marcan changed the topic of #asahi to: Asahi Linux: porting Linux to Apple Silicon macs | General project discussion | GitHub: https://alx.sh/g | Wiki: https://alx.sh/w | Topics: #asahi-dev #asahi-re #asahi-gpu #asahi-offtopic | Keep things on topic | Logs: https://alx.sh/l/asahi
odmir has quit [Remote host closed the connection]
odmir has quit [Remote host closed the connection]
marvin24 has quit [Ping timeout: 260 seconds]
marvin24 has quit [Ping timeout: 260 seconds]
marvin24 has joined #asahi
marvin24 has joined #asahi
Necrosporus_ has joined #asahi
Necrosporus_ has joined #asahi
Necrosporus is now known as Guest36649
Necrosporus is now known as Guest36649
Guest36649 has quit [Killed (orwell.freenode.net (Nickname regained by services))]
Necrosporus_ is now known as Necrosporus
Necrosporus_ is now known as Necrosporus
bbolli has quit [Remote host closed the connection]
bbolli has quit [Remote host closed the connection]
bbolli has joined #asahi
bbolli has joined #asahi
omoiti has quit [Ping timeout: 264 seconds]
omoiti has quit [Ping timeout: 264 seconds]
jeffmiw has joined #asahi
jeffmiw has joined #asahi
jeffmiw has quit [Ping timeout: 246 seconds]
jeffmiw has quit [Ping timeout: 246 seconds]
jeffmiw has joined #asahi
jeffmiw has joined #asahi
jeffmiw has quit [Ping timeout: 240 seconds]
jeffmiw has quit [Ping timeout: 240 seconds]
jeffmiw has joined #asahi
jeffmiw has joined #asahi
jeffmiw has quit [Ping timeout: 260 seconds]
jeffmiw has quit [Ping timeout: 260 seconds]
Necrosporus has quit [Ping timeout: 246 seconds]
Necrosporus has quit [Ping timeout: 246 seconds]
VinDuv has joined #asahi
VinDuv has joined #asahi
Necrosporus has joined #asahi
Necrosporus has joined #asahi
jeffmiw has joined #asahi
jeffmiw has joined #asahi
jeffmiw has quit [Read error: No route to host]
jeffmiw has quit [Read error: No route to host]
ephe_meral has joined #asahi
ephe_meral has joined #asahi
VinDuv has quit [Quit: Leaving.]
VinDuv has quit [Quit: Leaving.]
phiologe has quit [Ping timeout: 244 seconds]
phiologe has quit [Ping timeout: 244 seconds]
phiologe has joined #asahi
phiologe has joined #asahi
brandas has quit [Remote host closed the connection]
brandas has quit [Remote host closed the connection]
jeffmiw has joined #asahi
jeffmiw has joined #asahi
jeffmiw has quit [Ping timeout: 264 seconds]
jeffmiw has quit [Ping timeout: 264 seconds]
vimal has joined #asahi
vimal has joined #asahi
encounter[m] has quit [Quit: Idle for 30+ days]
encounter[m] has quit [Quit: Idle for 30+ days]
TellowKrinkle[m] has quit [Quit: Idle for 30+ days]
TellowKrinkle[m] has quit [Quit: Idle for 30+ days]
raster has joined #asahi
raster has joined #asahi
zkrx has quit [Ping timeout: 265 seconds]
zkrx has quit [Ping timeout: 265 seconds]
vimal has quit [Ping timeout: 244 seconds]
vimal has quit [Ping timeout: 244 seconds]
zkrx has joined #asahi
zkrx has joined #asahi
vimal has joined #asahi
vimal has joined #asahi
_whitelogger_ has quit [Remote host closed the connection]
_whitelogger_ has joined #asahi
_whitelogger_ has joined #asahi
brandas has joined #asahi
brandas has joined #asahi
rjeffman has joined #asahi
rjeffman has joined #asahi
JusticeEX has quit [Ping timeout: 246 seconds]
JusticeEX has quit [Ping timeout: 246 seconds]
Bublik has quit [Ping timeout: 245 seconds]
Bublik has quit [Ping timeout: 245 seconds]
Bublik has joined #asahi
Bublik has joined #asahi
crabbedhaloablut has quit [Ping timeout: 240 seconds]
crabbedhaloablut has quit [Ping timeout: 240 seconds]
crabbedhaloablut has joined #asahi
crabbedhaloablut has joined #asahi
crabbedhaloablut has quit [Ping timeout: 265 seconds]
crabbedhaloablut has quit [Ping timeout: 265 seconds]
crabbedhaloablut has joined #asahi
crabbedhaloablut has joined #asahi
<sven>
kettenis2: looks like the usb dart stream mapping isn't that simple. at least the event rings are translated with the DART I've also been using for device mode. the device context base array is also translated with that one, but the device contexts itself then seem to use the second DART.
<sven>
kettenis2: looks like the usb dart stream mapping isn't that simple. at least the event rings are translated with the DART I've also been using for device mode. the device context base array is also translated with that one, but the device contexts itself then seem to use the second DART.
<sven>
we probably really just want to mirror the pagetables across both DARTs
<sven>
we probably really just want to mirror the pagetables across both DARTs
alvaro_f1 has quit [Ping timeout: 240 seconds]
alvaro_f1 has quit [Ping timeout: 240 seconds]
<maz>
sven: do you have to mirror the PTs? or can you just point the two darts at the same PTs and issue TLB maintenance as a group?
<maz>
sven: do you have to mirror the PTs? or can you just point the two darts at the same PTs and issue TLB maintenance as a group?
<sven>
we can point them to the same PTs and just issue the TLB maintenance as a group
<sven>
we can point them to the same PTs and just issue the TLB maintenance as a group
alvaro_fe has joined #asahi
alvaro_fe has joined #asahi
<kettenis2>
that would simplify the binding isn't it?
<kettenis2>
that would simplify the binding isn't it?
<kettenis2>
you'd have a single node for the usb dart and a singly iommus reference in the usb controller node
<kettenis2>
you'd have a single node for the usb dart and a singly iommus reference in the usb controller node
<kettenis2>
and the dart driver would look at "regs" and notice that there are two sets of registers and decide to mirror the page tables?
<kettenis2>
and the dart driver would look at "regs" and notice that there are two sets of registers and decide to mirror the page tables?
<sven>
what's worse is that we need stream id 1 on the first and stream id 0 on the second dart
<sven>
what's worse is that we need stream id 1 on the first and stream id 0 on the second dart
<sven>
so we'd need an additional property in the dart node to configure that. and that still sounds like a hack to me tbh.
<sven>
so we'd need an additional property in the dart node to configure that. and that still sounds like a hack to me tbh.
<jn__>
what makes them "first" and "second"?
<jn__>
what makes them "first" and "second"?
<sven>
one of them starts at an earlier MMIO address than the other. but it doesn't matter really, I might've even swapped first and second there
<sven>
one of them starts at an earlier MMIO address than the other. but it doesn't matter really, I might've even swapped first and second there
<jn__>
first and second from the point of view of the dwc binding could just follow the stream IDs, i reckon
<jn__>
first and second from the point of view of the dwc binding could just follow the stream IDs, i reckon
<jn__>
rather than the order in which the IOMMUs' registers appear in MMIO
<jn__>
rather than the order in which the IOMMUs' registers appear in MMIO
maor26 has joined #asahi
maor26 has joined #asahi
JusticeEX has joined #asahi
JusticeEX has joined #asahi
<kettenis2>
correllium has a "sid-remap" option; the ADT has a "remap" option
<kettenis2>
correllium has a "sid-remap" option; the ADT has a "remap" option
<kettenis2>
and with option I mean property
<kettenis2>
and with option I mean property
<marcan>
sven: huh, xhci is split across two DARTs?
<marcan>
sven: huh, xhci is split across two DARTs?
<sven>
marcan: yuuup. great, isn't it?
<sven>
marcan: yuuup. great, isn't it?
<marcan>
butwhy.jpg
<marcan>
butwhy.jpg
<sven>
no idea. the split itself is also rather strange
<sven>
no idea. the split itself is also rather strange
<marcan>
something something virtualization?
<marcan>
something something virtualization?
<marcan>
but it still seems real odd
<marcan>
but it still seems real odd
<marcan>
like, either this is a hacky hardware thing where the controller has two bus masters and they did this for some reason
<marcan>
like, either this is a hacky hardware thing where the controller has two bus masters and they did this for some reason
<marcan>
or there is an underlying reason that that might dictate the right way to deal with it
<marcan>
or there is an underlying reason that that might dictate the right way to deal with it
<sven>
it sounds like some hack hw thing tbh
<sven>
it sounds like some hack hw thing tbh
<sven>
so there's this DCBAA thing which is really just a list of pointers to device contexts
<sven>
so there's this DCBAA thing which is really just a list of pointers to device contexts
<sven>
the list of pointers goes through the first DART, the device contexts through the second one
<sven>
the list of pointers goes through the first DART, the device contexts through the second one
<marcan>
I wrote qemu/hw/usb/hcd-xhci.c btw :)
<marcan>
I wrote qemu/hw/usb/hcd-xhci.c btw :)
<sven>
ah, nice :)
<sven>
ah, nice :)
<marcan>
but this sounds so weird
<marcan>
but this sounds so weird
<sven>
oh yes
<sven>
oh yes
<sven>
I checked it three times because I was convinced I messed something up
<sven>
I checked it three times because I was convinced I messed something up
<marcan>
context at 0x00000000ffff1000, +c0 is EP3 out isn't it?
<marcan>
context at 0x00000000ffff1000, +c0 is EP3 out isn't it?
<marcan>
is that really the device context? or some other structure?
<marcan>
is that really the device context? or some other structure?
<marcan>
since it looks like the device maxes out at ep2in
<marcan>
since it looks like the device maxes out at ep2in
<marcan>
and if it knows about endpoints it already completed some transactions, no?
<marcan>
and if it knows about endpoints it already completed some transactions, no?
<sven>
er.. wtf.. it looks like it got further this time
<sven>
er.. wtf.. it looks like it got further this time
<marcan>
bbiab, dinner and then let's see how awake I am :-)
<marcan>
bbiab, dinner and then let's see how awake I am :-)
<marcan>
(today was largely recovery day)
<marcan>
(today was largely recovery day)
<marcan>
but if it's not deterministic then that's one *real* horrible hardware implementation hack
<marcan>
but if it's not deterministic then that's one *real* horrible hardware implementation hack
<sven>
I'll try to reproduce the one where it failed earlier later
<sven>
I'll try to reproduce the one where it failed earlier later
Baughn has quit [Ping timeout: 246 seconds]
Baughn has quit [Ping timeout: 246 seconds]
jeffmiw has joined #asahi
jeffmiw has joined #asahi
mechpilotace has quit [Ping timeout: 245 seconds]
mechpilotace has quit [Ping timeout: 245 seconds]
jeffmiw has quit [Ping timeout: 240 seconds]
jeffmiw has quit [Ping timeout: 240 seconds]
<sven>
okay. weird. 0xc0 should actually not be used at all
<sven>
okay. weird. 0xc0 should actually not be used at all
mechpilotace has joined #asahi
mechpilotace has joined #asahi
omoiti has joined #asahi
omoiti has joined #asahi
Baughn has joined #asahi
Baughn has joined #asahi
omoiti has quit [Ping timeout: 245 seconds]
omoiti has quit [Ping timeout: 245 seconds]
omoiti has joined #asahi
omoiti has joined #asahi
Baughn has quit [Ping timeout: 256 seconds]
Baughn has quit [Ping timeout: 256 seconds]
bisko has joined #asahi
bisko has joined #asahi
odmir has joined #asahi
odmir has joined #asahi
macc24 has quit [Changing host]
macc24 has joined #asahi
macc24 has quit [Changing host]
macc24 has joined #asahi
ephe_meral has quit [Ping timeout: 264 seconds]
ephe_meral has quit [Ping timeout: 264 seconds]
JusticeEX has quit [Ping timeout: 260 seconds]
JusticeEX has quit [Ping timeout: 260 seconds]
<sven>
even more weird: my usb storage device seems to work with just a single dart enabled
<sven>
even more weird: my usb storage device seems to work with just a single dart enabled
Baughn has joined #asahi
Baughn has joined #asahi
VinDuv has joined #asahi
VinDuv has joined #asahi
<sven>
ugh, still not sure if this is just timing related. it tries to use the second dart directly after queueing a stop EP for slot id 1 ep idx 2 which lines up with ffff10c0
<sven>
ugh, still not sure if this is just timing related. it tries to use the second dart directly after queueing a stop EP for slot id 1 ep idx 2 which lines up with ffff10c0
<sven>
it does the same for my usbstorage device but that one takes longer to get to that point
<sven>
it does the same for my usbstorage device but that one takes longer to get to that point
Baughn has quit [Quit: ZNC 1.6.2+deb1 - http://znc.in]
Baughn has quit [Quit: ZNC 1.6.2+deb1 - http://znc.in]
dacre[m] has joined #asahi
dacre[m] has joined #asahi
krbtgt_ is now known as k|r|b|t|g|t
krbtgt_ is now known as k|r|b|t|g|t
Baughn has joined #asahi
Baughn has joined #asahi
Baughn has quit [Quit: ZNC 1.6.2+deb1 - http://znc.in]
Baughn has quit [Quit: ZNC 1.6.2+deb1 - http://znc.in]
Baughn has joined #asahi
Baughn has joined #asahi
antekghost has joined #asahi
antekghost has joined #asahi
zopieux has quit [Ping timeout: 258 seconds]
zopieux has quit [Ping timeout: 258 seconds]
zopieux has joined #asahi
zopieux has joined #asahi
antekghost has left #asahi ["WeeChat 3.1"]
antekghost has left #asahi ["WeeChat 3.1"]
sungold has joined #asahi
sungold has joined #asahi
sungold has left #asahi ["User left"]
sungold has left #asahi ["User left"]
omoiti has quit [Ping timeout: 240 seconds]
omoiti has quit [Ping timeout: 240 seconds]
raster has quit [Quit: Gettin' stinky!]
raster has quit [Quit: Gettin' stinky!]
VinDuv1 has joined #asahi
VinDuv1 has joined #asahi
VinDuv1 has quit [Read error: Connection reset by peer]
VinDuv has quit [Read error: Connection reset by peer]
VinDuv1 has quit [Read error: Connection reset by peer]
VinDuv has quit [Read error: Connection reset by peer]
VinDuv has joined #asahi
VinDuv has joined #asahi
JusticeEX has joined #asahi
JusticeEX has joined #asahi
VinDuv has quit [Quit: Leaving.]
VinDuv has quit [Quit: Leaving.]
omoiti has joined #asahi
omoiti has joined #asahi
yrlf has quit [Quit: Ping timeout (120 seconds)]
yrlf has quit [Quit: Ping timeout (120 seconds)]
yrlf has joined #asahi
yrlf has joined #asahi
ifthenelse has quit [Quit: Ping timeout (120 seconds)]
ifthenelse has quit [Quit: Ping timeout (120 seconds)]