RedW has joined #maemo-leste
cockroach has quit [Quit: leaving]
zhxt has quit [Remote host closed the connection]
Guest40 has joined #maemo-leste
Guest40 has quit [Client Quit]
xmn has quit [Quit: ZZZzzz…]
xmn has joined #maemo-leste
Danct12 has joined #maemo-leste
belcher has quit [Ping timeout: 260 seconds]
belcher has joined #maemo-leste
joerg has quit [Ping timeout: 264 seconds]
joerg has joined #maemo-leste
mardy has joined #maemo-leste
pere has quit [Ping timeout: 260 seconds]
xmn has quit [Quit: ZZZzzz…]
pere has joined #maemo-leste
RedW has quit [Quit: huh upgrades]
joerg has quit [Read error: Connection reset by peer]
joerg has joined #maemo-leste
RedW has joined #maemo-leste
Pali has joined #maemo-leste
<freemangordon> omg, SW rendering works :D
uvos has joined #maemo-leste
<uvos> freemangordon: yay :)
<uvos> freemangordon: hows performance? comperable to ddk1.9?
<uvos> (with -video-omap)
<Wizzup> freemangordon: \o/
<Wizzup> MVP ;)
<Wizzup> of the month or somethin
<Wizzup> g
<uvos> yeah this is massive boon for leste
<uvos> soon we can have just one kernel for everything
<Wizzup> mhm
xmn has joined #maemo-leste
<freemangordon> Wizzup: no idea, it is 2d only still
<freemangordon> would you recommend 2d benchmark?
<Wizzup> freemangordon: I recall one, just a minute...
<Wizzup> x11perf?
<freemangordon> ok, thanks
<freemangordon> could you run it on d4 with omap driver (and no h-d running)?
<Wizzup> yeah, sure, I can do that
<Wizzup> now?
<freemangordon> no, when you have time
<freemangordon> I want to have some values to compare to
<Wizzup> ok
<Wizzup> trying to find xperf in debian atm
<Wizzup> looks like it is x11-apps, but my d4 cannot find it
<Wizzup> ok, the repo cache was unhappy or something, installing no
<Wizzup> w
<Wizzup> any idea what in particular to benchmark?
<freemangordon> x11perf -scroll500
<Wizzup> funny, it's worse without h-d running
<Wizzup> with h-d:
<Wizzup> 600 reps @ 8.7288 msec ( 115.0/sec): Scroll 500x500 pixels
<Wizzup> without h-d:
<Wizzup> 400 reps @ 13.7097 msec ( 72.9/sec): Scroll 500x500 pixels
<uvos> im not sure how usefull xperf is
<uvos> no one draws xorgs primitives anymore
<uvos> mostly pixmaps are blited
<Wizzup> what do you suggest then?
<uvos> how fast you can blit and transform a pixmap is more relevant
<Wizzup> gtkperf?
<uvos> that would work yeah (gtk2 3 would use gl), but there was a
<uvos> nother utility for this..
<Wizzup> running gtkperf atm
<uvos> xperf copywin* actually should be ok
<Wizzup> yes that is also what I was thinking
<Wizzup> but gtkperf is running now
<uvos> also cpyplane and putimage
<Wizzup> it has a scroll test
<uvos> not sure how its implemented tho
<uvos> i would shy away from that one
<Wizzup> freemangordon: gtkperf https://dpaste.com/58WAC4Z7Q
<uvos> but yeah xperf has more relevant tests than i tought
<Wizzup> how about -comppixwin500
<uvos> yeah
<Wizzup> x11perf -comppixwin500 = 4000 trep @ 10.2475 msec ( 97.6/sec): Composite 500x500 from pixmap to window
<Wizzup> without h-d
<Wizzup> with h-d running:
<Wizzup> 4000 trep @ 10.7185 msec ( 93.3/sec): Composite 500x500 from pixmap to window
<Wizzup> freemangordon: ^^
<uvos> how dose x11perf -copypixwin500
<uvos> do
<uvos> interestingly on desktop (glamor) this perfroms the same as comp
<freemangordon> lemme check
<freemangordon> what is x11perf -rect500 ?
<freemangordon> 3600 reps @ 1.4711 msec ( 680.0/sec): Composite 500x500 from pixmap to window
<freemangordon> 3600 reps @ 1.4348 msec ( 697.0/sec): Copy 500x500 from pixmap to window
<freemangordon> but, this is without sync
<Wizzup> freemangordon: -rect500 is 10000 trep @ 2.9374 msec ( 340.0/sec): 500x500 rectangle
<Wizzup> without h-d
_inky has joined #maemo-leste
<Wizzup> so whatever you have already seems faster
<Wizzup> (by a lot)
<freemangordon> 20000 reps @ 0.4609 msec ( 2170.0/sec): 500x500 rectangle
<freemangordon> yeah
<freemangordon> and this is without pvr 2d accelleration
<Wizzup> so the code you have now, it can also render 3d in x?
<Wizzup> as in, h-d can work and stuff?)
<freemangordon> 3d is trough mesa as of now
<freemangordon> on d4 and 5.15 there is some issue with 3d
<freemangordon> hmm, why TS does not work?
<uvos> you built a kernel without touchscreen-buttons
<uvos> i presume
<uvos> wayland 3d works fine on d4
<uvos> but yeah with chromeos mesa
<uvos> on d4 has issues
<freemangordon> chromeos mesa here too
<freemangordon> yeah
<Wizzup> freemangordon: I guess you don't know for sure about n900 and you're testing on d4 for speed reasons?
<freemangordon> yes
<Wizzup> ok
<freemangordon> but, I don;t see a reason for 3d to work ATM as I still have to implemnt the code that renders PRESENT bos :)
<uvos> freemangordon: if you have a kernel without ts-buttons
<uvos> since it makes libinput ignore the other ts driver
<freemangordon> uvos: ok, thanks. I have to reboot the device after that, right?
<uvos> udevadm --trigger
<uvos> should be enough
<freemangordon> ok
<uvos> its udevadm trigger
<freemangordon> wait, I alredy have 85-input-devices.rules -> 85-input-devices.rules.leste
<uvos> you have to remove it
<freemangordon> ah
<uvos> "you have to drop ...."
<freemangordon> ah, *drop* :D
<freemangordon> I read that as 'drop there"
<uvos> ah ok
<freemangordon> ok, works now
<freemangordon> well, h/d is trough llvmpipe
<Wizzup> because of the 3d problem?
<freemangordon> mhm
xmn has quit [Ping timeout: 260 seconds]
cockroach has joined #maemo-leste
<freemangordon> Wizzup: re n900 - so far I saw no isses
<freemangordon> but until I implement EGL rendering of bo, 3d will simply not be visible
<freemangordon> unless through llvmpipe ofc
<freemangordon> cool, rotation works :)
<Wizzup> right @ n900
<freemangordon> Wizzup: https://pastebin.com/kMJjJ1CP
<freemangordon> :)
<freemangordon> 10x improvement without HW acceleration is not that bad I would say
<bencoh> indeed :)
<uvos> aslo means it will do atleast ok on n900
pere has quit [Ping timeout: 258 seconds]
<freemangordon> mhm
<uvos> since n900 is about 1/5 the speed on threaded loads
<freemangordon> yeah, but lets see how it will behave when we have both 2d and 3d, implementing ATM
<_inky> hey hey, will your gfx improvements affect pinephone?
<freemangordon> they should
<freemangordon> but we can;t be sure until we try it
DPA has quit [Ping timeout: 258 seconds]
<_inky> heh
<_inky> should i update from -devel?
xes has quit [Ping timeout: 264 seconds]
DPA has joined #maemo-leste
cockroach has quit [Quit: leaving]
DPA has quit [Ping timeout: 252 seconds]
pere has joined #maemo-leste
DPA has joined #maemo-leste
DPA has quit [Ping timeout: 252 seconds]
<lyubov> /close
lyubov has left #maemo-leste [WeeChat 3.0]
<uvos> _inky: no its not something installable yet
<_inky> okay (:
xes has joined #maemo-leste
<Wizzup> freemangordon: how will that improve pinephone? with glamor disabled you mean?
<freemangordon> no, I mean I am not using pvr2d and still have 10x faster than omap driver we use currently
<freemangordon> ah
<freemangordon> I think rendering in 3d text and such is stupid
<Wizzup> right, but pinephone uses glamor with lima
<uvos> rendering 3d text ans sutch is not supid on modern desktop hardware
<freemangordon> so? it has 4 cores that I am sure can perform 2d ops faster
<uvos> not by any strech
<freemangordon> on mobile
<uvos> Wizzup: it helps pp by avoiding bugy opengl on it
<freemangordon> this one too
<uvos> so it gets slower but is useable
<Wizzup> yeah ok
<freemangordon> is it? I remember terrible rendering artifacts last time I tried it
<Wizzup> I think for lima artifacts and stability are more of a problem than scrolling perf
<freemangordon> but thing could have improved since then, yeah
<Wizzup> and others *might* fix tha tt for us
<freemangordon> oh, it is GL, not GLES?
<Wizzup> freemangordon: yeah mali driver was better than lima in terms of stability and perf, but eh
<Wizzup> yes lima driver does both
<freemangordon> ok
<freemangordon> because I aim es2 only
<uvos> thats fine
<uvos> its ogl thats broken
<freemangordon> ok
<freemangordon> good
<uvos> (according to others)
<uvos> i dont have a pp ofc
<freemangordon> so, we should see an improvement, because X renders directly on mmaped video memory
<freemangordon> instead of rendering to memory pixmap and then bliting
<uvos> ok
<uvos> depends on kernel bahvior
<uvos> ofc
<uvos> hw wise the kernel should be able to just ajust the mmu to move the ram from cpu to gpu area and back
<freemangordon> and for 3d/present, we (will) have a simple shader that just renders a texture
<uvos> no idea if the kernel dose so in practice
<freemangordon> it does
<uvos> ok
<uvos> great
<freemangordon> at least thta's what I think happens
<freemangordon> for SGX
<freemangordon> because I see driver traces setting SGX mmu in chunks of a page
inky_ has quit [Remote host closed the connection]
<freemangordon> even if we are hit by a perf penalty because of cache flushes and such, we still should be faster than copying
<freemangordon> in theory (tm)
<uvos> yeah ofc
<tmlind> nice
<uvos> that shouldbe very performant
<freemangordon> and we see it
<tmlind> freemangordon: any luck getting droid4 to update the display?
<freemangordon> 10x improvement for 2d
<freemangordon> tmlind: the issue is that SGX recovery gets triggered
<tmlind> ah
<freemangordon> not that display is not updated
<tmlind> ok
<freemangordon> see backscroll, with my (half-baked) xf86-video-modesetting-module-gbm we see ~10x improvement in 2d ops compared to omap driver
<uvos> are you calling the drm display update thing
<uvos> there is an ioctl to update comannd mode displays
<freemangordon> this is modesetting with a glamor replacement which performs all 2d in SW (so far, I plan to implement 2d in SGX)
<uvos> iirc
<freemangordon> uvos: no, I guess modesetting does that for me
<freemangordon> mind you, I am replacing glamor, not the whole ddx
<freemangordon> glamors just renders on BOs provided by MS
<freemangordon> drm interaction is out of scope
<freemangordon> but, obviously something updates the display, will record a video of gtkperf running when I get tired
<freemangordon> though my n90 won;t be able to capture it I guess (30fps)
<freemangordon> *n900
uvos has quit [Ping timeout: 260 seconds]
inky_ has joined #maemo-leste
Twig has quit [Ping timeout: 260 seconds]
<freemangordon> hmm, is it possible that NEON core cannot read mmaped video memory?
<freemangordon> tmlind: ^^^?
<freemangordon> hmm, no, it is not that
akossh has joined #maemo-leste
<Wizzup> freemangordon: debugging the d4 problem?
Danct12 has quit [Ping timeout: 264 seconds]
<freemangordon> no, Xorg was segfaulting
<freemangordon> but I found the reason
<Wizzup> ok
Danct12 has joined #maemo-leste
<freemangordon> gtkperf on n900:
<freemangordon> Total time: 72.87
<Wizzup> faster than our current d4 path
<freemangordon> mhm
inky_ has quit [Read error: Connection reset by peer]
inky has joined #maemo-leste
<freemangordon> glmark2 Score: 21
<freemangordon> but! this is with memcpy of 3d memory to 2d
DPA has joined #maemo-leste
<sicelo> Magic :-)
DPA has quit [Ping timeout: 260 seconds]
<Wizzup> freemangordon: so 3d render shows on screen now?
<freemangordon> yes
<Wizzup> :-o
<freemangordon> going to implement the fast (hopefully) path - GL render of Bo texture
<Wizzup> check
<Wizzup> with the simple shader?
<freemangordon> I hope so
* Wizzup too
DPA has joined #maemo-leste
DPA has quit [Ping timeout: 258 seconds]
DPA has joined #maemo-leste
mardy has quit [Quit: WeeChat 2.8]
avoidr has quit [Ping timeout: 258 seconds]
avoidr has joined #maemo-leste
<Wizzup> freemangordon: unrelated but I am getting this: string "Method "SendReceive" with signature "" on interface "com.nokia.modest" doesn't exist
<Wizzup> oh
<Wizzup> I think I know what the problem is
<Wizzup> I think dns is fubar on my device
<Wizzup> note if the files in /etc/dnsmasq.d/00_leste_dns seem to exist
<Wizzup> parazyd: ^^
<Wizzup> uhhh dnsmasq 5377 0.0 0.2 8172 2076 ? S 22:33 0:00 /usr/sbin/dnsmasq -x /run/dnsmasq/dnsmasq.pid -u dnsmasq -r /run/dnsmasq/resolv.conf -7 /etc/dnsmasq.d,.dpkg-dist,.dpkg-old,.dpkg-new --local-service --trust-anchor=.,20326,8,2,e06d44b80b8f1d39a95c0b0d7c65d08458e880409bbc683457104237c7f8ec8d
<Wizzup> /run/dnsmasq/resolv.conf is empty
<Wizzup> I think this happens because resolvconf is installed
<Wizzup> (per the dnsmasq initscript)
<Wizzup> yup, that's it...
<lel> MerlijnWajer opened an issue: https://github.com/maemo-leste/bugtracker/issues/583 (Installing resolvconf breaks our dnsmasq setup)
<lel> MerlijnWajer assigned an issue: https://github.com/maemo-leste/bugtracker/issues/583 (Installing resolvconf breaks our dnsmasq setup)
<Wizzup> pls fix
cockroach has joined #maemo-leste
cockroac1 has joined #maemo-leste
xmn has joined #maemo-leste
Langoor_ has quit [Ping timeout: 268 seconds]
akossh has left #maemo-leste [#maemo-leste]
Langoor_ has joined #maemo-leste
Pali has quit [Ping timeout: 260 seconds]
Langoor_ has quit [Ping timeout: 260 seconds]
Langoor_ has joined #maemo-leste