ChanServ changed the topic of #panfrost to: Panfrost - FLOSS Mali Midgard & Bifrost - Logs https://oftc.irclog.whitequark.org/panfrost - <macc24> i have been here before it was popular
paulk has quit [Read error: Connection reset by peer]
paulk-bis has joined #panfrost
cphealy has quit [Quit: Leaving]
cphealy has joined #panfrost
Daaanct12 has quit [Remote host closed the connection]
ckileeeeeeeeeeeeeeeeeeeeeeetb^ has joined #panfrost
hanetzer has quit [Quit: WeeChat 3.6]
davidlt_ has joined #panfrost
davidlt_ has quit [Ping timeout: 480 seconds]
davidlt_ has joined #panfrost
Dylanger has quit [Write error: connection closed]
stebler[m] has quit [Write error: connection closed]
JulianGroOld[m] has quit [Write error: connection closed]
thecycoone[m] has quit [Remote host closed the connection]
toggleton[m] has quit [Write error: connection closed]
strongtz[m] has quit [Write error: connection closed]
mairacanal[m] has quit [Write error: connection closed]
zhxt[m] has quit [Write error: connection closed]
underpantsgnome[m] has quit [Write error: connection closed]
CalebFontenotHaileysCuteNerdyB has quit [Write error: connection closed]
frytaped[m] has quit [Write error: connection closed]
Guest4044 has quit [Write error: connection closed]
nergzd723 has quit [Write error: connection closed]
jenneron[m] has quit [Write error: connection closed]
Przemek[m] has quit [Write error: connection closed]
psydroid[m] has quit [Write error: connection closed]
thaytan[m] has quit [Write error: connection closed]
robertfoss[m] has quit [Write error: connection closed]
sharonmary6[m] has quit [Write error: connection closed]
nlhowell has joined #panfrost
CalebFontenotHaileysCuteNerdyB has joined #panfrost
guillaume_g has joined #panfrost
rasterman has joined #panfrost
ckileeeeeeeeeeeeeeeeeeeeeeetb^ has quit [Remote host closed the connection]
Dylanger has joined #panfrost
frytaped[m] has joined #panfrost
Guest4154 has joined #panfrost
jenneron[m] has joined #panfrost
mairacanal[m] has joined #panfrost
JulianGroOld[m] has joined #panfrost
nergzd723 has joined #panfrost
Przemek[m] has joined #panfrost
psydroid[m] has joined #panfrost
robertfoss[m] has joined #panfrost
sharonmary6[m] has joined #panfrost
stebler[m] has joined #panfrost
strongtz[m] has joined #panfrost
thaytan[m] has joined #panfrost
thecycoone[m] has joined #panfrost
underpantsgnome[m] has joined #panfrost
toggleton[m] has joined #panfrost
zhxt[m] has joined #panfrost
nlhowell has quit [Ping timeout: 480 seconds]
nlhowell has joined #panfrost
rasterman has quit [Read error: Connection reset by peer]
rasterman has joined #panfrost
floof58 is now known as Guest4182
floof58 has joined #panfrost
Guest4182 has quit [Ping timeout: 480 seconds]
camus has joined #panfrost
camus1 has quit [Remote host closed the connection]
atler is now known as Guest4184
atler has joined #panfrost
Guest4184 has quit [Ping timeout: 480 seconds]
nlhowell has quit [Ping timeout: 480 seconds]
guillaume_g has quit []
guillaume_g has joined #panfrost
guillaume_g has quit []
alyssa has quit [Quit: leaving]
<greenjustin>
robclark: I'll give that a go I suppose. The bug does look a whole lot like a synchronization issue
<greenjustin>
Thanks for the help everyone!
rasterman has quit [Quit: Gettin' stinky!]
davidlt_ has quit [Ping timeout: 480 seconds]
moa has joined #panfrost
bluebugs has quit [Ping timeout: 480 seconds]
sigmaris_ has joined #panfrost
sigmaris has quit [Ping timeout: 480 seconds]
anholt has quit [Remote host closed the connection]
<alyssa>
If you recall, about a year ago you reported a performance problem with glmark2 -bbuffer:update-method=subdata
<alyssa>
and my fix was to flush when shadowing
<alyssa>
As it turns out, that flush is hurting performance on other workloads... a LOT!
<alyssa>
That MR replaces the flush with better bookkeeping
<alyssa>
On my RK3399 system, that actually makes the glmark2 case faster... in addition to the motivating 10x improvement in Duckstation
<alyssa>
I recall your hardware has different performance characteristics than mine (with the little CPUs) ... I would appreciate if you could retest glmark2 with and without that MR to confirm that your hardware is helped (or at the very least, not hurt)
<alyssa>
Running CTS now to make sure I didn't hose everything
<alyssa>
deqp-gles2 passes, probably fine :p
<alyssa>
bbrezillon: ^^ Would appreciate a review on the top patch of that MR
<alyssa>
because it is SUBTLE and I think you wrote the original code and given I got this wrong once I don't trust myself to have gotten it right the second time