alyssa changed the topic of #panfrost to: Panfrost - FLOSS Mali Midgard & Bifrost - Logs https://freenode.irclog.whitequark.org/panfrost - <daniels> avoiding X is a huge feature
stikonas has quit [Remote host closed the connection]
janrinze has quit [Remote host closed the connection]
mixfix41 has quit [Ping timeout: 264 seconds]
mixfix41_ has joined #panfrost
mixfix41 has joined #panfrost
vstehle has quit [Ping timeout: 258 seconds]
Cruft has quit [Ping timeout: 240 seconds]
Cruft has joined #panfrost
Stary has quit [Ping timeout: 246 seconds]
Stary has joined #panfrost
davidlt has joined #panfrost
kaspter has quit [Quit: kaspter]
kaspter has joined #panfrost
Cruft has quit [Ping timeout: 256 seconds]
Cruft has joined #panfrost
vstehle has joined #panfrost
buzzmarshall has quit [Remote host closed the connection]
_whitelogger has joined #panfrost
nerdboy has joined #panfrost
rhyskidd has joined #panfrost
NeuroScr has quit [Quit: NeuroScr]
Cruft has quit [Quit: Leaving]
mixfix41_ has quit [Quit: Lost terminal]
mixfix41_ has joined #panfrost
raster has joined #panfrost
ckeepax1 has quit [Quit: WeeChat 2.2]
stikonas has joined #panfrost
robmur01 has quit [Read error: Connection reset by peer]
robmur01_ is now known as robmur01
icecream95 has quit [Ping timeout: 256 seconds]
robmur01 has quit [Quit: robmur01]
xdarklight has quit [Quit: ZNC - http://znc.in]
xdarklight has joined #panfrost
robmur01 has joined #panfrost
<tomeu> tested here on the N2 because that's what I use for building mesa and I share the nfsroot with all my other arm64 boards, and kmscube and glmark2 seem to work to the same extent as on the g31
<tomeu> well, if one ignores the random faulting during the first few frames
<tomeu> alyssa: btw, came up with this: http://paste.debian.net/1142100/
<tomeu> alyssa: so not much, next step is to make sure that the hierarchy mask is a hierarchy mask as we understand it, and check whether the address and size are really for the heap, or the polygon list
tgall_foo has joined #panfrost
<robmur01> uncompressed compressed format... sounds so very useful :)
<tomeu> sure they only added it to aid in REing...
rhyskidd has quit [Quit: rhyskidd]
rhyskidd has joined #panfrost
<tomeu> alyssa: traced the whole of glmark2 (well, at least until it crashed in the ideas scene)
<tomeu> and got again all those 4 values
<tomeu> (00001010, 00010100, 00101000, 01010000, 10100000)
<tomeu> well, five :)
kaspter has quit [Quit: kaspter]
kaspter has joined #panfrost
robmur01_ has joined #panfrost
robmur01_1 has joined #panfrost
robmur01 has quit [Ping timeout: 260 seconds]
robmur01_1 is now known as robmur01
buzzmarshall has joined #panfrost
robmur01_ has quit [Ping timeout: 256 seconds]
tomboy64 has quit [Ping timeout: 240 seconds]
<tomeu> alyssa: ran out of time before tackling the blend descriptor, can do that tomorrow
<tomeu> pushed a rebased "bifrost" branch, seems to work
tomboy64 has joined #panfrost
<alyssa> tomeu: Not sure what to make of those values.
<alyssa> 0b1010 << l
<alyssa> for l = 0 ... 4
<alyssa> daniels: AFBU
<alyssa> :P
<HdkR> Excuse you
<alyssa> tomeu: Also - on mdg it's not draw *count* but rather *vertex* count
<alyssa> A single draw with huge #s of vertices (e.g. -bshading:model=bunny, or -brefract) will hit the highest mode
<HdkR> oh btw, I got told recently that the min/max calculation was removed with Valhall. So that no longer needs to happen there
<HdkR> :P
<alyssa> Finally :D
<Lyude> alyssa: if I get time today could you help me setup my n2?
camus1 has joined #panfrost
kaspter has quit [Ping timeout: 258 seconds]
camus1 is now known as kaspter
<raster> alyssa: https://termbin.com/1fah
<raster> th discard there (discard based on smul result before?)
<raster> that will just exit the frag shader? nothing more gets done? how is depth written? only in writeout?
mixfix41 has quit [Ping timeout: 256 seconds]
mixfix41_ has quit [Ping timeout: 256 seconds]
mixfix41 has joined #panfrost
cwabbott has quit [Ping timeout: 256 seconds]
cwabbott has joined #panfrost
<alyssa> raster: yes, yes, I don't think it is?, that I would espect
<raster> hmmm
<raster> well something is writing to z buffer after that discard - and it shouldn't
<raster> jus was making sure the dicard was doing what it seems to be
karolherbst has quit [Quit: duh 🐧]
adjtm has quit [Remote host closed the connection]
karolherbst has joined #panfrost
adjtm has joined #panfrost
<alyssa> maybe there are unknown interactions b/w discard and gl_FragDepth
<HdkR> vertex depth passthrough perhaps?
<HdkR> Does a bit of magic on discard?
Venemo has quit [Quit: ZNC 1.7.3 - https://znc.in]
Venemo has joined #panfrost
<raster> well there is definitely something...
<raster> it's not seemingly in the shader asm tho
davidlt has quit [Ping timeout: 258 seconds]
mixfix41 has quit [Quit: Lost terminal]
icecream95 has joined #panfrost
robmur01 has quit [Read error: Connection reset by peer]
robmur01 has joined #panfrost
vstehle has quit [Ping timeout: 265 seconds]
vstehle has joined #panfrost
buzzmarshall has quit [Remote host closed the connection]
buzzmarshall has joined #panfrost
NeuroScr has joined #panfrost
jernej has quit [Read error: Connection reset by peer]
jernej_ has joined #panfrost
jernej_ is now known as jernej
nerdboy has quit [Ping timeout: 264 seconds]