<Tony_>
I'm sure it is hardware problem, because other device can works fine. But I want to know where is the problem ?
<Tony_>
rk3188 or rk616 or LCD ?
<Tony_>
actually, LCD is fine because it can works fine with other board.
<naobsd>
then rk3188 or rk616 will be wrong
<Tony_>
from the good capture screen by adb, I think the rk3188 is fine. right ?
<Tony_>
do you think so ?
<Tony_>
by the way, the /dev/graphic/fb0 is fine.
<naobsd>
memory is fine
<Tony_>
actually, I'm not sure the flow a picture from CPU to LCD.
<naobsd>
sorry, I have no idea why you think you can get something from me with a picture... you already tried to exchange components, you already know answer
<Tony_>
memory -> fb0 -> rk616 -> LCD ?
<naobsd>
fb0 is memory
<naobsd>
do you have block diagram of your hardware?
<naobsd>
you can see connection among lcdc[01], rk616, and lcd
<Tony_>
naobsd, yes, I have block diagram.
<Tony_>
I'm clear than before, thank you. CPU ->(problem maybe) -> RK616(problem maybe) ->(problem maybe) -> LCD
<naobsd>
you said LCD is fine
<naobsd>
is it worth to find out which one is wrong?
<Tony_>
Yes, I can replace it with a good one. by the way, I can be more clear about flow that picture from CPU to LCD. ;)
<Tony_>
(what I mean is I can replace bad IC with a good one)
<naobsd>
unless your board has something like test points between RK3188 and RK616, you need to replace RK3188 and/or RK616 one by one
<Tony_>
there are test points between RK616 and LCD.
<Tony_>
but Actually, I don't know well about LVDS wave.
<Tony_>
;(
<Tony_>
I only know they are Differential signaling, but I don't know what signaling wave will let the display like that problem one.
<Tony_>
naobsd, would you please tell me that the problem's name ?
<Tony_>
like the picture.
<naobsd>
sorry, I don't know. I have no hardware knowledge.
gani has quit [Quit: Leaving]
<Tony_>
naobsd, oh, me too, we always give a custom name for that. I even don't know how to Google it.;(
<Tony_>
naobsd, custom such as "halo". ;)
salv0 has quit [Ping timeout: 264 seconds]
salv0 has joined #linux-rockchip
<Tony_>
naobsd, I want to talk about software. I'm confusion about GPU memory LCDcontroler CPU.
<Tony_>
GPU is only a calculator for picture. right ?
<Tony_>
Whether there is GPU or not, the picture is in memory.
<naobsd>
Tony_: please read "Chapter 1 Introduction" in RK3188 TRM first
<Tony_>
naobsd, okay, thank you.
<naobsd>
and "Chapter 22 LCDC"
<Tony_>
Yes, I seen. I think this is not rk3188's feature, it is all the CPU's feature.
<naobsd>
?
<naobsd>
RK3188 is CPU
<naobsd>
RK3188 TRM describes RK3188 features
<Tony_>
Yes, I know. RK3188 is CPU, My question is a normally concept.
<Tony_>
I think .
<naobsd>
if you think RK3188 TRM is useless, you cannot get any information
markm has quit [Ping timeout: 244 seconds]
<Tony_>
;) I'm seeing it.
<naobsd>
if you said "CPU" as "all CPUs in this world", no, RK3188 TRM describes on only about RK3188
<naobsd>
there might be similar controller in this world, but RK3188 TRM doesn't describe anything about another CPU
<rperier>
hi all
<Tony_>
naobsd, okay. ;)
<Tony_>
rperier, hi.
<Tony_>
long time no see. rperier
<rperier>
yeah a very long time... something like yesterday :D
<rperier>
;)
RayFlower_ has joined #linux-rockchip
<rperier>
I started to investigate arc_emac issues... the code is very complicated compared to other ethernet driver I think... (I am talking about the core not about emac_rockchip)
<rperier>
I will investigate packets lost and perhaps refactore some stuffs, don't know yet
<Tony_>
oh, I don't know about arc_emac. ;) maybe you have to ask help from naobsd.
<rperier>
no because I don't like to ask naobsd for all things I do on my rockchip boards, and as this is me who developped emac_rockchip (based on arc_emac) it should be okay
<rperier>
this is how you learn ;)
<rperier>
s/this is/that's/
<PaowZ_>
hi rperier ! about packets loss, this is something you got aware yourself ?
<PaowZ_>
..or user feedbacks..
<naobsd>
Tony_: <rperier> when you have a question, don't ask a specific person (except if you want to ask him something he did or related to his work), just ask publically here if you have any questions, people will answer
<naobsd>
rperier: "the code is very complicated" should be the cause of problem. it should be more simple.
<naobsd>
(about tx buffer and descriptor handling, if my memory is correct)
Astralix|away is now known as Astralix
<Tony_>
;-) Noted, ask publically here.
selfbg has joined #linux-rockchip
<rperier>
naobsd: yeah it might help, I agree
<rperier>
PaowZ_: I already had few issues using NFS but it was very ramdom, but it was usable and I had more important bugs at this time. But these issues are seriously very boring because I use ethernet for all things I do (uploading my kernel image on the fly to the board, boot over NFS etc..)