alyssa changed the topic of #panfrost to: Panfrost - FLOSS Mali Midgard & Bifrost - https://gitlab.freedesktop.org/panfrost - Logs https://freenode.irclog.whitequark.org/panfrost - <daniels> avoiding X is a huge feature
BenG83 has quit [Ping timeout: 252 seconds]
stikonas has quit [Remote host closed the connection]
BenG83 has joined #panfrost
somy has quit [Ping timeout: 245 seconds]
vstehle has quit [Ping timeout: 244 seconds]
NeuroScr has joined #panfrost
NeuroScr has quit [Quit: NeuroScr]
BenG83 has quit [Remote host closed the connection]
BenG83 has joined #panfrost
BenG83 has quit [Remote host closed the connection]
BenG83 has joined #panfrost
vstehle has joined #panfrost
davidlt has joined #panfrost
indy has quit [Ping timeout: 246 seconds]
adjtm has quit [Read error: Connection reset by peer]
adjtm has joined #panfrost
BenG83 has quit [Ping timeout: 244 seconds]
guillaume_g has joined #panfrost
guillaume_g has quit [Ping timeout: 245 seconds]
shadeslayer[m] has quit [Remote host closed the connection]
thefloweringash has quit [Write error: Connection reset by peer]
travankor has quit [Remote host closed the connection]
sphalerit has quit [Read error: Connection reset by peer]
guillaume_g has joined #panfrost
thefloweringash has joined #panfrost
travankor has joined #panfrost
sphalerit has joined #panfrost
stikonas has joined #panfrost
<cwabbott> robclark: no, but what it can do is create a list of huge triangles that get added to every tile
<cwabbott> as well as every power of two level inbetween
stikonas has quit [Remote host closed the connection]
<cwabbott> that's what hierarchical tiling is -- it shrinks the size of the tile list when there are big triangles
<cwabbott> the fragment backend still works on a single 16x16 tile, but it fetches the list of triangles to rasterize from the current tiles list as well as the list of triangles in a 2x2 grouping of tiles, 4x4 grouping, 8x8 grouping, etc.
pH5 has joined #panfrost
rhyskidd has quit [Quit: rhyskidd]
rhyskidd has joined #panfrost
yann has joined #panfrost
adjtm has quit [Quit: Leaving]
tgall_foo has quit [Ping timeout: 268 seconds]
_whitelogger has joined #panfrost
cwabbott has quit [Quit: cwabbott]
cwabbott has joined #panfrost
tgall_foo has joined #panfrost
davidlt has quit [Remote host closed the connection]
davidlt has joined #panfrost
memeka has joined #panfrost
<memeka> alyssa: just saw the article on phoronix :) congrats on getting mutter to work!
<memeka> Last time I checked, T628 32bit was not working, is this still the case?
<HdkR> Midgard 1st gen still isn't operational
adjtm has joined #panfrost
davidlt_ has joined #panfrost
davidlt has quit [Ping timeout: 272 seconds]
davidlt_ is now known as davidlt
<tomeu> alyssa: we have a green CI again \o/
<memeka> HdkR: what’s missing for Midgard 1st gen?
<HdkR> Someone to fix some bugs in command stream generation and maybe some SBFD fixes?
<tomeu> and most important, to keep it working
adjtm has quit [Quit: Leaving]
<alyssa> It's the keep it working that's hard..
<tomeu> alyssa: btw, have you thought about the need for a BO cache?
adjtm has joined #panfrost
<alyssa> tomeu: Somewhat.
<alyssa> For transient allocations (the day-to-day, good for one frame then tossed), it shouldn't matter, since we cache all that ourselves anyway
<alyssa> (With pools)
<alyssa> For honest-to-goodness BOs, I mean, I would hope there's not so much traffic that it shouldn't matter either but I haven't profiled it. It's low-prio I think regardless, unless you have numbers I don't
<tomeu> yeah, the problem is mainly with imported BOs, as we currently create 2 or more BOs for the same kernel-side BO, and when the first gets destroyed, we also close the GEM handle
<tomeu> affecting the second allocation as well
afaerber has quit [Quit: Leaving]
<tomeu> don't know if there's any performance implications as well
pH5 has quit [Quit: bye]
afaerber has joined #panfrost
<tomeu> alyssa: and I guess I should do that before I can continue work on modifiers, because we are hitting assertions
<alyssa> :+1:
<alyssa> So.... reference counting?
<tomeu> yep
BenG83 has joined #panfrost
<alyssa> :+1:
<alyssa> Woop, got NPOT divisors working
<alyssa> Looks like there's some odd interactions between indexing and instancing
adjtm has quit [Remote host closed the connection]
guillaume_g has quit [Quit: Konversation terminated!]
yann has quit [Ping timeout: 272 seconds]
davidlt has quit [Read error: Connection reset by peer]
davidlt has joined #panfrost
_whitelogger has joined #panfrost
stikonas has joined #panfrost
davidlt has quit [Remote host closed the connection]
davidlt has joined #panfrost
stikonas has quit [Read error: Connection reset by peer]
stikonas has joined #panfrost
stikonas has quit [Ping timeout: 268 seconds]
davidlt has quit [Remote host closed the connection]
davidlt has joined #panfrost
davidlt has quit [Remote host closed the connection]
davidlt has joined #panfrost
afaerber has quit [Quit: Leaving]
NeuroScr has joined #panfrost
somy has joined #panfrost
NeuroScr has quit [Quit: NeuroScr]
somy has quit [Remote host closed the connection]
davidlt has quit [Ping timeout: 268 seconds]
somy has joined #panfrost
somy has quit [Ping timeout: 252 seconds]
<alyssa> Sweet, my instancing branch is now passing 100% of:
<alyssa> dEQP-GLES3.functional.draw.*
* alyssa crowd goes wild
<Lyude> alyssa: this is all to get g-s working right?
<alyssa> Lyude: Nah, I just like Es3 :p
* alyssa is naughty ;P
afaerber has joined #panfrost
BenG83 has quit [Read error: Connection reset by peer]
rhyskidd has quit [Ping timeout: 245 seconds]
adjtm has joined #panfrost