#asahi-re · 2022-04-08 06:52
rqou_fwiw bluetooth has 2 BARs so it's not "simple"#asahi-re · 2022-04-08 06:50
rqou_i'm guessing wifi was made working without involving much tracing?#asahi-re · 2022-04-08 06:49
rqou_i'm still not super familiar with all of the trace tooling#asahi-re · 2022-04-08 06:49
rqou_so what do i have to put into the HV to trace it?#asahi-re · 2022-04-08 06:48
svenmaybe that bluetooth function just exposes a simple UART over pcie :D#asahi-re · 2022-04-08 06:48
rqou_afaict d221ap/iphoneX/A11 has bluetooth on a uart while d331ap/iphoneXS/A12 has it on pcie#asahi-re · 2022-04-08 06:46
rqou_interestingly, bluetooth has a compatible of "bluetooth,n88" in the m1 adt, but there exists idevice DTs that have "bluetooth,n88" attached to a UART#asahi-re · 2022-04-08 06:46
svenbluetooth will essentially just be a MMIO region inside the pcie one#asahi-re · 2022-04-08 06:39
jannauHCI is just host controller interface. it's just a name from the standards. sata as ahci, nvme has nvmhci, firewire has one too#asahi-re · 2022-04-08 06:35
jannauchadmed: macos ioreg doesn't look like there's usb involved for bt#asahi-re · 2022-04-08 06:32
rqou_i usually see that explicitly specified as {x|E|U|O}HCI though#asahi-re · 2022-04-08 06:32
rqou_oh right lol#asahi-re · 2022-04-08 06:31
Jamie[m]1isn't usb over pcie also called HCI? :P#asahi-re · 2022-04-08 06:31
rqou_HCI over USB over PCIe also seems pretty reasonable tbh#asahi-re · 2022-04-08 06:30
rqou_i didn't see any references to USB#asahi-re · 2022-04-08 06:30
rqou_chadmed: are you sure?#asahi-re · 2022-04-08 06:30
rqou_there's a pcie driver, but all it seems to do is issue function level resets#asahi-re · 2022-04-08 06:29
rqou_and i definitely never figured out what driver is even involved in macos, since the macos stack seems really overcomplicated#asahi-re · 2022-04-08 06:29
rqou_i have no idea how to trace PCIe though#asahi-re · 2022-04-08 06:28
rqou_yes, function 1 vs function 0#asahi-re · 2022-04-08 06:28
chadmedi would imagine so but dont quote me on that#asahi-re · 2022-04-08 06:26
Jamie[m]1is it in the same broadcom chip as the wifi?#asahi-re · 2022-04-08 06:25
chadmedbut i dont think anyone figured anything out beyond that information #asahi-re · 2022-04-08 06:25
chadmedits like HCI over USB over PCIe or something silly like that#asahi-re · 2022-04-08 06:25
jannauI'm not aware of anyone looking into bluetooth#asahi-re · 2022-04-08 06:21
rqou_i assume it's been made way more complicated than just HCI over PCIe?#asahi-re · 2022-04-08 06:20
rqou_has anyone taken any kind of look at bluetooth?#asahi-re · 2022-04-07 19:00
rqou_the scaler and prores blocks have a _huge_ pile of dart tunables#asahi-re · 2022-04-07 18:58
rqou_that's why my current guess is that it might be something about caching or something about transaction queue limits#asahi-re · 2022-04-07 18:57
svenas long as ignoring it works 🤷🏻♂️#asahi-re · 2022-04-07 18:57
rqou_but i don't know what that is either lol#asahi-re · 2022-04-07 18:57
rqou_the tunables code has a debug error message mentioning something about "client partitions"#asahi-re · 2022-04-07 18:57
sven*or t6000#asahi-re · 2022-04-07 18:56
svenyeah, there’s no need for any dart tunables on t8103 or the either#asahi-re · 2022-04-07 18:53
rqou_btw for reference: not applying the t8110 DART tunables still seems to work (consequences not yet determined), but not applying the prores tunables seems to make the block completely stop working#asahi-re · 2022-04-07 18:52
rqou_feels like it#asahi-re · 2022-04-07 18:51
svensounds like they intend to keep that version for a while then#asahi-re · 2022-04-07 18:47
rqou_bits [15:8] and [7:0] for major/minor#asahi-re · 2022-04-07 18:46
rqou_it's found at register +0x8#asahi-re · 2022-04-07 18:46
jannauversion is unfortunately not annotated in the ADT#asahi-re · 2022-04-07 18:41
rqou_t8110 also seems to have settings related to controlling or otherwise affecting the number of outstanding transactions, but i don't understand anything about that either#asahi-re · 2022-04-07 18:39
rqou_e.g. "STT" "CTC" "STC"#asahi-re · 2022-04-07 18:39
svenfun!#asahi-re · 2022-04-07 18:38
rqou_e.g. invalidating caches with an address range, extra page table levels, and a bunch of other acronym soup i don't understand#asahi-re · 2022-04-07 18:36
rqou_version 2 supports a lot more features#asahi-re · 2022-04-07 18:36
rqou_yes, what i mean is that the m2 might have t8110 darts with version 2 instead of version 1#asahi-re · 2022-04-07 18:36
jannaunot sure about that, t8110 is used in a15 so it will probably be in m2 (if that doesn't skip a Axy generation)#asahi-re · 2022-04-07 18:33
svenyeah#asahi-re · 2022-04-07 18:33
rqou_the m2 might be much more "exciting"#asahi-re · 2022-04-07 18:33
rqou_the t8110 dart in the t600x feels like a weird transitionary implementation with most of the features missing