#asahi · 2021-04-10 19:29
marcan\o/#asahi · 2021-04-09 17:27
marcanarnd: fyi, here's your signed key: https://hub.marcan.st/t/arnd-signed.gpg#asahi · 2021-04-09 17:26
marcanand I guess we could use a kernel feature matrix, yeah#asahi · 2021-04-09 17:24
marcanarnd: I do plan on keeping some sort of tree that people can use with the various bits and pieces merged in some fashion#asahi · 2021-04-09 10:59
marcanarnd: fyi, I think you applied the fixup patch to arm/apple-m1 and merged it into for-next, but didn't push arm/apple-m1 itself#asahi · 2021-04-09 10:43
marcan(fiq is also in arm64 already)#asahi · 2021-04-09 10:43
marcanvhe is going via arm64#asahi · 2021-04-09 10:43
marcanfiq is in the merge#asahi · 2021-04-09 10:41
marcaneta: that's an artifact of a merge we did from tty-next, because the Samsung UART patches in the series took a different path via that tree#asahi · 2021-04-09 08:47
marcanI can do it now :)#asahi · 2021-04-09 08:44
marcanwhat's a good time?#asahi · 2021-04-09 08:43
marcanworks for me#asahi · 2021-04-09 08:42
marcannot sure what makes the most sense#asahi · 2021-04-09 08:42
marcanor I could just get 3 sigs and get hosted on kernel.org I guess#asahi · 2021-04-09 08:41
marcanarnd: you mentioned issues with github earlier, I'm happy to set up a git.asahilinux.org if that's helpful, or do you think it's not necessary?#asahi · 2021-04-09 08:26
marcanso about infra, I guess I should have the key signed? :)#asahi · 2021-04-09 08:26
marcanright#asahi · 2021-04-09 08:20
marcanah, so linux-next would pull from my tree then before I submit#asahi · 2021-04-09 08:20
marcanI saw that failure earlier from some branch maz had, but I failed to do a quick repro and wasn't sure what was going on there; is there a way to get that kbuild robot to run on my branches before I submit? I've seen it run on ML patches, but not this particular failure.#asahi · 2021-04-09 08:19
marcancool#asahi · 2021-04-09 08:18
marcan(more generally: how often do kernel folks hate it when the build is broken somewhere in the middle of history? :))#asahi · 2021-04-09 08:17
marcangot it#asahi · 2021-04-09 08:16
marcane.g. linux-next is rebased, right?#asahi · 2021-04-09 08:15
marcanarnd: I'm curious how the patch cycle usually works for fixes like this (especially when they break the build, so could annoy people who bisect); at what point do things usually switch from rebasing and rewriting history to fix-ups on top?#asahi · 2021-04-09 05:43
marcanI can ack the patch#asahi · 2021-04-09 05:43
marcansven: and yes, feel free to put me down as maintainer for all those drivers#asahi · 2021-04-09 05:21
marcan(maybe it could've used a tag to reference that?)#asahi · 2021-04-09 05:21
marcanit's a fixup for asm-generic/io.h: Add a non-posted variant of ioremap()#asahi · 2021-04-09 05:20
marcanarnd: sent#asahi · 2021-04-09 05:18
marcanwill do#asahi · 2021-04-09 04:52
marcanhow do we handle this? tack it on top, or does it still make sense to fix up the original series at this point?#asahi · 2021-04-09 04:51
marcanarnd: fix is here: https://github.com/AsahiLinux/linux/commit/736b638f5e73e992dc670b8279c3697d2223ae74#asahi · 2021-04-09 04:48
marcanah yes, it was#asahi · 2021-04-09 04:48
marcanwait, no, it wasn't?#asahi · 2021-04-09 04:48
marcanI nested an ifdef wrong#asahi · 2021-04-09 04:47
marcanah, fuck, it was a typo in my series#asahi · 2021-04-09 04:39
marcanwait, no, it's not __KERNEL__... what is this?#asahi · 2021-04-09 04:30
marcanI think this is also partially e12b090e's fault#asahi · 2021-04-09 04:27
marcanso it does not include the generic definitions when building the asm-offsets tool#asahi · 2021-04-09 04:26
marcanit's because powerpc has the entire asm/io.h wrapped in ifdef __KERNEL__#asahi · 2021-04-09 04:26
marcanargh, it's an asm-offsets thing, and it only happens with PM_SUSPEND I think, which is why I didn't see it in my quick ppc64 test#asahi · 2021-04-09 04:20
marcanoh, it's a .config thing apparently#asahi · 2021-04-09 04:10
marcangetting ppc64 kbuild failures again... let's see if I can repro#asahi · 2021-04-08 18:46
marcanin this case you can probably just repeat my merges (for the conflict, delete the entire conflict section, it was a removal) and verify that the result is the same, or something like that#asahi · 2021-04-08 18:45
marcanright#asahi · 2021-04-08 16:09
marcananyway, I'll get dinner too :) (shows you how offset I am...)#asahi · 2021-04-08 16:05
marcanfor future reference, would using v5.12-rc3 have been preferable?#asahi · 2021-04-08 16:04
marcanboth sounded vaguely wrong to me :-)#asahi · 2021-04-08 16:04
marcanright#asahi · 2021-04-08 16:04
marcanso using that as a base should work