ianka / android-vnc

Automatically exported from code.google.com/p/android-vnc
GNU General Public License v2.0
0 stars 0 forks source link

Problem in fbvncserver referesh screen in and1.5 #7

Open GoogleCodeExporter opened 8 years ago

GoogleCodeExporter commented 8 years ago
What steps will reproduce the problem?
1. Problem in fbvncserver referesh screen in and1.5 

What is the expected output? What do you see instead?
   Dirty page is not identified in update_screen(void)in every keypress.
   only in two keypresses Dirty page  happends

What version of the product are you using? On what operating system?
Kernel Patch kernel worked ok
build fbvncserver worked finr

Please provide any additional information below.

Original issue reported on code.google.com by mnis...@marvell.com on 22 Aug 2009 at 6:14

GoogleCodeExporter commented 8 years ago
There are known problems with the image available from /dev/graphics/fb0 in 
recent
versions of Android, which may explain this problem.

This is because /dev/graphics/fb0 on Android does not accurately reflect the 
true
screen image due to double-buffering. Specifically it's caused by the
"PARTIAL_UPDATES" flag inside SurfaceFlinger, which appears to indicate that the
display hardware can be informed that only a certain area of the screen needs
updating - there is therefore no need for SurfaceFlinger to update the rest of 
the
fb0 device.

Similar problems are visible when using DDMS to grab screenshots, since that too
refers to /dev/graphics/fb0.

The true screen image is only guaranteed to be available inside SurfaceFlinger 
using
OpenGL calls; some future Android devices may not even have /dev/graphics/fb0.

We (at RealVNC) soon plan to submit some patches to the Android open-source 
project
which will fix this problem, by providing a new API from SurfaceFlinger to grab 
the
screen. It's yet to be determined exactly how this API will be secured, though, 
so
it's not clear whether it will be accessible to fbvncserver.

Original comment by adetay...@gmail.com on 24 Feb 2010 at 10:29