« "Look Ma, No Busywaits!" | Home | Comments Enabled »
November 15, 2009
Internal Memory Bottlenecks and Their Removal
While debugging something different late last night, I noticed some flags in one of Glamo's registers which looked interesting: FIFO settings for the LCD engine. This reminded me of an observation by Lars a few weeks ago that the LCD engine seems to conflict with Glamo's 2D engine on memory accesses, leading to slower performance of accelerated 2D operations when the screen is switched on. So I turned the FIFO up to "8 stages" (from 1) to see what happened. The result was much faster 2D operations - literally twice the speed!
At "8 stages", the price of this was that the display became jittery and unstable. However, the same speed improvement is seen at the "4 stages" setting. I've also seen some occasional artifacts with this setting, so I'm using 2 stages at the moment, where the speed is still right up there. I'll be testing some more and seeing if things can be tuned even more.
Because we don't make the maximum use possible of the 2D engine, this doesn't immediately translate into a huge increase in the UI speed. But the differences are very obvious with x11perf and some of my test programs. The program I showed in the screenshot recently jumped from 45-48fps right up to 95-98fps!
At "8 stages", the price of this was that the display became jittery and unstable. However, the same speed improvement is seen at the "4 stages" setting. I've also seen some occasional artifacts with this setting, so I'm using 2 stages at the moment, where the speed is still right up there. I'll be testing some more and seeing if things can be tuned even more.
Because we don't make the maximum use possible of the 2D engine, this doesn't immediately translate into a huge increase in the UI speed. But the differences are very obvious with x11perf and some of my test programs. The program I showed in the screenshot recently jumped from 45-48fps right up to 95-98fps!
Hello, i have got one other question, does this could accelerate mplayer mpeg4 decode???? (with -vo glamo )?
This modification should make MPEG decoding go faster, but it hasn't been tested yet. A "DRI-ised" version of "-vo glamo" hasn't been made yet, but the FIFO fix should apply to the current kernel to make it faster. Note that since the X server and mplayer will end up trying to use the same command queue, by rights things should all fall apart. I don't know why it works at all without DRM...
Thanks for your answer, and very thanks for your work on glamo. :)
Thank you very much for your energies! We are looking forward to see the final result. So just one question about Glamo. Is it have any camera interface? If yes, what parameters it have and what kind of interface is it? (I hope this info is not under DRM agreement).
Is it possible to attach any CMOS chip to it?
thanks a lot one more time! you must open some donations wallet I think ;)
There's a CMOS camera interface According to the FR's schematics, the connections are grounded through four resistor packs. I don't know where they are on the physical board - possibly not anywhere useful, but I dont' know...