ChanServ changed the topic of #asahi-dev to: Asahi Linux: porting Linux to Apple Silicon macs | Non-development talk: #asahi | General development | GitHub: https://alx.sh/g | Wiki: https://alx.sh/w | Logs: https://alx.sh/l/asahi-dev
mniak has quit [Quit: Leaving]
darkapex1 has quit [Remote host closed the connection]
darkapex1 has joined #asahi-dev
ah- has quit [Read error: Connection reset by peer]
ah- has joined #asahi-dev
gabuscus has quit []
john-cabaj has quit [Ping timeout: 480 seconds]
pg12 has quit [Ping timeout: 480 seconds]
pg12 has joined #asahi-dev
delroth has quit [Ping timeout: 480 seconds]
pg12_ has joined #asahi-dev
pg12 has quit [Ping timeout: 480 seconds]
pg12_ is now known as pg12
crabbedhaloablut has joined #asahi-dev
delroth has joined #asahi-dev
gabuscus has joined #asahi-dev
tristan2_ has joined #asahi-dev
tristan2 has quit [Ping timeout: 480 seconds]
darkapex1 has quit [Remote host closed the connection]
HardWall has quit [Quit: This computer has gone to sleep]
kidplayer666 has quit [Quit: Connection closed for inactivity]
dylanchapell is now known as dylanchapell_
dylanchapell_ is now known as dylanchapell__
dylanchapell__ is now known as dylanchapell
AnuthaDev has quit []
i509vcb has joined #asahi-dev
cza has joined #asahi-dev
<cza>
hello, i was wondering if anyone could tell me what's the recommended way to hack on m1n1? im trying to add a few things to the hypervisor but im not sure what would be the fastest way to test out my changes. im looking into handling some missing system registers like PMCCNTR_EL0 (reading it gives me a sync abort), is there perhaps a way to handle that using the proxy? im not very well versed with aarch64 and apple platforms but i'm
<j`ey>
cza: the apple cpu's have a non archtiectural PMU, so I guess that register just isnt there
<cza>
yeah i figured that out, the esr was weird but there seems to be an apple PMU equivalent
<cza>
so i wanted to somehow work around that and return the value from one of these registers
<j`ey>
not sure if the proxy can handle that, but you could edit m1n1 itself to do it
<cza>
is there an easy way to run the new build of m1n1 as the hypervisor?
<cza>
so i can run the uefi os under the newly built hypervisor
<cza>
or do i have to put it on the efi partition for stage 1 to chainload? that sounds like it will take a lot of time to test a single build
<j`ey>
you can chainload m1n1 via the proxy, yeah
<j`ey>
m1n1 stage1 -> proxy mode -> chainload stage2 m1n1 and then load u-boot
<cza>
so the proxy mode i get by using -v boot args comes from stage 1?
<j`ey>
yeah
<cza>
alright, what im wondering about is after loading my new payload with newly built m1n1 and uboot how do i get this *whole thing* to run under the m1n1 i just chainloaded?
<cza>
sorry if i cant explain what im talking about
<cza>
but i want the newly built m1n1 to act as the hypervisor
<j`ey>
cza: take a look at the proxyclient/tools/run_guest.py and proxyclient/tools/run_guest_kernel.sh scripts
<j`ey>
as well as the wiki, it has some info about this
<cza>
im using the run_guest.py script already, but i think im just misunderstanding the boot process
<cza>
it seems to me that when i use the run_guest.py script, whatever payload i provide is what is run under the hypervisor, is that correct?
<j`ey>
yep
<cza>
so now, can the chainloaded m1n1 that's running under the hypervisor run as a hypervisor itself?
<cza>
because if not, then any exceptions from the underlying os will be caught by stage 1 m1n1
<cza>
which does not contain my changes
<j`ey>
yes it can
<j`ey>
if you use proxyclient/tools/chainload.py
<cza>
ah, that makes a lot more sense
<cza>
i knew i was doing something wrong because none of my changes were actually there
<cza>
i'll give that a go, thanks!
<j`ey>
if you see the output of the chainload, it should say something about being at EL2
<cza>
yep that's exactly what it says after it reconnects to the proxy, thank you for that!
<cza>
ill give that a go now that my code is actually being run :^)
<j`ey>
is this windows or something
<cza>
the thing im trying to run under m1n1? yep lol
<cza>
i worked on porting a bootloader to m1 a few days ago and now i wanted to see how far windows can get
<cza>
though the bootloader was light work compared to this, some minor memory related stuff + weird framebuffer format