[kernel][ICS][23/05/2012] mnics (LR/CM9)
- Thread starter mialwe
- Start date
- Status
That's right. The cpufreq_stats.ko module in /system/lib/modules does not match the kernel version.
- if you did install it vie Heimdall/Odin: please provide /data/user.log
- if you did it install in any other way: please reflash using Heimdall/Odin.
EDIT: Just to save some time later: Can you chack if you have /system/lib/modules/ansi_cprng.ko and what timestamp (= created ) this file has?
Sorry, this information should be in the user.log, too.
Did you use this version before the new nightly, too, in that case without this issue?
0510-nightly downoading right now, have to go afk for some hours now but I'll check this and try to reproduce it with latest nightly ASAP if you can provide the log.
- if you did install it vie Heimdall/Odin: please provide /data/user.log
- if you did it install in any other way: please reflash using Heimdall/Odin.
Sorry, this information should be in the user.log, too.
Did you use this version before the new nightly, too, in that case without this issue?
0510-nightly downoading right now, have to go afk for some hours now but I'll check this and try to reproduce it with latest nightly ASAP if you can provide the log.
Last edited:
Upgrade from nighlies 09.05 to 10.05.On 09.05 nightlies was ok(yesterday )
Just reflashed nightlies and flashed kernel with odin.
************************************************
MIDNIGHT-ICS BOOT LOG
************************************************
Thu May 10 14:41:29 CEST 2012
Kernel: 3.0.8-MNICS-20120509-7-CM9
PATH: /sbin:/vendor/bin:/system/sbin:/system/bin:/system/xbin
ROM: ro.build.display.id=cm_galaxysmtd-userdebug 4.0.4 IMM76I eng.jenkins.20120510.002908 test-keys
BusyBox:BusyBox v1.20.0-cm9 bionic (2012-05-06 01:47 +0200) multi-call binary.
BusyBox is a multi-call binary that combines many common Unix
link to busybox for each function they wish to use and BusyBox
----------------------------------------
Thu May 10 14:41:29 CEST 2012 init bootlog
Thu May 10 12:41:26 GMT 2012
generating mount directories...
generating device nodes...
mounting devices...
version: 3.0.8-MNICS-20120509-7-CM9
grep: /system/lib/modules/ansi_cprng.ko: No such file or directory
updating modules...
rm: can't remove '/system/lib/modules/*': No such file or directory
cp: can't create '/system/lib/modules/ansi_cprng.ko': No such file or directory
cp: can't create '/system/lib/modules/cpufreq_stats.ko': No such file or directory
cp: can't create '/system/lib/modules/fuse.ko': No such file or directory
cp: can't create '/system/lib/modules/gspca_main.ko': No such file or directory
cp: can't create '/system/lib/modules/logger.ko': No such file or directory
cp: can't create '/system/lib/modules/nls_utf8.ko': No such file or directory
cp: can't create '/system/lib/modules/scsi_wait_scan.ko': No such file or directory
/data/local/logger.ko, not found, skipping...
coping kernel app
rm: can't remove '/system/app/MidnightControl.apk': No such file or directory
rm: can't remove '/data/dalvik-cache/*MidnightControl.apk*': No such file or directory
updating lights.aries.so...
starting bootmode detection...
cleaning up...
Thu May 10 12:41:26 GMT 2012
----------------------------------------
Thu May 10 14:41:29 CEST 2012 modules
/system/lib/modules: No such file or directory
Thu May 10 14:41:29 CEST 2012 modules loaded
Module Size Used by Not tainted
Thu May 10 14:41:29 CEST 2012 MidnightControl settings parsing
APP: no blocker file present, proceeding...
APP: checking app preferences...
APP: preferences file found, parsing...
APP: IO sched -> sio
APP: limit800 ->
APP: cpugov -> ondemand
APP: uv at boot ->
APP: uv1000 -> 0
APP: uv800 -> 0
APP: uv400 -> 0
APP: uv200 -> 0
APP: uv100 -> 0
APP: readahead -> 256
APP: vibration intensity -> 36450
APP: touchwake -> false
APP: touchwake timeout->
Thu May 10 14:41:30 CEST 2012 mount
mount: mounting none on /acct failed: Device or resource busy
mount: mounting none on /dev/cpuctl failed: Device or resource busy
rootfs / rootfs ro,noatime 0 0
tmpfs /dev tmpfs rw,nosuid,noatime,mode=755 0 0
devpts /dev/pts devpts rw,noatime,mode=600 0 0
proc /proc proc rw,noatime 0 0
sysfs /sys sysfs rw,noatime 0 0
none /acct cgroup rw,relatime,cpuacct 0 0
tmpfs /mnt/asec tmpfs rw,noatime,mode=755,gid=1000 0 0
tmpfs /mnt/obb tmpfs rw,noatime,mode=755,gid=1000 0 0
none /dev/cpuctl cgroup rw,relatime,cpu 0 0
/dev/block/mtdblock2 /system yaffs2 ro,noatime 0 0
/dev/block/mtdblock3 /cache yaffs2 rw,nosuid,nodev,noatime 0 0
/dev/block/mtdblock5 /radio yaffs2 rw,noatime 0 0
/dev/block/mmcblk0p2 /data ext4 rw,nosuid,nodev,noatime,nodiratime,barrier=1,data=ordered,noauto_da_alloc 0 0
/dev/block/mtdblock6 /datadata yaffs2 rw,noatime 0 0
/dev/block/mtdblock4 /efs yaffs2 rw,noatime 0 0
/sys/kernel/debug /sys/kernel/debug debugfs rw,noatime 0 0
Thu May 10 14:41:31 CEST 2012 cpu
configfile /data/local/midnight_options.conf found, checking values...
oc1128 not selected, using 1Ghz max...
insmod: can't open '/system/lib/modules/cpufreq_stats.ko': No such file or directory
800Mhz limit deactivated, skipping...
cpu: found vaild cpugov: <ondemand>
CPU: values after parsing: 0, 0, 0, 0, 0
max : 1000000
gov : ondemand
UV_mv : 0 0 0 0 0
states_enabled: 1 1 1 1 1
freq/voltage :
1000000 1275 1275
800000 1200 1200
400000 1050 1050
200000 950 950
100000 950 950
Thu May 10 14:41:32 CEST 2012 vm
swappiness: 0
dirty_writeback_centisecs: 1500
dirty_expire_centisecs: 1500
dirty_background_ratio: 5
dirty_ratio: 15
page-cluster: 3
laptop_mode: 0
oom_kill_allocating_task: 0
panic_on_oom: 0
overcommit_memory: 1
Thu May 10 14:41:33 CEST 2012 sec
SEC: ip_forward :0
SEC: rp_filter :0
SEC: use_tempaddr :2
SEC: accept_source_route :0
SEC: send_redirects :0
SEC: icmp_echo_ignore_broadcasts :1
Thu May 10 14:41:33 CEST 2012 prop
wifi.supplicant_scan_interval (is this actually used?): 180
Thu May 10 14:41:33 CEST 2012 kernel
sched_features: NO_GENTLE_FAIR_SLEEPERS START_DEBIT WAKEUP_PREEMPT AFFINE_WAKEUPS NO_NEXT_BUDDY LAST_BUDDY CACHE_HOT_BUDDY NO_ARCH_POWER NO_HRTICK NO_DOUBLE_TICK LB_BIAS OWNER_SPIN NONIRQ_POWER TTWU_QUEUE NO_FORCE_SD_OVERLAP
sem: 500 512000 64 2048
sched_latency_ns: 3000000
sched_wakeup_granularity_ns: 500000
sched_min_granularity_ns: 500000
panic_on_oops: 0
panic: 0
Thu May 10 14:41:33 CEST 2012 read_ahead_kb
default: 128
read_ahead: found vaild value: <256>
179.0: 256
179.8: 256
Thu May 10 14:41:33 CEST 2012 io
/sys/block/mtdblock0/queue/scheduler: noop deadline cfq [sio]
/sys/block/mtdblock0/queue/rotational: 0
/sys/block/mtdblock0/queue/iostats: 0
/sys/block/mtdblock0/queue/read_ahead_kb: 128
/sys/block/mtdblock0/queue/rq_affinity: 1
/sys/block/mtdblock0/queue/nr_requests: 1024
/sys/block/mtdblock1/queue/scheduler: noop deadline cfq [sio]
/sys/block/mtdblock1/queue/rotational: 0
/sys/block/mtdblock1/queue/iostats: 0
/sys/block/mtdblock1/queue/read_ahead_kb: 128
/sys/block/mtdblock1/queue/rq_affinity: 1
/sys/block/mtdblock1/queue/nr_requests: 1024
/sys/block/mtdblock2/queue/scheduler: noop deadline cfq [sio]
/sys/block/mtdblock2/queue/rotational: 0
/sys/block/mtdblock2/queue/iostats: 0
/sys/block/mtdblock2/queue/read_ahead_kb: 16
/sys/block/mtdblock2/queue/rq_affinity: 1
/sys/block/mtdblock2/queue/nr_requests: 1024
/sys/block/mtdblock3/queue/scheduler: noop deadline cfq [sio]
/sys/block/mtdblock3/queue/rotational: 0
/sys/block/mtdblock3/queue/iostats: 0
/sys/block/mtdblock3/queue/read_ahead_kb: 16
/sys/block/mtdblock3/queue/rq_affinity: 1
/sys/block/mtdblock3/queue/nr_requests: 1024
/sys/block/mtdblock4/queue/scheduler: noop deadline cfq [sio]
/sys/block/mtdblock4/queue/rotational: 0
/sys/block/mtdblock4/queue/iostats: 0
/sys/block/mtdblock4/queue/read_ahead_kb: 128
/sys/block/mtdblock4/queue/rq_affinity: 1
/sys/block/mtdblock4/queue/nr_requests: 1024
/sys/block/mtdblock5/queue/scheduler: noop deadline cfq [sio]
/sys/block/mtdblock5/queue/rotational: 0
/sys/block/mtdblock5/queue/iostats: 0
/sys/block/mtdblock5/queue/read_ahead_kb: 128
/sys/block/mtdblock5/queue/rq_affinity: 1
/sys/block/mtdblock5/queue/nr_requests: 1024
/sys/block/mtdblock6/queue/scheduler: noop deadline cfq [sio]
/sys/block/mtdblock6/queue/rotational: 0
/sys/block/mtdblock6/queue/iostats: 0
/sys/block/mtdblock6/queue/read_ahead_kb: 64
/sys/block/mtdblock6/queue/rq_affinity: 1
/sys/block/mtdblock6/queue/nr_requests: 1024
/sys/block/mtdblock7/queue/scheduler: noop deadline cfq [sio]
/sys/block/mtdblock7/queue/rotational: 0
/sys/block/mtdblock7/queue/iostats: 0
/sys/block/mtdblock7/queue/read_ahead_kb: 128
/sys/block/mtdblock7/queue/rq_affinity: 1
/sys/block/mtdblock7/queue/nr_requests: 1024
/sys/block/mmcblk0/queue/scheduler: noop deadline cfq [sio]
/sys/block/mmcblk0/queue/rotational: 0
/sys/block/mmcblk0/queue/iostats: 0
/sys/block/mmcblk0/queue/read_ahead_kb: 256
/sys/block/mmcblk0/queue/rq_affinity: 1
/sys/block/mmcblk0/queue/nr_requests: 1024
/sys/block/mmcblk1/queue/scheduler: noop deadline cfq [sio]
/sys/block/mmcblk1/queue/rotational: 0
/sys/block/mmcblk1/queue/iostats: 0
/sys/block/mmcblk1/queue/read_ahead_kb: 256
/sys/block/mmcblk1/queue/rq_affinity: 1
/sys/block/mmcblk1/queue/nr_requests: 1024
/sys/block/loop0/queue/scheduler: none
/sys/block/loop0/queue/rotational: 0
/sys/block/loop0/queue/iostats: 0
/sys/block/loop0/queue/read_ahead_kb: 128
/sys/block/loop0/queue/rq_affinity: 0
/sys/block/loop0/queue/nr_requests: 0
/sys/block/loop1/queue/scheduler: none
/sys/block/loop1/queue/rotational: 0
/sys/block/loop1/queue/iostats: 0
/sys/block/loop1/queue/read_ahead_kb: 128
/sys/block/loop1/queue/rq_affinity: 0
/sys/block/loop1/queue/nr_requests: 0
/sys/block/loop2/queue/scheduler: none
/sys/block/loop2/queue/rotational: 0
/sys/block/loop2/queue/iostats: 0
/sys/block/loop2/queue/read_ahead_kb: 128
/sys/block/loop2/queue/rq_affinity: 0
/sys/block/loop2/queue/nr_requests: 0
/sys/block/loop3/queue/scheduler: none
/sys/block/loop3/queue/rotational: 0
/sys/block/loop3/queue/iostats: 0
/sys/block/loop3/queue/read_ahead_kb: 128
/sys/block/loop3/queue/rq_affinity: 0
/sys/block/loop3/queue/nr_requests: 0
/sys/block/loop4/queue/scheduler: none
/sys/block/loop4/queue/rotational: 0
/sys/block/loop4/queue/iostats: 0
/sys/block/loop4/queue/read_ahead_kb: 128
/sys/block/loop4/queue/rq_affinity: 0
/sys/block/loop4/queue/nr_requests: 0
/sys/block/loop5/queue/scheduler: none
/sys/block/loop5/queue/rotational: 0
/sys/block/loop5/queue/iostats: 0
/sys/block/loop5/queue/read_ahead_kb: 128
/sys/block/loop5/queue/rq_affinity: 0
/sys/block/loop5/queue/nr_requests: 0
/sys/block/loop6/queue/scheduler: none
/sys/block/loop6/queue/rotational: 0
/sys/block/loop6/queue/iostats: 0
/sys/block/loop6/queue/read_ahead_kb: 128
/sys/block/loop6/queue/rq_affinity: 0
/sys/block/loop6/queue/nr_requests: 0
/sys/block/loop7/queue/scheduler: none
/sys/block/loop7/queue/rotational: 0
/sys/block/loop7/queue/iostats: 0
/sys/block/loop7/queue/read_ahead_kb: 128
/sys/block/loop7/queue/rq_affinity: 0
/sys/block/loop7/queue/nr_requests: 0
Thu May 10 14:41:34 CEST 2012 vibration intensity
found sensitivity value, setting...
/sys/class/timed_output/vibrator/duty: 36450
Thu May 10 14:41:34 CEST 2012 touchwake
deactivated, nothing to do...
Thu May 10 14:41:34 CEST 2012 init.d/userinit.d
configfile /data/local/midnight_options.conf found, checking values...
Thu May 10 14:41:34 CEST 2012 USER INIT START from /system/etc/init.d
Thu May 10 14:41:34 CEST 2012 USER INIT DONE from /system/etc/init.d
Thu May 10 14:41:34 CEST 2012 USER INIT START from /data/local/userinit.d
Thu May 10 14:41:34 CEST 2012 USER INIT DONE from /data/local/userinit.d
Last edited:
Ok, you don't seem to have the directory itself:
Can you check this? In the downloaded 0510-nightly it does exist so that's a bit strange.
If the directory is not there you can try to create it (/system/lib/modules) and reflash the kernel again. Not sure if this will work as the new initramfs chechs for the version of ansi_cprng.ko in this directory which should exist on all kernels.
Code:
rm: can't remove '/system/lib/modules/*': No such file or directoryIf the directory is not there you can try to create it (/system/lib/modules) and reflash the kernel again. Not sure if this will work as the new initramfs chechs for the version of ansi_cprng.ko in this directory which should exist on all kernels.
I didn't have that folder so i ectracted form cm9 zip file and copy it to system/lib.
Reflash kernel boot and folder is gone.
Create empty folder reflash kernel and now it's ok.
Yesterday i was on nightlies 0905 i flashed your kernel.
After that simple flash of nightlies 1005 then your kernel and this happens.
Never mind i have to go home now real life first
Last edited:
Nothing to regret, I completely understand that you bear the full responsibility!
Nevertheless, I'll keep using fast_charge and will report if something bad happens!
Just tested this (dentist let me go earlyI didn't have that folder so i ectracted form cm9 zip file and copy it to system/lib.
Reflash kernel boot and folder is gone.
Create empty folder reflash kernel and now it's ok.
Yesterday i was on nightlies 0905 i flashed your kernel.
After that simple flash of nightlies 1005 then your kernel and this happens.
Never mind i have to go home now real life first
):- was using MNICS @ CM9 0507-nightly
- downloaded 0510-nightly
- flashed 0510-nightly
- booted once
- rebooted once to check if I have a bootloop - nope (still CM9 kernel)
- downloaded MNICS ...-07.tar
- extracted zImage
- flashed with Heimdall
Code:
Thu May 10 14:34:37 GMT 2012
generating mount directories...
generating device nodes...
mounting devices...
version: 3.0.8-MNICS-20120509-7-CM9
updating modules...
found /data/local/logger.ko, updating...
coping kernel app
...
...Strange.
What were your exact steps for flashing CM9 + MNICS?
Just tested this (dentist let me go early
- was using MNICS @ CM9 0507-nightly
- downloaded 0510-nightly
- flashed 0510-nightly
- booted once
- rebooted once to check if I have a bootloop - nope (still CM9 kernel)
- downloaded MNICS ...-07.tar
- extracted zImage
- flashed with Heimdall
= > All fine, worked perfectly, /system/lib/modules are there (MNICS does *not* remove the directory itself), kernel modules are copied correctly, CPUSpy works as it should.Code:Thu May 10 14:34:37 GMT 2012 generating mount directories... generating device nodes... mounting devices... version: 3.0.8-MNICS-20120509-7-CM9 updating modules... found /data/local/logger.ko, updating... coping kernel app ... ...
Strange.
What were your exact steps for flashing CM9 + MNICS?
I will try again tommorow with new nightly.Since nobody else have a problem no need to get worried.Plus hangover from last night
Edit:diference is you flashed zimage with Heimdall and i flashed tar with odin.
Last edited:
Just tested this (dentist let me go early
- was using MNICS @ CM9 0507-nightly
- downloaded 0510-nightly
- flashed 0510-nightly
- booted once
- rebooted once to check if I have a bootloop - nope (still CM9 kernel)
- downloaded MNICS ...-07.tar
- extracted zImage
- flashed with Heimdall
= > All fine, worked perfectly, /system/lib/modules are there (MNICS does *not* remove the directory itself), kernel modules are copied correctly, CPUSpy works as it should.Code:Thu May 10 14:34:37 GMT 2012 generating mount directories... generating device nodes... mounting devices... version: 3.0.8-MNICS-20120509-7-CM9 updating modules... found /data/local/logger.ko, updating... coping kernel app ... ...
Strange.
What were your exact steps for flashing CM9 + MNICS?
*dentist*, eh? that's alright, you don't have to tell us her name ...
loaded on Cm9 of May 8th this morning using Odin and still working fine after 14 hours no problem, except for the missing instructions that I had to figure out miself.
Now my question, and maybe this is stupid: what WAS the reason people got bootloops? 'cause I don't think I saw an explanation anywhere .....
tx in advance and best regards,
Jacobus
---------- Post added at 07:32 PM ---------- Previous post was at 07:26 PM ----------
Well, I think you misunderstood something here:
- for MNICS (and earlier Midnight for Gingerbread) I give *all* my spare time. There's nothing left. Priority is from 1) family to 2) job, 3) kernel, 4) sleep.
- I'm doing this for free because I enjoy working with code for 20 years now, nobody has to pay for it
- between releases there is a *lot* of work behind the scenes to get this running without too many bugs
- I daily try to find the time to answer questions and give information in this thread
- last: I cannot provide detailed instructions for a software (Odin/Heimdall) I have not developed myself and for which instructions are posted all over the web + there's a question forum for things like that here at XDA (and for Heimdall a dedicated thread).
I just don't have the time to do this research for you + it will clutter the OP/second post which I try to have clean and with kernel related information only (and not too much technical stuff) and credits.
=> this kernel thread contains information related to using this kernel, not other software. That's why it is a kernel tread. If any other user wants to help and explain thats absolutely ok, If I happen to find the time I'll answer myself, too. But you cannot request that.
Besides that I don't even own a Windows machine (used laptop + Linux only) and cannot provide any information on Odin/Windows.
No offense from my side, too, but if you want to customize your phone you have to search, read and learn about it as everybody has to. I cannot prepare everything in detail and serve it to you.
Just as I had to learn - as I like to try myself befor asking somwhere - how to compile kernels, compile recovery, work with initramfs, source patching and finally implementing own changes. I had to ask a few questions to other devs but 95% information is out there and I just had to use a search engine, read and learn.
Sorry if this sounds too harsh, not intended. But in general there seems to be a tendency to request things from others without doing anything yourself. Not on this thread I have to say - I always liked the way users are communicating here - big thanks to all for this
You are not serious, are you
actually yes, I was serious
I'm a programmer, albeit on a totally different platform and language, but I always insist on detailed instructions in my job ('foolproof'). It gets to be part of your personality.but I get your point, none taken and grateful for your effort, for sure.
"if it's worth doin' in production, it's worth documentin'"
Last edited:
LOL because of this --> *dentist*, eh? that's alright, you don't have to tell us her name ...
Personally, I believe some people who flash on CWM with their USB cable plugged in that got the bootloop problem. I've never had any bootloop when flashing custom kernel. And I did try (once) switching between kernels, no bootloop.
And you can see how many people on CM9 thread say that they got into bootloop when flashing CM9 ROM (with their USB cable plugged on).
I flashed the newest nightly build coming from CM9 with mnics kernel also no bootloop.
The only time I got bootloop (since using ICS) is when I flash theme.
actually yes, I was serious
but I get your point, none taken and grateful for your effort, for sure.
"if it's worth doin' in production, it's worth documentin'"
So how about writing up a well structured, detailed step-by-step guide about the usage of Odin in order to flash a custom kernel?
This way you could help others that run into similar problems with your experience in production and documentation
And finally, what you missed in the OP of this kernel would be fixed by yourself.
... tapatalked
I never flashed with the usb cable plugged in and those bootloop issues quite often. But let's not fill this thread with bootloop stories ;-) There's an extra thread only about this problem.
http://xdaforums.com/showthread.php?p=25496241
... tapatalked
I give up explaining this. There's a post about 2 pages back which explains why I will not do what you just suggested.So how about writing up a well structured, detailed step-by-step guide about the usage of Odin in order to flash a custom kernel?
This way you could help others that run into similar problems with your experience in production and documentation
And finally, what you missed in the OP of this kernel would be fixed by yourself
You can open a thread and do this if you want, no problem.
Now I'm getting really angry. I missed a detailed explanation about how Odin works and I can fix it myself this way ???
Before I have to go to work, some Initramfs info as requested (this is copied from our sources):
# CM9 currently uses an init system with two separate initramfs
# images which are packed into the kernel image at compile time.
# During every boot process the needed initramfs image (recovery
# or Android) gets extracted and the boot process continues using
# the freshly extracted initramfs.
We are not sure but we thought packing and extracting separate boot images for recovery/normal boot images at every boot could be the reason, maybe something goes wrong withing this process. So we decided to build an initramfs/init that does *not* have to unpack initramfs images at every boot. I'm not sure about it but until now I did not see new bootloop reports.
I think you misunderstood Fenvarien post. I believe his post was a response to jacobuscuracao. It was not meant for you....
Mialwe, you completely misunderstood my intention - I tried to support your position! I answered to Jacobuscuracao's posting. Please - read my text again without anger
Grüße aus dem wilden Süden,
Fenvarien
... tapatalked
Last edited:
I tried flashing midnight on the latest CM9 nightly with Odin too (usually I take Heimdall).
Odin works flawlessly and I encountered no further problems with the modules folder.
... tapatalked
So sorry Fenvarien, really, really sorry.Mialwe, you completely misunderstood my intention - I tried to support your position! I answered to Jacobuscuracao's posting. Please - read my text again without anger
Yes, I completely misunderstood your post. That happens if I'm angry anyways and just dont read the text you quoted and then write too fast.
I'm really sorry and apologize for this.
Not a good day today, now I'm waiting for the next bad thing to happen, started this morning, continued at work and now I'm waiting...
So sorry Fenvarien, really, really sorry.
Yes, I completely misunderstood your post. That happens if I'm angry anyways and just dont read the text you quoted and then write too fast.
I'm really sorry and apologize for this.
Not a good day today, now I'm waiting for the next bad thing to happen, started this morning, continued at work and now I'm waiting...
Kudos for being such a gentleman!
(I'm having one of those days too. One of my main work tools, a really expensive bass guitar, fell over from its stand and its neck cracked. Later I spilled a cup of coffee on my keyboard. Great. So you're not alone!)
Sent from my GT-I9000 using Tapatalk 2
- Status
Similar threads
- Locked
Top Liked Posts
-
There are no posts matching your filters.
-
181The MNICS (Midnight-ICS) kernel is a personal project started JUL/2011 with the GB version because a kernel with the features I needed did not exist.
Main focus: stability and smooth experience using a modified Conservative CPU governor. Secondary: useful additions + easy to use app.
Big thanks to XDA, the community, testers, bug reporters and hardcore, chainfire, lukiqq, stratosk, zach, exitz, supercurio, koush, neldar, madrooster, pikachu01, pawtip & teamhacksung, ezekeel, renaudallard, toadlife, Tk-Glitch, krarvind, derTeufel1980, F4k and all the other devs. Everything I have learned about SGS I have learned reading XDA posts, sources or from conversation with XDA members - maybe I can give something back with this kernel.
Defaults:
- 1Ghz, Midnight Conservative governor, stock voltages
- SIO IO scheduler, 256Kb sdcard readahead
- Logcat + init.d disabled (MNICS does not install any init.d scripts)
Installation:
- boot to download mode
- Install kernel zip via Odin or Heimdall (CWM for older versions)
Sources: Github/mialwe, worklog always updated live
FAQ/already answered: nightmode, no 1.2Ghz? / DIDLE, LiveOC, displayed values in kernel app differ / minfreq 200Mhz? / settings not saved/"]]" error in log? + solution / compass / root, adb remount, TUN? / touchwake? / high AndroidOS % with ICS 4.0.3? / fast_charge?
Known issues:
- Phone-off/charging animation, screen stays off @ button press. Phone charges fine to 100%, thenn screen will switch on.
Changelog:
Code:[B]next[/B] - not sure [B]20120523-1[/B] - bugfix: module installer failed without /system/lib/modules directory - bugfix: sdcard wipe while switching GB->ICS [B]20120516-1[/B] - readded fast_charge [B]20120509-7[/B] - changed naming (date-version-[stock]/LR/CM9) - changed to [URL="http://xdaforums.com/showpost.php?p=25875892&postcount=900"]new initramfs system[/URL] (download mode flashable) - changed to Odin tar package (Heimdall: extract + flash zImage) - removed fast_charge patch - freed 2 Mb more RAM, testing - CM9/i9000 kernel source sync - reworked logfile /data/user.log [B]202 - 2012/04/30[/B] - app: fixed wrongly displayed default undervolting preset - app: fixed colors not beeing applied at boot (v190 bug) - app: fixed 800Mhz CPU limit not beeing applied at boot (v190 bug) - zip: added missing MidnightControl.apk to CM9 version [B]190 - 2012/04/29[/B] - kernel: sync with CM9 sources (incl. sdcard mountpoints) - kernel: USB [URL="http://xdaforums.com/showthread.php?t=1586844"]fast_charge[/URL] by chad0989 (toggle widget [URL="https://play.google.com/store/apps/details?id=com.incredicontrol.fastchargewidget"]available[/URL]) [default: deactivated] - kernel: enabled deadline/cfq IO schedulers [default: still SIO] - kernel: 2 bugfixes (cpufreq), thanks to [URL="http://xdaforums.com/showthread.php?t=1445214"]derTeufel1980[/URL] for the hints - kernel: touchwake sync with Ezekeels [URL="https://github.com/Ezekeel/GLaDOS-nexus-prime/commit/bb4ebd593be8022c21a94c7f49b36b608ec005af"]source[/URL] - app: added fast_charge toggle - initramfs: 256Kb sdcard read_ahead default - initramfs: added /data/local/userinit.d execution - initramfs: executes init.d/userinit.d scripts starting with "S" only [B]143 - 2012/04/08[/B] - app: updated to version 1.4 - adjusted default min_brightness from 25 to 20 (stock?) - app: fixed bug with old nightmode sysfs setting at startup [B]142 - 2012/04/08[/B] - kernel: fix green screen on some occasions [B]141 - 2012/04/08[/B] - app: added missed call service toggle (reboot required) - app: updated help file - initramfs: fixed issues with some busyboxes/$PATH (SlimICS, grep not found) [B]133 - 2012/04/07[/B] - changed name: MNICS-<buildnr>[LR (large RAM for 386Mb)] - kernel: 5 frequencies, switchable maxfreq 1/1.128Ghz - kernel: adjusted smooth conserv. triggers, mp3@100Mhz again - kernel: fixed internal divider bug, thanks to [URL="http://xdaforums.com/showthread.php?t=1445214"]theDevil1980[/URL] for the hint - kernel: removed old recovery backup (-700Kb) - CWM: added manual zipalign option - CWM: moved init.d/overclocking/Logcat toggles to CWM - app: service for missed calls BLN - app: cleanup [B]0.9.0-2 - 2012/03/30[/B] - 340Mb RAM - 386Mb for HUGERAM version \w patched ROMs - fixed /datadata MidnightControl settings bug (for. e.g. ICSSGS, Bluelightning) - BLN switch both MidnightControl and ROM->settings->display->notification LED toggles at the same time (one-click BLN enable) [B]0.7.9 - 2012/03/28[/B] - initramfs: fixed 0.7.7 bug - initramfs: adjusted vm tweaks - initramfs: cleaned up boot script and log output - kernel: BLN off as default setting - kernel: adjusted build flags - app: updated help file [B]0.7.7 - 2012/03/25 - [COLOR="Red"]removed 03/26[/COLOR] (apps-don't-open-bug)[/B] - kernel: fixed freeze on 800Mhz max. freq. - kernel: adjusted int.volt 1100->1125mV for 1.2Ghz (stability) - kernel: added touchwake, thanks to [URL="http://xdaforums.com/member.php?u=4143554"]ezekeel[/URL] (dev) and [URL="http://xdaforums.com/showthread.php?t=1445214"]devil[/URL] (porting support via pm) - kernel: added vibration intensity sysfs interface - MidnightControl: rearranged menus - MidnightControl: removed unused option - MidnightControl: added touchwake toggle - MidnightControl: added vibration intensity slider - CWM: added delete init.d contents option - CWM: added delete NStools settings option [B]0.7.4 - 2012/03/17[/B] - first public upload22v190, 190LR, 190CM9
190 - 2012/04/29
- kernel: sync with CM9 sources (incl. sdcard mountpoints)
- kernel: USB fast_charge by chad0989 (toggle widget available) [default: deactivated]
- kernel: enabled deadline/cfq IO schedulers [default: still SIO]
- kernel: 2 bugfixes (cpufreq), thanks to derTeufel1980 for the hints
- kernel: touchwake sync with Ezekeels source
- app: added fast_charge toggle
- initramfs: 256Kb sdcard read_ahead default
- initramfs: added /data/local/userinit.d execution
- initramfs: executes init.d/userinit.d scripts starting with "S" only
Download in OP, v190=340Mb RAM, v190LR=386Mb RAM for ROMs with large RAM patch, v190CM9=~170Mb RAM for CM9.220.9.0-2 + 0.9.0-2-hugeram
0.9.0-2
- 340Mb RAM
- fixed /datadata MidnightControl settings bug (for. e.g. ICSSGS, Bluelightning)
- BLN switch toggles both MidnightControl and ROM->settings->display->notification LED at the same time (one-click BLN enable)
0.9.0-2-HUGERAM for patched ROMs only
- 386Mb RAM
Big thanks to test123456789012 for testing on Slim-ICS, krarvind for notifying us and KalimochoAz for working on this. Works fine on patched CM9, too.2020120509-6-CM9
- changed naming (date-version-[stock]/LR/CM9)
- changed to new initramfs system (download mode flashable)
- changed to Odin tar package (Heimdall: extract + flash zImage)
- removed fast_charge patch
- freed 2 Mb more RAM, testing
- CM9/i9000 kernel source sync
- reworked logfile /data/user.log
Download in first post.
For now CM9 version only, stock-RAM and LR will follow.
As a lot of users experienced bootloops with all ICS/i9000 kernels Stratosk opened a thread to investigate this and after a few posts and pm we decided to work on an initramfs which can be used to build Odin/Heimdall flashable kernel images.
This initramfs has several advantages:
- faster boot
- no need to extract separate sub-images (that's how ICS kernels with default initramfs work)
- kernel flashable via Heimdall/Odin/CWM-zip
- included kernel modules
After the first posts about the Semaphore test version including the new initramfs in Stratosks' bootloop thread the new kernels will
- not cure existing bootloop issues but
- maybe prevent new bootloops while using this kernels
While testing different versions of the new initramfs including flashing lots of ROMs and modems and wiping cache and dalvik-cache I was not able to get a single bootloop again.
Suggestion:
- flash your desired ROM (if needed)
- reboot to make sure it does not bootloop with original kernel
- enter download mode
- flash Semaphore/MNICS new-initramfs kernels (tested MNICS/Heimdall only)
Please report if bootloops occur while using the test kernels (NOT after e.g. flashing another ROM which will install it's default kernel), if possible please provide /data/last_user.log for MNICS which will contain the init stage log messages (new) and the known initramfs/script/init.d log messages for your LAST boot.
Big thanks to Stratosk, I enjoyed working on this with him *a lot*
