Let me quote the GLINT 500TX PRM: ;)
2.1.4 Task Switching
Where multiple applications wish to make simultaneous access to GLINT,
it is the responsibility of the software driving the chip to handle the
loading of the correct state. GLINT has been designed to support a
number of different software architectures. For instance some of the
facilities available are:
* Synchronous operation means that a new task can load its context
without waiting for current rendering to complete
* All loadable state can be read back
* Sync command to flush all rendering which can be polled or returned an
interrupt
If there was a kernel hook that could do this when context switching
between the X server and the OpenGL rendering process, they could access
the card as where they both alone on the computer.
I wont say yet that this is the best approach yet. But at least it is
possible. The Permedia chipset family can also do this. There are other
issues.
For example, in some circumstances it is best to do PIO using the
graphics pipeline FIFO, in other circumstances, it is better to use DMA.
Only experimentation will reveal what is the best way to do different
things.
-- Torgeir Veimo, Vertech AS,email: Torgeir.Veimo@vertech.no, mobile: 90673881, office: +47 55563755
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu