<ValPackett[m]>
has anyone actually apitraced ngl?
<barni2000[m]>
There was a ticket about it but on gtk side and not mesa
<barni2000[m]>
I will need some reasons why we separate msm8937.dtsi from msm8917.dtsi, I have some ideas but maybe you know more
<barni2000[m]>
I need to tell some reasons to Konrad
<barni2000[m]>
nvm i have answered, i hope i can upstream 8937 in 2 or 3 round
<barni2000[m]>
* have answered it, i
<barni2000[m]>
lets send SDM429 rpm upstream
<minecrell[m]>
<barni2000[m]> "I will need some reasons why..." <- If you based 8937 on 8917.dtsi then the argument with the different license seems a bit moot, since deriving something from GPL would make the derivative GPL as well.
<minecrell[m]>
But well, DT licensing is a mess upstream
<Danct12[m]>
<barni2000[m]> "I will need some reasons why..." <- as for this, my idea is to make msm89x7.dtsi which acts as a base, and then add 8917, 8937, sdm439 stuffs into it's own dtsi
<right_front[m]>
Danct12[m]: agree
<barni2000[m]>
<minecrell[m]> "But well, DT licensing is a mess..." <- pmi8950 is gpl msm8953 is bsd-3 a lotf of device dt what uses pmi8950 is bsd-3, ig they should not be but i dont care
<barni2000[m]>
<Danct12[m]> "as for this, my idea is to..." <- we have many common stuff with other SoCs like msm8953 8916 8939 etc. my plan is SDM439 and 8940 will based 8937.dtsi maybe 429 and 8920 should based 8917
<barni2000[m]>
<minecrell[m]> "If you based 8937 on 8917.dtsi..." <- if you start to make it from scratch you will get the same dt so BSD-3 is fine anyway
<M0xCAFEBABE[m]>
<barni2000[m]> "we have many common stuff with..." <- 429 is stripped down 439
<M0xCAFEBABE[m]>
what about qm215 btw
<barni2000[m]>
it is supported
<barni2000[m]>
ik but with 4 core is similar for 8917
<barni2000[m]>
* 4 core it is similar
<barni2000[m]>
anyway i will not make support for 429 because i don't have one
<M0xCAFEBABE[m]>
barni2000[m]: except for the cpu most of the rest are copy for 439?
<M0xCAFEBABE[m]>
* except for the cpu most of the rest are copy of 439?
<barni2000[m]>
does it have dual dsi support?
<M0xCAFEBABE[m]>
even if it doesn't, it's minor
<M0xCAFEBABE[m]>
sdm439 have many difference to msm89xx you know
<barni2000[m]>
it seems 429 differences only -4 core
<barni2000[m]>
and gpu
<barni2000[m]>
Adreno 504
<barni2000[m]>
439 have 505
<M0xCAFEBABE[m]>
<barni2000[m]> "ik but with 4 core is similar..." <- meanwhile 8917: a306
<barni2000[m]>
yes
<barni2000[m]>
so msm8917->qm215,msm8920 and msm8937->msm8940,sdm439,sdm429
<barni2000[m]>
after 8937 upstreaming i will work on 439
<M0xCAFEBABE[m]>
cool
<M0xCAFEBABE[m]>
<barni2000[m]> "we have many common stuff with..." <- btw when talking about similarities why not to mention 8952/8956 xd
<M0xCAFEBABE[m]>
u maintain 8956 too and u should know how similar they are
<barni2000[m]>
8956 boots on different cluster
<barni2000[m]>
M0xCAFEBABE[m]: 8956 is upstreamed i only added some small patches too it
<barni2000[m]>
and i have ported only 1 device, kenzo
<barni2000[m]>
also 8956 have A72 cores
<M0xCAFEBABE[m]>
M0xCAFEBABE[m]: seems like 8937's father is 8952, and 8956 is 8952's brother
<barni2000[m]>
8937 is more like 8939
<barni2000[m]>
there are more similarities between them than 8952
<barni2000[m]>
but i have not checked 8952 too much