nanogui: nano-X, correct implementation of touchpanel driver?


Previous by date: 31 Oct 2005 01:04:01 +0000 viewml linker errors, krishna_mca52.sify.com
Next by date: 31 Oct 2005 01:04:01 +0000 Nano-X stays at command prompt, Patrick Boyd
Previous in thread:
Next in thread:

Subject: nano-X, correct implementation of touchpanel driver?
From: Sean Machin ####@####.####
Date: 31 Oct 2005 01:04:01 +0000
Message-Id: <43656CF7.1020305@yahoo.com>

Hi Greg & others,

I'm trying to get a touchpanel driver for my custom hardware to work 
correctly, and with nano-X.

My Linux driver creates an entry in /dev which can be read.  The driver 
read implementation
returns  -EAGAIN if there's no new data to read, else returns an array 
of 5 bytes
(up/down flag & x,y coords as 16 bit values) if the screen had been 
pressed or released and there
was thus new data to read.  This seems to be the correct way to behave 
according to my
Linux device driver book for a char device opened with O_NONBLOCK.
Is this the correct way data should be returned for a touchpanel
driver for nano-X?

I am using mou_touchscreen.c to interface the above to nano-X.  I only 
had to change
some code in PD_Read to interpret.  As per the original  code it will 
return 0 if
there was no new data to read (driver returned -EAGAIN), else will return
2 if the touchpanel press was down, else 3 if the touchpanel was released.
I take it this is also the correct behavior?
I have verified that PD_Read returns the coordinates and 2 and then 3 
when the touchpanel
is pressed then released.

The problem I currently see is that the mouse pointer never moves or 
responds to
click events at all (my mou_touchscreen code did not hide the mouse 
pointer).

Can anyone advise what might be going wrong?

Thanks,
Sean


Previous by date: 31 Oct 2005 01:04:01 +0000 viewml linker errors, krishna_mca52.sify.com
Next by date: 31 Oct 2005 01:04:01 +0000 Nano-X stays at command prompt, Patrick Boyd
Previous in thread:
Next in thread:


Powered by ezmlm-browse 0.20.