rajkosto has quit [Read error: Connection reset by peer]
<wens>
anyone attending LPC or ELCE two weeks from now?
<jernej>
wens: me
<jernej>
Linux Media Summit and ELCE
<wens>
cool
moteen has joined #linux-sunxi
moteen has quit [Ping timeout: 480 seconds]
<paulk>
me too
<paulk>
but had to cancel attending the media summit in the end :(
JohnDoe_71Rus has joined #linux-sunxi
<MoeIcenowy>
paulk: it should be similar to the one of R329
hentai has quit [Ping timeout: 480 seconds]
<MoeIcenowy>
or maybe I should do it on T113 first
hentai has joined #linux-sunxi
moteen has joined #linux-sunxi
moteen has quit [Ping timeout: 480 seconds]
<wens>
paulk: :( feels like you should be participating in the video encoder discussion
<paulk>
wens: there's that and camera topics as well
<libv>
i found lpc/elce an overrated waste of money
<libv>
way too corporate, and way too self-entitled
<paulk>
well it's definitely an industry event, barely a community-oriented one
<libv>
oh, dublin again
<libv>
that was its only plus side when i went in 2015
<paulk>
if I had to pay the ticket I would certainly not go ;)
<libv>
paulk: as i stated when you came up with the hw enablement devroom, i do not get why "the kernel people" do not visit fosdem, apart from the self-entitlement bit that is
<libv>
or to be fawned over by product managers and sales reps perhaps
<paulk>
libv: one conference I attendent for the first time recently is kernel/embedded recipes
<paulk>
attended*
<paulk>
and I found it to be a good mix of community+kernel devs, without the corporate bullshit
<libv>
oh, i had that jotted down as "same shit, different name
<apritzel>
moteen: can you coordinate with smaeul if you want to upstream D1 U-Boot support? And then please follow the U-Boot process, so post patches to the mailing list.
cnxsoft has quit []
cnxsoft has joined #linux-sunxi
<moteen>
ok will do that thankyou!
bauen1 has quit [Ping timeout: 480 seconds]
<apritzel>
moteen: I have some patches for R528/T113 support, including pinctrl, so ideally some D1 patches would not be needed anymore
<moteen>
are those merged with upstream u-boot yet?
cnxsoft has quit []
<apritzel>
moteen: no, they just exist on my disk for now
<moteen>
alright will co ordinate this once I start sending the series of patches
<apritzel>
if someone wants to help: the T113 (and D1/D1s?) DRAM code in szemzoa's repo would need a rework. And the licensing should be clarified: I doubt that this is genuine GPLv2 code ...
moteen has quit [Remote host closed the connection]
moteen has joined #linux-sunxi
moteen has quit [Ping timeout: 480 seconds]
cnxsoft has joined #linux-sunxi
cnxsoft has quit []
evgeny_boger has quit [Ping timeout: 480 seconds]
bauen1 has joined #linux-sunxi
moteen has joined #linux-sunxi
evgeny_boger has joined #linux-sunxi
buZz has quit [Ping timeout: 480 seconds]
moteen has quit [Remote host closed the connection]
buZz has joined #linux-sunxi
moteen has joined #linux-sunxi
evgeny_boger1 has joined #linux-sunxi
evgeny_boger has quit [Read error: Connection reset by peer]
moteen has quit [Remote host closed the connection]
moteen has joined #linux-sunxi
JohnDoe_71Rus has quit []
moteen has quit [Remote host closed the connection]
moteen has joined #linux-sunxi
hentai has quit [Ping timeout: 480 seconds]
evgeny_boger1 has quit [Ping timeout: 480 seconds]
moteen has quit [Remote host closed the connection]
<smaeul>
from an upstreaming perspective, boot0 (aka "sun20i_d1_spl") is dead and can be ignored now that U-Boot SPL works.
<smaeul>
apritzel: the D1 DRAM init C code is derived from an assembly dump released as "GPL-2.0+". the code in szemzoa's repo is the same or based on it, so GPLv2 is correct
<apritzel>
smaeul: it looks like an assembly dump, but who released that? And can you just stick a GPL2 sticker on this kind of disassembled code?
<smaeul>
the assembly dump came from Allwinner, in the BSP
<apritzel>
I mean at the end it's still AW's proprietary code, isn't it? And you would need something closer to clean room reverse engineering to make it GPL
<apritzel>
so it was assembly there, with a GPL2 sticker, and not disassembled?
<smaeul>
correct. it's a build artifact explicitly generated by AW's build system
<paulk>
writing matching source code from a GPL binary/assembly is absolutely fine AFAIK
<paulk>
here it seems to be the case :)
<paulk>
but obviously that's way too crappy to be merged in u-boot
<paulk>
but could be useful to match it with an existing proper implementation
<paulk>
there's a good chance this controller is used on some other platform
<smaeul>
yes, that was MoeIcenowy's thought with R329
<apritzel>
paulk: yes, writing matching code (based on disassembly) is what we did in the past, but that requires some significant work
<apritzel>
which we probably still need to do for upstreaming, but not to the full extent
<smaeul>
moteen: all of those "Hack up the driver" commits are not upstreamable, because they break other platforms. they need to be rewritten so they add D1 support without breaking any of the other 150+ Allwinner boards
<smaeul>
apritzel: could you please push your t113 patches somewhere?
<apritzel>
smaeul: of course, but first they need to work and be in a presentable shape ;-)
<apritzel>
smaeul: for instance I am not sure I reworked too much of the GPIO/pinctrl code
<paulk>
apritzel: yeah I'm not saying it's a low-hanging fruit :(
rajkosto has quit [Read error: Connection reset by peer]
hentai has joined #linux-sunxi
<jernej>
ELCE has hobbyist tickets, which are still not cheap, but at least reachable for hobbyists. Still can't beat FOSDEM, since it's free :)
cnxsoft has quit []
<apritzel>
the price for virtual attendance is only 25 USD
<jernej>
ah, yes
moteen has joined #linux-sunxi
<jernej>
apritzel: btw, I'm already halfway having decent AC200 codec driver. Unsurprisingly, some small tweaks to mfd driver are needed.
<apritzel>
ah, nice
<apritzel>
any big changes, that we want to have in the first drop already?
<jernej>
yes, another property is needed, nvmem cell reference for bandgap value
<jernej>
it's used by TVE and codec
<jernej>
and I would rather just remove header from includes
<jernej>
as it turns out, struct doesn't need to be public, you can get regmap from device
<jernej>
and unsurprisingly, I missed some registers and misnamed some others
<jernej>
it's best if each subdriver defines it's own
<apritzel>
do you mean include/linux/mfd/ac200.h?
<jernej>
yes
<apritzel>
in general I am all for removing those kind of header files, when they don't describe an interface. It's the job of a driver to abstract those details
ftg has joined #linux-sunxi
vagrantc has joined #linux-sunxi
bauen1 has quit [Ping timeout: 480 seconds]
apritzel has quit [Ping timeout: 480 seconds]
moteen has quit [Remote host closed the connection]
<wens>
jernej: yeah, I used to use the hobbyist tickets