#asahi-dev · 2021-08-28 16:56
svenurgh.. this is gonna be even more complicated now :(#asahi-dev · 2021-08-28 16:50
sven... ofc i only figure this out just after i submitted the series#asahi-dev · 2021-08-28 16:50
sventhat code only aligns allocations to a PAGE_SIZE boundary afaict#asahi-dev · 2021-08-28 16:50
svenhrm, now that i'm reading the swiotlb code again it seems that the untrusted stuff won't work correctly#asahi-dev · 2021-08-28 16:20
svenalyssa: https://github.com/torvalds/linux/blob/master/drivers/gpu/drm/drm_gem_cma_helper.c#L407 👀#asahi-dev · 2021-08-28 16:19
svenso i *think* that function actually doesn't need any change, but i'd rather wait until something runs into that and fails in an obvious way than just hoping it's correct#asahi-dev · 2021-08-28 16:02
svenhah, and the commit message that introduced (well, move it there) mentions "The comments are spot on and should be near the central API"#asahi-dev · 2021-08-28 16:01
sven"The whole dma_get_sgtable() idea is fundamentally unsafe" is already a good start, yes :D#asahi-dev · 2021-08-28 15:59
svenand then there's also https://github.com/torvalds/linux/blob/master/kernel/dma/mapping.c#L317#asahi-dev · 2021-08-28 15:57
svenrobin suggested to do it that way *shrug*#asahi-dev · 2021-08-28 15:57
svenand their pcie driver might also break because m1n1 already did the tunables#asahi-dev · 2021-08-28 15:56
sventhey support 64 streams instead of 16#asahi-dev · 2021-08-28 15:56
svensmall variant of the current one#asahi-dev · 2021-08-28 15:56
sventhe thunderbolt darts aren't even supported by my code right now#asahi-dev · 2021-08-28 15:55
svenor at least not without much effort#asahi-dev · 2021-08-28 15:55
svennope, won't work#asahi-dev · 2021-08-28 15:55
sventhe most interesting changes happen in two commits#asahi-dev · 2021-08-28 15:48
svennow.. mailbox next i guess#asahi-dev · 2021-08-28 15:47
svenalyssa: just submitted it, you should've gotten it as well :)#asahi-dev · 2021-08-28 13:59
svenwill probably send them later today and make sure to cc you. then you can even try yourself ;)#asahi-dev · 2021-08-28 13:54
sventhe changes are outside of the dart driver#asahi-dev · 2021-08-28 13:54
sveni don't see why they wouldn't#asahi-dev · 2021-08-28 13:54
svenoh, work on 4k kernels?#asahi-dev · 2021-08-28 13:54
svendo what too?#asahi-dev · 2021-08-28 10:57
sventhe good news is that nvme works just fine with 4k pages, but that was kinda expected anyway :D#asahi-dev · 2021-08-28 10:55
svenhrm, i guess i should finally submit that 4k fixes. i just noticed i've been using them for the past month or so with 4k pages and nothing broke :D#asahi-dev · 2021-08-28 10:52
svenhttps://github.com/tihmstar/img4tool#asahi-dev · 2021-08-28 10:51
svenmaybe it's a kernelcache wrapped in img4 or something like that?#asahi-dev · 2021-08-28 10:47
svenshould be dev.kc.macho.development: Mach-O 64-bit filetype=12 arm64e#asahi-dev · 2021-08-28 10:47
svenj_ey: ^--#asahi-dev · 2021-08-28 10:47
svenwhat does file ~/kernelcache report?#asahi-dev · 2021-08-28 08:43
svenespecially (well, on SMC at least) the syslog does it correctly: it gets an address and never replies.#asahi-dev · 2021-08-28 08:43
svenyeah, i dunno how the managed to mess that up#asahi-dev · 2021-08-28 08:12
sveni can just not reply and crashlogs will still work fine#asahi-dev · 2021-08-28 08:12
svenmarcan: i think that physical address reply to the crashlog request is a bug as well btw.#asahi-dev · 2021-08-26 06:52
sventl;dr: yes, it's a size#asahi-dev · 2021-08-26 06:46
svensee -re#asahi-dev · 2021-08-26 06:28
svenand yeah, if i send it to hibernate and wake it up again it just sends the same initial message#asahi-dev · 2021-08-26 06:23
sven(iova == physaddr because SART can't do any remapping)#asahi-dev · 2021-08-26 06:23
sventhat's for ANS#asahi-dev · 2021-08-26 06:23
sven> 0x10800809d68000 (TYPE=0x1, SIZE=0x8, IOVA=0x809d68000) / 0x1 (EP=0x1)#asahi-dev · 2021-08-26 06:23
sven< 0x10800000000000 (TYPE=0x1) / 0x20350000000001 (EP=0x1)#asahi-dev · 2021-08-26 06:19
svenweird#asahi-dev · 2021-08-26 06:19
svenit *asks* for one from you#asahi-dev · 2021-08-26 06:19
svenat least i don't think so#asahi-dev · 2021-08-26 06:18
svenmarcan: you just setup a new one every time you make ANS go through the hibernate/wake cycle#asahi-dev · 2021-08-25 19:11
sventhis is taken straight from the hv trace#asahi-dev · 2021-08-25 19:09
svenit's really just the common rtkit bootstrap code + asc.send(0x4e54415000013011, 0x20) and then asc.recv()#asahi-dev · 2021-08-25 19:09
svenbut i think most of what i have should be in m1n1 already