try the very latest version. is has slightly higher voltage for 900.
950 work for me too but 1GHz crashes.

About DSP clock - we are seeing reports of noted difference between kernels with 430 and with 500 DSP in watching videos in BBC iplayer. While this reports are not from blind tests, and might be affected by placebo effect, I tend to accept them as proof that DSP clock does change.

1000mhz same voltage as stock 600mhz and 1100mhz overvolted

I wonder whether the file drivers/dsp/bridge/rmgr/drv_interface.c
is not used on OMAP? it seems to contain the DSP clocks.

so guys, what i understood from reading the posts that the LV and ULV kernels give the performance of 900Mhz (for ex) with lower voltage? Or do you need to increase the clock speed because you are using lower voltage? can anyone explain plz.

It is used, but all it does is write the table in shared memory where the DSP can find it. This allows the DSP to request certain frequency, but it is still the Linux kernel that actually sets the frequency. We don't really know if the DSP program actually uses it.

In addition I recommend you to run this commands and test it again:

renice 1 `cat /syspart/applications/standby/background/tasks`
renice 1 `ps | grep modest | cut -c1-5`

It should save you a lot of battery over standby. I have increase more then 50%.

(Note: modest renice would work until you call e-mail GUI, after that it reverts back. I am working on it)

125-900 Mhz Kernel, Stock DSP:

Calculation of PI using FFT and AGM, ver. LG1.1.2-MP1.5.2a.memsave
Usage: ./pi_css5-vfp3 digits
Number of digits of pi to calculate? 1000000 initializing...
nfft= 262144
radix= 10000
error_margin= 0.00611182
calculating 1048576 digits of PI...
AGM iteration
precision= 48: 5.01 sec
precision= 80: 5.00 sec
precision= 176: 5.02 sec
precision= 352: 5.03 sec
precision= 688: 5.00 sec
precision= 1392: 5.02 sec
precision= 2784: 5.00 sec
precision= 5584: 5.04 sec
precision= 11168: 5.06 sec
precision= 22336: 4.99 sec
precision= 44688: 5.01 sec
precision= 89408: 5.00 sec
precision= 178816: 4.99 sec
precision= 357648: 4.99 sec
precision= 715312: 5.01 sec
precision= 1430640: 5.01 sec
writing pi1048576.txt...
93.72 sec. (real time)


ULV - 250-1000 Mhz Kernel (Running 1000 Max)

Calculation of PI using FFT and AGM, ver. LG1.1.2-MP1.5.2a.memsave
Usage: ./pi_css5-vfp3 digits
Number of digits of pi to calculate? 1000000 initializing...
nfft= 262144
radix= 10000
error_margin= 0.00611182
calculating 1048576 digits of PI...
AGM iteration
precision= 48: 4.63 sec
precision= 80: 4.65 sec
precision= 176: 4.63 sec
precision= 352: 4.66 sec
precision= 688: 4.66 sec
precision= 1392: 4.64 sec
precision= 2784: 4.64 sec
precision= 5584: 4.61 sec
precision= 11168: 4.64 sec
precision= 22336: 4.58 sec
precision= 44688: 4.57 sec
precision= 89408: 4.60 sec
precision= 178816: 4.60 sec
precision= 357648: 4.61 sec
precision= 715312: 4.61 sec
precision= 1430640: 4.58 sec
writing pi1048576.txt...
86.32 sec. (real time)

ULV - 250-1000 Mhz Kernel (Running 800 Max)
.
.

103.15 sec (real time)

Whats the diffrence between the ulv & lv kernels?
Do they have other Frequences? or really only other voltage use?

Titan

I don't have anything available under /sys/class/power_supply/bq27200-0/ with your latest update. The class power_supply doesn't exist on my device. What battery stats should be available in there? Rest seems to work fine with your kernel (max freq set to 810 MHz). I am curious whether my battery life would actually be better if I now idle, at a lower voltage at 125Mhz. Will see how it goes.

1200000 0
1100000 0
1000000 0
950000 0
900000 0
850000 0
810000 8963
750000 87
700000 264
600000 113
550000 22
500000 2260
250000 913
124999 19546