ChanServ changed the topic of #lima to: Development channel for open source lima driver for ARM Mali4** GPUs - Kernel has landed in mainline, userspace driver is part of mesa - Logs at https://people.freedesktop.org/~cbrill/dri-log/index.php?channel=lima and https://freenode.irclog.whitequark.org/lima - Contact ARM for binary driver support!
dddddd has quit [Remote host closed the connection]
Elpaulo has quit [Read error: Connection reset by peer]
Kwiboo has quit [Quit: .]
Elpaulo has joined #lima
_whitelogger has joined #lima
_whitelogger has joined #lima
Barada has joined #lima
_whitelogger has joined #lima
Barada1 has joined #lima
Barada has quit [Ping timeout: 260 seconds]
Barada1 is now known as Barada
_whitelogger has joined #lima
_whitelogger has joined #lima
_whitelogger has joined #lima
_whitelogger has joined #lima
yuq825 has joined #lima
dddddd has joined #lima
Barada has quit [Quit: Barada]
_whitelogger has joined #lima
_whitelogger has joined #lima
Barada has joined #lima
_whitelogger has joined #lima
yann has joined #lima
yuq825 has quit [Remote host closed the connection]
_whitelogger has joined #lima
_whitelogger has joined #lima
maciejjo has quit [Remote host closed the connection]
megi has joined #lima
Barada has quit [Quit: Barada]
yann|work has joined #lima
yann has quit [Ping timeout: 268 seconds]
_whitelogger has joined #lima
armessia has joined #lima
smaeul has quit [Quit: cya]
smaeul has joined #lima
smaeul has quit [Quit: cya]
smaeul has joined #lima
_whitelogger has joined #lima
smaeul has quit [Quit: cya]
smaeul has joined #lima
CounterPillow has left #lima ["Bye."]
<rellla> anarsoul: mali parser is finished, one more question though:
<rellla> we have textures at 0x10096800 and the first entry is the address of the texture descriptor -> 0x10080240
<rellla> is it possible, that we get more addresses in the list at 0x10096800? i'm not sure i do understand that correct, but i guess we can have more addresses here, because in lima there is the "lima_tex_list_size"
<rellla> if so, i have to implement that on top, what should be trivial though...
<rellla> anyway, now available to debug indexed draws :)
yann|work has quit [Ping timeout: 260 seconds]
<rellla> also, i guess http://imkreisrum.de/deqp/texture2d_lima/lima.dump.0001 and http://imkreisrum.de/deqp/texture2d_mali/mali.dump.0000 show the secrets to fix dEQP-GLES2.functional.shaders.texture_functions.fragment.texture2d
<rellla> :)
<anarsoul> rellla: maybe there can be more than one list of texture descriptors, but so far we haven't seen that in command stream
<anarsoul> rellla: I think for indexed draws we're not doing synchronization between PLBU and VS properly
<anarsoul> I can look into it later this week, but I'm OK if you want to take it over :)
<anarsoul> see my notes above
<rellla> anarsoul: i meant, is it possible to have more than one texture_desc?
<anarsoul> rellla: in theory yes, but I'm not sure if hardware can actually do that. Keep it as is, we can extend it later
<anarsoul> regarding VS and PLBU synchronization
<anarsoul> no
<anarsoul> I think we have to make sure that PLBU waits for VS to complete its draw command, then VS should wait for PLBU to complete its draw command
<anarsoul> also looks like we pass vertex arrays and indices via PLBU commands
<anarsoul> so VS should wait for PLBU to issue this commands before doing draw
<anarsoul> I'm still looking into blob dumps and trying to understand it though
<anarsoul> so I'm not sure that it's 100% correct
<rellla> i'm running it on master now and i get massive fails with shaders.loops.* ... so my guess is, that this commit broke sth. i did not bisect it though...
<rellla> anarsoul: about indexed draws, i think i can't look into it much earlier...
<rellla> [647110.723640] [drm:lima_sched_timedout_job [lima]] *ERROR* lima job timeout
<rellla> [647110.730567] lima 1e80000.gpu: gp task error int_state=0 status=8a
<rellla> i think, this is the reason why following test fail.
<rellla> yeah, definitely fries the driver. cancelled now...
<anarsoul> either VS shader issue
<anarsoul> or issue with semaphores
<anarsoul> sounds like shader issue
<rellla> the kernel message is with current master, not related to indexed draws