Initial M1 support merged into Linux SoC tree

Re:Initial M1 support merged into Linux SoC tree by LeonM on Friday April 09, 2021 @04:18AM (#26747956)

Why does this merge also delete all sorts of (seemingly) unrelated files? Like the 'rocket' and 'Cyclades' drivers in this case.

[ Parent | on HN ]

Re:Initial M1 support merged into Linux SoC tree by arnd on Friday April 09, 2021 @11:02AM (#26751399)

Some of the Apple M1 changes went into the tty/serial driver tree that GregKH maintains. The actual platform code requires those changes, so we based the branch on top of that, and the top-level merge commit contains all of these.

https://git.kernel.org/pub/scm/linux/kernel/git/soc/soc.git/... is the better commit to look at for the stuff that I just merged that was not already part of linux-next.

[ Parent | on HN ]

Re:Initial M1 support merged into Linux SoC tree by azernik on Friday April 09, 2021 @06:13AM (#26748662)

Lots of the ARM-related changes are cleanups/modernization of code that the Apple support uses, e.g. https://git.kernel.org/pub/scm/linux/kernel/git/soc/soc.git/...

Cyclades removal was about dead, unmaintained code: https://git.kernel.org/pub/scm/linux/kernel/git/soc/soc.git/...

This is a 60-commit branch, which had to decouple a lot of different bits of ARM-platform code. Specifically, they use a totally different boot method, have a weird and special power-management system, and other stuff I don't know. See this article: https://asahilinux.org/2021/03/progress-report-january-febru...

[ Parent | on HN ]

Re:Initial M1 support merged into Linux SoC tree by kelnos on Friday April 09, 2021 @05:30AM (#26748423)

I noticed that too. If you scroll through the entire diff, it looks like most of it is the removal of those seemingly unrelated things.

I also see some changes to what looks like common arm64 code, which isn't guarded to check for Apple or non-Apple; it seemed strange to me that those changes would be ok.

I wonder if both of those things are just changes from an unrelated merge that are showing up in this diff for no reason.

[ Parent | on HN ]

Re:Initial M1 support merged into Linux SoC tree by jychang on Friday April 09, 2021 @10:29AM (#26751007)

See here https://asahilinux.org/2021/03/progress-report-january-febru...

Mostly the section on IRQ.

[ Parent | on HN ]

Re:Initial M1 support merged into Linux SoC tree by Ericson2314 on Friday April 09, 2021 @11:01AM (#26751390)

Yeah all this is good not bad. This is an extremely high quality port in that general-purpose yak shaves were taking as necessary.

This is the exact opposite of your e.g. android phone "fork and abandon" where someone does the absolute minimal and ugliest hacks that upstream would never accept.

[ Parent | on HN ]

Re:Initial M1 support merged into Linux SoC tree by aulin on Friday April 09, 2021 @10:38AM (#26751119)

most probably the m1 stuff introduced some conflict in some dead and unmaintained driver and it was better to ditch them than work on a fix

[ Parent | on HN ]

Re:Initial M1 support merged into Linux SoC tree by duskwuff on Friday April 09, 2021 @04:05PM (#26755407)

Actually, it looks as those changes were unrelated! They were merged in from the tty/tty-next tree, which included some changes required to get the M1 UART working.

Here are the original commits:

https://git.kernel.org/pub/scm/linux/kernel/git/soc/soc.git/...

[ Parent | on HN ]