00:48
possiblemeatball has joined #asahi-gpu
00:57
possiblemeatball has quit [Ping timeout: 480 seconds]
00:57
possiblemeatball has joined #asahi-gpu
02:04
darkapex has joined #asahi-gpu
02:45
cr1901 has quit [Read error: Connection reset by peer]
02:47
cr1901 has joined #asahi-gpu
02:48
cr1901 has quit [Remote host closed the connection]
02:50
cr1901 has joined #asahi-gpu
03:01
c10l31 has joined #asahi-gpu
03:07
c10l3 has quit [Ping timeout: 480 seconds]
03:10
pyropeter1 has joined #asahi-gpu
03:12
PyroPeter_ has quit [Ping timeout: 480 seconds]
03:23
possiblemeatball1 has joined #asahi-gpu
03:30
possiblemeatball has quit [Ping timeout: 480 seconds]
03:47
bluetail has joined #asahi-gpu
04:06
possiblemeatball1 has quit [Ping timeout: 480 seconds]
04:07
possiblemeatball has joined #asahi-gpu
06:25
possiblemeatball has quit [Quit: Quit]
06:45
mairacanal has quit [Remote host closed the connection]
07:22
hightower2 has quit [Ping timeout: 480 seconds]
07:52
Guest4330 has joined #asahi-gpu
07:57
Guest4330 has quit [Quit: Page closed]
08:36
c10l has joined #asahi-gpu
09:00
mkurz has quit [Ping timeout: 480 seconds]
09:03
cr1901_ has joined #asahi-gpu
09:03
cr1901 has quit [Read error: Connection reset by peer]
09:11
nsklaus_ has joined #asahi-gpu
09:14
Nokurn has quit [Ping timeout: 480 seconds]
09:20
Dementor has quit [Quit: Ping timeout (120 seconds)]
09:21
Dementor has joined #asahi-gpu
09:32
hightower2 has joined #asahi-gpu
11:21
flatz has joined #asahi-gpu
11:23
mairacanal has joined #asahi-gpu
11:34
i509vcb has quit [Quit: Connection closed for inactivity]
11:41
jlco has joined #asahi-gpu
11:58
flatz has joined #asahi-gpu
12:25
jlco has joined #asahi-gpu
12:31
jlco has joined #asahi-gpu
12:38
rhysmdnz has quit [Ping timeout: 480 seconds]
12:38
Guest4202 has quit [Ping timeout: 480 seconds]
14:36
jlco has quit [Remote host closed the connection]
14:38
jlco has joined #asahi-gpu
15:37
cr1901_ is now known as cr1901
15:38
jlco_ has joined #asahi-gpu
15:41
jlco has quit [Ping timeout: 480 seconds]
15:48
jlco has joined #asahi-gpu
15:52
jlco has joined #asahi-gpu
16:46
bpye has quit [Ping timeout: 480 seconds]
16:59
alyssa has joined #asahi-gpu
16:59
<
alyssa >
yay, more fails to work through
16:59
<
alyssa >
thanks dEQP-EGL
17:04
possiblemeatball has joined #asahi-gpu
17:33
i509vcb has joined #asahi-gpu
17:50
bpye has joined #asahi-gpu
17:56
<
bluetail >
alyssa was that ironic?
18:02
possiblemeatball has quit [Remote host closed the connection]
18:04
jlco_ has joined #asahi-gpu
18:07
jlco has quit [Ping timeout: 480 seconds]
18:32
<
alyssa >
ok. one regression (fixed) and one pair of tests that are failing on wayland but not x11
18:32
<
alyssa >
not Asahi's bug but probably Mesa's
18:32
<
alyssa >
we can.. probably submit CTS with X11. although it feels wrong to do so if we intend to ship wayland..
18:35
jlco has joined #asahi-gpu
18:35
<
jannau >
does the pair of failing wayland tests fail on other hw/drivers?
18:35
<
alyssa >
jannau: yes, as far as I can see all Mesa drivers
18:36
<
alyssa >
dunno about non-Mesa drivers, I don't have any to test against :~)
18:38
<
jannau >
have you typoed window_8888_colorspace_default and window_888_colorspace_default? from the description I would expect that window_8888_colorspace_default fails
18:38
<
alyssa >
I have not
18:38
<
alyssa >
8888 passes, 888 fails.
18:39
<
alyssa >
but there's no 24-bit formats supported on our hardware, so Mesa is making up an alpha channel and then ignoring it
18:51
blazra has quit [Ping timeout: 480 seconds]
18:54
jlco has quit [Ping timeout: 480 seconds]
19:03
<
jannau >
problem for fp16 seems to be that the results are clamped to [0,1]. for
_888_ alpha is 3
19:15
blazra has joined #asahi-gpu
19:22
possiblemeatball has joined #asahi-gpu
19:25
possiblemeatball has quit []
19:30
possiblemeatball has joined #asahi-gpu
19:38
<
jannau >
it's using r10g10b10a2 for window_888_ so the problem is that it reads back alpha 1.0 instead of the expected 0.0
19:39
<
alyssa >
ooh. that's spicy.
19:49
nela1 has joined #asahi-gpu
19:53
nela has quit [Ping timeout: 480 seconds]
19:53
nela1 is now known as nela
20:22
jlco has joined #asahi-gpu
20:25
enick_149 has joined #asahi-gpu
20:26
rhysmdnz has joined #asahi-gpu
20:40
jlco has joined #asahi-gpu
20:46
jlco has joined #asahi-gpu
20:48
<
jannau >
dEQP-EGL.functional.wide_color.window_888_colorspace_default seems to be simply broken for r10g10b10a2
20:49
<
jannau >
how do I submit fixes for the cts? just merge request on github?
20:50
jlco has joined #asahi-gpu
20:51
jlco has joined #asahi-gpu
20:54
<
alyssa >
jannau: for you, yes
20:55
<
alyssa >
assuming you're not a Khronos member
20:55
<
alyssa >
actually I don't know who your employer is so I don't know if you're a Khronos member
20:55
<
alyssa >
(there are a lot of "interesting" members. like IKEA.)
20:55
<
alyssa >
sorry, on the board apparently???
20:56
<
alyssa >
Anyway, non-Khronos members submit patches through the github repo, yes.
20:56
<
jannau >
my employer is not a khronos member (to my knowledge)
20:56
<
alyssa >
sounds good
20:57
<
alyssa >
so PRing the github repo and signing their silly CLA should work. never done it myself though
21:03
<
jannau >
window_fp16_default_colorspace "passes" on x11/xwayland since fp16 is not supported
21:03
simon-perretta-img has joined #asahi-gpu
21:04
<
alyssa >
"pass" or skip?
21:04
<
alyssa >
skips count for conformance :~)
21:04
<
jannau >
but isn't that test broken as well? I would expect pixel values to be clamped to [0,1]
21:04
<
alyssa >
21:04 < jannau> but isn't that test broken as well? I would expect pixel values to be clamped to [0,1]
21:05
<
alyssa >
Why do you expect that?
21:05
<
alyssa >
The whole point of the HDR formats (including fp16) is that you can go outside the [0,1] range
21:06
<
jannau >
I failed to find anything which specified that
21:15
<
jannau >
ah, floating ppoint isn't clamped unless specified
21:17
<
alyssa >
Yeah, you would need a spec cite if you wanted it to clamp :p
21:28
darkapex has quit [Remote host closed the connection]
21:29
<
sven >
lol, apparently my employer is a khronos member :D
21:35
<
sven >
no idea why, but then again they have like 400k employees so I’m sure there are some people working on stuff where that makes sense :D
22:41
kaazoo has joined #asahi-gpu
22:55
Nokurn has joined #asahi-gpu
22:55
nsklaus_ has quit [Ping timeout: 480 seconds]
23:14
alyssa has quit [Quit: leaving]
23:24
nela has quit [Ping timeout: 480 seconds]
23:25
kaazoo has quit [Quit: Leaving.]
23:54
gi0 has joined #asahi-gpu