ChanServ changed the topic of #panfrost to: Panfrost - FLOSS Mali Midgard + Bifrost + Valhall - Logs https://oftc.irclog.whitequark.org/panfrost - I don't know anything about WSI. That's my story and I'm sticking to it.
q4a1 has joined #panfrost
q4a has quit [Ping timeout: 480 seconds]
hanetzer2 has quit [Quit: WeeChat 3.8]
hanetzer has joined #panfrost
falk689 has quit [Quit: No Ping reply in 180 seconds.]
falk689 has joined #panfrost
floof58 has quit [Remote host closed the connection]
floof58 has joined #panfrost
Leopold_ has quit [Ping timeout: 480 seconds]
Leopold_ has joined #panfrost
Leopold_ has quit [Remote host closed the connection]
Leopold_ has joined #panfrost
karolherbst_ is now known as karolherbst
karolherbst_ has joined #panfrost
floof58 has quit [Ping timeout: 480 seconds]
karolherbst has quit [Ping timeout: 480 seconds]
floof58 has joined #panfrost
davidlt has joined #panfrost
Leopold_ has quit []
davidlt has quit [Remote host closed the connection]
Leopold_ has joined #panfrost
floof58 has quit [Ping timeout: 480 seconds]
floof58 has joined #panfrost
kinkinkijkin is now known as Guest3224
Guest3224 has quit [Read error: Connection reset by peer]
kinkinkijkin has joined #panfrost
guillaume_g has joined #panfrost
kinkinkijkin has quit [Quit: Leaving]
rasterman has joined #panfrost
nlhowell has joined #panfrost
wicastC has quit [Remote host closed the connection]
wicastC has joined #panfrost
guillaume_g has quit [Ping timeout: 480 seconds]
guillaume_g has joined #panfrost
<bbrezillon>
alyssa: pancsf is definitely at fault here
<bbrezillon>
I'll try to reproduce locally
<alyssa>
bbrezillon: thanks :)
<alyssa>
bbrezillon: In general, given the state of brokennness on our mesa tree, deqp-runner is an excellent way to fuzz pancsf ;)
* alyssa
would be unsurprised if deqp-runner + broken mesa finds many bugs relating to fault handling, judging by our experiences with panfrost.ko
p0g0 has joined #panfrost
<bbrezillon>
alyssa: got a full deqp-gles2 run passing without a hang
<cphealy>
When comparing the Mali DDK with Mesa on Mali G52, I'm seeing that GL_SAMPLES supports 16, 8, and 4 as valid values with the Mali DDK while Mesa is only advertising support for 4. Is this due to a constraint with Mesa or just that 4x is the only one ever tested with and implemented?
<alyssa>
!!!
<alyssa>
wait, passing?
<bbrezillon>
no, not passing
<alyssa>
okay yeah that's what I thought, I saw a lot of flakes
<bbrezillon>
I mean, no platform hang
<alyssa>
:+1:
bluebugs has joined #panfrost
<alyssa>
(Unsure if the flakes are kernel or mesa, tbh.)
<alyssa>
(Although I think it was slightly less flaky on kbase..)
<bbrezillon>
don't know if that's related, but no devfreq in pancsf, didn't check what the default GPU clk freq is
<bbrezillon>
198MHz
<bbrezillon>
how long does it take on kbase?
<bbrezillon>
kbase being less flaky could be explained by the fact we consider the group as faulty/unusable as soon as one timeout or fatal fault happens, and we don't have code to recover in mesa (we should destroy the group and create a fresh one when that happens)
<alyssa>
Hmm, I think I was seeing flakiness even without faults
<alyssa>
Just running a few deqp test back to back, pass/fail would flake around
<alyssa>
missing synchronization is my guess, or maybe cache coherency
<bbrezillon>
hm, I see you had a CACHE_FLUSH instruction before the CALL in the kbase drm-shim
<bbrezillon>
I definitely don't have that kernel side
<bbrezillon>
might be worth trying to map all BOs GPU_UNCACHED
<alyssa>
yeah, the CACHE_FLUSH before the CALL was necessary
<bbrezillon>
(the CS buffers should be mapped UNCACHED already)
<bbrezillon>
well, the CACHE_FLUSH is needed if buffers are mapped GPU-cacheable, but I kept things simple and mapped those buffers uncacheable
<bbrezillon>
*if CS buffers are mapped GPU-cacheable
<alyssa>
I, uh, I do recommend caching the linear buffers :-p
<bbrezillon>
I'm not saying they shouldn't be cached, just pointing out that they were not in my branch
jdavidberger has joined #panfrost
Leopold_ has quit [Remote host closed the connection]
Leopold__ has joined #panfrost
jdavidberger has quit [Remote host closed the connection]
hanetzer has quit [Quit: WeeChat 3.8]
rasterman has quit [Quit: Gettin' stinky!]
karolherbst_ is now known as karolherbst
<cphealy>
Does the Panfrost shader compiler support the use of "fma" even though the GLSL version is 3.10?