Perflock Question - HTC Rezound

First, what does perflock do? Second, why can it not be turned off like on the DInc? These might be dumb questions, but I haven't seen anyone else ask or really go into detail. Seems pretty important that a workaround is found
Sent from my ADR6425LVW using xda app-developers app

---

Well because you guys are lacking source code for ics perflock can not be disabled
Once you have source you will be golden
Sent from my HTC One X using Tapatalk 2

---

superchilpil said:
Well because you guys are lacking source code for ics perflock can not be disabled
Once you have source you will be golden
Sent from my HTC One X using Tapatalk 2
Click to expand...
Click to collapse
Thanks for the answer, but what is perflock exactly? I've googled, but no really good answer
Sent from my ADR6425LVW using xda app-developers app

---

haliwa04 said:
Thanks for the answer, but what is perflock exactly? I've googled, but no really good answer
Sent from my ADR6425LVW using xda app-developers app
Click to expand...
Click to collapse
http://www.setcpu.com/documentation.html#disabling_perflock
Perflock attempts to maintain the default clock speed settings of the kernel. Programs such as SetCPU attempt to override that. Presumably the technique has varying success on newer kernels, depending on HTC's implementation. I know min/max values on ICS were still unstable for some that were trying the overclocked stock kernel. The values were in a tug-of-war between SetCPU and Perflock, causing jumps to the max value in the clock speed.
Custom kernels can be built without perflock; need kernel source first to make those for the most part.

---

This is probably a dumb question but why is source so important to kernels? Can't it be decompiled and in turn wouldn't that allow you to see the code? I'm trying to wrap my head around it. I've never asked why I just assumed that wouldn't be possible. Now I'm wanting to know why and if there isn't q way to circumvent needing source
sent from my rezound running AOKP

---

pwned3 said:
This is probably a dumb question but why is source so important to kernels? Can't it be decompiled and in turn wouldn't that allow you to see the code? I'm trying to wrap my head around it. I've never asked why I just assumed that wouldn't be possible. Now I'm wanting to know why and if there isn't q way to circumvent needing source
sent from my rezound running AOKP
Click to expand...
Click to collapse
From what I know its a one way thing, you build it but you can't unbuild it
Correct me if I'm wrong
There's a way to overclock but those are dirty kernels and not too popular
Sent from my HTC One X using Tapatalk 2

---

superchilpil said:
From what I know its a one way thing, you build it but you can't unbuild it
Correct me if I'm wrong
There's a way to overclock but those are dirty kernels and not too popular
Sent from my HTC One X using Tapatalk 2
Click to expand...
Click to collapse
I'm in the same boat I don't think it can be done but I'm trying to learn what prevents it from being reverse engineered
sent from my rezound running AOKP

---

That's I'm talkin about! Discussion, explanations, all of it.
Sent from my ADR6425LVW using xda app-developers app

---

haliwa04 said:
That's I'm talkin about! Discussion, explanations, all of it.
Sent from my ADR6425LVW using xda app-developers app
Click to expand...
Click to collapse
Some people have been using apps like setcpu to play a bit. I think you can change the i/o with it. Also, it seems at some point (check enable profiles) the 2nd core is turned on. You can also underclock and it doesn't seem to drive perflock crazy like it did before. I can see a direct correlation between benchmark and my set freq. Hopefully, it won't be too long before we get some custom kernels; then the fun begins. AND maybe we get 'quick charge'. That's what I miss the most from my Dinc.

---

superchilpil said:
From what I know its a one way thing, you build it but you can't unbuild it
Correct me if I'm wrong
There's a way to overclock but those are dirty kernels and not too popular
Sent from my HTC One X using Tapatalk 2
Click to expand...
Click to collapse
The way compiling works is it takes code, such as C++, Python and other programming languages, and converts them into binary.
Binary is what the hardware speaks, as cpu's, gpu's (etc..) dont understand programming languages. (BTW this is a very simplistic outlining of compiling, because more goes into it, such as hardware specific optimizations and so on)
So lets say you have a binary file. That file contains instructions for the CPU to perform (CPU Instruction Set). Those instructions are for example: Read Data, Move Data, Perform Subtraction/Addition/Multiplication etc..
These instuctions do not translate back into code very well, because there is a multitude of ways in a language to accomplish the same task. So going from a compiled binary to source code via reverse engineering tools like decompilers just doesnt work. Take this for example:
The binary instruction set for a file makes "Hello World" display on screen.
Here are the immediate problems you will encounter: What language was the source written in? Because if the correct language decompiler is used, bits and pieces of code can be seen, but sometimes its not even enough to give a hint of whats going on.
Lets say for the simplicity of the example its written in Powershell, or a similar scripting language...
I could display "Hello World" simply by using an echo command.
I could be calling individual variables in the echo such as "echo $h $w"
Hello World could be whats being displayed, AFTER a function successfully completes (if 2+2 = 4 then echo hello world)
So just from those three (There are plenty more ways to do this) you can see why decompiling and reverse engineering isnt an option. Its just a better decision to wait, rather than try to rebuild the wheel.