can it be done at kernel level itself... i mean to rotate the framebuffer itself and then calibrating the touch accordingly... is there any such option which we can specify in kernel commandline itelf (Intel based architecture)?? I remember one such option on omap boards omapfb.rotate=<angle>
regards, dhaval. On Sun, Nov 14, 2010 at 1:15 AM, Arjan van de Ven <[email protected]>wrote: > On 11/13/2010 11:37 AM, Carsten Munk wrote: > >> 2010/11/13 Arjan van de Ven<[email protected]>: >> >>> On 11/13/2010 11:29 AM, Javier wrote: >>> >>>> I'm trying the Handset 1.1 final-demo images on my N900, and I've found >>>> that in both landscape and portrait modes, the X11 root window is >>>> 800x480 >>>> (that is, always portrait). >>>> >>>> However, Qt applications (for example duicontrolpanel) render just fine >>>> in portrait. This means they (or Qt) must be doing some kind of rotation >>>> implemented on the X clients themselves. >>>> >>>> To be honest, this comes a bit of a surprise. Is it really expected that >>>> every X11 client application will have to implement their own means for >>>> rotation, or is the plan for rotation to be eventually implemented the >>>> XRandR way? >>>> >>> eventually it has to be done with XRandR for power/performance reasons; >>> the X server can do this more performance (and thus power) efficient than >>> the apps can, depending on what >>> exact hardware you have. (Some graphics hardware can use the new X 1.9 >>> stuff >>> for doing rotation centrally) >>> >> I've actually heard the opposite argument - not using xrandr for >> performance reasons. As it's faster to rotate a surface with the >> application on with GL/GLES than with xrandr. >> > > guess it depends on the (version of the) graphics driver..... > > > _______________________________________________ > MeeGo-dev mailing list > [email protected] > http://lists.meego.com/listinfo/meego-dev >
_______________________________________________ MeeGo-dev mailing list [email protected] http://lists.meego.com/listinfo/meego-dev
