[KERNEL] Sub501 Kernel, mods and control apk

[Guest Sub501]

that was it, thanks

sub501manager doesn't crash anymore

quick look at submanager: almost impossible to touch "txpower" tab with my fingers ;-)

but looks promising

No problemo, I'll make them larger.

[Guest woti17]

No problemo, I'll make them larger.

btw: nilfs2 is death for the phone. needs continously a daemon running, which sucks to much battery.

btrfs i use 10 days now for /data/data, its solid stable, maybe a little bit faster than ext4 on file (but not on sqlite) writes.

maybe i try reiserfs(3) as another option, if someone provides the kernel module.

[Guest Sub501]

btw: nilfs2 is death for the phone. needs continously a daemon running, which sucks to much battery.

btrfs i use 10 days now for /data/data, its solid stable, maybe a little bit faster than ext4 on file (but not on sqlite) writes.

maybe i try reiserfs(3) as another option, if someone provides the kernel module.

It will be available but I don't think this year. Waiting for my holiday...

[Guest David_P]

Only if Paul decides to use it, otherwise:

- flash Paul rom

- without rebooting flash this kernel then flash init mods

If already running r10, can we not go back to recovery and flash your kernel and mods? Does it have to be before the first boot?

David P

[Guest snapilica]

If already running r10, can we not go back to recovery and flash your kernel and mods? Does it have to be before the first boot?

David P

You can flash them just fine even after the first boot. Just make sure you do a nand backup just in case.

[Guest Sub501]

You can flash them just fine even after the first boot. Just make sure you do a nand backup just in case.

Sometimes a boot loop occurs doing this but why not trying. Again, do a nand before.

[Guest woti17]

@sub501

lsof static binary for arm architecture

maybe useful for you

http://code.google.com/p/android-group-kor...rm926ejs.static

[Guest Sub501]

btw: nilfs2 is death for the phone. needs continously a daemon running, which sucks to much battery.

btrfs i use 10 days now for /data/data, its solid stable, maybe a little bit faster than ext4 on file (but not on sqlite) writes.

maybe i try reiserfs(3) as another option, if someone provides the kernel module.

@woti17 , check the first post. Santa brought reiserfs to you.

[Guest woti17]

@woti17 , check the first post. Santa brought reiserfs to you.

tested now, reiserfs and jfs are both faster in sqlite db writes (TAB benchmark /data/data), but thats all, no overall improvement over ext4, for me at least with class 2 card

with the disadvantage of having to use ubuntu fs tools with cardreader to filesystemcheck, if nessesary

btw: if you make a new kernel sometime, could you enable swap support?

Edited December 27, 2010 by woti17

[Guest bpivk]

@Sub501: Does your kernel support swap? I'm trying swapper and see if it has any effect on some lag I experience if I open a lot of programs.

[Guest woti17]

@Sub501: Does your kernel support swap? I'm trying swapper and see if it has any effect on some lag I experience if I open a lot of programs.

no its not activated in the kernel

zcat /proc/config.gz |grep SWAP

# CONFIG_SWAP is not set

at least in kernel vers sub501@u50vg #8

[Guest Morgoth6]

It seems I have another problem with this kernel. First time I think it was one time problem, but it hits me uite frequently. When I woke up device sometimes it works extremaly slow and laggy and looking to SetCPU shows me it switch to powersave goventor and I need to switch it by hand to smartass or ondemand (I stop using SetCPU in this kernel as it seems to have everything in the scripts) and today when I wanted to use it it was powersave with frequencies set to 240MHz (both high and low) This is definitly not normal.

What can I do to help debug this kind of issue ? And is there a wat to just use smartass without any additional logic (low battery, sleeping, etc)

[Guest bpivk]

@Woti: Do you by any chance know of a Desire kernel with swap support. I was searching around but I can't find any.

@Sub501: If you can could you please add swap support to your next kernel?

[Guest woti17]

@Woti: Do you by any chance know of a Desire kernel with swap support. I was searching around but I can't find any.

@Sub501: If you can could you please add swap support to your next kernel?

this one should as described (and download link) here

http://android-roms.net/desire/2010/10/29/...4-sense-kernel/

and neophyte (at least the test kernel) and coutts seem to have it

Edited December 28, 2010 by woti17

[Guest woti17]

zcat /proc/config.gz |grep CGROUPS

# CONFIG_CGROUPS is not set

@sub501: could you please add that too?

CONFIG_CGROUP_SCHED

and it depends on

CONFIG_CGROUPS

as HTC has this bit of code in init.rc

# Create cgroup mount point for cpu accounting

mkdir /acct

mount cgroup none /acct cpuacct

mkdir /acct/uid

[Guest bpivk]

this one should as described (and download link) here

http://android-roms.net/desire/2010/10/29/...4-sense-kernel/

and neophyte (at least the test kernel) and coutts seem to have it

Thanks for this link. This seems like the closest rom to Sub501's so I guess it will do.

[Guest Sub501]

zcat /proc/config.gz |grep CGROUPS

# CONFIG_CGROUPS is not set

@sub501: could you please add that too?

CONFIG_CGROUP_SCHED

and it depends on

CONFIG_CGROUPS

as HTC has this bit of code in init.rc

# Create cgroup mount point for cpu accounting

mkdir /acct

mount cgroup none /acct cpuacct

mkdir /acct/uid

As I remember this setting is disabled by BFS. You can't have both. Don't have time to check now (vacation :-) ), this is what I remember. Correct me if I'm wrong (see the BFS doc).

[Guest woti17]

As I remember this setting is disabled by BFS. You can't have both. Don't have time to check now (vacation :-) ), this is what I remember. Correct me if I'm wrong (see the BFS doc).

in the couttstech kernel at xda

CONFIG_CGROUPS=y

# CONFIG_CGROUP_DEBUG is not set

CONFIG_CGROUP_NS=y

CONFIG_CGROUP_FREEZER=y

CONFIG_CGROUP_DEVICE=y

CONFIG_CGROUP_MEM_RES_CTLR=y

CONFIG_CGROUP_MEM_RES_CTLR_SWAP=y

# CONFIG_CGROUP_BFQIO is not set

# CONFIG_NET_CLS_CGROUP is not set

root@localhost://proc# >zcat config.gz |grep BFS

CONFIG_SCHED_BFS=y

# CONFIG_BFS_FS is not set

[Guest Sub501]

in the couttstech kernel at xda

CONFIG_CGROUPS=y

# CONFIG_CGROUP_DEBUG is not set

CONFIG_CGROUP_NS=y

CONFIG_CGROUP_FREEZER=y

CONFIG_CGROUP_DEVICE=y

CONFIG_CGROUP_MEM_RES_CTLR=y

CONFIG_CGROUP_MEM_RES_CTLR_SWAP=y

# CONFIG_CGROUP_BFQIO is not set

# CONFIG_NET_CLS_CGROUP is not set

root@localhost://proc# >zcat config.gz |grep BFS

CONFIG_SCHED_BFS=y

# CONFIG_BFS_FS is not set

Form BFS faq:

What features does BFS have and not have?

On top of the current scheduler design, it has a SCHED_IDLEPRIO which actually

does only schedule tasks when idle, and SCHED_ISO for unprivileged realtime

performance. BFS does NOT implement CGROUPS. A desktop user should not need

know about CGROUPS, nor should they need to use them. BFS also does not have

the feature of "lots of tunables I don't understand".

From BFS patch (kernel help):

+Features

+

+As the initial prime target audience for BFS was the average desktop user, it

+was designed to not need tweaking, tuning or have features set to obtain benefit

+from it. Thus the number of knobs and features has been kept to an absolute

+minimum and should not require extra user input for the vast majority of cases.

+There are precisely 2 tunables, and 2 extra scheduling policies. The rr_interval

+and iso_cpu tunables, and the SCHED_ISO and SCHED_IDLEPRIO policies. In addition

+to this, BFS also uses sub-tick accounting. What BFS does _not_ now feature is

+support for CGROUPS. The average user should neither need to know what these

+are, nor should they need to be using them to have good desktop behaviour.

Form the same BFS patch:

config CGROUP_CPUACCT

bool "Simple CPU accounting cgroup subsystem"

- depends on CGROUPS

+ depends on CGROUPS && !SCHED_BFS

help

No comment...

[Guest woti17]

Form BFS faq:

.....

Form the same BFS patch:

config CGROUP_CPUACCT

bool "Simple CPU accounting cgroup subsystem"

- depends on CGROUPS

+ depends on CGROUPS && !SCHED_BFS

help

No comment...

kernels with BFS i saw don't have CGROUP_CPUACCT set, so there is no depending conflict

but CONFIG_CGROUPS is sometimes enabled, also in backported kernels

btw: cgroups is in the mailnline kernel, not even experimental, at least in 2.6.37

bfs means always patching

but i don't really caree too much too try this, but SWAP would be nice if enabled, as it doesn change anything, if one doesn't use it

[Guest bpivk]

but i don't really caree too much too try this, but SWAP would be nice if enabled, as it doesn change anything, if one doesn't use it

I'm using swap now and I have a feeling that it's smoother but I can't compare because I'm also using a different kernel so it would be nice to compare it.

[Guest woti17]

I'm using swap now and I have a feeling that it's smoother but I can't compare because I'm also using a different kernel so it would be nice to compare it.

me too, i use neophytes kernel with swap now, but i miss vdd_level (neophyte has havs instead), and until now i got the best battery with sub501

swap - tested with starting random 12 programms

memory before starting was in RAM 240 used from 414MB, swap 0 from 768M

after programs started i switched around between programs, and it went to

320MB RAM used / 45MB used SWAP

doing nothing for a while

RAM down on 270MB / SWAP still 42MB

(i sat swapiness from 1 to 50 with sysctl before)

so i think this is quite useful

Edited December 29, 2010 by woti17

[Guest woti17]

@sub501

i try the sense Z rom 0.6 test from XDA

it is 1.72.405.2 FRG83D WWE based, and uses CGROUPS

without support in kernel it produces tons of messages

found neophytes kernel config, which produces no error

CONFIG_GROUP_SCHED=y

CONFIG_FAIR_GROUP_SCHED=y

CONFIG_RT_GROUP_SCHED=y

CONFIG_CGROUP_SCHED=y

CONFIG_CGROUPS=y

# CONFIG_CGROUP_DEBUG is not set

# CONFIG_CGROUP_NS is not set

# CONFIG_CGROUP_FREEZER is not set

# CONFIG_CGROUP_DEVICE is not set

# CONFIG_CGROUP_CPUACCT is not set

CONFIG_SCHED_AUTOGROUP=y

CONFIG_CGROUP_BFQIO=y

# CONFIG_NET_CLS_CGROUP is not set

Edited December 31, 2010 by woti17

[Guest Sub501]

errors or debug warning/info and so on? please share these errors with us.

[Guest Sub501]

errors or debug warning/info and so on? please share these errors with us.