azkali has quit [Read error: Connection reset by peer]
marvin24 has quit [Ping timeout: 258 seconds]
marvin24 has joined #tegra
hexdump0815 has quit [Ping timeout: 245 seconds]
hexdump0815 has joined #tegra
gouchi has joined #tegra
gouchi has quit [Client Quit]
jn_ has joined #tegra
jn_ has joined #tegra
jn_ has quit [Changing host]
<kwizart>
digetx, hi, how can I set the DRM_TEGRA_ORIG in a grate kernel tree ? the driver doesn't seem to play well on jetson-tk1 (probably because of vic disablement)
<kwizart>
it still tries to enforce GRATE_HOST1X
gavodavo has quit [Ping timeout: 260 seconds]
cyndis has quit [Ping timeout: 260 seconds]
jn has quit [Ping timeout: 260 seconds]
cyndis has joined #tegra
gavodavo has joined #tegra
<digetx>
kwizart: you can disable DRM_TEGRA and set DRM_TEGRA_ORIG, it will work then
<kwizart>
yep but it still tries to enable CONFIG_GRATE_HOST1X=y
<digetx>
GRATE_HOST1X is disabled automatically if DRM_TEGRA_ORIG is set
<digetx>
but it should work on tk1, if you have logs with a error, please show me them
<kwizart>
hah, got it, I need DRM_TEGRA_ORIG instead of DRM_TEGRA
<kwizart>
not both
<digetx>
yes
<kwizart>
digetx, main issue I have with grate on tk1 is that it's not using nouveau as a gl_renderer (uses llvmpipe), so it's probably discarded for that reason on wayland. Seems reproducible with DRM_TEGRA_ORIG on grate tree but not DRM_TEGRA on vanilla (fedora) 5.15-rc5 kernel
<kwizart>
well, I have an additional issue with usb keyboard not working with linux-next...
<digetx>
does usb work with grate kernel?
<kwizart>
nope I meant, I've only tested grate, but I expect it's because of linux-next...
<digetx>
do you have anything in kernel log when keyboard is connected?
<digetx>
ah, wait
<digetx>
tk uses xusb
<kwizart>
tegra-xusb: probe of 70090000.usb failed with error -22
<digetx>
that is the problem
<kwizart>
actually it's evern reproducible with 5.15-rc5
<digetx>
padctl node of t124 doesn't have interrupt, it's a bug introduced by that commit
<digetx>
I can type patch if you want
<digetx>
I'll type it, push it to grate later today and ping you to test
<digetx>
should be oneline fix
<digetx>
although, maybe not oneline
<kwizart>
digetx, thanks entering meeting, will test newer grate
<kwizart>
digetx, could it be that interrupt is missing on tegra124 padctl node ? (I had a similar issue with thermal driver)
<digetx>
yes, but older device-tree should continue to work
<digetx>
that should've been caught by tegra's and other CIs, wondering what those CIs are checking
<kwizart>
I think they are only checking ability to boot to userspace
<kwizart>
or I would wonder how a usb test would look like
<digetx>
tegra checks device driver presence, maybe it doesn't check usb
<digetx>
tagr: ^
<tagr>
I'll have to look into it, but yeah, that offending commit should've clearly made the new IRQ support optional
<tagr>
we typically check that after boot all devices have successfully been probed by their drivers, as well as unexpected warnings/errors in the boot log
<tagr>
both of those tests should've caught the XUSB pad controller issue
<cyndis>
looks like the devices test is not being run on TK1 for some reason
<tagr>
we might've never hooked it up
<tagr>
the data for Jetson TK1 looks to be all there
<tagr>
ah... yes, for some reason that test is explicitly disabled on Jetson TK1 =(
<tagr>
ah... looks like some issue with built-in USB driver vs. USB firmware not being available during early boot
<tagr>
which is something that we should be able to fix
<digetx>
cool
jn_ is now known as jn
dormito has joined #tegra
dormito has quit [Quit: WeeChat 3.3]
e1z0_ has quit [Ping timeout: 264 seconds]
e1z0 has joined #tegra
e1z0 has quit [Read error: Connection reset by peer]