Christian Gmeiner
2012-04-18 08:55:12 UTC
Hi all,
the company I am employed wants to support the further development of
the geode driver.
At the moment we are discusing this internaly how we could do this.
Sure the best way would
be to provide patches that gets included in next releases of the
driver, but our knowledge
in x11 driver development and the geode video hardware is quite
limited and there is also a
lack of man power. So we think the best way would be to sponsor some
money so that
a more suitable developer could work for full time as a freelancer on
the driver.
I am not sure how this developer could be at the moment.
As far as I can tell only one thing should be improved - the overall
performance of the driver.
--- Get rid of ping-pong pixmaps ---
"All these fallbacks would also be less costly, if we didn't end up
migrating the pixmaps out temporarily from video memory to system
memory. I believe that theoretically pixman based software fallbacks
could be done inside video memory, but I didn't get EXA classic mode
convinced of that yet to try it out. As in, video memory is really just
CPU memory too, just framebuffer memory set aside into which hardware
accelerated operations can put their result in; so if we were doing
software operations in there as well if there's enough room, without
migrating out of there, we could avoid the costly part of copying memory
around too much and fix it for all cases to not be SO big of a
performance hit, while it'd be still nice to do in hardware what is
possible." [1]
"All these fallbacks would also be less costly, if we didn't end up
migrating the pixmaps out temporarily from video memory to system
memory. I believe that theoretically pixman based software fallbacks
could be done inside video memory, but I didn't get EXA classic mode
convinced of that yet to try it out. As in, video memory is really just
CPU memory too, just framebuffer memory set aside into which hardware
accelerated operations can put their result in; so if we were doing
software operations in there as well if there's enough room, without
migrating out of there, we could avoid the costly part of copying memory
around too much and fix it for all cases to not be SO big of a
performance hit, while it'd be still nice to do in hardware what is
possible." [2]
I can not tell how much time this takes to implement it, but maybe
Martin can give some assessment.
What do you think about it? I am sorry that I do not write this mail from
my company mail address but gmail has the better spam filter :)
[1] http://lists.x.org/archives/xorg-driver-geode/2011-April/001209.html
[2] http://www.mail-archive.com/xorg-driver-***@lists.x.org/msg00971.html
--
Christian Gmeiner, MSc
the company I am employed wants to support the further development of
the geode driver.
At the moment we are discusing this internaly how we could do this.
Sure the best way would
be to provide patches that gets included in next releases of the
driver, but our knowledge
in x11 driver development and the geode video hardware is quite
limited and there is also a
lack of man power. So we think the best way would be to sponsor some
money so that
a more suitable developer could work for full time as a freelancer on
the driver.
I am not sure how this developer could be at the moment.
As far as I can tell only one thing should be improved - the overall
performance of the driver.
--- Get rid of ping-pong pixmaps ---
"All these fallbacks would also be less costly, if we didn't end up
migrating the pixmaps out temporarily from video memory to system
memory. I believe that theoretically pixman based software fallbacks
could be done inside video memory, but I didn't get EXA classic mode
convinced of that yet to try it out. As in, video memory is really just
CPU memory too, just framebuffer memory set aside into which hardware
accelerated operations can put their result in; so if we were doing
software operations in there as well if there's enough room, without
migrating out of there, we could avoid the costly part of copying memory
around too much and fix it for all cases to not be SO big of a
performance hit, while it'd be still nice to do in hardware what is
possible." [1]
"All these fallbacks would also be less costly, if we didn't end up
migrating the pixmaps out temporarily from video memory to system
memory. I believe that theoretically pixman based software fallbacks
could be done inside video memory, but I didn't get EXA classic mode
convinced of that yet to try it out. As in, video memory is really just
CPU memory too, just framebuffer memory set aside into which hardware
accelerated operations can put their result in; so if we were doing
software operations in there as well if there's enough room, without
migrating out of there, we could avoid the costly part of copying memory
around too much and fix it for all cases to not be SO big of a
performance hit, while it'd be still nice to do in hardware what is
possible." [2]
I can not tell how much time this takes to implement it, but maybe
Martin can give some assessment.
What do you think about it? I am sorry that I do not write this mail from
my company mail address but gmail has the better spam filter :)
[1] http://lists.x.org/archives/xorg-driver-geode/2011-April/001209.html
[2] http://www.mail-archive.com/xorg-driver-***@lists.x.org/msg00971.html
--
Christian Gmeiner, MSc