P500 performance

[kirurus]

Anything to increase the responsiveness/performance my friend. And there is slight a difference between your tweaked kernel and bob's original. I think a main factor was by switching the I/O scheduler from deadline to sio. I also have v6 supercharger and KAK running. zooM...zooM...

>.<

Disabled v6 and KAK and it seems a bit smoother.

 

[clodfelterac]

@tvall--- nice work... +1 to a thread for this kernel... I decided to give it test... BACKside+amphetamines/tvall v1.1...oc 480/600/smartv2... did notice that my dlspeed went from 30kbps to 60kbps... (my vm throttle speed)... ill post an update in a few days after everything's settled in... thanks man...

 

[anon(235100)]

tvall- where did you find the git for their p500 kernel from which you pulled your changes?

Also- can you take a look at this p500 kernel

 

[rukin5197]

tvall- where did you find the git for their p500 kernel from which you pulled your changes?

Also- can you take a look at this p500 kernel

I agree with this guy, what we really need is a .35 kernel. I believe, just believe, that tdm is trying to work with a 3.0 x kernel so that leaves out this bad boy.

 

[tvall]

tvall- where did you find the git for their p500 kernel from which you pulled your changes?

Also- can you take a look at this p500 kernel

https://github.com/thachtunganh/lge-kernel-thunderg

Was going to look at Franco.kernel later. Its .35 and I'm currently working with a .32 kernel, so it takes a little work to patch stuff in (I'm a noob at this, adding the ck patches was almost more that I could handle, and that was made for mainline 2.6.32)

Edit: current status.
Franco has a .32 kernel, but it has no commit history. Pretty useless for me. I cloned his .35 kernel and used git to make patches from the commits. There's over 300.. so today's project (after cleaning the kitchen) is to go through it and delete the ones that are reverted later, delete the ones that only apply to .35, then attempt to apply what's left

 

[anon(235100)]

https://github.com/thachtunganh/lge-kernel-thunderg

Was going to look at Franco.kernel later. Its .35 and I'm currently working with a .32 kernel, so it takes a little work to patch stuff in (I'm a noob at this, adding the ck patches was almost more that I could handle, and that was made for mainline 2.6.32)

Edit: current status.
Franco has a .32 kernel, but it has no commit history. Pretty useless for me. I cloned his .35 kernel and used git to make patches from the commits. There's over 300.. so today's project (after cleaning the kitchen) is to go through it and delete the ones that are reverted later, delete the ones that only apply to .35, then attempt to apply what's left

woah... I didn't expect all of that. You're amazing!
btw, Franco is a pretty cool guy when it comes to questions & answers. I'd asked him a long time ago about porting over to our phone. He conversed with blarf a bit on it but a direct port over would be real tricky.
I've been tempted to ask him "since you add & revert a lot on this kernel- what did you take away as some of the most beneficial commits?" I dunno.

 

[tvall]

Well I get a handful of patches ant they applied cleanly, but it wont compile...

Edit: I reverted the patches and it still ain't compiling... weird
Keep getting this:

  LD      .tmp_vmlinux1
init/built-in.o: In function `start_kernel':
calibrate.c:(.init.text+0x5d8): undefined reference to `kmem_cache_init_late'
make: *** [.tmp_vmlinux1] Error 1
tim@boxs:~/hdd2/android_kernel_thunderc$

 

[tvall]

Keep getting errors.. I just cloned Bob's kernel and am gonna apply my patches one by one to see if I can track down what happened

 

[tvall]

Well.. started over. Almost done. Just have to redo the config

Building with a lower timer frequency to save battery

 

[arnold1997]

Wow this kernel is fast and snappy on BACKside thanx!

 

[tvall]

Turns out it was the slqb patch. Reverted that and its all good. Back to work.

http://www.mediafire.com/?0xa10cycue7ea7c

 

[rebel69ization]

This thing is bad a! Just flashed and its faster than the last one! WOW

 

[anon(235100)]

This thing is bad a! Just flashed and its faster than the last one! WOW

It's actually just one revert in the commit history.

 

[tvall]

It's actually just one revert in the commit history.

actually.. the last one was at commit 91f83f8c93.
since then i pulled 3 patches from franco.kernel and fixed up some mistakes. and reverted one patch

 

[arnold1997]

So we can just boot in recovery without wiping anythung and flash?

 

[tvall]

So we can just boot in recovery without wiping anythung and flash?

the kernel? yes

 

[arnold1997]

the kernel? yes

Yes and thank you!

Edit
Just flashed kernel and its faster than the other one

 

[tvall]

i think i have a bad habit of hijacking threads...

 

[arnold1997]

i think i have a bad habit of hijacking threads...

Hahahaha

 

[tvall]

lets move discussion of my kernel over here -> http://forums.androidcentral.com/optimus-v-rooting-roms-hacks/162598-rom-tvalls-iho-roms-thread-hijacked-8.html