should read instructions more carefully :( |
| 18:43 | jnettlet[m]> | no worries |
| 18:44 | Ke> | wow, building u-boot is fast |
| 18:51 | vpeter> | jnettlet[m]: good news - it still boots. |
| 19:04 | jnettlet[m]> | vpeter: you probably want to go in and reset everything to the defaults. We have tweaked lots of settings for the thermal management. |
| 19:06 | vpeter> | Always first step after bios upgrade. Checked. |
| 20:37 | vpeter> | jnettlet[m]: is this error serious? mmc1: cache flush error -110 |
| 20:43 | Ke> | cache flush error sounds like the data could be on disk or not |
| 20:46 | vpeter> | That's not good then. Was wondering if emmc works with new bios but it is not :( |
| 20:51 | Ke> | fwiw error -110 is quite common, perhaps it's something like EINVAL, ENOENT or EIO |
| 20:52 | Ke> | useless error code that is there, because people writing specs hate you |
| 20:52 | Ke> | next revision will just have EERROR |
| 22:04 | jnettlet[m]> | vpeter: try going into the MMC config in the BIOS and set the speed to 52MHz instead of AUTO. |