fbdev hardware cursor support

fbdev hardware cursor support

[Jon Smirl]

What is the status of fbdev hardware cursor support? I see the ioctl
is disabled in the driver.

If a driver is using the softcursor support would the ioctl control that cursor?
Does softcursor work in a bitmap graphics mode?

The EGL based X server is going to need fbdev hardware cursors working.

-- 
Jon Smirl
jonsmirl@gmail.com


-------------------------------------------------------
This SF.Net email is sponsored by Yahoo.
Introducing Yahoo! Search Developer Network - Create apps using Yahoo!
Search APIs Find out how you can build Yahoo! directly into your own
Applications - visit http://developer.yahoo.net/?fr=offad-ysdn-ostg-q22005

Re: fbdev hardware cursor support

[Ville Syrjälä]

On Wed, May 25, 2005 at 01:40:57AM -0400, Jon Smirl wrote:
> What is the status of fbdev hardware cursor support? I see the ioctl
> is disabled in the driver.
> 
> If a driver is using the softcursor support would the ioctl control that cursor?
> Does softcursor work in a bitmap graphics mode?
> 
> The EGL based X server is going to need fbdev hardware cursors working.

Why doesn't it draw the cursor using OpenGL? On many cards the hardware 
cursor is too limited to allow nice looking cursors.

-- 
Ville Syrjälä
syrjala@sci.fi
http://www.sci.fi/~syrjala/


-------------------------------------------------------
This SF.Net email is sponsored by Yahoo.
Introducing Yahoo! Search Developer Network - Create apps using Yahoo!
Search APIs Find out how you can build Yahoo! directly into your own
Applications - visit http://developer.yahoo.net/?fr=offad-ysdn-ostg-q22005

Re: fbdev hardware cursor support

[Jon Smirl]

On 5/25/05, Ville Syrjälä <syrjala@sci.fi> wrote:
> On Wed, May 25, 2005 at 01:40:57AM -0400, Jon Smirl wrote:
> > What is the status of fbdev hardware cursor support? I see the ioctl
> > is disabled in the driver.
> >
> > If a driver is using the softcursor support would the ioctl control that cursor?
> > Does softcursor work in a bitmap graphics mode?
> >
> > The EGL based X server is going to need fbdev hardware cursors working.
> 
> Why doesn't it draw the cursor using OpenGL? On many cards the hardware
> cursor is too limited to allow nice looking cursors.

Software generated will be supported sooner or later but the code is
not written yet.

What about video overlays? When the cursor goes out of the OpenGL area
and into the video one it has to be hardware generated.

A hardware cursor is also easier for me in the unstable drawing
environment or a server under development.


-- 
Jon Smirl
jonsmirl@gmail.com


-------------------------------------------------------
SF.Net email is sponsored by: GoToMeeting - the easiest way to collaborate
online with coworkers and clients while avoiding the high cost of travel and
communications. There is no equipment to buy and you can meet as often as
you want. Try it free.http://ads.osdn.com/?ad_idt02&alloc_id\x16135&op=click

Re: fbdev hardware cursor support

[Michel Dänzer]

On Wed, 2005-05-25 at 10:10 -0400, Jon Smirl wrote:
> On 5/25/05, Ville Syrjälä <syrjala@sci.fi> wrote:
> > On Wed, May 25, 2005 at 01:40:57AM -0400, Jon Smirl wrote:
> > > What is the status of fbdev hardware cursor support? I see the ioctl
> > > is disabled in the driver.
> > >
> > > If a driver is using the softcursor support would the ioctl control that cursor?
> > > Does softcursor work in a bitmap graphics mode?
> > >
> > > The EGL based X server is going to need fbdev hardware cursors working.
> > 
> > Why doesn't it draw the cursor using OpenGL? On many cards the hardware
> > cursor is too limited to allow nice looking cursors.
> 
> Software generated will be supported sooner or later but the code is
> not written yet.

These claims of yours have been refuted on the dri-egl list. Why are you
repeating them here?


> What about video overlays? When the cursor goes out of the OpenGL area
> and into the video one it has to be hardware generated.

We can worry about that once we've figured out how to expose video
overlays in Xgl, especially with compositing...


-- 
Earthling Michel Dänzer      |     Debian (powerpc), X and DRI developer
Libre software enthusiast    |   http://svcs.affero.net/rm.php?r=daenzer



-------------------------------------------------------
SF.Net email is sponsored by: GoToMeeting - the easiest way to collaborate
online with coworkers and clients while avoiding the high cost of travel and
communications. There is no equipment to buy and you can meet as often as
you want. Try it free.http://ads.osdn.com/?ad_idt02&alloc_id\x16135&op=click

Re: fbdev hardware cursor support

[Jon Smirl]

On 5/25/05, Michel Dänzer <michel@daenzer.net> wrote:
> > Software generated will be supported sooner or later but the code is
> > not written yet.
> 
> These claims of yours have been refuted on the dri-egl list. Why are you
> repeating them here?

I'm writing the Xegl server and I can tell you for sure that I haven't
written the software cursor support yet. How can you know the status
of the Xgl code when I haven't checked it into CVS yet?

Personally I'd prefer to start with the hardware one which I know will
work. Drawing in Xgl is not stable and there is a good chance a
software generated cursor isn't going to work half of the time until
the code is fixed.

> > What about video overlays? When the cursor goes out of the OpenGL area
> > and into the video one it has to be hardware generated.
> 
> We can worry about that once we've figured out how to expose video
> overlays in Xgl, especially with compositing...
> 
> 
> --
> Earthling Michel Dänzer      |     Debian (powerpc), X and DRI developer
> Libre software enthusiast    |   http://svcs.affero.net/rm.php?r=daenzer
> 
> 
> 
> -------------------------------------------------------
> SF.Net email is sponsored by: GoToMeeting - the easiest way to collaborate
> online with coworkers and clients while avoiding the high cost of travel and
> communications. There is no equipment to buy and you can meet as often as
> you want. Try it free.http://ads.osdn.com/?ad_idt02&alloc_id\x16135&opclick
> _______________________________________________
> Linux-fbdev-devel mailing list
> Linux-fbdev-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/linux-fbdev-devel
> 


-- 
Jon Smirl
jonsmirl@gmail.com


-------------------------------------------------------
SF.Net email is sponsored by: GoToMeeting - the easiest way to collaborate
online with coworkers and clients while avoiding the high cost of travel and
communications. There is no equipment to buy and you can meet as often as
you want. Try it free.http://ads.osdn.com/?ad_idt02&alloc_id\x16135&op=click

Re: fbdev hardware cursor support

[Michel Dänzer]

On Wed, 2005-05-25 at 11:08 -0400, Jon Smirl wrote:
> On 5/25/05, Michel Dänzer <michel@daenzer.net> wrote:
> > > Software generated will be supported sooner or later but the code is
> > > not written yet.
> > 
> > These claims of yours have been refuted on the dri-egl list. Why are you
> > repeating them here?
> 
> I'm writing the Xegl server and I can tell you for sure that I haven't
> written the software cursor support yet. How can you know the status
> of the Xgl code when I haven't checked it into CVS yet?

David Reveman pointed out that Xglx has working software cursor code. I
can't see that caring about whether glitz uses a GLX or EGL backend.

> Personally I'd prefer to start with the hardware one which I know will
> work. Drawing in Xgl is not stable and there is a good chance a
> software generated cursor isn't going to work half of the time until
> the code is fixed.

And for the hardware cursor, you only need to:

      * enable/implement/improve (does what is disabled right now even
        remotely fit the needs of an X server? I didn't think so) the
        fbdev hardware cursor support
      * design and implement an EGL extension for it
      * add support for that to glitz
      * add support for that to Xegl

Indeed, that seems much easier and less prone to problems...


PS: Please don't add random e-mail addresses to the CC: list.

-- 
Earthling Michel Dänzer      |     Debian (powerpc), X and DRI developer
Libre software enthusiast    |   http://svcs.affero.net/rm.php?r=daenzer



-------------------------------------------------------
SF.Net email is sponsored by: GoToMeeting - the easiest way to collaborate
online with coworkers and clients while avoiding the high cost of travel and
communications. There is no equipment to buy and you can meet as often as
you want. Try it free.http://ads.osdn.com/?ad_idt02&alloc_id\x16135&op=click

Re: fbdev hardware cursor support

[Jon Smirl]

On 5/25/05, Michel Dänzer <michel@daenzer.net> wrote:
> On Wed, 2005-05-25 at 11:08 -0400, Jon Smirl wrote:
> > On 5/25/05, Michel Dänzer <michel@daenzer.net> wrote:
> > > > Software generated will be supported sooner or later but the code is
> > > > not written yet.
> > >
> > > These claims of yours have been refuted on the dri-egl list. Why are you
> > > repeating them here?
> >
> > I'm writing the Xegl server and I can tell you for sure that I haven't
> > written the software cursor support yet. How can you know the status
> > of the Xgl code when I haven't checked it into CVS yet?
> 
> David Reveman pointed out that Xglx has working software cursor code. I
> can't see that caring about whether glitz uses a GLX or EGL backend.

Xglx gets it software cursor support from the nested X server it runs
inside of. I don't have this server and consequentially I don't have a
software cursor.

Same reason I'm having to build an event pump into Xegl. Xglx used the
nested X server for it's event pump.

-- 
Jon Smirl
jonsmirl@gmail.com


-------------------------------------------------------
SF.Net email is sponsored by: GoToMeeting - the easiest way to collaborate
online with coworkers and clients while avoiding the high cost of travel and
communications. There is no equipment to buy and you can meet as often as
you want. Try it free.http://ads.osdn.com/?ad_idt02&alloc_id\x16135&op=click

Re: fbdev hardware cursor support

[Michel Dänzer]

On Wed, 2005-05-25 at 11:30 -0400, Jon Smirl wrote:
> On 5/25/05, Michel Dänzer <michel@daenzer.net> wrote:
> > On Wed, 2005-05-25 at 11:08 -0400, Jon Smirl wrote:
> > > On 5/25/05, Michel Dänzer <michel@daenzer.net> wrote:
> > > > > Software generated will be supported sooner or later but the code is
> > > > > not written yet.
> > > >
> > > > These claims of yours have been refuted on the dri-egl list. Why are you
> > > > repeating them here?
> > >
> > > I'm writing the Xegl server and I can tell you for sure that I haven't
> > > written the software cursor support yet. How can you know the status
> > > of the Xgl code when I haven't checked it into CVS yet?
> > 
> > David Reveman pointed out that Xglx has working software cursor code. I
> > can't see that caring about whether glitz uses a GLX or EGL backend.
> 
> Xglx gets it software cursor support from the nested X server it runs
> inside of.

You didn't even read David's correction of this. Great.


-- 
Earthling Michel Dänzer      |     Debian (powerpc), X and DRI developer
Libre software enthusiast    |   http://svcs.affero.net/rm.php?r=daenzer



-------------------------------------------------------
SF.Net email is sponsored by: GoToMeeting - the easiest way to collaborate
online with coworkers and clients while avoiding the high cost of travel and
communications. There is no equipment to buy and you can meet as often as
you want. Try it free.http://ads.osdn.com/?ad_idt02&alloc_id\x16135&op=click

Re: fbdev hardware cursor support

[Jon Smirl]

On 5/25/05, Michel Dänzer <michel@daenzer.net> wrote:
> On Wed, 2005-05-25 at 11:30 -0400, Jon Smirl wrote:
> > On 5/25/05, Michel Dänzer <michel@daenzer.net> wrote:
> > > On Wed, 2005-05-25 at 11:08 -0400, Jon Smirl wrote:
> > > > On 5/25/05, Michel Dänzer <michel@daenzer.net> wrote:
> > > > > > Software generated will be supported sooner or later but the code is
> > > > > > not written yet.
> > > > >
> > > > > These claims of yours have been refuted on the dri-egl list. Why are you
> > > > > repeating them here?
> > > >
> > > > I'm writing the Xegl server and I can tell you for sure that I haven't
> > > > written the software cursor support yet. How can you know the status
> > > > of the Xgl code when I haven't checked it into CVS yet?
> > >
> > > David Reveman pointed out that Xglx has working software cursor code. I
> > > can't see that caring about whether glitz uses a GLX or EGL backend.
> >
> > Xglx gets it software cursor support from the nested X server it runs
> > inside of.
> 
> You didn't even read David's correction of this. Great.

If there is a working software cursor implementation in Xgl I'm yet to
find it. Of course I am looking at 5,000+ lines of code without a
single comment other than the copyright notice. On the other hand I
can definitely see how the xnest cursor implementation is working.

Feel free to help with Xegl. I have stated several times that I know
nothing about the X input system. I've identified exactly where to
hook up hardware cursor support and that would let me proceed from
where I am at.

-- 
Jon Smirl
jonsmirl@gmail.com


-------------------------------------------------------
SF.Net email is sponsored by: GoToMeeting - the easiest way to collaborate
online with coworkers and clients while avoiding the high cost of travel and
communications. There is no equipment to buy and you can meet as often as
you want. Try it free.http://ads.osdn.com/?ad_idt02&alloc_id\x16135&op=click

Re: fbdev hardware cursor support

[Jon Smirl]

On 5/25/05, Michel Dänzer <michel@daenzer.net> wrote:
> > > David Reveman pointed out that Xglx has working software cursor code. I
> > > can't see that caring about whether glitz uses a GLX or EGL backend.
> >
> > Xglx gets it software cursor support from the nested X server it runs
> > inside of.
> 
> You didn't even read David's correction of this. Great.

Three days later I have software cursors working. It wasn't anywhere
near as simple as turning on a flag. They're ugly and they lag, but
they can probably be fixed with more effort.  I would have had the
hardware one up in a day at most since I knew exactly what was doing.

-- 
Jon Smirl
jonsmirl@gmail.com


-------------------------------------------------------
This SF.Net email is sponsored by Yahoo.
Introducing Yahoo! Search Developer Network - Create apps using Yahoo!
Search APIs Find out how you can build Yahoo! directly into your own
Applications - visit http://developer.yahoo.net/?fr=offad-ysdn-ostg-q22005

Re: fbdev hardware cursor support

[James Simmons]

On Wed, 25 May 2005, Jon Smirl wrote:

> What is the status of fbdev hardware cursor support? I see the ioctl
> is disabled in the driver.
> 
> If a driver is using the softcursor support would the ioctl control that cursor?
> Does softcursor work in a bitmap graphics mode?
> 
> The EGL based 

Some time ago we had userland cursor support but it was removed :-( Now 
that we have sysfs it could make a come back. The software cursor code 
doesn't behave proprely from userlane. It was design to work with the 
console system only. 



-------------------------------------------------------
This SF.Net email is sponsored by: NEC IT Guy Games.  How far can you shotput
a projector? How fast can you ride your desk chair down the office luge track?
If you want to score the big prize, get to know the little guy.  
Play to win an NEC 61" plasma display: http://www.necitguy.com/?r=20

Re: fbdev hardware cursor support

[Jon Smirl]

On 6/6/05, James Simmons <jsimmons@www.infradead.org> wrote:
> Some time ago we had userland cursor support but it was removed :-( Now
> that we have sysfs it could make a come back. The software cursor code
> doesn't behave proprely from userland. It was design to work with the
> console system only.

I have software cursors going in Xegl now so I am ok for the moment.
At some point we need to resurrect the hardware cursor support in
fbdev.

-- 
Jon Smirl
jonsmirl@gmail.com


-------------------------------------------------------
This SF.Net email is sponsored by: NEC IT Guy Games.  How far can you shotput
a projector? How fast can you ride your desk chair down the office luge track?
If you want to score the big prize, get to know the little guy.
Play to win an NEC 61" plasma display: http://www.necitguy.com/?r