• After 15+ years, we've made a big change: Android Forums is now Early Bird Club. Learn more here.

Root [Help] Kernel Testers

roadkill42 said:
what is these two down loads and did u ever get the 2ooo mhz oc done to for testing?
Click to expand...
One is the boot.IMG and other is zram that can be used with earlier test build.

I stopped working on that for the next few days (OC) to get the source code cleaned up and pretty. Making small commits here and there. Getting it ready for primetime

@spock1134: what's the abootimg command for making a boot image, my zImage is getting too big
 
Hey! :P (So yeah, I'm SLOWLY coming back, but of course after all the good work has been done in my absence >.<)

Honestly, this sounds like my 2.1 kernel way back, when it turned out I was just really modifying the 'display text' essentially. The 17.5 k bench sounds like a 1.6 clock speed.
 
cooldudezach said:
Hey! :P (So yeah, I'm SLOWLY coming back, but of course after all the good work has been done in my absence >.<)

Honestly, this sounds like my 2.1 kernel way back, when it turned out I was just really modifying the 'display text' essentially. The 17.5 k bench sounds like a 1.6 clock speed.
Click to expand...

Any idea what the IPC on our snapdragons are? His kernel actually could be holding a steady 2.1 and not show much improvement beyond a point simply for this reason. OC, doesn't help IPC one bit. So, coupled with such high OC, it could actually give poorer results than expected due to making more errors, and having to double back to correct them. I've had this happen on PC CPU's. This is why we have "sweet spots". If its even slightly unstable at that speed, it can do much more harm than good. The errors increase exponentially once instability is hit.
 
modhatter6811 said:
Any idea what the IPC on our snapdragons are? His kernel actually could be holding a steady 2.1 and not show much improvement beyond a point simply for this reason. OC, doesn't help IPC one bit. So, coupled with such high OC, it could actually give poorer results than expected due to making more errors, and having to double back to correct them. I've had this happen on PC CPU's. This is why we have "sweet spots". If its even slightly unstable at that speed, it can do much more harm than good. The errors increase exponentially once instability is hit.
Click to expand...
Well, considering I have personally recorded benchmarks at 19500+ while running at almost 2.0, I'm inclined to believe that this is an error due to something else. Especially because I have been there before, as I was able to 'clock' my device up to 2.4 before I realized what was going on. To put it extremely simply, basically all that was updating was the 'display text' of the clock speed, not the clock speed itself. So my 2.4 was really closer to 1.4/1.5 or thereabouts.

Also, I had no idea what IPC meant, and had to google it. ;) So yeah, no idea what it is for this chipset.
 
cooldudezach said:
Well, considering I have personally recorded benchmarks at 19500+ while running at almost 2.0, I'm inclined to believe that this is an error due to something else. Especially because I have been there before, as I was able to 'clock' my device up to 2.4 before I realized what was going on. To put it extremely simply, basically all that was updating was the 'display text' of the clock speed, not the clock speed itself. So my 2.4 was really closer to 1.4/1.5 or thereabouts.

Also, I had no idea what IPC meant, and had to google it. ;) So yeah, no idea what it is for this chipset.
Click to expand...

I'm inclined to believe that you must be right about that. I was pushing 17k+, but not anywhere near 19k. I was also running the bench while severely debloated, just for some extra umph. And I'm pretty sure that someone on these forums (might have been you actually) topped out over an older "HTC One" benchmark.
 
vestalpaul said:
So umm...I'm really excited about this. What's up with it? Thanks for all the hard work. This little $50 phone is hardcore!
Click to expand...

I believe he's working on adding z-ram. It could be a min before we see any sort of progress though, I've read that z-ram can be a real pain in the arse to implement depending on what API and phone you're working with. And there's no telling what else he maybe adding with the next release. I'm sure work is in progress though.
 
modhatter6811 said:
I believe he's working on adding z-ram. It could be a min before we see any sort of progress though, I've read that z-ram can be a real pain in the arse to implement depending on what API and phone you're working with. And there's no telling what else he maybe adding with the next release. I'm sure work is in progress though.
Click to expand...
Zram is done, fully working......as well as zcache (latest 3.4 version)
The OC is wrong like cooldudezach said. I found my error and looking into getting it really 2+.
But so far, has anyone had a reboot at all?
 
rbheromax said:
Zram is done, fully working......as well as zcache (latest 3.4 version)
The OC is wrong like cooldudezach said. I found my error and looking into getting it really 2+.
But so far, has anyone had a reboot at all?
Click to expand...

were can I down load the oc kernal
 
rbheromax said:
Zram is done, fully working......as well as zcache (latest 3.4 version)
The OC is wrong like cooldudezach said. I found my error and looking into getting it really 2+.
But so far, has anyone had a reboot at all?
Click to expand...

Happy to report that I have not had a single reboot. I ran it for 6 days without any issues. Did you update the OP with the zram kernel? Cause if so, I haven't tested that one yet, just the original. (I really want to stomp an Htc one benchmark.)
 
ok the earlier kernel was not going as fast as i thought
i fixed that, got up to 2GHz, runs hot, reboots under heavy load if you force 2GHz min and max
im uploading a version for people to play with through fastboot (no zips, and wifi will not work)
i just want to see if it boots for others and what other weirdness occurs
 
WIFI DOES NOT WORK!!! while booted into this kernel

seems stable enough on L awe 2.0

http://tempsend.com/7237A95181

changes:
whatever ramdisk spock had
FULL cifs support
hfs/hfs+ support (mac os stuff)
xfs
supports SD cards above 64 GB
2.0 GHz cpu (WILL reboot for some at high freqs)
i added some more stuff, cant thing of it atm

fastboot boot, or write with dd to boot partition... DO NOT WIPE CACHE, kernel boots at 384 MHz so it takes a while to boot with cache already built, imagine having to build cache on 384 MHz single core

antutu bench fc's after 3d test, which gets above average scores.

WIFI DOES NOT WORK!!! while booted into this kernel
 
i flashed it as a kernal seems to b working good so far
running for 30 mins booth cores wide open no ishues no WiFi but we all ready no this:):):)
 
2GHz is a lil better now, few reboots here and there BARELY, i had raised FSB speed to 600 MHz for 1.9+ (thanks to cooldudezach for adding 600 MHz tick but never using it)

still running hot but i guess its supposed to. maybe 2.1 can be achieved on some awe's *shrugs*
 
You must log in or register to reply here.
Back
Top Bottom