Akari` has joined #zink
Akari has quit [Ping timeout: 480 seconds]
orbea has quit [Remote host closed the connection]
orbea has joined #zink
eukara has joined #zink
sadlerap has quit [Remote host closed the connection]
sadlerap has joined #zink
LexSfX has quit [Ping timeout: 480 seconds]
LexSfX has joined #zink
LexSfX has quit [Ping timeout: 480 seconds]
LexSfX has joined #zink
quantum5 has joined #zink
LexSfX has quit [Remote host closed the connection]
LexSfX has joined #zink
LexSfX has quit [Remote host closed the connection]
LexSfX has joined #zink
<
zmike>
literally anyone
<
zmike>
rc1 will have broken zink without this
<
kusma>
Is this why I get that Unknown sequence number while processing queue pretty much all the time when I resize glxgears?
<
kusma>
Because, uh. It seems pretty broken ATM to me...
<
zmike>
no, that's different
<
zmike>
this is more like glxgears crashes on startup
<
kusma>
Hmm, I'm not seeing that one, neither on ANV nor RADV...
<
kusma>
NV-related, perhaps?
eukara has quit [Read error: No route to host]
eukara has joined #zink
sadlerap1 has joined #zink
sadlerap has quit [Ping timeout: 480 seconds]
<
anholt>
zmike: I assume zink wouldn't be able to handle a demote_if or terminate_if nir intrinsic, and would need lowering, right?
<
zmike>
anholt: yeah sounds right
<
zmike>
there's already one for discard_If being used
<
anholt>
I'm extending that pass, and seeing what flags you want
<
anholt>
(apparently nouveau nir gets demote_ifs, so presumably at some point you'll see those in zink.
<
zmike>
hm maybe from nine or clover or something?
<
zmike>
but yes, that would be appreciated!
<
anholt>
EXT_demote_to_helper_invocation, I guess?