On Sun, 2003-09-28 at 21:57, Jacek Rosik wrote:
> W liście z nie, 28-09-2003, godz. 20:24, Keith Whitwell pisze: 
> > Jacek Rosik wrote:
> > > 
> > > 1) About CP. As I understand it it's some kind of queue for commands.
> > > So, is there such posibility that if I write some value into register
> > > using it and after that another value is written using RADEON_WRITE the
> > > second value will be overwritten by first one? 

Yes.

> > > Is it flushed on context change? For example when server grabs a lock 
> > > after client.

Not necessarily AFAIK; gratuitous flushing of the CP ring could affect
performance.

> > If you want to write registers in the hardware while the CP is active, it's 
> > typically necessary to use CP commands to do so.  However, it does depend 
> > which register you are talking about.
> 
> RADEON_CRTC_OFFSET/RADEON_CRTC2_OFFSET in order to flip buffers.

Those aren't FIFO'd AFAIK, so writing to them directly should be safe
even while the CP is running. Writing to a FIFO'd register directly
while the CP is running can lock up if the CP happens to have filled the
FIFO.


> > > 2) Can I block radeon interrupt (swi and vbl) and if so how can I do it?
> > > What I mean is to block them for a while and not to loose any.

Would disabling the interrupts on the chip (by writing to the
GEN_INT_CNTL register) be enough for your purpose? Could lose or at
least delay interrupts though...


-- 
Earthling Michel Dänzer   \  Debian (powerpc), XFree86 and DRI developer
Software libre enthusiast  \     http://svcs.affero.net/rm.php?r=daenzer



-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
Dri-devel mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/dri-devel

Reply via email to