[CM7][KANG][N1] But our Princess is in another castle! (CLOSED)

Status
Not open for further replies.
Search This thread

daveid

Senior Member
Jan 5, 2009
1,892
356
Samsung Galaxy S21 Ultra
One thing you should add to your first post (if you really want to keep the mindless blabber to a minimum) is that "advanced" users should leave the moderating to the moderators. If someone notices a post from a "newb", try to ignore it (or simply hit that report button - call it spam) and let the moderators clean the thread properly.
 

redstar3894

Retired Recognized Developer
Jun 3, 2010
428
119
Chicago, IL
redstar.euroskank.com
Woohoo! :D
Fits returns to cook Bacon and print money! Glad to see the return of gigglebread ... :p

FWIW, I agree that the moderating should he left to the moderators.... that's what they are there for, after all.... :)
No matter how much someone may deserve to get flamed, its probably better if we stay on topic here... sometimes actions (like ignoring someone) speak much louder than anything that we can type out here.... :rolleyes:


Now, about that wonking problem.... :p

Sent from my Nexus One
 

chillwill11

Senior Member
Oct 18, 2010
212
39
Jerzey
I got my highest quadrant score from this rom, it runs as smooth as butter 1820
d0365a4f-b419-0222.jpg


Sent from my Nexus One using XDA Premium App
 
Last edited by a moderator:
  • Like
Reactions: nischalnischal

chillwill11

Senior Member
Oct 18, 2010
212
39
Jerzey
Yeah actually this is the only rom that it works 8/10 times. Other roms it would force close.

Sent from my Nexus One using XDA Premium App
 

redstar3894

Retired Recognized Developer
Jun 3, 2010
428
119
Chicago, IL
redstar.euroskank.com
If fits doesn't object, I'll engage in a little shameless self-promotion here.... :p


While my kernels don't fix the wonking problem (yet), they do run especially well on this ROM and others like it.... :D

So if anyone is looking for something different, try flashing any of the "-CM7" kernels in my mediafire folder or the latest one is usually in my signature....

I'm either running them with fits' gigglebread or on a build I compile myself from github, so they should all be working with CM7 based stuff.... :)

I don't want to clutter up the thread too much, so if you have a kernel related question, please post it here only if it relates to this ROM, any general kernel questions, PM me and I'll get back to you as soon as I can...


Oh, and did I mention that my kernel increases the snapdragon's Bacon cooking efficiency by over 1000%? ;) :p

Sent from my Nexus One
 

chillwill11

Senior Member
Oct 18, 2010
212
39
Jerzey
I use redstars latest kernel, Ive used wildmonks latest cm7 kernel, but redstar kernel seems superior even through wildmonks is clocked higher. Not sure why, but in my opinion redstars kernel better
d0365a41-d503-a7a9.jpg


redstar's kernel
Sent from my Nexus One using XDA Premium App
 
Last edited:

dc211

Senior Member
Jul 21, 2010
572
203
Thanks so much for doing this Fit.

I have a really dumb question. I've sat here trying to figure it out for awhile now with no success. What does the 0817 in the thread title mean?
 

Garner

Inactive Recognized Contributor
Oct 25, 2009
2,344
2,371
C:\windows\system32\
I use redstars latest kernel, Ive used wildmonks latest cm7 kernel, but redstar kernel seems superior even through wildmonks is clocked higher. Not sure why, but in my opinion redstars kernel better
d0365a41-d503-a7a9.jpg


redstar's kernel
Sent from my Nexus One using XDA Premium App

Can you trmell me which one of the redstar kernels you are using please
 

dc211

Senior Member
Jul 21, 2010
572
203
With the 2.3.3 code merged I got a big bump in speed. On the 2.3.2 build I sat around the 1400's or 1500's in quadrant. With the 2.3.3 builds and IR's Kernel I get low 1800's.

I saw the screen shot and figured why not. Redstar's downloaded and installed Redstar's kernel and got my new high score with 1113mhz overclock.



Thanks Redstar! Now if I could only get up to maybe 1152mhz I can finally break 2k.

I just sorted his files by date and used 2.6.37.2_AVS-950mV-CM7_CFS-exp_20110303_0017 if anyone is interested.

Previous high score was at using CM 6.1 with Wildmonk's kernel and an 1190mhz overclock (or 1228mhz i forget) I got a 1904.
 
  • Like
Reactions: ilos

fitsnugly

Inactive Recognized Developer
Mar 10, 2010
2,454
2,208
Indianapolis, IN
Profiles just got merged... you guys should give Martin a thank you email :)

(the icons I made for him aren't too shabby, either ;) )
 
Status
Not open for further replies.

Top Liked Posts

  • There are no posts matching your filters.
  • 50
    Due to personal reasons, I have closed this thread an decided to no longer post builds here.

    I appreciate all the people who enjoyed testing new and excited mods to CM with me. I'm glad I earned a few friends out of doing this, you know who you are.
    17
    I've put the patch on gerrit under the commit "Fix for a1026 suspend".

    Hopefully this is the end of the wonk. It was an annoying bug to find. I've been spending the past couple weeks comparing versions of files used in froyo to the gingerbread, and not finding anything useful. Turns out the froyo version was just lucky to work and none of the changes between versions mattered.

    I finally noticed a difference after giving up and installing 6.1.1 yesterday. After looking at a dmesg after a call, I noticed that the a1026 driver was switching into mode 0, which is the suspend mode. I had thought that it was weird that the driver in gingerbread would switch to mode 13 instead, but I couldn't find any differences in the code to explain it.

    A bunch of debug print statements later, and I tried this.
    9
    Kernel Stuff... per earlier requests....

    Alrighty, apologies for this being a little long and for the double post, I'll try my best to be succinct here, but those of you who remember me from MicroMod's threads know that I have a tendency to be overly verbose.... :p

    Differences between my kernels/iR's kernels/stock CM:

    Cyanogen generally likes to keep things a little more on the stable side as far as kernel hacking goes... there could be a lot of reasons for this so I'm not going to speculate on why they choose do so... probably more for stability's sake IMO...

    With that being said, iR and I (and to a somewhat lesser extent wildmonks when he's around...) tend to be a little closer to the bleeding edge of the kernel world... to put it this way: he's willing to try more things than CM, and I'm usually willing to try more things than iR... though this isn't always the case.... :p
    I'm referring to iR more here because I base my kernels off his source and usually add a few more things, but as far as source code is concerned, our kernels are 99.9% the same... I just like to configure things slightly differently than he does... :p

    With that in mind, here are some of the major differences between my/our kernels and CM:
    Note: this is not an exhaustive list... I don't have enough space here to fully describe ALL the differences... :rolleyes:
    If you want more information, I suggest you learn how to interpret source code and take a look at my github... everything is available for consumption there... ;)


    Hybrid Adaptive Voltage Scaling: will dynamically determine the appropriate voltage to feed the CPU under certain conditions.... it considers clock speed and CPU load and will select the lowest stable voltage for the given conditions.... for more info on how it works, see iR's thread as he's done the most work with it....
    AVS generally gives better battery life over the stock voltage scaling, but doesn't work well for everyone and some phones don't respond well to AVS.... whether it gives you better battery life will depend on YOUR usage patterns, it works for most people but the only way to tell if you will benefit is by trying it for yourself.... everyone uses their phone differently.... ;)

    More aggressive compilation options: these tell the compiler to use different methods when building the kernel that *usually* improve performance but can cause situational instabilities and/or not work well for everyone.... historically the Linux kernel doesn't always respond well to increased optimizations as they can cause instability in the kernel runtime, but in this case we are just trying to best utilize the available hardware (scorpion CPU) and wring every last possible bit of performance out of the Snapdragon SoC.....

    New "improved and optimized" toolchain: I'll try not to get too technical here, but this has been my major project over the last 2-3 months, and for those of you who don't know, the toolchain is essentially what builds the code into binary form.... so the rationale is that as long as we're optimizing the kernel, why not optimize what builds it? ;)
    I've been working extensively with iR on this, and between the two of us, we've been able to come up with a way for the kernel to be compiled to utilize more hardware floating point instructions.... what that means is less code is emulated in software and as I'm sure most of you know from full-size computers, hardware rendering is always faster than software emulation.... so the same principle applies here, we're essentially reducing the complexity and overhead of the kernel and allowing it to execute considerably faster.... to put it as simply as I can.... :p
    Those of you who have been using my kernels or iR's kernels probably noticed a major increase in responsiveness about a month ago.... that's when we were finally able to get the "hard floating point instructions" fully implemented.... :D
    So to sum all that up, what all this means to you guys, the end users, is that you get a kernel that is significantly more responsive across all speeds.... it may not necessarily benchmark higher, but it will "feel" much faster... :p
    I was showing one of my friends the improvements (he's got a Nexus S) and he was amazed at how quick and responsive it had become while being underclocked to 844Mhz, usually I had to overclock in order to get smooth GUI action, but no longer, as it seems to stay nice and quick even at slower clock speeds... ;)


    Like I said, I won't get into speculation on why this stuff isn't used in CM's kernel, but most of these hacks are just that, hacks, and they are usually intended to make things faster but sometimes unfortunately have the opposite effect... :p


    Okay, long winded post over... hopefully this will settle any more questions regarding my kernels... :rolleyes:
    If anyone has any other questions/comments/concerns/requests for clarification, please PM me and that way we can try to reduce the clutter in this thread and save more of it for 'Giggle Chatter'... ;)
    Fits, you're welcome to link to this post on the OP if you wish, and also the link to my mediafire folder for those who are interested is as follows: http://goo.gl/Xbu3M

    Next I'll also try and get a write up done of what we know about the "wonk" problem and what has been tried so far/why it isn't an easy fix...etc... that's if no one has any objections to another somewhat lengthy (but informative :D) post... :p


    EDIT: I have no intentions whatsoever of trying to take over this thread... just wanted to clear some things up and have this all out there... :)
    I'm just here to help all you degenerates out however I can with kernel stuff.... ;)
    7
    Redstar's N1 CM7 AVS kernels: http://redstar.euroskank.com/


    You should probably go DOWNLOAD some wallpapers too.....


    Looking for the morph? Try using an awesome CM7 theme engine apk like HONEYBREAD! But it's on my server if you really want it and look real hard...
    6
    Kernel Update!

    Drum Roll Please... :D

    2.6.38.2 Kernels

    My newest kernel is now based on 2.6.38.2! With 10% more #win and lucky charms than .37, you can't go wrong with upgrading! :p


    Finally found some time to get this stuff done... :rolleyes:

    Oh, and happy April Fools' Day! :D
    Was going to claim that I fixed the wonk, but that would just be too cruel... :rolleyes: