<FLHerne>
which changes the soname and vendor strings to not conflict with current Mesa releases
<FLHerne>
and disables drivers that are provided by current Mesa
fab has quit [Ping timeout: 480 seconds]
aravind has quit [Ping timeout: 480 seconds]
danvet has joined #dri-devel
MajorBiscuit has joined #dri-devel
mclasen_ has joined #dri-devel
gawin has quit [Ping timeout: 480 seconds]
itoral has quit []
mclasen has quit [Ping timeout: 480 seconds]
heat_ has joined #dri-devel
danvet has quit [Ping timeout: 480 seconds]
gawin has joined #dri-devel
nchery has joined #dri-devel
eukara has quit [Remote host closed the connection]
cengiz_io has joined #dri-devel
Guest6943 has quit []
DragoonAethis has joined #dri-devel
oneforall2 has quit [Remote host closed the connection]
jewins has joined #dri-devel
sdutt has joined #dri-devel
oneforall2 has joined #dri-devel
lkw has joined #dri-devel
<jekstrand>
bl4ckb0ne: I said that was theoretically possible. I don't think the current dmabuf APIs allow for it, though.
<bl4ckb0ne>
you also mentionned blits, would it be doable to do cubemap blit to multiple dmabuf texture, send that and blit them into a cubemap on the other process
<jekstrand>
Yes
<bl4ckb0ne>
gonna do the second one as a poc and see what i can do to ge the first one done, thanks
<bl4ckb0ne>
what would be a good starting point? extending the eglcreateimage spec?
<jekstrand>
Not sure. It's going to require tweaking modifiers for any drivers that support arrays (or cubes) to also exactly specify array layout.
<bl4ckb0ne>
the array layout is how the cubemap is stored in the buffer?
<jekstrand>
Yes. There's some computation necessary to figure out the stirde between planes and we need to ensure that's the same on both ends.
<jekstrand>
s/planes/array slices/
<bl4ckb0ne>
sounds like a lot of fun ahead, ill start with opening an issue on the egl registry issue tracker to see where it goes
<bl4ckb0ne>
thanks!
kts has quit [Ping timeout: 480 seconds]
fxkamd has quit [Remote host closed the connection]
fxkamd has joined #dri-devel
danvet has joined #dri-devel
Jeremy_Rand_Talos_ has quit [Remote host closed the connection]
<jekstrand>
bl4ckb0ne: Ok, that sounds like as good a start as any.
kts has joined #dri-devel
danvet has quit [Ping timeout: 480 seconds]
kts has quit []
fxkamd has quit []
kts has joined #dri-devel
gawin has quit [Remote host closed the connection]
fab has joined #dri-devel
Jeremy_Rand_Talos has joined #dri-devel
fab has quit [Quit: fab]
fahien has joined #dri-devel
nehsou^ has quit [Ping timeout: 480 seconds]
aravind has joined #dri-devel
oneforall2 has quit [Remote host closed the connection]
saurabhg has joined #dri-devel
oneforall2 has joined #dri-devel
anholt has quit [Remote host closed the connection]
gawin has joined #dri-devel
ybogdano has joined #dri-devel
lkw has quit [Ping timeout: 480 seconds]
oneforall2 has quit [Remote host closed the connection]
lkw has joined #dri-devel
alanc has quit [Remote host closed the connection]
alanc has joined #dri-devel
anholt has joined #dri-devel
oneforall2 has joined #dri-devel
lkw has quit [Ping timeout: 480 seconds]
oneforall2 has quit [Remote host closed the connection]
lkw has joined #dri-devel
aravind has quit [Ping timeout: 480 seconds]
mclasen_ has quit [Ping timeout: 480 seconds]
nicholasbishop has joined #dri-devel
<robclark>
kusma: maybe someone should actually create/push a branch called "amber".. not having that seems to cause confusion
iive has joined #dri-devel
Jeremy_Rand_Talos_ has joined #dri-devel
Jeremy_Rand_Talos has quit [Remote host closed the connection]
<sravn>
airlied: Any opinion on the openchrome stuff? The current driver targets x86-openchrome and aims to add 2D and maybe later some 3D acceleration. tzimmermann (offline atm) and I pushes for a driver that uses dumb buffers and thus a more generic approach
<sravn>
I know next to nothing on the user space side of things, but expects that we can have a generic openchrome driver that just-works
<sravn>
And it seems most beneficial from a long time perspective to have a more minimal driver that works, so we have something we can maintain the day the current maintainer looses interest (in case that happens)
<sravn>
But on the other hand, I feel bad trying to stop the initiative to add 2D and 3D acceleration as I do not see it adding complexity to anythig else
oneforall2 has quit [Remote host closed the connection]
jeeeun841 has joined #dri-devel
gawin has joined #dri-devel
frankbinns has quit [Remote host closed the connection]
oneforall2 has joined #dri-devel
eukara has joined #dri-devel
Duke`` has joined #dri-devel
tursulin has quit [Ping timeout: 480 seconds]
stuart has joined #dri-devel
ngcortes has joined #dri-devel
gawin has quit [Ping timeout: 480 seconds]
MajorBiscuit has joined #dri-devel
oneforall2 has quit [Remote host closed the connection]
oneforall2 has joined #dri-devel
nehsou^ has joined #dri-devel
danvet has joined #dri-devel
fahien has quit [Ping timeout: 480 seconds]
mclasen_ has joined #dri-devel
mclasen has quit [Ping timeout: 480 seconds]
rasterman has quit [Quit: Gettin' stinky!]
danvet has quit [Ping timeout: 480 seconds]
gawin has joined #dri-devel
Duke`` has quit [Read error: Network is unreachable]
danvet has joined #dri-devel
ybogdano has quit [Ping timeout: 480 seconds]
gouchi has quit [Remote host closed the connection]
oneforall2 has quit [Remote host closed the connection]
fab has joined #dri-devel
kts has quit [Ping timeout: 480 seconds]
heat_ has quit [Remote host closed the connection]
heat_ has joined #dri-devel
fab has quit [Quit: fab]
fab has joined #dri-devel
frieder has quit [Remote host closed the connection]
oneforall2 has joined #dri-devel
fab has quit [Ping timeout: 480 seconds]
danvet has quit [Ping timeout: 480 seconds]
ybogdano has joined #dri-devel
heat has joined #dri-devel
heat_ has quit [Read error: No route to host]
mvlad has quit [Quit: Leaving]
nicholasbishop has quit [Ping timeout: 480 seconds]
mbrost has quit [Read error: Connection reset by peer]
<zmike>
dcbaker: !
<dcbaker>
uh oh
<zmike>
congrats
<dcbaker>
oh
<dcbaker>
thanks?
<zmike>
this was a tough one
* dcbaker
has just kept getting distracted all week :/
<zmike>
I know this feeling
<dcbaker>
if anyone ever tells me memory management is easy in C again…
<zmike>
people have told you that?
<dcbaker>
all the time
<dcbaker>
usually when criticizing my choice to use C++ for something
<zmike>
huh
<zmike>
weird
<zmike>
the go-to for c++ criticism is spaghetti code, not memory management
<zmike>
everyone should know that
<ccr>
pasta management?
<Sachiel>
memory management in C is easy
<dcbaker>
mmmm, pasta
<Sachiel>
just call malloc when you need any, and when the program ends it all gets freed
<zmike>
are you a genius?
ngcortes has quit [Ping timeout: 480 seconds]
<dcbaker>
Sachiel: ... that is the problem, yes
<ccr>
"just valgrind until there are no leaks (or you've added enough suppressions)!"
Rayyan_ has quit [Remote host closed the connection]
Rayyan has joined #dri-devel
angerctl has joined #dri-devel
Namarrgon has quit [Ping timeout: 480 seconds]
ngcortes has joined #dri-devel
ybogdano has quit [Ping timeout: 480 seconds]
MajorBiscuit has quit [Ping timeout: 480 seconds]
pcercuei has quit [Quit: dodo]
iive has quit [Quit: They came for me...]
gawin has quit [Quit: Konversation terminated!]
Haaninjo has quit [Quit: Ex-Chat]
JohnnyonFlame has quit [Ping timeout: 480 seconds]
<dcbaker>
valgrind (archane weapon): loose 2 sanity, and roll two dice, if the combined value is >= 10 fix two memory bugs, if it is >= 6 and < 10 fix one memory bug, if >= 4 and < 6 fix one bug, add one new memory bug, < 4 and >= 2 add two new memory bugs, < 2 you lose all sanity and are lost in time and space
<airlied>
though in the scale of fixing memory allocation, not sure C++ is much better answer :-P
<airlied>
just gives you multiple guns to shoot at yourself with
<HdkR>
Just do it the FEX way, allocate all the memory. Then you can't even run ASAN.
<HdkR>
:>
MajorBiscuit has joined #dri-devel
<zmike>
what bitness is armhf?
<Sachiel>
19 bits middle endian
<zmike>
what the
<airlied>
zmike: le
<zmike>
no I mean 32 or 64
<HdkR>
32
<zmike>
hmmm
<airlied>
ah sorry, yeah 32le
<zmike>
well that sucks, I think there's a vk header bug
<HdkR>
nice
<zmike>
dcbaker: can we get a meson/mesa build macro for determining 32/64
<zmike>
or is there one already
<zmike>
really frustrating to not have something reliable and universal for this
<dcbaker>
what the bitness of the platform is?
<zmike>
yeah
<HdkR>
haha, I was asking about the same thing a couple of days ago
<airlied>
just stick a #define sizeof(void *) == 8 somewhere
<airlied>
zmike: new vk header bug or old one?
<zmike>
new
<dcbaker>
zmike: that seems like a pretty reasonable thing for us to provide, though we'll have to do a compiler check probably, since you can run 32bit python on a 64bit OS... hmmm
<zmike>
or...I guess it's possible that struct packing could just work totally differently on this arch?
<airlied>
yeah I don't think it needs meson, just util/compiler.h or someat
<HdkR>
Struct packing can change depending on arch, regardless of bitness
<zmike>
ugh
jkrzyszt has quit [Ping timeout: 480 seconds]
<HdkR>
quite a few 64-bit things land on Itanium packing but not quite
<airlied>
32-bit packing is the wierd one around alignment of 64-bit
<airlied>
always be padding the 64-bits manually
nicholasbishop has joined #dri-devel
MajorBiscuit has quit [Ping timeout: 480 seconds]
<zmike>
is this going to be the same as 32bit arm on godbolt
<airlied>
yeah should be, don't think hf matters
<HdkR>
Yea, godbolt runs armhf abi by default, but regardless of which arm ABI you choose the struct packing is the same
<HdkR>
(soft, softfp, hardfp)
<HdkR>
-mfloat-abi={soft,softfp,hard}
ybogdano has joined #dri-devel
nicholasbishop has quit [Ping timeout: 480 seconds]
oneforall2 has quit [Remote host closed the connection]
<mangix>
jimjams: my evil plan is to flash an EDID on my steam deck with fixed color values. Surprisingly, the stock EDID has all zeroes.
oneforall2 has joined #dri-devel
<mangix>
second evil idea: connect my dual link DVI monitor to an active displayport adapter to avoid driver problems with high pixel clock resolutions (overclocked)
<jimjams>
Mangix: that's awesome, would love to know how that goes (and if you have any cool tricks for getting a good color calibration)
<mangix>
The values I derived are: Color Characteristics: Red : 0.6064, 0.3447 Green: 0.3427, 0.5644 Blue : 0.1630, 0.1201 White: 0.2988, 0.3056 but I'm not sure they're correct.
<mangix>
That thread has instructions to use reshade, but it's probably better to edit the EDID so that it works everywhere (I hope?)
<mangix>
In the case of color values of all zeroes, I'm assuming some default color profile gets used. I have no idea where to look for such a thing.
oneforall2 has quit [Remote host closed the connection]