#
✕ nick:sid-linux
#asahi-dev · 2023-08-01 10:10
sid-linuxso kvm would have to emulate that
#asahi-dev · 2023-08-01 10:10
sid-linux(if accessed in the VM)
#asahi-dev · 2023-08-01 10:09
sid-linuxthanks sven ! regarding the broken virtual GIC -- yes I was looking at the kernel source code. all accesses to the interrupt ack register ICH_IAR1_EL1 are trapped
#asahi-dev · 2023-08-01 09:49
sid-linuxfigured it out -- the `33:` etc. numbers in /proc/interrupts corresponds to the logical irq number in linux
#asahi-dev · 2023-08-01 06:11
sid-linuxNow I'm trying to understand that this is something that GIC offers or is this a property of the CPU (that may or may not have GIC as apple doesn't seem to have)
#asahi-dev · 2023-08-01 06:10
sid-linuxbasically I'm trying to understand what namespace these numbers belong to. I read somewhere that aarch64 cpus have a certain number of PPIs (private peripheral interrups) and SPIs (Shared peripheral interrupts)
#asahi-dev · 2023-08-01 06:09
sid-linuxif you look at /proc/interrupts there are numbers in the first column e.g. `33:` `34:` etc.
#asahi-dev · 2023-08-01 06:03
sid-linuxcpu interrupt pins
#asahi-dev · 2023-08-01 06:02
sid-linuxSo instead of GICv2/GICv3/GICv4 etc. we have AIC2 directly connected to the interrupt pins then?
#asahi-dev · 2023-08-01 06:01
sid-linuxchadmed: thanks ! In m1n1 source code we have various exception handlers for IRQs, FIQs. How is the cpu core triggering these exception handlers then? FIQ would probably be triggered by timer interrupts but then what about IRQs
#asahi-dev · 2023-08-01 05:57
sid-linuxUltimately every interrupt controller feeds into the GIC correct?
#asahi-dev · 2023-08-01 05:50
sid-linuxwhat is the parent interrupt controller for the aic controller ? is it the gic ?
#asahi-dev · 2023-08-01 05:49
sid-linuxwe have the aic interrupt controller
#asahi-dev · 2023-08-01 05:49
sid-linuxHi ! I am looking at the apple dtsi, specifically t8112.dtsi in the asahi source
#asahi-dev · 2023-07-27 08:46
sid-linuxmystery resolved
#asahi-dev · 2023-07-27 08:46
sid-linuxfirecracker also creates a fdt in `fn create_fdt` !
#asahi-dev · 2023-07-27 08:15
sid-linuxj`ey: checking out -- thanks
#asahi-dev · 2023-07-27 08:08
sid-linuxis this correct?
#asahi-dev · 2023-07-27 08:07
sid-linuxactually wait a sec ... i think I know how -- virtio mmio passed via the kernel commandline ??
#asahi-dev · 2023-07-27 08:07
sid-linuxthere are no device trees (dtbs) anywhere. So how does the linux kernel know how to access the console and block device ?
#asahi-dev · 2023-07-27 08:05
sid-linuxin firecracker you can for instance simply provide a standard VM kernel image and initrd taken from a cloud image (e.g. ubuntu cloud image)
#asahi-dev · 2023-07-27 08:05
sid-linuxit seems to work well
#asahi-dev · 2023-07-27 08:05
sid-linuxwhen linux is running in EL2 via a normal boot, you can run VM via `firecracker` or `qemu -M virt -cpu host`
#asahi-dev · 2023-07-27 08:04
sid-linuxI have been checking out the intricate setup that is performed in EL1 by the m1n1 hypervisor in EL2. Its really awesome. However it leads me to a question...
#asahi-dev · 2023-07-25 17:53
sid-linuxmarcan: awesome :) thanks!
#asahi-dev · 2023-07-25 13:49
sid-linux(thanks for the update ! I figured out mesa corresponds to upstream https://gitlab.freedesktop.org/asahi/mesa/-/tags/asahi-20230725 but we should get PKGBUILDs.git updated too)
#asahi-dev · 2023-07-25 13:37
sid-linux`paman -Syu` just gave me a new `mesa-asahi-edge` package. Things work fine but `AsahiLinux/PKGBUILDs.git` (dev branch) seems to have been forgotten to get updated.
#asahi-dev · 2023-07-24 11:55
sid-linuxi mean D1.3.2
#asahi-dev · 2023-07-24 11:55
sid-linuxmaz: thanks -- D1.2.2 answered my question !
#asahi-dev · 2023-07-24 11:24
sid-linuxQuestion regarding `daif`. If i enter EL2 what is the daif ? (a) The daif when the core was last time in EL2 or (b) the daif of the level you entered _from_ ?
#asahi-dev · 2023-07-24 09:12
sid-linuxthanks both !
#asahi-dev · 2023-07-24 09:12
sid-linuxyes that makes sense
#asahi-dev · 2023-07-24 09:10
sid-linuxsame for the physical timer register CNTVCT_EL0 -- each core would have its own count i would guess ?
#asahi-dev · 2023-07-24 09:03
sid-linuxIn general all system registers are on a per core basis rather for the whole system as I understand, correct?
#asahi-dev · 2023-07-24 09:00
sid-linuxAn aarch64 conceptual question. System register, say, `HCR_EL2` -- is this register on a per core level or the setting is valid all the cores? My understanding is that setting HCR_EL2 is only on per core basis.
#asahi-dev · 2023-07-16 06:43
sid-linuxsid-linux: thanks for the link to the firmware overview ChaosPrincess -- very useful..
#asahi-dev · 2023-07-16 06:41
sid-linuxsorry wrong username !
#asahi-dev · 2023-07-16 06:41
sid-linuxchadmed: thanks again :)
#asahi-dev · 2023-07-15 19:31
sid-linuxin `/chosen/memory-map`
#asahi-dev · 2023-07-15 19:30
sid-linuxi guess it must be iboot because SEPFW appears in the ADT
#asahi-dev · 2023-07-15 19:29
sid-linuxin other words who adds the SEF Firmware to the memory during stage 1?
#asahi-dev · 2023-07-15 19:27
sid-linuxdoes iboot add SEP Firmware in stage 1 boot ? Or does Asahi Linux package the SEP firmware with m1n1 stage 1 during installation of Asahi Linux?
#asahi-dev · 2023-07-15 18:27
sid-linux*never overwritten by m1n1 as part of the chainloading to the stage 2 m1n1
#asahi-dev · 2023-07-15 18:27
sid-linux(figured it out also. adt is never overwritten by m1n1 and it lies at a lower address to _base where m1n1 is loaded)
#asahi-dev · 2023-07-15 18:13
sid-linuxwow -- very useful. thanks ChaosPrincess
#asahi-dev · 2023-07-15 18:11
sid-linuxif you look at chainload_image() in m1n1 src/chainload.c , there is a copying of SEPFW but no copying of the adt
#asahi-dev · 2023-07-15 18:09
sid-linuxso while the pointer to the adt is in bootargs, it has to lie outside it
#asahi-dev · 2023-07-15 18:09
sid-linuxmax bootargs size is 16K. However devtree size is 0x5C00 or ~376K
#asahi-dev · 2023-07-15 17:59
sid-linuxbut I would have thought that we treat the memory as "fresh" when m1n1 stage 2 restarts
#asahi-dev · 2023-07-15 17:58
sid-linuxaccording to my calculations ADT would be at something like 0x802_B30_000

plus ›