#
…

2022-07-22

00:01gladiac has quit [Quit: Ping timeout (120 seconds)]
00:02gladiac has joined #asahi-dev
00:02Catyre has joined #asahi-dev
00:03yrlf has quit [Quit: Ping timeout (120 seconds)]
00:03yrlf has joined #asahi-dev
00:04doggkruse has joined #asahi-dev
00:10Catyre has quit [Ping timeout: 480 seconds]
00:30doggkruse has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
00:31hays has quit []
00:44hays has joined #asahi-dev
01:12Catyre has joined #asahi-dev
01:27Catyre has quit [Remote host closed the connection]
01:28kaseyreed has quit [Quit: Konversation terminated!]
02:01Catyre has joined #asahi-dev
02:33jluthra has quit [Remote host closed the connection]
02:33jluthra has joined #asahi-dev
02:49PhilippvK has joined #asahi-dev
02:53phiologe has quit [Ping timeout: 480 seconds]
03:02Catyre has quit [Remote host closed the connection]
03:26pyropeter3 has joined #asahi-dev
03:27pyropeter2 has quit [Ping timeout: 480 seconds]
03:50kov has quit [Quit: Coyote finally caught me]
03:54Catyre has joined #asahi-dev
04:02Catyre has quit [Ping timeout: 480 seconds]
04:08dethklok_x has joined #asahi-dev
04:09dethklok_x has quit []
04:16skipwich has quit [Ping timeout: 480 seconds]
04:39skipwich has joined #asahi-dev
05:34Catyre has joined #asahi-dev
05:43Catyre has quit [Ping timeout: 480 seconds]
07:19MajorBiscuit has joined #asahi-dev
07:29Major_Biscuit has joined #asahi-dev
07:31MajorBiscuit has quit [Ping timeout: 480 seconds]
07:40pcc has joined #asahi-dev
07:45pccmarcan: I wonder if this is your linux tty race condition: https://michael.stapelberg.ch/posts/2021-04-27-linux-usb-virtual-serial-cdc-acm/
07:46pcc has quit [Quit: leaving]
08:19fossdd has joined #asahi-dev
09:37systwi has joined #asahi-dev
09:42the_lanetly_052 has joined #asahi-dev
10:14marcanpcc: no, it's not about tty cooking stuff, it just outright buffers/does not deliver some USB packets
10:15marcanas a workaround I duplicated the tty kick in the ACM driver with a memory barrier in between and that fixed it
10:15marcanremains to be seen where the real problem lies
10:15marcancould even be an issue in my AIC driver, but then I'd be surprised it doesn't break something else
10:16marcanI suspect it might have something to do with the workqueue getting scheduled before the buffer kick is visible to the other CPU, but I'm not sure if workqueue scheduling is supposed to implicitly have a memory barrier
10:16marcananyway, too busy with random stuff right now, I'll revisit this one later
10:24the_lanetly_052__ has joined #asahi-dev
10:25marcanactually, this might be AIC...
10:25marcanthe GIC driver has this
10:25marcan * Ensure that stores to Normal memory are visible to the
10:25marcan * other CPUs before they observe us issuing the IPI.
10:25marcan dmb(ishst);
10:26marcanmaz: I don't think we do anything in AIC to ensure memory ordering of regular stores with IPIs, we only do all the fine grained ordering of the AIC driver's internal stuff
10:26marcanthat sounds like a bug...
10:29the_lanetly_052 has quit [Ping timeout: 480 seconds]
10:31marcanah, but we do have an atomic release which shold be creating this ordering
10:31marcanso unless I got something subtly wrong there, this isn't the bug
10:32marcanback to suspecting the tty layer then
10:33mazmarcan: ISTR that we had that discussion at the time, and came to the conclusion that the atomics were enough to create the ordering.
10:34marcanyeah, that rings a bell
10:34maznote that this doesn't by itself order with the sysreg access that generates the IPI, but the control dependency should be enough here.
10:34maz(though it took me some time to convince myself)
10:35marcanthe tty layer does some stuff then queues some work in a workqueue, and doing that twice with a barrier in between seems to fix the bug. but I suspect the problem is the stuff, not the workqueue ordering now.
10:35marcandeep in the tty layer bowels there's a ton of atomics/acquire/release nonsense which smells to me like it probably has a bug somewhere
10:36marcanand it wouldn't at all be surprising if it only hits M1s with their massive reorder buffers
10:37marcanat least I have an easy reproducer now (repeatedly readmem()ing a certain size in m1n1 triggers it in seconds)
10:37marcanjust need to set up a silly chain of 3 systems when I get home to debug this (host to test, target to run kernels on, second target to run proxyclient against :))
10:37mazyup, doesn't seem unlikely. I've been chasing a similar issue with QEMU's implementation of virtio that goes in the weeds when the machine gets properly hammered.
10:38mazand only M1 triggers it.
10:41kov has joined #asahi-dev
10:41mazbut it is really hard to trigger (8 parallel VMs on a mini, and only one failure per 6 or 7 runs)
10:41marcanouch, that sounds painful
10:41marcanat least this one is easy
10:42marcanit was killing most of my hypervisor runs...
10:42mazyeah, I'm not actively looking, too many things going on ATM.
10:42marcanthe annoying thing was tracing it through the stack, because it would just hang and I was at first convinced it was my fault (e.g. something in the HV speculating into bad memory and crashing)
10:42mazhypervisors are crap. let's go shopping! :D
10:42marcanwent through the chain, HV -> usbmon -> cdc-acm driver and now I've arrived at tty as the likely culprit...
10:42kode54annoying M1 virtualization feature Apple will probably never fix
10:43kode54VMs are useless for development if you want to run from the latest beta but want to dev in a VM running stable
10:43kode54can't log into iCloud in a VM
10:43marcanbecause VMs are insecure, no real SEP/etc
10:44marcanpretty sure that's by design
10:44kode54guess I'll just stop developing for Apple computers
10:44kode54I can develop for Windows in a VM just fine
10:45marcanmacos VMs are offtopic though
10:46kode54yeah, I realized this wasn't the offtopic channel
10:51bisko has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
11:46dabaum|2er0 has quit [Ping timeout: 480 seconds]
12:00dabaum|2er0 has joined #asahi-dev
12:04ChaosPrincess has quit [Quit: WeeChat 3.5]
12:05ChaosPrincess has joined #asahi-dev
12:10dabaum|2er0 has quit [Quit: http://quassel-irc.org - Chat comfortably. Anywhere.]
12:10dabaum|2er0 has joined #asahi-dev
12:48yuyichao_ has joined #asahi-dev
14:14Manouchehri has joined #asahi-dev
14:15ManouchehriQuestion, how does / and /boot/efi get a new UUID with the installer script? I see these UUIDs are manually set, but don't match what was installed for me. https://github.com/AsahiLinux/asahi-alarm-builder/blob/main/build.sh#L14-L16
14:22the_lanetly_052__ has quit [Ping timeout: 480 seconds]
14:25bisko has joined #asahi-dev
14:32bisko has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
14:43bisko has joined #asahi-dev
14:49bisko has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
14:55doggkruse has joined #asahi-dev
14:57doggkruse_ has joined #asahi-dev
15:03doggkruse has quit [Ping timeout: 480 seconds]
15:07kaseyreed has joined #asahi-dev
15:10Gaspare has joined #asahi-dev
15:26linxz_ has joined #asahi-dev
15:27Chainsaw_ has joined #asahi-dev
15:27h_ has joined #asahi-dev
15:27hays has quit [synthon.oftc.net graviton.oftc.net]
15:27JTL has quit [synthon.oftc.net graviton.oftc.net]
15:27Chainsaw has quit [synthon.oftc.net graviton.oftc.net]
15:27djorz has quit [synthon.oftc.net graviton.oftc.net]
15:27linxz has quit [synthon.oftc.net graviton.oftc.net]
15:27Chainsaw_ is now known as Chainsaw
15:27linxz_ is now known as linxz
15:28JTL has joined #asahi-dev
15:43jluthra has quit [Remote host closed the connection]
15:43jluthra has joined #asahi-dev
15:43Gaspare has quit [Quit: Gaspare]
16:26Catyre has joined #asahi-dev
16:34Major_Biscuit has quit [Ping timeout: 480 seconds]
16:35Gaspare has joined #asahi-dev
16:37Major_Biscuit has joined #asahi-dev
16:47Catyre has quit [Remote host closed the connection]
17:02Catyre has joined #asahi-dev
17:23doggkruse has joined #asahi-dev
17:26doggkruse_ has quit [Ping timeout: 480 seconds]
17:28Major_Biscuit has quit [Ping timeout: 480 seconds]
17:51marcanManouchehri: the firstboot script changes them
17:51marcansee asahi-scripts
17:53doggkruse_ has joined #asahi-dev
17:54doggkruse has quit [Ping timeout: 480 seconds]
18:01___nick___ has joined #asahi-dev
18:01___nick___ has quit []
18:03___nick___ has joined #asahi-dev
18:18doggkruse_ has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
18:22doggkruse has joined #asahi-dev
18:34doggkruse has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
18:57AoV has joined #asahi-dev
19:20doggkruse has joined #asahi-dev
19:28doggkruse_ has joined #asahi-dev
19:30doggkruse has quit [Ping timeout: 480 seconds]
19:32djorz has joined #asahi-dev
19:45kaseyreed has quit [Quit: Konversation terminated!]
20:02Major_Biscuit has joined #asahi-dev
20:07___nick___ has quit [Ping timeout: 480 seconds]
20:25Major_Biscuit has quit [Ping timeout: 480 seconds]
21:42yamii has quit [Quit: WeeChat 3.6]
21:42yamii has joined #asahi-dev
22:30doggkruse_ has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
22:36doggkruse has joined #asahi-dev
22:40Gaspare has quit [Quit: Gaspare]
…