Root [STOCK][LINARO 4.9.3] Volt Uno v3.3

[dvdavid888]

I have been having trouble enabling swap with beta v1

 

[spock1104]

For those having trouble, please try turning down the clock to 1200 and see if that helps. I saw something in the overclock code I'm not sure of.

 

[bg4m3r]

For those having trouble, please try turning down the clock to 1200 and see if that helps. I saw something in the overclock code I'm not sure of.

According to past attempts to overtrick this, it's not possible (it will say it is, but doesn't actually run faster), so you could just remove the OC function if it's going to cause a problem.

 

[bg4m3r]

Still no joy. First installed and wiped cache and dalvik. Booted OK, used Kernel Adiutor to set CPU frequency to 1190, intellidemand governor, and boost frequency to 1190. As soon as I turned on the tab, it crashed. Booted into safe mode and cleared data for KA and rebooted. Crashed again while I was putting my widgets back. Never enabled swap or any tweaks. Is there a log I can upload to help?

 

[spock1104]

Hopefully /proc/last_kmsg or failing that a logcat

 

[spock1104]

As for changing clocks, on the awe it turned out both mpdecision and thermald would automatically set the max clock. That might be the case here as well.

In those cases I chose Arco68's in-kernel msm_mpdecision and thermal controls as replacement. I'll have to see how intelli compares.

 

[bg4m3r]

Hopefully /proc/last_kmsg or failing that a logcat

Here's a kernel log...
https://drive.google.com/file/d/0B1vczMQ-9a-wOElrNk9Ka3M4elE/view?usp=docslist_api

 

[spock1104]

Here's a kernel log...
https://drive.google.com/file/d/0B1vczMQ-9a-wOElrNk9Ka3M4elE/view?usp=docslist_api

Thanks, however just like ShapeShifter499 there's no direct indication of an error, it just stops. Only thing of remote interest is that about two minutes before the last entry, several more times in between, and the very last entry, is an selinux audit denial. Similar actually showed up in ShapeShifter499's log. For grins, try switching selinux to permissive http://forum.xda-developers.com/showthread.php?t=2524485

 

[ShapeShifter499]

Thanks, however just like ShapeShifter499 there's no direct indication of an error, it just stops. Only thing of remote interest is that about two minutes before the last entry, several more times in between, and the very last entry, is an selinux audit denial. Similar actually showed up in ShapeShifter499's log. For grins, try switching selinux to permissive http://forum.xda-developers.com/showthread.php?t=2524485

I flashed beta 1 again and tried setting to permissive and it crashed. Heres the log from /proc/last_kmsg http://pastebin.com/MpcQEk0h No kernel panic again, can you see anything of interest?

 

[spock1104]

Just want to get your wording clarified, was the crash right after you set permissive? I switched to permissive on my phone as a test and no issues.

 

[ShapeShifter499]

Just want to get your wording clarified, was the crash right after you set permissive? I switched to permissive on my phone as a test and no issues.

Oh, the issue wasn't setting SELinux to permissive. It was swap memory again that crashed the device. Build 494 is what I'm currently on and swap works fine.

 

[bg4m3r]

Oh, the issue wasn't setting SELinux to permissive. It was swap memory again that crashed the device. Build 494 is what I'm currently on and swap works fine.

Still having issues downloading these... Seems related to the captcha bull... What browser are you guys using?

 

[ShapeShifter499]

Still having issues downloading these... Seems related to the captcha bull... What browser are you guys using?

Using firefox here. Here is a mirror of the 494 build https://www.androidfilehost.com/?fid=24269982086999074

EDIT: Firefox on Arch Linux 64 bit.

 

[bg4m3r]

Using firefox here. Here is a mirror of the 494 build https://www.androidfilehost.com/?fid=24269982086999074

EDIT: Firefox on Arch Linux 64 bit.

Cool... I've been trying to download on my phone. Installed Firefox and it downloads. Wish Firefox supported Chromecast better so I could do away with Chrome completely.

 

[bg4m3r]

Got 494 installed, working well. Noticed there aren't many features in this build other than the multi core power save. Is it supposed to be that way or did something go wrong?

 

[spock1104]

Got 494 installed, working well. Noticed there aren't many features in this build other than the multi core power save. Is it supposed to be that way or did something go wrong?

That's correct, it's pure stock zv6 code compiled with linaro 4.9.4. I wanted to see what the behavior was before and after some of the tweaks. Thanks, this helps narrow down the reboot issue. Almost convinced it's the overclocking, but I could be wrong.

 

[ShapeShifter499]

No matter what I do I cannot enable any extra features using your kernel source @spock1104, these do not boot on my device. I found out I needed the dtbTool, I used the CM version from https://github.com/xiaolu/mkbootimg_tools. I implemented the fixes from https://github.com/czobor/android_kernel_msm/commit/4cc5db7b7410b25f05ae3581f4e45feb657f26ec AND http://freetz.org/ticket/2759. I used the toolchain 4.9.4 for a7 from here, http://forum.xda-developers.com/showthread.php?t=2098133If If I just compile as it appears in your source with the two fixes mentioned above without triggering any extra .config options it compiles and boots fine.

 

[spock1104]

No matter what I do I cannot enable any extra features using your kernel source @spock1104, these do not boot on my device. I found out I needed the dtbTool, I used the CM version from https://github.com/xiaolu/mkbootimg_tools. I implemented the fixes from https://github.com/czobor/android_kernel_msm/commit/4cc5db7b7410b25f05ae3581f4e45feb657f26ec AND http://freetz.org/ticket/2759. I used the toolchain 4.9.4 for a7 from here, http://forum.xda-developers.com/showthread.php?t=2098133If If I just compile as it appears in your source with the two fixes mentioned above without triggering any extra .config options it compiles and boots fine.

Odd that I don't need those two patches, but it might be something about my build environment vs yours. In any case, make sure you switched to the dos branch (git pull; git checkout dos) since I've abandoned master.

 

[spock1104]

New test build: http://www.mediafire.com/download/va3wkvhnwg119bv/volt_dos_beta1g.zip

I'm hoping GPU OC is working, please test it out

looking good so far...

 

[ShapeShifter499]

Odd that I don't need those two patches, but it might be something about my build environment vs yours. In any case, make sure you switched to the dos branch (git pull; git checkout dos) since I've abandoned master.

Turkey must have gave me brain fog, I forgot to checkout the 'dos' branch when I redid downloads. Oops, will report back.

EDIT: This command usually does it for me in one shot (git clone -b dos --single-branch https://github.com/spock1104/android_kernel_lge_voltdos)

EDIT2: What Operating System are you using for compile? I'm on Arch Linux 64 bit which usually gives me the latest updates of most Linux stuff.

 

[spock1104]

I'm on ubuntu since that's what's officially supported by Google

 

[bg4m3r]

Just gave the new beta a test. Installed the new version of Curtis's stock ROM along with the kernel. Ran a benchmark with 3Dmark and the physics test tanked, with nothing else showing significant improvement over stock. For control purposes, I installed the stock kernel and ran another test, getting expected results. It looks like, according to the graph, the CPU bottomed out several times during the test, while it spiked on stock. There's interesting differences in the battery and temp lines too. Here's screenshots, beta first, then stock.
(The scores in parenthesis are stock baseline scores provided by the app)

 

[spock1104]

Just gave the new beta a test. Installed the new version of Curtis's stock ROM along with the kernel. Ran a benchmark with 3Dmark and the physics test tanked, with nothing else showing significant improvement over stock. For control purposes, I installed the stock kernel and ran another test, getting expected results. It looks like, according to the graph, the CPU bottomed out several times during the test, while it spiked on stock. There's interesting differences in the battery and temp lines too. Here's screenshots, beta first, then stock.
(The scores in parenthesis are stock baseline scores provided by the app)

Make sure you disable mpdecision and intelliplug, enable all cores (intelliplug won't turn them on once it's disabled) and switch to the performance governor, before you bench.

 

[spock1104]

Okay, so GPU OC is yet again ineffective. *sigh*

Also ran AnTuTu at 19138, so it's more or less where Uno was.

 

[bg4m3r]

Also ran AnTuTu at 19138, so it's more or less where Uno was.

How do you recommend setting up governors and such? Right now I'm getting numbers equivalent to stock. I've never gotten over 18xxx in Antutu.