qole
|
2008-04-22
, 17:09
|
|
Moderator |
Posts: 7,109 |
Thanked: 8,820 times |
Joined on Oct 2007
@ Vancouver, BC, Canada
|
#161
|
|
2008-04-23
, 11:02
|
Posts: 30 |
Thanked: 2 times |
Joined on Jan 2007
@ Minnesota
|
#162
|
The Following User Says Thank You to Zhe For This Useful Post: | ||
|
2008-04-23
, 16:10
|
|
Moderator |
Posts: 7,109 |
Thanked: 8,820 times |
Joined on Oct 2007
@ Vancouver, BC, Canada
|
#163
|
The Following 3 Users Say Thank You to qole For This Useful Post: | ||
|
2008-04-24
, 20:23
|
|
Posts: 198 |
Thanked: 273 times |
Joined on Jan 2006
@ Helsinki, Finland
|
#164
|
I wouldn't say nothing, igor. Fanoush has done some very good things with the kernel, to name one example. If I understand correctly, though, one thing holding back community kernel work in general has been the frustration over getting improvements pushed upstream. Can anyone comment on that? I may be wrong.
|
2008-04-24
, 21:20
|
|
Posts: 11,700 |
Thanked: 10,045 times |
Joined on Jun 2006
@ North Texas, USA
|
#165
|
From my side I would have been happy to see some improvement to what was done for DVFS. Provided the cpufreq driver with all the settings defined - so that there is no risk of trashing the device - it would have been interesting to see some tuning on the cpufreq governor. I don't know if it can actually be significantly improved, but AFAIK nobody has really tried.
|
2008-04-24
, 21:26
|
|
Posts: 198 |
Thanked: 273 times |
Joined on Jan 2006
@ Helsinki, Finland
|
#166
|
If we're gonna overclock these things, we're gonna need itty bitty water-cooling systems.
|
2008-04-24
, 22:00
|
|
Posts: 11,700 |
Thanked: 10,045 times |
Joined on Jun 2006
@ North Texas, USA
|
#167
|
No, as i said no change of used OP.
But for example, as i've already explained several times, op selection is really hindred by the unrealistic constraints we had on MIPS available for audio playback.
The cpufreq policies would be quite easy to play with.