#
✕ nick:sven
#asahi-dev · 2022-04-15 14:36
svenlike, what am i supposed to do with that information?
#asahi-dev · 2022-04-15 14:36
svenyeah, but i don't even get the intention behind peek data
#asahi-dev · 2022-04-15 14:35
svenbut uhhh.... i don't really see where that would help at all if you don't poll those messages as well
#asahi-dev · 2022-04-15 14:35
sveni'm also very confused what mbox_client_peek_data is actually supposed to be used for. most other drivers just seem to return true if there's a pending message
#asahi-dev · 2022-04-15 14:33
svenjust sent nvme v2 https://lore.kernel.org/linux-nvme/20220415142055.30873-1-sven@svenpeter.dev/T/
#asahi-dev · 2022-04-15 14:31
sven*sigh*
#asahi-dev · 2022-04-15 12:26
svenyeah, though for us the mux will be inside the phy driver
#asahi-dev · 2022-04-14 17:34
sveniommu api it is then ;)
#asahi-dev · 2022-04-14 17:33
svenattaching a locked dart together with another one to the same domain sounds like pain
#asahi-dev · 2022-04-14 17:33
svenalso needs changes to the dart driver I think
#asahi-dev · 2022-04-14 17:32
svenyeah, I don’t think the lazy approach is a good idea
#asahi-dev · 2022-04-14 17:27
svenyeah… I got a bit too familiar with that iommu subsystem when I worked on that 4K patch I keep abandoning
#asahi-dev · 2022-04-14 17:13
svenhttps://lore.kernel.org/linux-iommu/74621c69-ef68-c12a-3770-319cb7a0db73@arm.com/
#asahi-dev · 2022-04-14 17:12
sveniirc robin commented on that in one of the 4K iommu patches I sent a while ago
#asahi-dev · 2022-04-14 17:11
svenright now it’s doing pretend dma buf exports
#asahi-dev · 2022-04-14 12:22
svenwdt=-1 should be enough
#asahi-dev · 2022-04-14 07:56
svenI’ll wait if any of those CI bots complain and submit v2 otherwise then
#asahi-dev · 2022-04-14 07:50
svennice!
#asahi-dev · 2022-04-13 17:08
svendunno if I want to keep that apple_rtkit_send_message api change, that's left from my first approach that I then discarded
#asahi-dev · 2022-04-13 17:08
svenjannau: https://github.com/AsahiLinux/linux/commit/dcc9c30aa40607a6c8762f3a6a6b7a0471ea7605 i'm a bit concerned because it compiled and worked on the first try though ;)
#asahi-dev · 2022-04-13 15:49
svenanyway, let's see if I can make rtkit play nicely with this atomic mailbox crap
#asahi-dev · 2022-04-13 15:46
svenand any further requests (i.e. "get fingerprint") will then be encrypted with a key they negotiated
#asahi-dev · 2022-04-13 15:46
sventhe sensor will likely only reply to pairing requests which expect to get data back signed by some private key hidden inside the SEP
#asahi-dev · 2022-04-13 15:45
svenif apple designed this thing well enough we can't get any unencrypted data out of it
#asahi-dev · 2022-04-13 15:45
sveni'm not sure that's even possible
#asahi-dev · 2022-04-13 12:16
sveneuhm, it has a weird name
#asahi-dev · 2022-04-13 12:15
svenI know that it's fingerprint sensor <--SPI--> CPU <--> SEP sbio endpoint but i don't know much about the protocol itself
#asahi-dev · 2022-04-13 12:15
svennote that i'm not 100% sure on everything here, I just briefly looked SEP
#asahi-dev · 2022-04-13 12:14
svenpresumably the magic keyboard one just has the same protocol but over bluetooth
#asahi-dev · 2022-04-13 12:14
svenbut it's an encrypted channel afaik
#asahi-dev · 2022-04-13 12:13
svenand the CPU talks to SEP
#asahi-dev · 2022-04-13 12:13
sventhe fingerprint talks to the CPU
#asahi-dev · 2022-04-13 12:13
svenno
#asahi-dev · 2022-04-13 12:13
svenbut it's married to the SEP and the CPU just forwards communication between them afaik
#asahi-dev · 2022-04-13 12:13
sventhe fingerprint sensor is on a SPI bus
#asahi-dev · 2022-04-13 11:00
svenand likely also v3 and possibly v4 after that ;)
#asahi-dev · 2022-04-13 10:30
svenj`ey: :D
#asahi-dev · 2022-04-13 10:27
sven:-P
#asahi-dev · 2022-04-13 10:26
svenbut right now i first want to focus on getting rtkit/nvme upstream
#asahi-dev · 2022-04-13 10:26
sveni have a wip branch for usb3 support and made very early progress with thunderbolt
#asahi-dev · 2022-04-13 06:51
sven*that
#asahi-dev · 2022-04-13 06:51
svenif that doesn’t work I can at least send out rtkit the way it is now and we can “fix” then mailbox crap once you upstream SMC
#asahi-dev · 2022-04-13 06:50
svenmarcan: yeah, I’m gonna give work queues one more try
#asahi-dev · 2022-04-13 05:17
svenI think I’m staring to understand that now
#asahi-dev · 2022-04-12 19:06
svenyeah, that’s what I expected. It’s the spinlock around that mailbox data received that would cause the schedule-while-atomic bugs
#asahi-dev · 2022-04-12 18:34
svenyeah, it’ll only do that with the atomic patches because messages are received inside a spinlock there
#asahi-dev · 2022-04-12 18:17
sven:D
#asahi-dev · 2022-04-12 18:07
svenanyway, thanks for testing!
#asahi-dev · 2022-04-12 18:07
svenso much for my plan to submit nvme v2 today
#asahi-dev · 2022-04-12 18:01
svenbut looks like i'll have to revisit that worker thread issue anyway :(

plus ›