00:06
Wizzup has quit [Ping timeout: 240 seconds]
02:23
megi has quit [Ping timeout: 276 seconds]
04:02
_whitelogger has joined #lima
05:11
_whitelogger has joined #lima
05:41
<
bshah|matrix >
Ah right. I'll be able to test this next Monday
05:41
<
bshah|matrix >
Right now I don't have access to device
05:42
<
anarsoul >
I guess it can wait till Monday :)
05:42
<
anarsoul >
I've been running q3a demo in a loop for more than an hour and so far no pp errors
05:43
<
bshah|matrix >
ping MartijnBraam ^^^ if you want to test this out
05:43
<
anarsoul >
it fixes cursor lag in X11
06:22
monstr has joined #lima
06:25
<
anarsoul >
well, I got ppmmu fault on read eventually
06:25
<
anarsoul >
I'll check if I get the same without my MR
06:58
jernej has joined #lima
06:59
jernej has quit [Read error: Connection reset by peer]
07:05
jernej has joined #lima
07:07
adjtm has quit [Read error: Connection reset by peer]
07:07
adjtm has joined #lima
07:12
adjtm_ has joined #lima
07:12
adjtm has quit [Read error: Connection reset by peer]
07:24
<
Danct12[m] >
<bshah|matrix "ping MartijnBraam ^^^ if you wan"> i can test it out
07:26
<
Danct12[m] >
will be compiling it in a few minutes
07:54
<
Tofe >
anarsoul: I'll try it yes
07:56
dddddd has quit [Remote host closed the connection]
08:00
jernej has joined #lima
08:24
jernej has joined #lima
08:37
<
rellla >
i don't know, if it's really complete, because glmark didn't return after the run.
08:38
megi has joined #lima
08:42
<
rellla >
please ping me, if you need more information
08:43
<
rellla >
btw, i haven't seen the display output, so i can not say, if it's correct.
08:45
<
rellla >
anarsoul: and yes, i can confirm now, that VS CMD is much larger than in my example yesterday :p
09:02
<
rellla >
used glmark version is commit 71b3274, because since introduction of GLAD i had issues with EGL
09:20
chewitt has joined #lima
09:55
BenG83 has joined #lima
10:50
jailbox has quit [Ping timeout: 265 seconds]
11:13
jailbox has joined #lima
11:25
<
MoeIcenowy >
rellla: can I download it with `wget -r` ?
11:26
<
MoeIcenowy >
oh seems that a simple sequence wget works
11:28
<
rellla >
MoeIcenowy: for /glmark you have to wait, since upload hasn't finished yet
11:30
<
MoeIcenowy >
rellla: ah I have no interest
11:30
<
MoeIcenowy >
just interested in ideas
11:31
<
MoeIcenowy >
why is the data divided to multiple segments?
11:31
<
rellla >
is the divider
11:32
<
MoeIcenowy >
are they divided by the tracker?
11:34
<
rellla >
it's one per frame
12:20
<
Danct12[m] >
<anarsoul "it fixes cursor lag in X11"> doesn't seem to fix cursor lag for me on the pinetab though
12:20
<
Danct12[m] >
still somewhat feeling it
12:23
<
asriel >
I guess the framerate should be better..?
12:23
<
asriel >
Since "optimization"
12:25
<
Danct12[m] >
I haven't tried timedemo yet but appears that quake 3 is still lagging
12:26
<
Danct12[m] >
and that only happens on x11, wlroots doing it just fine
12:26
<
Danct12[m] >
even if the patch isn't applied
12:28
paulk-leonov has quit [Ping timeout: 264 seconds]
12:33
paulk-leonov has joined #lima
12:39
jernej has quit [Ping timeout: 246 seconds]
12:45
jernej has joined #lima
12:48
wiewo has quit [Ping timeout: 250 seconds]
12:51
wiewo has joined #lima
12:55
Barada has quit [Quit: Barada]
13:02
wiewo has quit [Ping timeout: 252 seconds]
13:08
wiewo has joined #lima
13:14
dddddd has joined #lima
13:58
monstr has quit [Remote host closed the connection]
14:09
enunes has joined #lima
15:13
<
anarsoul >
Danct12[m]: are you sure that you're using my branch?
15:15
<
Danct12[m] >
just the MR patch
15:34
<
anarsoul >
Danct12: you need whole branch, not one patch. There're 3 extra commits in this branch
15:37
<
Danct12[m] >
dumb little ol' me, anyway will try again and let you know
16:40
<
Danct12[m] >
anarsoul: just tried, can't tell the difference about the cursor lag
16:40
<
Danct12[m] >
and with/without the patch
16:45
<
Tofe >
on my side I'll have to rebuild the image, it'll take a bit of time
16:51
<
anarsoul >
Danct12[m]: I wonder if you have cursor plane patches in your kernel
16:56
<
anarsoul >
Danct12: I double checked, I can definitely tell the difference
16:57
<
anarsoul >
asriel: there won't be much difference in q3a since it doesn't do partial draws
16:58
<
anarsoul >
but cursor is a lot smoother and I can move terminal window without visible lag
17:03
<
anarsoul >
anyway, I just hit pp task error
17:07
<
anarsoul >
rellla: that's mali450 I guess?
17:07
<
anarsoul >
(plbu_array_address is 0x0)
17:15
<
rellla >
mali450, yes
17:32
<
anarsoul >
darn, I just have no idea why partial update may result in pp error
17:33
<
anarsoul >
looking at the pattern it's clear that only one PP did its job and it was full or almost full redraw
17:38
<
Danct12[m] >
<anarsoul "Danct12: I wonder if you have cu"> if that's the kermel, nope
17:38
<
Danct12[m] >
but would happy to patch it up
17:39
<
bshah|matrix >
Danct12: in pmOS kernel it's included
17:42
<
Danct12[m] >
<bshah|matrix "Danct12: in pmOS kernel it's inc"> ah then i have no idea
17:43
<
Danct12[m] >
on fvwm, i can move the window fine but it'll lag behind
17:43
<
anarsoul >
Danct12: try openbox
17:43
<
Danct12[m] >
in a way which shows delays
17:44
<
anarsoul >
fvwm unlikely to have a patch that submits another draw when previous is complete
17:46
<
Danct12[m] >
i can notice it a little bit, but otherwise moves perfectly fine
18:37
warpme_ has quit [Quit: Connection closed for inactivity]
20:23
<
anarsoul >
rellla: enunes: I checked ideas dump and if it has 2 shaders it sets bit 12 (1 << 12) in aux0 for first shader
20:23
<
anarsoul >
MoeIcenowy: ^^
20:40
<
anarsoul >
also aux1 doesn't have bit 13 set
21:16
<
anarsoul >
enunes: I found the bug. It's trivial one liner.
21:17
<
anarsoul >
I'll send MR tonight
21:19
<
anarsoul >
ideas now render correctly
21:20
<
anarsoul >
I have strong suspicion that it works with flush because stack ends up in cache and with more shaders we have to actually spill
23:31
<
enunes >
anarsoul: aha, nice one
23:32
<
enunes >
anarsoul: flush doesnt fix ideas actually, it was the other reported issue
23:32
<
enunes >
but it might be the same issue
23:34
<
anarsoul >
I also discovered how to change register that's used for storing gl_FragColor
23:35
<
anarsoul >
I tried to adjust some bits according to the dump and accidentally changed wrong one and got interesting effect
23:36
<
anarsoul >
try changing "render->multi_sample = 0x0000F807" to 0x1111F807 :)
23:36
<
anarsoul >
or to 0x0010F807 as I did
23:41
<
anarsoul >
also setting bit 0x1000 in aux0 has another interesting effect
23:42
<
anarsoul >
it slightly misrenders ideas when it's set
23:42
<
anarsoul >
but it yields 2 extra fps
23:42
<
anarsoul >
likely it enables some optimization
23:42
<
anarsoul >
but I have no idea when we can enable it
23:43
<
anarsoul >
blob sets this bit in ideas for some draws but I don't see the pattern