<\x>
KotCzarny: i do have to use openwrt toolchain for this right? and load the lkm at boot
<KotCzarny>
yeah, should be compiled using toolchain/config used by your kernel
<\x>
i will
<\x>
im more interested in the xr819 tho
<\x>
well, i guess ill put the zero plus out of service now
<\x>
been my homeserver for like a month now
<KotCzarny>
or use armbian
<KotCzarny>
and setup firewall yourself
<KotCzarny>
and routing
<KotCzarny>
no need to stick to openwrt specifically
<\x>
ah yeah i mean thats easy
<\x>
but i want openwrt
<\x>
mesh meme
<KotCzarny>
i would expect that's opensource, which means should work with 4.20 etc, no?
<\x>
it should
<\x>
i think next version will have 4.19
<KotCzarny>
you can compile kernel yourself
<\x>
my current router has 4.14 kernel on it
popolon has joined #linux-sunxi
<\x>
thanks for the help. ill tell ya if i get it werking. tho i dont even know if ill do it kek. it does work nicely for now.
apritzel has joined #linux-sunxi
iamfrankenstein has quit [Quit: iamfrankenstein]
f0xx has quit [Ping timeout: 272 seconds]
reinforce has joined #linux-sunxi
Amit_T has quit [Quit: Leaving]
<karlp>
\x it's not jsut newer variants, it's _never_ worked....
<karlp>
8189ftv, hauke has a tree with some attempts at it, but it's ~worthless in openwrt first.
IgorPec has quit [Ping timeout: 244 seconds]
<karlp>
xr819 looks to have even less chance of a future IMO, but could change :)
lkcl has quit [Ping timeout: 245 seconds]
typedef_pil has quit [Ping timeout: 245 seconds]
chewitt has quit [Max SendQ exceeded]
chewitt has joined #linux-sunxi
lkcl has joined #linux-sunxi
xerpi has quit [Ping timeout: 246 seconds]
typedef_pil has joined #linux-sunxi
xerpi has joined #linux-sunxi
vagrantc has joined #linux-sunxi
AneoX has joined #linux-sunxi
AneoX_ has quit [Ping timeout: 246 seconds]
Putti has quit [Ping timeout: 258 seconds]
typedef_pil has quit [Ping timeout: 245 seconds]
AneoX_ has joined #linux-sunxi
AneoX has quit [Ping timeout: 258 seconds]
xerpi has quit [Ping timeout: 246 seconds]
NeuroScr has joined #linux-sunxi
Rafael1980 has quit [Quit: Konversation terminated!]
jbrown has joined #linux-sunxi
<vagrantc>
any relatively simple suggestions for the pinebook to backport to 4.19.x kernel in Debian? I'm working on the battery support ... not sure how complicated sound or HDMI support would be to backport
<vagrantc>
i saw some fix for the RTC clock ... is that related to the bug on a64 causing the clock to leap into the future?
<vagrantc>
how significant are the .dts changes: add SID node/efuse, cd-inverted, L2 cache nodes, data strobe pin from mmc2 ...
<vagrantc>
backlight regulator...
jbrown has quit [Ping timeout: 252 seconds]
<apritzel>
vagrantc: RTC clock is not related to the timer, that's two different things
<vagrantc>
debian will be shipping 4.19.x in the next stable release, but there seem to be many improvements in 4.20/4.21
<vagrantc>
so trying to pull in at least the low-hanging fruit
<vagrantc>
the timer issue is really weird ... i can run for a week or two without it getting triggered, and then sometimes it gets triggered several times in a single day
<catphish>
this bare metal thing is hard work, no matter what i do, i can't seem to get my USB PHYs working, root hub just won't detect anything connected to the port, it's definitely my code to blame because u-boot works fine, learned so much about usb, but ever so frustrated
<vagrantc>
but been using the pinebook for my primary computer since early december without too many other issues ... the timer seems to be the biggest frustration
<apritzel>
vagrantc: well, the timer fix will still take more time, I guess it won't even land in 4.21
Mr__Anderson has quit [Remote host closed the connection]
<apritzel>
but you can use the FSL workaround meanwhile
<vagrantc>
apritzel: tell me more! :)
xerpi has joined #linux-sunxi
<apritzel>
vagrantc: just add fsl,erratum-a008585 to the arch timer node in the DT
<apritzel>
it's frowned upon by the maintainers, we should have at least a separate property for that
<vagrantc>
e.g. the same erratta fixes it, but it's just "named wrong" ?
<apritzel>
vagrantc: somewhat, it's a bit more complex
<vagrantc>
fair enough
<apritzel>
the erratum is probably not the same, but related
<apritzel>
and the FSL workaround is more expensive than we probably need for Allwinner
<vagrantc>
ah. so something to document as a workaround, but probably not good to enable out of the box
<apritzel>
the neat thing about the FSL property is that's in the kernel for ages
<apritzel>
so just adding it to the DT makes the problem go away on older kernels as well
<vagrantc>
i don't forsee using on anything older than 4.19.x at this point
jbrown has joined #linux-sunxi
<apritzel>
we should probably do that: add a new allwinner,timer-erratum property to the bindings and the DT
<apritzel>
and wire it up to the FSL workaround, for now
<apritzel>
then we can later refine this
<apritzel>
that should simplify your backport
<vagrantc>
yeah, if you can land it even in linux-next *soon* and it's not a huge linecount, i can probably get it into debian's next stable release
<apritzel>
and we solve the problem, although not in the most efficient way
xerpi has quit [Remote host closed the connection]