[ECOS] assabet LCD problem,

[Jonathan Larmour]

Andrew Xiang wrote:
> 
> When I write to the frame buffer, if I write too much
> without a delay in between, it will crash the whole
> system and reboot. I notice I had to put a
> cyg_thread_delay(1) in the middle of the continous
> write to the frame buffer in order to keep it from
> crashing. But it makes the display a new image very
> low. Anyone one had the same problem?

Not here, and we've used graphical demos, although I can't guarantee they
were hitting the LCD particularly hard.

> ex.
> 
> for(row=0;row<320;row++)
> {
> for(col=0;col<240;col++)
> {
> frame[row][col]=0;
> }
> cyg_thread_delay(1); //must have that to keep it from
> not crash.
> }
> 
> Note: I tried the same code on vxWorks, I did not have
> the problem. Is that mean problem with ECOS?

Or vxWorks is intrinsically slower? ;-)

Perhaps it is cache bursts. Lots of writes would mean that the cache could
spew them out really quickly. Try checking the MMU config to be sure the
frame buffer is definitely in uncacheable memory. If so, patches welcome
:-).

Jifl
-- 
Red Hat, Rustat House, Clifton Road, Cambridge, UK. Tel: +44 (1223) 271062
Maybe this world is another planet's Hell -Aldous Huxley || Opinions==mine