15:00:19 <pwhalen> #startmeeting Fedora ARM and AArch64 Status Meeting
15:00:19 <zodbot> Meeting started Tue Oct 18 15:00:19 2016 UTC.  The chair is pwhalen. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:19 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:00:19 <zodbot> The meeting name has been set to 'fedora_arm_and_aarch64_status_meeting'
15:00:19 <pwhalen> #chair pwhalen pbrobinson dgilmore hrw dmarlin yselkowitz jonmasters ahs3 msalter
15:00:19 <zodbot> Current chairs: ahs3 dgilmore dmarlin hrw jonmasters msalter pbrobinson pwhalen yselkowitz
15:00:29 <pwhalen> morning folks, whos here?
15:01:04 * pbrobinson o/
15:02:04 <pbrobinson> pwhalen: lets get this done while everyone else isn't here ;-)
15:02:13 <pwhalen> heh
15:02:33 <pwhalen> #topic 1) ==== Userspace Status  ====
15:03:16 <pwhalen> i think we're looking pretty good here, anyone aware of any issues?
15:05:59 <pbrobinson> so mono 4.6 was boostrapped and done in F-26/rawhide last week
15:06:25 <pbrobinson> I also fixed up a bunch of the matplotlib leftovers so I think that's all good as well
15:06:50 <pwhalen> #info Mono 4.6 was boostrapped in F-26/rawhide last week.
15:07:50 <pbrobinson> that was the biggest block left over from the merge, there's a few other stragglers but nothing large in terms of pure packages
15:08:03 <pbrobinson> if anyone wants a couple of small projects let me know
15:08:58 <pwhalen> cool, anyone have anything else to mention?
15:09:19 <pwhalen> #topic 2) ==== Kernel Status ====
15:09:31 <pwhalen> #info Latest Rawhide - kernel-4.9.0-0.rc1.git0.2.fc26
15:09:31 <pwhalen> #link http://koji.fedoraproject.org/koji/buildinfo?buildID=810642
15:09:46 <pwhalen> #info Please test and report any issues to the list or #fedora-arm.
15:10:26 <pbrobinson> 4.8.2 is built for Fedora 25 too, please test as that should be the one headed to stable releases shortly
15:11:19 <pwhalen> #info Latest F25 kernel - kernel-4.8.2-300.fc25
15:11:30 <pwhalen> #link http://koji.fedoraproject.org/koji/buildinfo?buildID=810665
15:13:06 <pwhalen> any kernel issues to mention?
15:13:53 <pbrobinson> none off hand from my PoV but this is normally when testers pop their head up
15:15:01 <pwhalen> no issues thus far here.
15:15:17 <pwhalen> #topic 3) ==== Bootloader Status ====
15:16:27 <pbrobinson> so 2016.11rc2 landed upstream but it's got issues with openssl 1.1, I've reported it upstream
15:16:36 <pbrobinson> that's about it from my PoV
15:17:07 <pwhalen> #info Latest - uboot-tools-2016.11-0.1.rc1.fc26
15:17:13 <pwhalen> #link http://koji.fedoraproject.org/koji/buildinfo?buildID=808089
15:17:44 <ahs3> sorry to bring up a kernel question late, but is CPPC enabled?  we may want to turn that off due to upstream problems
15:17:56 <pwhalen> working well here, also fixes the ext4 /boot partition issue
15:19:48 <pbrobinson> ahs3: yes, it is, the APM people were testing it, but that was back in 4.7.x I think
15:20:29 <pbrobinson> ahs3: but I think that was internal firmware or something, any particular kernels it should be turned off on?
15:21:43 <ahs3> pbrobinson: any arm64; upstream has already found 3 bugs in 4.8
15:22:15 <masta> ahs3, is this related to 1372408 ?
15:22:56 <ahs3> lemme look
15:24:20 <ahs3> that's one of them; there are some basic arithmetic errors in the calculations, too
15:26:01 <pwhalen> anything else for bootloaders?
15:26:11 <pbrobinson> ahs3: is that purely the ACPI_CPPC_CPUFREQ option, or are there others we should disable too?
15:26:11 <masta> So in a nutshell disable aarch64 performance governors for now, try again with future kernel?
15:26:44 <ahs3> yeah, that would be my recommendation
15:26:54 * pbrobinson slaps masta with a wet fish for using the marketing term nutshell
15:27:13 <pbrobinson> ahs3++ awesome thanks, will push that shortly
15:27:13 <zodbot> pbrobinson: Karma for ahs3 changed to 1 (for the f24 release cycle):  https://badges.fedoraproject.org/tags/cookie/any
15:27:14 <masta> oh man, now I smell fishy
15:27:35 <pbrobinson> masta: no further comment needed ;-)
15:28:00 <pwhalen> #topic 4) ==== F25 Final ====
15:28:17 <pwhalen> #info Testing so far:
15:28:17 <pwhalen> #info ARMhfp - https://fedoraproject.org/wiki/Test_Results:Fedora_25_Branched_20161017.n.0_Summary
15:28:17 <pwhalen> #info AArch64 - https://fedoraproject.org/wiki/Architectures/AArch64/QA/Fedora-25-20161017.0
15:28:50 <pwhalen> #info Please test and add results to the wiki.
15:29:05 <pwhalen> looks good so far
15:29:29 <pbrobinson> So for final I'm still planning on getting some form of Pine64 support and disk images working. I'm hoping I should have some cycles late this week to focus on that
15:29:38 <pwhalen> #info Fedora 25 Schedule:
15:29:38 <pwhalen> #info 2016-11-01 - Final Freeze
15:29:39 <pwhalen> #info 2016-11-15 - Final Release
15:30:02 <masta> pwhalen, your test matrix is a thing of beauty. Seriously, that is well organized table setup.
15:30:33 <jlinton> Its not F25 related, so much as the linaro 16.9 firmware for juno, seems to have problems booting from USB devices (and likely PXE from the PCIe nic)...
15:30:34 <pwhalen> yea, not mine, likely Adam did it
15:31:01 <jlinton> So, that causes problems if anyone else tries to install F25 on it.
15:31:23 <pbrobinson> jlinton: thanks for the heads up
15:32:26 <masta> pbrobinson, does the pine64 have any special needs for the disk image?
15:32:41 <pbrobinson> masta: what do you mean?
15:33:15 <masta> well some boards look for a loader at specific byte offset of the disk, stuff like that
15:33:48 * masta just wondering if pine64 is *special* in some way
15:33:50 <pbrobinson> masta: yes, it's the standard combo of AllWinner plus aarch64 trusted boot requirements
15:34:18 <masta> cool
15:34:19 <pbrobinson> masta: so it's a little different to standard allwinner but in a standard ARMv8 difference
15:35:15 <pwhalen> #topic 5) == Open Floor ==
15:35:32 <pwhalen> I think thats all we had for this week, anyone have anything for open floor?
15:35:58 <masta> who here is getting one of the aarch64 laptops ?
15:36:34 <pbrobinson> masta: which ones? I've ordered the kickstarter one based on the A64 SoC (same as pine64)
15:36:52 <pbrobinson> do you mean the announced Chromebooks though?
15:37:34 <masta> pbrobinson, oh cool, I don't even know what that is... but I want one now that I'm aware. I was not being clear, I was implying aarch64 chromebook/android  laptops
15:38:20 <masta> so yeah, chromebooks
15:38:55 <pbrobinson> masta: https://www.kickstarter.com/projects/hybrxpc/ultra-slim-laptop-with-android-51-os-and-remix-os
15:39:41 <masta> pbrobinson, thank you!
15:40:28 <pwhalen> cool, looks like you can expect yours soon pbrobinson.. let us know how it is
15:40:29 <masta> ok, well I'm going to see about exploring the fedora aarch64 workstation story.
15:40:35 <pbrobinson> masta: I would like to have some coordination (I know imcleod is doing a bunch) of getting any changes needed into Fedora for the chromebooks so standard images etc can be used
15:41:10 <pbrobinson> I know at least the exynos chromebooks can now boot a vanilla Fedora kernel except I need to do a patch around the MMC module
15:41:53 <masta> pbrobinson, yes, coordination.
15:42:36 <pbrobinson> masta: I know also the tegra ones we can write a u-boot for chaining into the "legacy" area so then should be able to boot a normal OS. It needs to be wrapped as a FIT image ec
15:42:37 <pbrobinson> etc
15:43:17 <masta> yeah, I both love and hate the FIT image format
15:44:25 <masta> ok, well I just wanted to gauge who is going to be having what aarch64 laptops, I'm goign to get as many as I can afford
15:44:41 <pbrobinson> so we should be able to produce  that as part of the u-boot process for each device so that the end users wouldn't have to mess with generating that bit
15:44:55 <masta> pbrobinson, that is a good idea
15:45:15 <masta> actually, that is a great idea.
15:45:42 <pbrobinson> masta: I have lots of bad ones for each good one ;-)
15:45:44 <masta> in theory we could generate just one FIT image for all the devices
15:46:18 <pbrobinson> masta: the idea would be that they can then grab the FIT image and write it to the legacy ROM and then boot direct from a SD card or usb stick
15:47:13 <masta> pbrobinson, or chain boot the FIT image
15:47:14 <pbrobinson> masta: well given we produce one u-boot for each device I'd just do a FIT image per device. I'm not sure how much space on the legacy bit we have and a single image would get large/complex quickly
15:47:33 <masta> pbrobinson, true
15:47:46 <masta> keep it simple
15:48:15 <masta> okay folks, eof
15:48:29 <pwhalen> anything else for open floor?
15:49:37 <pwhalen> #endmeeting