Re: Building meta-arago daisy

["Devin Hughes" <devin@jd2.com>]

> -----Original Message-----
> From: Denys Dmytriyenko [mailto:denys@ti.com]
> Sent: Friday, August 29, 2014 12:46 PM
> To: Devin Hughes
> Cc: 'Cooper Jr., Franklin'; meta-arago@arago-project.org
> Subject: Re: [meta-arago] Building meta-arago daisy
> 
> On Fri, Aug 29, 2014 at 11:58:09AM -0400, Devin Hughes wrote:
> > > > [Devin Hughes] Is the daisy configured build stable? I rebuilt the
> > > > image yesterday after the servers returned and it seems to run
> > > > unexpectedly. No boot errors, but the touchscreen and application
> > > > performance seem worse than the SDK release.
> > > [Franklin] Can you explain what you mean in terms of application
> > performance
> > > issues?
> > [Devin Hughes] Performance was the wrong choice of words now that I
> > know about the touchscreen trouble. It's more the permanent title bar
> > and lack of screen updates after closing an application. For instance,
> > after running one of the 3d demos in matrix gui, the title bar shows
> > the last opengl rendering on the top of the lcd.
> 
> 3D demos don't talk to Wayland stack and they take over fbdev completely
and
> previously Qt4-based refresh-screen app was used to update the screen
> afterwards.
> 
[Devin Hughes] This makes sense. 
> 
> > > I am going to build a customized image as a test next,
> > > > and I was thinking arago-base from daisy would be a good starting
> > > > point since I would like qt 5.2 or greater.
> > > [Franklin] So by default we are using Wayland/Weston and for Qt 5 we
> > > are
> > using
> > > the Wayland compositor. However, there are several touch screen
> > > related
> > bugs
> > > in Weston and Qt 5 that we are working with the Qt 5 and Weston team
> > > to address. It will take some time to fix so if acceptable I would
> > > recommend
> > just
> > > using a usb mouse as a work around for now if your running into
> > > those
> > problems
> > > also.
> > [Devin Hughes] USB mouse would be fine for me, but libusb is failing
> > to initialize. In dmesg I can see musb_init_controller failures, and
> > the platform driver reports a deferred probe. The only similar
> > information I can find on google describes improperly configured
> > device tree sources, but the fixes are already included in the daisy
> > referenced kernel.
> 
> Hmm, you said that is with 335x SK, right? I haven't seen this on
BeagleBone
> Black or 335x EVM. I'll try to see if I can reproduce it with SK...
[Devin Hughes] Yes, it's the 335x SK, and a vanilla arago-daisy conf with
gcc-4.7. No custom layers have been added yet.
> 
> --
> Denys