ChanServ changed the topic of #lima to: Development channel for open source lima driver for ARM Mali4** GPUs - Kernel driver has landed in mainline, userspace driver is part of mesa - Logs at https://oftc.irclog.whitequark.org/lima/
adjtm is now known as Guest3314
Guest3314 has quit [Read error: Connection reset by peer]
adjtm has joined #lima
drod has quit [Remote host closed the connection]
uis has quit []
uis has joined #lima
uis has quit []
uis has joined #lima
anarsoul|2 has joined #lima
anarsoul has quit [Read error: Connection reset by peer]
anarsoul|2 has quit [Ping timeout: 480 seconds]
anarsoul has joined #lima
uis has quit []
drod has joined #lima
uis has joined #lima
uis has quit []
uis has joined #lima
alyssa has joined #lima
* alyssa
might need some help with the Lima side of !16201
<alyssa>
would appreciate an ack on the lima change from there then
<alyssa>
The commit messages there are pretty panfrost centric (because the problem is magnified when you throw AFBC into the mix), but the same issue applies to lima
<alyssa>
namely, the quantity `width * bytes per pixel` is meaningless for a tiled texture
<alyssa>
the meaningful quantity is `width * bytes per pixel * 16` for 16x16 tiled
<alyssa>
"bytes between adjacent rows"
<alyssa>
that MR reworks Panfrost to use the latter quantity exclusively (except for WSI for legacy compatibility with older Mesa)
<alyssa>
including in the tiling routines
<alyssa>
so without changing lima's data structures, that just requires scaling an argument to the tiling routines
uis has joined #lima
<enunes>
alyssa: great to see the CI getting some use, thanks!
<enunes>
I can take a look at the MR a bit later today
<alyssa>
enunes: thanks!
uis has quit []
uis has joined #lima
erle has quit [Ping timeout: 480 seconds]
erle has joined #lima
<anarsoul>
alyssa: if enunes doesn't get to it earlier I can review it by the end of the week. Can't do it earlier -- I just returned from a short vacation and I'm trying to deal with email backlog at work