ChanServ changed the topic of #etnaviv to: #etnaviv - the home of the reverse-engineered Vivante GPU driver - Logs https://oftc.irclog.whitequark.org/etnaviv
tlwoerner has quit [Quit: Leaving]
tlwoerner has joined #etnaviv
DPA has quit [Read error: Connection reset by peer]
DPA has joined #etnaviv
alarumbe has quit []
chewitt has joined #etnaviv
alarumbe has joined #etnaviv
lynxeye has joined #etnaviv
pcercuei has joined #etnaviv
pjakobsson has quit [Ping timeout: 480 seconds]
pjakobsson has joined #etnaviv
pjakobsson has quit [Read error: No route to host]
pjakobsson has joined #etnaviv
mvlad has joined #etnaviv
tlwoerner has quit [Quit: Leaving]
tlwoerner has joined #etnaviv
<marex> austriancoder: so, started digging again ... AXI error still present, now cleaning up KMD and mesa patches
lynxeye has quit [Quit: Leaving.]
<marex> austriancoder: and to wrap it up, if I comment out pm_runtime_put_autosuspend() in event_free(), axi error goes away
<austriancoder> \o/
<marex> austriancoder: the question is , why ... does some power domain get shut down ?
<marex> I guess so
<marex> it cannot be clock, those are controlled separately
<marex> [ 84.020524] etnaviv-gpu 48280000.gpu: MMU fault status 0x00000001
<marex> [ 84.021087] etnaviv-gpu 48280000.gpu: MMU 0 fault (slave not present) addr 0xfee25000
<marex> [ 84.029687] etnaviv-gpu 48280000.gpu: recover hung GPU!
<marex> [ 84.034433] etnaviv-gpu 48280000.gpu: offending task: weston (weston -Bdrm -i 0)
<marex> [ 84.053526] etnaviv-gpu 48280000.gpu: MMU fault status 0x00000002
<marex> [ 84.054082] etnaviv-gpu 48280000.gpu: MMU 0 fault (page not present) addr 0x00000000
<marex> [ 84.062003] etnaviv-gpu 48280000.gpu: recover hung GPU!
<marex> this is new
<marex> kmscube works, weston blows up
<marex> austriancoder: that is for weston I assume ?
<marex> austriancoder: yes, that indeed helps
<marex> austriancoder: feel free to add my TB on that one
<marex> oh wait ... no
<marex> does not help, I forgot to insmod the etnaviv.ko module
<austriancoder> k
<marex> sorry
<marex> austriancoder: it could be related to this SEC_KERNEL stuff maybe ?
f_ has quit [Ping timeout: 480 seconds]
f_ has joined #etnaviv
<marex> austriancoder: yes, the AXI error is caused by OpTee-OS/SCMI being catastrophically broken on the STM32MP25xx, again, siiiigh ...
mvlad has quit [Remote host closed the connection]
<marex> austriancoder: is it actually possible the GPU is left in some botched state before the power domain is stopped, and when it is started again, the GPU goes boom and crashes the AXI bus ?