ChanServ changed the topic of #dri-devel to: <ajax> nothing involved with X should ever be unable to find a bar
yoslin has quit [Quit: WeeChat 3.6]
yoslin has joined #dri-devel
rotationmatrix has quit [Remote host closed the connection]
rotationmatrix has joined #dri-devel
Company has quit [Quit: Leaving]
reductum has joined #dri-devel
Leopold has joined #dri-devel
co1umbarius has joined #dri-devel
columbarius has quit [Ping timeout: 480 seconds]
mwk_ has joined #dri-devel
mwk has quit [Ping timeout: 480 seconds]
camus has joined #dri-devel
Jeremy_Rand_Talos has quit [Remote host closed the connection]
Jeremy_Rand_Talos has joined #dri-devel
Leopold has quit []
orbea has quit [Quit: You defeated orbea! 2383232 XP gained!]
heat has quit [Ping timeout: 480 seconds]
<kurufu>
is this a fine place to ask about src/intel/perf? Im trying to follow the perfetto instructions but it seems pps-producer with IntelDriver is finding 0 queries and aborting. Is this expected to work on DG2 or would I need to use something like the backports driver to test this, there seems to be some commits related to dg2 from a few months ago.
chip_x has joined #dri-devel
chipxxx has quit [Ping timeout: 480 seconds]
<Frogging101>
This is a good place to ask, but it's the middle of the night :)
kts has joined #dri-devel
kts has quit [Quit: Leaving]
kts has joined #dri-devel
alanc has quit [Remote host closed the connection]
alanc has joined #dri-devel
Duke`` has joined #dri-devel
orbea has joined #dri-devel
fab has joined #dri-devel
danvet has joined #dri-devel
pa- has joined #dri-devel
pa has quit [Ping timeout: 480 seconds]
pa has joined #dri-devel
pa- has quit [Ping timeout: 480 seconds]
mwk_ has quit []
mwk has joined #dri-devel
kts has quit [Quit: Leaving]
Namarrgon has quit [Ping timeout: 480 seconds]
Namarrgon has joined #dri-devel
pcercuei has joined #dri-devel
JohnnyonFlame has quit [Ping timeout: 480 seconds]
Lucretia has quit [Remote host closed the connection]
Lucretia has joined #dri-devel
fab has quit [Quit: fab]
fab has joined #dri-devel
Company has joined #dri-devel
fab has quit [Quit: fab]
fab has joined #dri-devel
fab has quit []
fab has joined #dri-devel
Haaninjo has joined #dri-devel
fab has quit [Read error: No route to host]
fab has joined #dri-devel
gouchi has joined #dri-devel
gouchi has quit [Remote host closed the connection]
tonyk has quit [Quit: Ping timeout (120 seconds)]
tonyk has joined #dri-devel
warpme___ has joined #dri-devel
heat has joined #dri-devel
<jenatali>
Frogging101: There's people from a lot of time zones here, it's not the middle of the night for everyone
pa has quit [Ping timeout: 480 seconds]
pa- has joined #dri-devel
rasterman has joined #dri-devel
<ishitatsuyuki>
also sometimes people work at the middle of night *cough*
pa has joined #dri-devel
pa- has quit [Ping timeout: 480 seconds]
pa- has joined #dri-devel
pa has quit [Ping timeout: 480 seconds]
srslypascal is now known as Guest3779
srslypascal has joined #dri-devel
srslypascal is now known as Guest3780
srslypascal has joined #dri-devel
Guest3779 has quit [Ping timeout: 480 seconds]
Guest3780 has quit [Ping timeout: 480 seconds]
FireBurn has quit [Ping timeout: 480 seconds]
<tango_>
or can't sleep
camus1 has joined #dri-devel
camus has quit [Ping timeout: 480 seconds]
<Frogging101>
Activity patterns tend to follow time zones closer to mine. I was just trying to convey that he might wait a bit for an answer :)
<Frogging101>
And I was almost as tired when I posted that as I am now. (I'm not a morning person)
<jenatali>
I think it being the weekend is probably the bigger factor. I often wake up to a lot of overnight activity
<Frogging101>
Hmm, that might be true
FireBurn has joined #dri-devel
montjoie has joined #dri-devel
<montjoie>
hello I use an old laptop and upgrading my buildroot on it, lead to a new mesa3d package which breaks "acceleration" see https://paste.debian.net/1258022/
djbw has quit [Remote host closed the connection]
<HdkR>
montjoie: You need to use the amber mesa branch for that old hardware
<montjoie>
but it state that iris should work, and it dont...
<HdkR>
Phrasing there could probably be clarified, `852GM/855GM` is too old for that I believe
<montjoie>
so either iris/crocus/i915g are unusable for me right ?
<dv_>
with tiled formats, do plane strides actually have any meaning?
<dv_>
I mean in general - I guess the details vary from vendor to vendor
digetx is now known as Guest3787
digetx has joined #dri-devel
sarnex has quit [Quit: Quit]
Guest3787 has quit [Ping timeout: 480 seconds]
<kisak>
montjoie: Intel 852GM hardware is Intel Generation 2. i915g is Intel Gen 3 (gallium variant of the driver), crocus is Intel gen 4-8 (Gallium variant of the driver). Iris is Intel Gen 8-12 gallium-based. What got retired are the "classic" mesa drivers. That included mesa/i830 (I could be wrong on the name).
sarnex has joined #dri-devel
sarnex has quit []
lemonzest has quit [Quit: WeeChat 3.6]
sarnex has joined #dri-devel
<montjoie>
kisak: so no choice, I am stuck on mesa 21.3.x
<kisak>
If you want the ... OpenGL 1.3? era hardware acceleration. Given the major feature disparity between the hardware and the expectations of modern applications, I'd consider just running the hardware with llvmpipe to render on the CPU
<montjoie>
kisak: I just want ffplay to be "normal", (=like on mesa 21.3)
FireBurn has quit [Quit: Konversation terminated!]
The_Company has joined #dri-devel
Company has quit [Ping timeout: 480 seconds]
caef^ has joined #dri-devel
heat has quit [Remote host closed the connection]
ahuillet has quit [Quit: ZNC]
slattann has joined #dri-devel
kem has quit [Ping timeout: 480 seconds]
<kurufu>
Frogging101: thanks, yea it might not have been the best time to ask. It seems i found the answer in patchwork that there are pending commits and a referenced MR in mesa for these counters i missed.
kem has joined #dri-devel
djbw has joined #dri-devel
slattann has quit []
srslypascal is now known as Guest3795
srslypascal has joined #dri-devel
Guest3795 has quit [Ping timeout: 480 seconds]
srslypascal is now known as Guest3797
srslypascal has joined #dri-devel
Guest3797 has quit [Ping timeout: 480 seconds]
srslypascal is now known as Guest3799
srslypascal has joined #dri-devel
Guest3799 has quit [Ping timeout: 480 seconds]
gouchi has joined #dri-devel
kem has quit [Ping timeout: 480 seconds]
Surkow|laptop has quit [Ping timeout: 480 seconds]
Surkow|laptop has joined #dri-devel
kem has joined #dri-devel
tarceri has quit [Ping timeout: 480 seconds]
deathmist1 has joined #dri-devel
heat has joined #dri-devel
tarceri has joined #dri-devel
sarnex has quit [Quit: Quit]
sarnex has joined #dri-devel
sravn has quit []
apinheiro has joined #dri-devel
rasterman has quit [Quit: Gettin' stinky!]
apinheiro has quit [Ping timeout: 480 seconds]
gouchi has quit [Remote host closed the connection]
danvet has quit [Ping timeout: 480 seconds]
Duke`` has quit [Ping timeout: 480 seconds]
fab has quit [Quit: fab]
srslypascal is now known as Guest3805
srslypascal has joined #dri-devel
Guest3805 has quit [Ping timeout: 480 seconds]
<illwieckz>
jannau, I improved the patches to get embree running with LuxMark (previously it only built, but was crashy), on raspberry pi running OpenCL on CPU with PoCL is 35 time faster than C++ with Embree: https://twitter.com/illwieckz/status/1584296569775353856
<illwieckz>
I wonder what you get on Apple M1
<illwieckz>
I guess that's what we get with Embree on non-Intel CPUs without things like AVX and others.
<illwieckz>
If you use my script to build luxmark3, you better want to entirely purge the workspace as I have changed things deeply (pulling in already cloned trees will not work)
srslypascal is now known as Guest3809
srslypascal has joined #dri-devel
Guest3809 has quit [Ping timeout: 480 seconds]
columbarius has joined #dri-devel
co1umbarius has quit [Ping timeout: 480 seconds]
mhenning has joined #dri-devel
<jannau>
illwieckz: using commit sha1 as branches doesn't work, replaced them with '-' for now
jewins has joined #dri-devel
co1umbarius has joined #dri-devel
columbarius has quit [Ping timeout: 480 seconds]
Haaninjo has quit [Quit: Ex-Chat]
<jannau>
embree seems to be built as debug with '-O0' by default
<illwieckz>
ah
<illwieckz>
damn
<illwieckz>
ah, wait
<illwieckz>
I added new options to the build script
<illwieckz>
you can pass --type=Release
<illwieckz>
(it takes cmake options, like Release, RelWithDebInfo, and Debug, but default it is RelWithDebInfo)
<jannau>
illwieckz: symbol selection in embree seems to be broken with Release build, tons of "embree/kernels/common/scene.cpp:626:60: error: ‘createTriangleMesh’ is not a member of ‘embree::neon’" for AVX*
<jannau>
embree repo seems to claim it supports Apple M1 on macos. I guess that is with Rosetta2
<illwieckz>
pull my repo
<illwieckz>
I forgot to push a fix for the embree patch
<illwieckz>
on thing is that luxmark is old and this is the embree I made working with old luxmark
<illwieckz>
so maybe newer embree has M1 support, but I'm basically using an embree I know to be compatible with luxmark3
<illwieckz>
don't forget to purge the embree repo or the patch will not be applied
<illwieckz>
(alternatively you can manually checkout the branch prior to patching and delete the 'patched' branch)
dakr has quit [Ping timeout: 480 seconds]
<jannau>
illwieckz: result is 213 so pocl is only 11 times faster. I still used the internal tasking system since my TBB is not compatible (probaby too new)
<illwieckz>
ah yes
<illwieckz>
I just turned on TBB on my script because it worked on the raspberry pi
<illwieckz>
but maybe on some other distros it doesn't
<illwieckz>
anyway perf is the same with or without TBB on my end
<illwieckz>
so I may disable it
dakr has joined #dri-devel
<illwieckz>
jannau, thank you for your test and the results! That means OpenCL on CPU isn't a bad idea.
<illwieckz>
It brings the portability, and outsources to the optimization to the OpenCL platform itself.