15:01:02 #startmeeting Fedora ARM and AArch64 Status Meeting 15:01:02 #chair pwhalen pbrobinson jlinton coremodule 15:01:02 Meeting started Tue Nov 10 15:01:02 2020 UTC. 15:01:02 This meeting is logged and archived in a public location. 15:01:02 The chair is pwhalen. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:01:02 Useful Commands: #action #agreed #halp #info #idea #link #topic. 15:01:02 The meeting name has been set to 'fedora_arm_and_aarch64_status_meeting' 15:01:02 Current chairs: coremodule jlinton pbrobinson pwhalen 15:01:11 Good morning folks, who's here today? 15:01:17 * pbrobinson o/ 15:01:26 * coremodule is here. Good morning/evening! 15:01:48 morning 15:03:10 #topic 1) ==== Userspace Status ==== 15:03:53 Anyone aware of any new problems? 15:04:58 #info No new issues reported. 15:05:08 #topic 2) ==== Kernel Status ==== 15:05:17 #info kernel-5.10.0-0.rc3.68.fc34 15:05:23 #link https://koji.fedoraproject.org/koji/buildinfo?buildID=1639995 15:05:51 Started testing this yesterday, seems ok so far. 15:06:02 The coresight/module patches landed in 5.10, so that should allow us to enable the coresight modules. 15:06:02 I will give this some cycles today 15:06:43 jlinton: nice 15:07:40 Any one having issues with 5.9? 15:08:05 I would guess that goes stable this week. 15:08:17 Nothing here, I ran the kernel regression suite on it last week and everything passed 15:08:23 on an RPi4 15:08:33 5.9 that is 15:08:45 Awesome, thanks. 15:10:17 Oh, thats good, the regression suite was failing on the rpi4 due to running out of ram/being slow for a long time. 15:10:59 it was on a 4gb pi with aarch64 so maybe that helped... it ran well! :) 15:12:39 back, just had a power outage here 15:12:55 oh wow, is that common? 15:13:22 no, they're doing some digging in the neighbourhood 15:13:28 ah gotcha 15:13:47 #info Please test and report any issues to the list or #fedora-arm. 15:13:57 #topic 3) ==== Bootloader Status ==== 15:14:34 #info Latest U-Boot available for testing - uboot-tools-2020.10-3.fc34 (building now) 15:14:41 #link https://koji.fedoraproject.org/koji/buildinfo?buildID=1640445 15:15:19 Looks like that has a couple enhancements for the Pinebook Pro 15:15:40 #info Please test and report any issues to the list or #fedora-arm. 15:15:43 very cool. thanks pbrobinson 15:16:13 next up is me documenting it all 15:16:28 everyone loves documentation :) 15:16:55 do you need any help? 15:16:58 then people can stop bothering me about that and continue bothering me about other randeom devices 15:17:04 LOL 15:17:12 it will never end. Always something new 15:17:13 coremodule: depends on how well you know the low levels of firmware and SPI flash 15:18:18 hmmm, I don't know either... I understand some bits, but I want to learn more... if I can help somehow, I'm glad to 15:18:50 coremodule: if you have a PBP once I've posted details feedback would be great 15:19:11 sounds good 15:20:00 #topic 4) ==== Fedora 33 ==== 15:20:33 Checking F33, looks liek the alt website still needs updating to point to the new images. Will ping on the ticket. 15:20:59 Anyone aware of any other issues in F33? 15:21:10 thanks pwhalen, do you have the ticket handy? 15:22:57 https://pagure.io/fedora-web/websites/issue/158 15:24:36 Anything else for F33? 15:24:48 #topic 4) ==== Fedora 34 ==== 15:25:32 Any plans for F34? I think we plan on moving to the uefi images as the default for armhfp 15:26:05 I've been considering MTE lately, although I think F34 might be to early to enable it 15:26:39 (thats the arm v8.5 memory tagging extension) 15:27:05 An early version of the kernel code landed in 5.10 too. 15:27:23 yes, I've got on my list to look at UEFI on ARMv7 for everything 15:27:44 also looking at GPT as I think now rpi firmware supports loading from GPT 15:28:31 and adjusting where the partitions start to give us some room at the beginning for rockchips U-Boot to make it easier to support that 15:28:34 I also want to figure out how to re-enable https boot. 15:28:43 that might be a better F34 feature. 15:28:46 jlinton: where was it disabled? 15:29:24 Well, it could be edk2, but when I test https booting the grub.cfg isn't right, and the kernel/initrd need to be part of the initial load. 15:29:45 its 99% working at the moment, but needs a couple of slight tweaks. 15:29:57 to be able to boot directly from the mirrors. 15:30:25 jlinton: that would be fab. let me know if i can help 15:31:51 that would explain the issues I had when testing in f33. 15:32:49 #topic 5) == Open Floor == 15:33:00 Anything else for today? 15:33:13 just fyi 15:33:15 I spent some time playing with hexxeh's upstream new .dtb for the RPi400 and aarch64 rawhide yesterday, couldn't make it boot. It complains about not being able to access the phy, then dumps me to a uboot prompt. But I have the hardware for testing for when it makes it's way to us. 15:35:14 coremodule: there's a new firmware for it, I've not pushed it yet because I've not tested that on anything I have 15:35:22 and it'll probably need U-boot work 15:36:16 I don't have the device and I have little urge to get one myself, I'm mostly burned out on the rpi in general so it'll likely have to wait until stuff is upstream I can consume 15:36:17 pbrobinson, is there any documentation that you know of on properly installing/testing the firmware? just so I can run some tests to try to save you some time 15:36:36 coremodule: I can do you a scratch build 15:36:44 that would be great 15:36:50 I have most of that side of things done, I just haven't done any testing 15:37:07 I'm trying to share the load, so the burden of rpi on fedora doesn't fall solely on you 15:38:17 ok, lets continue in channel. 15:38:22 #endmeeting