<ssvb>
vector80: you are having tearing when doing what?
akaizen_ has joined #linux-sunxi
<ssvb>
vector80: you should have no tearing on video playback and in 3D OpenGL ES applications
Andy-D has joined #linux-sunxi
<ssvb>
vector80: if you want to have no tearing on moving windows or scrolling, then you need to have a GLES compositing window manager
<vector80>
ssvb
<vector80>
ssvb, thank you very much
<vector80>
I have tearing during window display
<vector80>
How can I have a GLES compositing window manager?
<vector80>
Does LXDE is not a compositing window manager?
<ssvb>
install something like kwin_gles
akaizen has quit [Ping timeout: 260 seconds]
<vector80>
let me check what it is...
<vector80>
kwin_gles seems to be a kde window manager ?
<ssvb>
yes
<vector80>
I use LXDE for the best performance, if I use KDE, my performance will not be worse than before ?
<ssvb>
please note, than having a compositing window manager incurs a significant performance loss
Andy-D_ has joined #linux-sunxi
<vector80>
So, you mean, currently the best way is to use LXDE without GLES support ?
<ssvb>
it's always a tradeoff
<vector80>
Ok, let me explain what I needed. I made a kiosk application with my product, the board is Interra-3
<vector80>
I use midori as the browser,
<vector80>
but I feel it is not HW accelerated
<vector80>
So, how to have a browser with hw acceleration ?
Andy-D has quit [Ping timeout: 272 seconds]
<ssvb>
you can't have a "regular" web browser with hw acceleration
<vector80>
very clear, than, can you recommend one with hw accel ?
<ssvb>
midori is using xrender api, which is supposed to be accelerated in the x server
<vector80>
Does my x server accelerated ? How I can confirm this?
<ssvb>
the problem is that all attempts to implement accelerated drivers on any hardware resulted in something that is slower than software rendering, the only exception is the Intel SNA driver
<ssvb>
on ARM hardware everything is particularly ugly
<vector80>
waww... that is really bad news
<vector80>
I captured one video, let me share
<vector80>
pls wait
<ssvb>
the Chromium browser is basically using software rendering (or at least that was the case in the past), it is the reason why it is generally considered to be faster than Firefox
<arokux>
ssvb: I started. yesterday I've built everything from scratch and got to the point where I saw switching between host and gadget, but nothing more than that, it was too late. I'll continue. I'm gathering experiences. there are some messages on ML
<ssvb>
arokux: and the usb otg driver in android works on my tablet (in gadget mode with adb, and in host mode when I connect a keyboard to it)
<ssvb>
arokux: so it definitely can work in principle
<arokux>
ssvb: now, I do not understand why you patch is a hack, it is just a config?
<arokux>
ssvb: and why change tristate to bool?
<ssvb>
arokux: because it does not want to be compiled statically otherwise
<arokux>
ssvb: ok, so this is the hack, I see
<ssvb>
arokux: and I use it for NFS root without initramfs
<arokux>
ssvb: :) my style
megal0maniac has quit [Ping timeout: 244 seconds]
<ssvb>
I still could not resolve the problem with using FEL to load u-boot/kernel and then hand over to g_cdc/g_ether for mounting the rootfs over NFS
wingrime has quit [Read error: Connection reset by peer]
wingrime has joined #linux-sunxi
<ssvb>
FEL does not seem to leave the USB connection in a clean state and everything fails later
<arokux>
ssvb: the FEL uses normal USB port?
<ssvb>
when the u-boot/kernel is loaded from the SD card, everything kinda works
megal0maniac has joined #linux-sunxi
<ssvb>
arokux: FEL uses the same USB OTG connector
<ssvb>
arokux: it's some sort of a simple USB protocol implemented by BROM
wingrime1 has joined #linux-sunxi
wingrime has quit [Read error: Connection reset by peer]
<ssvb>
arokux: but don't worry about FEL right now
<arokux>
ssvb: I'm totally new to the gadget story. i didn't know about it. it is very cool. so you basically use g_cdc to unite two things g_ethernet and g_serial. with g_serial you can talk with device.
<arokux>
ssvb: and with g_ether you have it on the network?
<ssvb>
arokux: I use composite, specifically for having both ethernet and console on /dev/ttyACM0
<arokux>
ssvb: I do not quite understand how the network thing is working. it should be some kind of peer-to-peer connection, right? the other machines on the network cannot access the tablet
<ssvb>
arokux: it's a network connection with your PC, and you can configure NAT to provide Internet support for the device
<arokux>
ssvb: it is bending my mind :). with NAT you mean you set up routing on your PC?
<ssvb>
arokux: yes, when the device wants to connect to some random Internet address, your PC needs to be forwarding these network requests on behalf of your device
wingrime1 has quit [Read error: Connection reset by peer]
<ssvb>
hramrach_: "A typical MHL source will be shared with USB 2.0 on a standard 5-pin micro-usb receptacle, which switches from USB to MHL when it recognizes an MHL-qualified sink detected on the control wire."
<ssvb>
hramrach_: now I wonder how much bandwidth can it provide
esperegu has joined #linux-sunxi
<ssvb>
hramrach_: "pixel clock rate of up to 75 MHz for MHL 2.0, sufficient for 1080i and 720p"
<ssvb>
hramrach_: and for MHL 3.0 "Support of 4K formats up to 3840 × 2160 at 30 Hz"
<ssvb>
hramrach_: but indeed, that's an interesting idea
<ssvb>
hramrach_: looks like hansg has f%$^ed up https://github.com/linux-sunxi/u-boot-sunxi with one of his submarine unreviewed commits since the middle of Summer and now it is supposed to be urgent
<ssvb>
hramrach_: btw, the patch does not apply cleanly, but it is just a single number in a single line to change
<hramrach_>
ok. I hope I get to testing the a10s tomorrow
<ssvb>
or maybe somebody should just push it as-is without testing :)
<hramrach_>
there is a queue of stuff to test on it already but the board is really hard to work with
akaizen has joined #linux-sunxi
<hramrach_>
I was wondering why they do not maek a case for it but besides it being quite rare it has so insane layout building a case for it is pretty much impossible
akaizen has quit [Ping timeout: 260 seconds]
<hramrach_>
From all the descriptions it appears that it's either USB or mhl
<hramrach_>
and it's not clear at all that the mhl link can embed usb packets
<hramrach_>
or pretty much anything about MHL except connecting two MHL certified high end devices with a straight MHL cable is a good idea
<hramrach_>
well, you can use BT keyboard and mouse I guess but it's kind of lame then
<hramrach_>
I even wonder what those MHL to HDMI cables do
bonbons has quit [Quit: Leaving]
<hramrach_>
like we had too few display connection standards. HDMI, DP, Thunderbolt, MHL, ...
akaizen has joined #linux-sunxi
<hramrach_>
TB is at least clear on including everything and the kitchensink but then you cannot implement fully compliant TB on anything less than Tegra. it's supposed to have at least DP and PCIe multiplexed on the connector
Andy-D has joined #linux-sunxi
<hramrach_>
hmm, I wonder how long before ARM chip makers coming from mobile start including PCIe
Renard has quit [Ping timeout: 272 seconds]
<ssvb>
hramrach_: the ARM chip makers are already including PCIe
<ssvb>
hramrach_: NVIDIA and Qualcomm
akaizen_ has joined #linux-sunxi
<hramrach_>
Samsung in Exynos 5440
<hramrach_>
nVidia was making PCIe stuff before it started in the ARM business so it does not count
akaizen has quit [Ping timeout: 272 seconds]
<libv>
last i heard, amd still has broken pcie
<libv>
on their arm preproduction boards
wickwire has joined #linux-sunxi
<hramrach_>
not that I care
<hramrach_>
or is there anything interesting coming from AMD?
konradoo77 has quit [Ping timeout: 240 seconds]
<hramrach_>
hmm, so maybe we will see tablets with TB now ;-)
Black_Horseman has joined #linux-sunxi
<hramrach_>
but maybe not. who is in the TB business besides Intel and Apple? I guess neither of those would sell TB patent to random SoC or tablet makers to include on their devices
Renard has joined #linux-sunxi
konradoo77 has joined #linux-sunxi
nove has quit [Quit: nove]
tomboy65 has joined #linux-sunxi
tomboy64 has quit [Ping timeout: 264 seconds]
arokux2 has joined #linux-sunxi
jinzo has quit [Quit: Leaving]
hypothalamus has joined #linux-sunxi
sehraf has quit [Ping timeout: 260 seconds]
<ssvb>
libv: do the fex patches for sunxi-boards need review before pushing?
konradoo77 has quit [Ping timeout: 240 seconds]
<libv>
ssvb: we haven't done so
skaag has joined #linux-sunxi
<ssvb>
libv: I guess, I'll just push the primo73 & primo81 fex files to sunxi-boards
<ssvb>
libv: do you still want a dump of a31s dram registers?
<libv>
yes, and yes
<froese>
has anyone used the ps2 ports found on the A20? what are the electrical specs of those pins? are they 5V resistant or is some special circuitry required to connect a regular ps2 keyboard?
<froese>
s/resistant/tolerant/
<arokux2>
guys, please explain me how it works. a soft core is sold and then each and every client is writing his own buggy driver?
<froese>
silly, isn't it ;-) i guess, the programmer who writes the driver is cheaper than bying a software license.
<arokux2>
isn't it the case, that manufactures of the soft cores need to write a driver anyway to test it?
<froese>
arokux2: yes, and if a client wants that driver he has to pay for it.
<froese>
(don't know for sure - just my understanding of how that "IP-business" works)
hramrach_ has quit [Remote host closed the connection]
hramrach_ has joined #linux-sunxi
bengal has quit [Quit: Leaving]
petr has quit [Ping timeout: 240 seconds]
RzR is now known as rZr
deasy has quit [Quit: Nom d'un quark, c'est Edmonton !]
focus has quit [Read error: Connection reset by peer]
skaag has quit [Quit: Leaving.]
<arokux2>
there is even something like usb multipoint communication