Our experience is that froyo CAN support mouse with no/little extra effort.
I recalled that I spent a little time to work through the path and might
have some small fixes, then it worked.
Please give us a tip ^_^
Sorry for the long delay. I would like to recommend this: http://blog.csdn.net/myarrow/article/details/7091061
for understanding of android input system.

I have looked at the old froyo code, no bug fixes are needed for the mouse to work. So just try to understand the input flow, find out what is wrong. e.g. does getevent(run it in adb shell) output something when you move the mouse? if so, does EventHub get the event? (enable or add some debugging information in frameworks/base/services/input for ics, in frameworks/base/libs/ui, EventHub.cpp, InputReader.cpp, InputDispatcher.cpp etc.)

ICS repository for loongson3 laptop is now uploaded to github.com/foxsen, the input system need the following patch(I don't see why it will not cause problem for other platform...)
     commit d7acbb5f2e4180eb5777dd055488f8276ff5abce
Author: repo sync <zhan...@lemote.com>
Date:   Sat Feb 4 14:53:31 2012 +0800

    fix wrong size of keybit/relbit etc. It causes failures of mouse

diff --git a/services/input/EventHub.h b/services/input/EventHub.h
index 048083c..a50de33 100644
--- a/services/input/EventHub.h
+++ b/services/input/EventHub.h
@@ -289,12 +289,12 @@ private:

         uint32_t classes;

-        uint32_t keyBitmask[(KEY_MAX + 1) / 32];
-        uint32_t absBitmask[(ABS_MAX + 1) / 32];
-        uint32_t relBitmask[(REL_MAX + 1) / 32];
-        uint32_t swBitmask[(SW_MAX + 1) / 32];
-        uint32_t ledBitmask[(LED_MAX + 1) / 32];
-        uint32_t propBitmask[(INPUT_PROP_MAX + 1) / 32];
+        uint32_t keyBitmask[(KEY_MAX + 32) / 32];
+        uint32_t absBitmask[(ABS_MAX + 32) / 32];
+        uint32_t relBitmask[(REL_MAX + 32) / 32];
+        uint32_t swBitmask[(SW_MAX + 32) / 32];
+        uint32_t ledBitmask[(LED_MAX + 32) / 32];
+        uint32_t propBitmask[(INPUT_PROP_MAX + 32) / 32];

         String8 configurationFile;
         PropertyMap* configuration;

The array size will be zero for relBitmask etc. with original code.

To get all code:
 mkdir loongson-ics
 cd loongson-ics
repo init --repo-url=git://github.com/foxsen/repo.git --repo-branch=stable -u git://github.com/foxsen/manifests.git -b loongson-ics-r0 -m loongson-ics-r0.xml
 repo sync

Will check the code when have time。
For 4.0, there is a bug prevent a mouse being recognized as CURSOR
device,will post the code(together with the other fixes) to my github
later.

BTW:
I have finally got some free time in last weekend to continue. Now
android ics for loongson is in a good shape for LS3 laptop:
opengl, alsa sound, mouse/keyboard, wifi, fake sdcard (to make lots of
apps happy), some mips apk from Junzheng nova system including QQ/金山
wps etc. partly works.
With some more work, lots of apps are hopefully to work with decent
performance:0
Nice news. If someday I can have a LS3 laptop, I will be glad to join you.



http://farm8.staticflickr.com/7208/6907265731_7b65a53f59.jpg

BTW, our project is up. Visit us at: https://github.com/lone-/manifests/wiki


--
You received this message because you are subscribed to the Google Groups 
"loongson-dev" group.
To post to this group, send email to loongson-dev@googlegroups.com.
To unsubscribe from this group, send email to 
loongson-dev+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/loongson-dev?hl=en.

Reply via email to