#
✕ nick:fl0_id
#asahi-dev · 2025-03-18 21:53
fl0_idoh this is probably the real fix for M3 adt dumping (and M4) instead of just ignoring the error. @mischa85 did you do this or just ignore the error? https://github.com/AsahiLinux/m1n1/pull/418/commits/69f3e4d24bf6c1be686fc1a94e27c5c5b4ef9df0
#asahi-dev · 2025-03-18 21:49
fl0_idsven well, it would look like that, because it just put the macros/constants back into midr.h / removed them from the file where the original author duplicated them. so the only remaining part of that is https://github.com/AsahiLinux/m1n1/pull/450/commits/fa3cd4b96fd6acaeb4eca7eeccb2e4ec595b7f24#diff-81896a3d888013ed9999381fea7e22d11b2de48149444a059a54bc37de4814e5
#asahi-dev · 2025-03-18 17:38
fl0_idsven squashed that one commit. rebase -i <3
#asahi-dev · 2025-03-18 14:03
fl0_idthat's why I wasn't sure what was going on
#asahi-dev · 2025-03-18 14:01
fl0_idnickchan ok. yeah I just checked commits just before that and those still worked and I didn't see any workflow changes
#asahi-dev · 2025-03-18 13:57
fl0_idsven sure that's no problem.
#asahi-dev · 2025-03-18 13:47
fl0_idnickchan that issue seems pretty helpful and they have a link to docs
#asahi-dev · 2025-03-18 13:47
fl0_idsven ok cool :)
#asahi-dev · 2025-03-18 13:44
fl0_idsven it should. someone on #asahi also said they would test for t6030 but they haven't been online yet again, maybe can tell them later
#asahi-dev · 2025-03-18 13:36
fl0_idthe different clock frequencies i mean
#asahi-dev · 2025-03-18 13:35
fl0_idassuming this counterscaling things is also present on m4, maybe mischa can also contribute to that
#asahi-dev · 2025-03-18 13:34
fl0_idthis is based on this: https://github.com/AsahiLinux/m1n1/pull/426 so yes it would have to be dealt with. But t6031 was also merged before doing that, so I think there's no reason not to add t6030
#asahi-dev · 2025-03-18 13:30
fl0_idthough force pushing is a good idea in this case, didn't even think of that
#asahi-dev · 2025-03-18 13:29
fl0_idnickchan ok true. Though I still think I'm first waiting for some feedback
#asahi-dev · 2025-03-18 11:02
fl0_id https://github.com/AsahiLinux/m1n1/pull/418/files#diff-b5fdbbf0714bc26b69cb2a8418c84f4ffd187f5cfb05faca3d79730e483e2dd7
#asahi-dev · 2025-03-18 11:02
fl0_idother question - this file is from a not accepted PR due to binary RE-ing of the author (on other drivers) - but as it seems to just rename mcc things to make t6031 things also apply to t6030 things, would it be acceptable to work off of it to enable mcc for t6030?
#asahi-dev · 2025-03-18 10:43
fl0_idI would just rerun it, but for obvious reasons I cannot do that
#asahi-dev · 2025-03-18 10:43
fl0_iddo you by any chance know about this rust fmt error in my PR? seems more like a CI fluke as I didn't touch any rust code.
#asahi-dev · 2025-03-18 10:42
fl0_idnickchan good to know
#asahi-dev · 2025-03-17 14:34
fl0_idI kind of think this rustfmt error is just a CI fluke or because the image used changed or sth
#asahi-dev · 2025-03-17 13:56
fl0_idonly somehow rust fmt check seems broken in CI
#asahi-dev · 2025-03-17 13:55
fl0_idapart from that I made the t6030 bringup commits a bit prettier and merged them with t6031 (like f.e. have t6030 in order before t6031) https://github.com/AsahiLinux/m1n1/pull/450
#asahi-dev · 2025-03-17 13:52
fl0_idci cargo fmt fails with "error: no such command: `fmt`" is that normal? I didn't change anythin there
#asahi-dev · 2025-03-16 17:19
fl0_idm4 chip designer be like "i have this fancy new way to do this, finally have time to do it for m4"
#asahi-dev · 2025-03-16 17:17
fl0_idmischa85 yeah just wanted to say 0x88000 is same as m3. does m3 do any tickling?
#asahi-dev · 2025-03-15 23:19
fl0_idmischa85 nice
#asahi-dev · 2025-03-15 17:35
fl0_idlol
#asahi-dev · 2025-03-15 17:30
fl0_idyeah maybe it has to be brute forced, no idea how else ppl found it
#asahi-dev · 2025-03-15 16:51
fl0_idwhoop whoop
#asahi-dev · 2025-03-15 08:53
fl0_idnickchan thx
#asahi-dev · 2025-03-14 21:35
fl0_idbut no matter
#asahi-dev · 2025-03-14 21:35
fl0_idmischa85 yeah but I assume in general, better C, architecture and stuff probably based on your github, but ofc just guessing. I feel like I wouldn't even know what to throw or would a lot more trial and error and reading up on architecture etc maybe
#asahi-dev · 2025-03-14 21:28
fl0_idsven ok yeah i can do that, I think
#asahi-dev · 2025-03-14 21:27
fl0_idand thanks
#asahi-dev · 2025-03-14 21:27
fl0_idmischa85 yeah but you probably know more in general
#asahi-dev · 2025-03-14 21:20
fl0_id(which was basically just adding the t6030 define there)
#asahi-dev · 2025-03-14 21:19
fl0_idThen could I resubmit it for merging? (a-ramses did most of that work, I just resolved the conflicts by moving macros/defines to midr.h)
#asahi-dev · 2025-03-14 21:18
fl0_idI was just comparing it to other socs and everywhere else init_common is used but not there
#asahi-dev · 2025-03-14 21:18
fl0_idmmh ok sure
#asahi-dev · 2025-03-14 21:14
fl0_idand I cannot just test t6030 as I only have t6031
#asahi-dev · 2025-03-14 21:14
fl0_iddoes someone know if the order of chicken bits usually needs to be in order? because I would like to simplify the previous (unmerged) t6030 bringup code by moving everything that is common to init_common_everest() but not all common parts right now are directly after another
#asahi-dev · 2025-03-14 21:09
fl0_idoh ok good to know, just saw that this was their commit I think
#asahi-dev · 2025-03-14 21:09
fl0_idalyssa haha
#asahi-dev · 2025-03-14 18:03
fl0_id(there's a comment saying "it is new but no idea what it does")
#asahi-dev · 2025-03-14 18:02
fl0_idif daniel berlin is here, do you have any idea what reg_set(s3_0_c15_c2_3, BIT(3)); (and the next line) does now?
#asahi-dev · 2025-03-14 18:02
fl0_idinteresting comparing existing code for t6030 to t6031. a lot is the same but some bits are not, and not sure if that actually has a reason or not lol
#asahi-dev · 2025-03-14 18:01
fl0_idah right
#asahi-dev · 2025-03-14 17:53
fl0_idif there is UNUSED(rev); (or similar) does it mean it is actually unused or shouldn't be used or what's the dea? (yes I checked where that is defined, it didn't help)
#asahi-dev · 2025-03-14 17:20
fl0_idmmh, 6031 chip id is in hv/__ini__.py on main, but not sure if that is enough for hv
#asahi-dev · 2025-03-14 17:15
fl0_idI cleaned the t6030-bringup commits up a little by moving constants to midr.h - but I couldn't test this yet and don't know if a-ramses would be fine with me continuing it. (though ofc all their commits are still by them) https://github.com/step21/m1n1/tree/t6031-bringup (I renamed it t6031-bringup as I don't have a t6030 to test)

plus ›