ChanServ changed the topic of #dri-devel to: <ajax> nothing involved with X should ever be unable to find a bar
rasterman has quit [Quit: Gettin' stinky!]
gawin has quit [Ping timeout: 480 seconds]
ybogdano has quit [Ping timeout: 480 seconds]
mclasen has quit []
mclasen has joined #dri-devel
ybogdano has joined #dri-devel
tursulin has quit [Read error: Connection reset by peer]
mclasen has quit []
mclasen has joined #dri-devel
columbarius has joined #dri-devel
co1umbarius has quit [Ping timeout: 480 seconds]
pnowack has quit [Quit: pnowack]
mbrost has quit [Ping timeout: 480 seconds]
sdutt_ has joined #dri-devel
sdutt has quit [Read error: Connection reset by peer]
mattst88 has quit [Quit: leaving]
mattst88 has joined #dri-devel
mattst88 has quit []
mattst88 has joined #dri-devel
camus has joined #dri-devel
mclasen has quit [Ping timeout: 480 seconds]
camus has quit [Remote host closed the connection]
camus has joined #dri-devel
boistordu_ex has joined #dri-devel
boistordu_old has quit [Ping timeout: 480 seconds]
dsrt^ has joined #dri-devel
dsrt^ has quit [Ping timeout: 480 seconds]
dsrt^ has joined #dri-devel
Daanct12 is now known as Danct12
Duke`` has joined #dri-devel
Company has quit [Quit: Leaving]
sdutt has joined #dri-devel
sdutt_ has quit [Ping timeout: 480 seconds]
jernej has quit [Quit: Free ZNC ~ Powered by LunarBNC: https://LunarBNC.net]
linearcannon has quit [Read error: Connection reset by peer]
linearcannon has joined #dri-devel
jernej has joined #dri-devel
jernej_ has joined #dri-devel
jernej_ has quit []
jernej_ has joined #dri-devel
jernej has quit [Ping timeout: 480 seconds]
jewins has quit [Ping timeout: 480 seconds]
camus1 has joined #dri-devel
camus has quit [Read error: Connection reset by peer]
linearcannon has quit [Remote host closed the connection]
camus1 has quit [Remote host closed the connection]
camus has joined #dri-devel
hfink_ has quit []
hfink has joined #dri-devel
itoral has joined #dri-devel
agd5f_ has joined #dri-devel
mdnavare_ has quit []
mdnavare has joined #dri-devel
alanc has quit [Remote host closed the connection]
alanc has joined #dri-devel
agd5f has quit [Ping timeout: 480 seconds]
fxkamd has quit []
danvet has joined #dri-devel
agd5f has joined #dri-devel
agd5f_ has quit [Ping timeout: 480 seconds]
f11f12 has quit [Ping timeout: 480 seconds]
dsrt^ has quit [Ping timeout: 480 seconds]
dsrt^ has joined #dri-devel
lplc has quit []
lplc has joined #dri-devel
tzimmermann has joined #dri-devel
f11f12 has joined #dri-devel
camus has quit [Remote host closed the connection]
camus has joined #dri-devel
<danvet> daniels, just realized: getfb2 doesn't pass num_planes around
<danvet> so it just fails if you have additional ccs planes or how does this work?
<danvet> or is it handle != 0
<danvet> ?
* danvet not awake and no idea
sdutt has quit [Ping timeout: 480 seconds]
jkrzyszt has joined #dri-devel
pnowack has joined #dri-devel
<danvet> daniels, also did you ever write your mythical blog about how to modifier?
tursulin has joined #dri-devel
zf` has joined #dri-devel
zf has quit [Ping timeout: 480 seconds]
<emersion> it works fine with multiple planes
<emersion> yeah need to check handles != 0
<emersion> hm, drm_info checks pitch > 0
<emersion> because it's not necessarily drm master
<danvet> emersion, I think would be good to document that all in uapi headers?
<danvet> if you feel like ...
<emersion> hm
<emersion> i think some people wanted to use pitch=0 with a real plane…
<danvet> emersion, tbh I'm with daniels comment on this
<danvet> fake pitch sounds silly
<danvet> maybe bnieuwenhuizen has an update
<emersion> we get these at a steady rate :S
<danvet> yeah I just recently shot down some nonsense from arm
<emersion> okay
<emersion> let's document this then
itoral has quit [Remote host closed the connection]
itoral has joined #dri-devel
<danvet> emersion, I commented on the etna thread too
<emersion> thx
<danvet> stride is well defined
* danvet needs a tshirt with that
<danvet> with arm the bikeshed is that they also consider stride well-defined for ccs aux planes
<danvet> but have an incompatible definition of that stride compared to drm
<danvet> and get really sad every time they screw up
boistordu has joined #dri-devel
boistordu_ex has quit [Ping timeout: 480 seconds]
rgallaispou has joined #dri-devel
itoral has quit [Read error: Connection reset by peer]
itoral has joined #dri-devel
rasterman has joined #dri-devel
anholt_ has joined #dri-devel
anholt has quit [Ping timeout: 480 seconds]
Duke`` has quit []
pcercuei has joined #dri-devel
dsrt^ has quit [Ping timeout: 480 seconds]
Duke`` has joined #dri-devel
<pq> I only now noticed: KMS property "pixel blend mode" value "None" means add, not none. O_o
<pq> by the doc
tobiasjakobi has joined #dri-devel
tobiasjakobi has quit [Remote host closed the connection]
<emersion> doesn't it force the alpha channel to 1 for all pixels?
<emersion> much like ARGB → XRGB?
<pq> emersion, the doc says the result from blend none is: fg.rgb + bg.rgb
<pq> not what I would expect: fg.rgb
<emersion> hm yeah that's pretty weird
<pq> oh but the plane_alpha is doing tricks
<emersion> that doesn't really make sense to me
<emersion> yeah there's that too…
<pq> yeah, there is no problem, I just brain fart
<emersion> btw -- if you're using AMD, don't use the overlay https://gitlab.freedesktop.org/drm/amd/-/issues/1769
<pq> heh
<pq> sounds like igt is lacking
<pq> or maybe AMD has no way to check the result?
<emersion> yeah igt tests are too easy to pass :)
<emersion> AMD has checksum support and passes the igts
<emersion> there's just one test failing, but unclear if that's about this bug
mclasen has joined #dri-devel
gawin has joined #dri-devel
soreau has quit [Ping timeout: 480 seconds]
<tzimmermann> danvet, i built linus' tree for aarch64 with allmodconfig. i cannot see any problems with drm vs dma-buf
ppascher has quit [Quit: Gateway shutdown]
JohnnyonFlame has quit [Ping timeout: 480 seconds]
ppascher has joined #dri-devel
flacks has quit [Quit: Quitter]
flacks has joined #dri-devel
<danvet> tzimmermann, thx for checking, I guess it's all good then and gregkh was simply confused
jfalempe has quit [Ping timeout: 480 seconds]
<tzimmermann> np
mbrost has joined #dri-devel
itoral has quit []
Duke`` has quit []
tarceri has quit [Remote host closed the connection]
nchery has joined #dri-devel
tarceri has joined #dri-devel
jfalempe has joined #dri-devel
gawin has quit [Ping timeout: 480 seconds]
<kisak> "Handle non-infinite wait timeouts > 49.7 days as infinite" Application: Hold my tea for me, I'll be back in a month and a half ... Driver: They're not coming back :D
alyssa has left #dri-devel [#dri-devel]
Company has joined #dri-devel
vivijim has joined #dri-devel
mbrost_ has joined #dri-devel
sneil has quit [Remote host closed the connection]
mbrost__ has joined #dri-devel
sneil has joined #dri-devel
gawin has joined #dri-devel
mbrost has quit [Ping timeout: 480 seconds]
mbrost_ has quit [Ping timeout: 480 seconds]
jfalempe has quit []
jfalempe has joined #dri-devel
sdutt has joined #dri-devel
co1umbarius has joined #dri-devel
sdutt has quit []
sdutt has joined #dri-devel
jfalempe has quit []
columbarius has quit [Ping timeout: 480 seconds]
camus1 has joined #dri-devel
soreau has joined #dri-devel
nchery has quit [Ping timeout: 480 seconds]
rando25892 has quit [Remote host closed the connection]
camus has quit [Ping timeout: 480 seconds]
nchery has joined #dri-devel
rando25892 has joined #dri-devel
jfalempe has joined #dri-devel
shadeslayer[m] is now known as shadeslayer
camus has joined #dri-devel
mbrost__ has quit []
mbrost has joined #dri-devel
shadeslayer has quit [Quit: Reconnecting]
shadeslayer has joined #dri-devel
camus1 has quit [Ping timeout: 480 seconds]
kazlaus has joined #dri-devel
<karolherbst> mhh, do we concider s-o-b tags to be acks?
columbarius has joined #dri-devel
<tzimmermann> karolherbst, if the s-o-b comes from a tree maintainer
<karolherbst> okay
co1umbarius has quit [Ping timeout: 480 seconds]
<karolherbst> but yeah, that's what I meant
cphealy has quit [Remote host closed the connection]
Duke`` has joined #dri-devel
mbrost has quit [Remote host closed the connection]
camus1 has joined #dri-devel
camus has quit [Read error: Connection reset by peer]
fxkamd has joined #dri-devel
nchery is now known as Guest5436
nchery has joined #dri-devel
Guest5436 has quit [Remote host closed the connection]
jewins has joined #dri-devel
dviola has quit [Quit: WeeChat 3.3]
Bennett has joined #dri-devel
camus has joined #dri-devel
camus1 has quit [Ping timeout: 480 seconds]
X-Scale` has joined #dri-devel
jernej_ is now known as jernej
zf` is now known as zf
tolszak has quit [Read error: No route to host]
cphealy has joined #dri-devel
X-Scale has quit [Ping timeout: 480 seconds]
<karolherbst> tzimmermann: I guess it also needs an updated MAINTAINERS file if new person appear there :/
ybogdano has quit [Remote host closed the connection]
<karolherbst> totally forgot abvout that detail
JohnnyonFlame has joined #dri-devel
dwlsalmeida has joined #dri-devel
gawin has quit [Ping timeout: 480 seconds]
markyacoub___ has quit []
<danvet> karolherbst, dim checks for either author != committer or an a-b or r-b tag
<danvet> so just plain committer for drm-misc also good enough as implied ack
markyacoub has joined #dri-devel
Daanct12 has joined #dri-devel
mbrost has joined #dri-devel
<karolherbst> danvet: right... but I am thinking here more from a gitlab workflow pov
<karolherbst> although I could add checks in my new dim commands for that
<karolherbst> danvet: also.. I am wondering if we should add the link to gitlab merge_qurests as "L:" in MAINTAINERS, any thoughts?
<karolherbst> or Q: ?
<daniels> I always thought it was more of a W
<karolherbst> well, I think for the link to the merge_request Q seems to be the perfect fit "Q: *Patchwork* web based patch tracking system site"
<karolherbst> it says patchwork, but...
Danct12 has quit [Ping timeout: 480 seconds]
<karolherbst> the issue is just, that this entire file is very centric around email/mailing lists :/
Lucretia has quit []
Daanct12 is now known as Danct12
camus1 has joined #dri-devel
<daniels> karolherbst: that was a weak joke where 'W' and 'L' are used as contractions for win/loss
<karolherbst> .......
<karolherbst> just be ashamed :p
<daniels> I just had to admit in a work channel that I used to maintain KDE for Debian; shame is beyond me
<karolherbst> :D
Lucretia has joined #dri-devel
f11f12 has quit [Quit: Leaving]
<karolherbst> I think I will write a patch to MAINTAINERS and then we can discuss it the "established" way of discussing changes :p
camus has quit [Ping timeout: 480 seconds]
<dwlsalmeida> Sending this again because apparently it did not go through the first time:
<dwlsalmeida> drm_gem_shmem_purge in the shrinker callback to get rid of the "DONTNEED" bos in this case? I wonder whether that's bad, since I noticed that GEM objects are refcounted.
<dwlsalmeida> Hello! :) I am currently working on a GEM shrinker for virtio_gpu. Calling drm_gem_shmem_purge leads to an oops, probably because virtio_gpu_object_shmem_init doesn't call drm_gem_shmem_get_pages_sgt and also dma_map_sgtable is behind an if statement in that function. I wonder whether it's ok to simply call virtio_gpu_free_object instead of
<dwlsalmeida> danvet do you happen to know anything about my question? :)
rgallaispou has quit [Remote host closed the connection]
<danvet> dwlsalmeida, sry was busy
<danvet> dwlsalmeida, so this is maybe not the answer you're looking for
<danvet> but I think before we can really ramp up more common code around shrinker, we need to sort out the locking a bit in these helpers
<danvet> or well, at least in shmem helper
<danvet> dwlsalmeida, https://dri.freedesktop.org/docs/drm/gpu/todo.html#move-buffer-object-locking-to-dma-resv-lock <- essentially doing this, but for just the shmem helpers
fxkamd has quit []
<danvet> except this isn't even close to "just" some patches, since each driver has their own locking, and unifying all the per-bo locking under dma_resv_lock is some serious refactoring
<danvet> but once we have that we can essentially write helpers for shrinker logic and all that
<danvet> because there's only 1 lock per bo involved
<danvet> and we can now trylock that
<danvet> because the thing is that you have to trylock in the shrinker or it just doesn't work
mbrost_ has joined #dri-devel
<danvet> and probably also means we need to lift the trylock into the bo release path a la ttm
<danvet> maybe share that code too
<danvet> karolherbst, imo use W: and side-step the public discussion
<danvet> there is no winning move in discussing gitlab on lkml
<danvet> plus the W: vs L: joke rocks
<danvet> dwlsalmeida, the not-so-much-refactoring approach is to invent all that locking in virtio and call the right functions in the right cases
<karolherbst> danvet: sad
<danvet> but that just makes the "move towards dma_resv_lock" refactor step another driver harder
mbrost__ has joined #dri-devel
mbrost has quit [Ping timeout: 480 seconds]
<dwlsalmeida> danvet well tbh this is my first time dabbling around DRM code, so I didn't know this was an issue. My understanding was that e.g. the panfrost implementation was something I could base myself off of. Bummer :/
<danvet> dwlsalmeida, yeah it should work
mbrost_ has quit [Ping timeout: 480 seconds]
<danvet> but I'm also not super happy about yet another driver baking in not-so-great locking
<danvet> hence the todo
<danvet> plus without that locking fixed we can't really put more of this code into helpers
<danvet> and shrinker code is tricky, it'd be good to have it in helpers
<dwlsalmeida> I see, thanks for chiming in, man.
gawin has joined #dri-devel
gouchi has joined #dri-devel
ybogdano has joined #dri-devel
X-Scale has joined #dri-devel
X-Scale` has quit [Ping timeout: 480 seconds]
<daniels> danvet: trying to think of a way to approach this by building up something that looks suspiciously similar, converging a driver at a time towards it, then flipping the big switch later on …
<danvet> daniels, copypaste
<danvet> because unless everyone is using dma_resv_lock you can't rely on it
<danvet> nor can you just take it
<danvet> so either copypaste new helpers
<danvet> or fix the locking first and then start using it
<danvet> in some cases you can't even do the _locked vs _unlocked thing and otherwise not touch drivers
<daniels> mmm
<danvet> because taking the dma_resv_lock while the driver is holding it's own per-bo lock (which they tend to nest underneath dma_resv_lock) can get you a recursion already
<danvet> or at least that's where I got stuck last time around
<danvet> maybe you can start the new helpers with something based on msm
<daniels> I think there’s something deeply stupid you can do as a bridge …
<daniels> but I need to look at the code first which is difficult from my phone :)
aperezdc has left #dri-devel [#dri-devel]
JohnnyonFlame has quit [Ping timeout: 480 seconds]
dianders has joined #dri-devel
tzimmermann has quit [Quit: Leaving]
camus has joined #dri-devel
camus1 has quit [Ping timeout: 480 seconds]
gawin has quit [Remote host closed the connection]
ybogdano has quit [Ping timeout: 480 seconds]
ybogdano has joined #dri-devel
agx has quit [Ping timeout: 480 seconds]
agx has joined #dri-devel
dsrt^ has joined #dri-devel
<danvet> daniels, I guess I'll get surprised?
ppascher has quit [Ping timeout: 480 seconds]
jfalempe has quit [Ping timeout: 480 seconds]
<evadot> so libdrm is still using wayland/ci-templates while other projects like wayland and mesa are using freedesktop/ci-templates, anyone have plan to move libdrm to use freedesktop/ci-templates ?
<evadot> I'm looking at adding FreeBSD to libdrm CI (and after mesa)
<evadot> and since some FreeBSD stuff were added to freedesktop/ci-templates I want to use them for libdrm
<evadot> but I don't know shit about gitlab CI thing and freedesktop/ci-templates isn't a drop replacement for wayland/ci-templates
camus has quit [Remote host closed the connection]
camus has joined #dri-devel
f11f12 has joined #dri-devel
lyudess has quit []
Lyude has joined #dri-devel
kazlaus has quit [Quit: Leaving]
JohnnyonFlame has joined #dri-devel
jkrzyszt has quit [Ping timeout: 480 seconds]
pcercuei has quit [Ping timeout: 480 seconds]
pcercuei has joined #dri-devel
alatiera has quit [Quit: Ping timeout (120 seconds)]
alatiera has joined #dri-devel
alatiera is now known as Guest5458
Guest5458 is now known as alatiera
Duke`` has quit [Ping timeout: 480 seconds]
nchery has quit [Quit: Leaving]
camus1 has joined #dri-devel
danvet has quit [Ping timeout: 480 seconds]
camus has quit [Ping timeout: 480 seconds]
gouchi has quit [Remote host closed the connection]
f11f12 has quit [Quit: Leaving]
pcercuei has quit [Quit: dodo]
<idr> Eh... NIR txb with bias = constant 0 is just... tex, right?
<anholt_> sounds right to me
<idr> Okay... I just saw a shader that has a small pile of those, and I thought, "Wha... ???"
<imirkin> i'd sorta figure that by now you're not as surprised by weird things shaders do...
<idr> You would think.
<idr> Every now and then I see something that makes me question my understanding of things.
<imirkin> kudos to the shader writer then
<imirkin> for reaching such a high bar of confusion