[KERNEL] Bricked-Kernel Mako

[g5abiera]

I've been using the same uv settings as before. -187500 on all voltages.

Sent from my Nexus 4 using xda app-developers app

 

[daishou]

Didnt change anything. Had 2 crashes because s2w didnt function properly and the whole system wasnt smooth. Restored back to CM kernel.

Sent from my Nexus 4 using xda premium

 

[show-p1984]

Hey. Can someone link me to the recommended settings for this Kernel?
Thx

Recommended = defaults

Has anyone noticed a reduction in undervolting ability in 0.75? In 0.7 I was able to UV 192MHz to 700mV, 384 to 712mV, going one step up in voltage levels for each clock step. but with 0.75 I cannot have 192 below 800mV; anything lower and I cannot unlock the device after locking it, I get no response from the power button.

There was no change to the voltages between 0.7 and 0.75.

Didnt change anything. Had 2 crashes because s2w didnt function properly and the whole system wasnt smooth. Restored back to CM kernel.

Didn't see one log in there, or any helpful information for that matter. Thanks for nothing I guess.

 

[FSRBIKER]

I'm getting great performance & very good battery life with 0.7 kernel. No data drops at all, I do get them on Faux kernel so I'm sticking with what works.

What app is best for tweaks, I have been using the Faux app.

Sent from my Nexus 4

 

[romeoliny]

Recommended = defaults

There was no change to the voltages between 0.7 and 0.75.


Didn't see one log in there, or any helpful information for that matter. Thanks for nothing I guess.

Show-p, I have a lot of respect for you as a dev. I think you're very good at what you do but sometimes the way you answer people is a bit demeaning.

I realize this is xda and most feel the need to ridicule people who don't read or follow threads properly but in the end we are all here to support and appreciate the devs work and help each other.

I know you were responding to someone else but that's my 2 cents anyway. Just something to consider.

Sent from my Nexus 4 using xda premium

 

[show-p1984]

Show-p, I have a lot of respect for you as a dev. I think you're very good at what you do but sometimes the way you answer people is a bit demeaning.

I realize this is xda and most feel the need to ridicule people who don't read or follow threads properly but in the end we are all here to support and appreciate the devs work and help each other.

I know you were responding to someone else but that's my 2 cents anyway. Just something to consider.

If you are a dev you have no time to pry the information out of somebody. You filter out messages that are incomplete or just plain stupid, ruthlessly. You are blunt, because you just have no frakking time to say please and maybe and could you every time you explain something, which happens 50 times a day though the frakking information is already in the first post or on google!

eg: Person X asks: "How do I get a kmsg?"
What you think about that person: "Ok, a newbie, np I will tell him."
What mods think about that person: "Why the frak do we have a search function?"
What hackers (aka devs) think about that person: *throws the post into the imaginary trash can*

eg: Person X says: "Crashed x times because y didn't function properly and it overall sucked! I reverted back to stock!"
What you think about that person: "Valid feedback."
What mods think about that person: "Somewhat lacking information but valid."
What hackers (aka devs) think about that person:
a) That person just assumes that feature y is broken while he has no clue what he is doing or about anything he just assumes would be broken. Strange dude #ignore or be blunt. (depending on time available)
b) That person has done some digging and found out that feature y is broken but didn't include the frakking log. #ignore or assume a) because if he had done some digging he surely would have included the log.
c) What the... #imaginary-trash-can

If I would be polite and pry out the necessary information out of everybody this wouldn't be a full time job, this would be impossible to accomplish on a 24 hour day.
I highly suggest that you read the link in my signature, which explains this situation and the misconception that someone is angry just because he is blunt quite good.

For example this would be a hole-in-one issue report which never, ever occurred (this issue is just a dummy):
##############################################
Let's start with the place: The ISSUE TRACKER, not the thread!
Device: mako
Android version: 4.2.1
Kernel version: Bricked 0.75
Settings: STOCK
Title: Kernel panic and hard reboot. Null pointer oops in do_sweep2wake()
Content:
Hi,
I recently experienced hard reboots due to kernel panics when using s2w on my mako with Bricked 0.75. It usually happens after resume from deep sleep and seems to be late_resume related. I have attached a log in which you can see a null pointer oops in do_sweep2wake().
[OPTIONAL, IF YOU HAVE PROGRAMMING KNOWLEDGE:]
My research showed that the power device pointer is not checked if it is valid and therefore is NULL sometimes.
I do not know why that pointer is not set on rare occasions, but I have added this patch:

#simplified

+ if (bad_device_ptr->bla)
  bad_device_ptr->bla();

This seems to fix the issue and keeps the function intact.

[IF YOU HAVE NO PROGRAMMING KNOWLEDGE:]
I googled this issues and it seems that there are source code changes required which I have no knowledge of.
I would be happy to provide any more information needed.
##############################################

I can deal with someone who has no experience, but I can't/won't deal with someone who made ignorance his/her life philosophy.

 

[benjimatt]

I wouldn't let these people get to you because they can't read an op. I figure if your not going to post a log cat for a dev to fix a problem then you may as well not even say anything.

It's my biggest pet peeve when someone says something about a Rom or kernel then says I went back to bla bla bla. Nobody cares what rom/kernel your on lol.

Anyways I can't wait for sweep2wake 2!!!! It's my favorite and most used kernel feature.

 

[Rafase282]

Well I'm back to this kernel for version .75 if not new version is coming up I will report back after a couple of day/battery cycles.

 

[FirelordAzula]

Recommended = defaults

There was no change to the voltages between 0.7 and 0.75.

Alright. Thanks Show.

 

[iKarido]

If you are a dev you have no time to pry the information out of somebody. You filter out messages that are incomplete or just plain stupid, ruthlessly. You are blunt, because you just have no frakking time to say please and maybe and could you every time you explain something, which happens 50 times a day though the frakking information is already in the first post or on google!

eg: Person X asks: "How do I get a kmsg?"
What you think about that person: "Ok, a newbie, np I will tell him."
What mods think about that person: "Why the frak do we have a search function?"
What hackers (aka devs) think about that person: *throws the post into the imaginary trash can*

eg: Person X says: "Crashed x times because y didn't function properly and it overall sucked! I reverted back to stock!"
What you think about that person: "Valid feedback."
What mods think about that person: "Somewhat lacking information but valid."
What hackers (aka devs) think about that person:
a) That person just assumes that feature y is broken while he has no clue what he is doing or about anything he just assumes would be broken. Strange dude #ignore or be blunt. (depending on time available)
b) That person has done some digging and found out that feature y is broken but didn't include the frakking log. #ignore or assume a) because if he had done some digging he surely would have included the log.
c) What the... #imaginary-trash-can

If I would be polite and pry out the necessary information out of everybody this wouldn't be a full time job, this would be impossible to accomplish on a 24 hour day.
I highly suggest that you read the link in my signature, which explains this situation and the misconception that someone is angry just because he is blunt quite good.

For example this would be a hole-in-one issue report which never, ever occurred (this issue is just a dummy):
##############################################
Let's start with the place: The ISSUE TRACKER, not the thread!
Device: mako
Android version: 4.2.1
Kernel version: Bricked 0.75
Settings: STOCK
Title: Kernel panic and hard reboot. Null pointer oops in do_sweep2wake()
Content:
Hi,
I recently experienced hard reboots due to kernel panics when using s2w on my mako with Bricked 0.75. It usually happens after resume from deep sleep and seems to be late_resume related. I have attached a log in which you can see a null pointer oops in do_sweep2wake().
[OPTIONAL, IF YOU HAVE PROGRAMMING KNOWLEDGE:]
My research showed that the power device pointer is not checked if it is valid and therefore is NULL sometimes.
I do not know why that pointer is not set on rare occasions, but I have added this patch:

#simplified

+ if (bad_device_ptr->bla)
  bad_device_ptr->bla();

This seems to fix the issue and keeps the function intact.

[IF YOU HAVE NO PROGRAMMING KNOWLEDGE:]
I googled this issues and it seems that there are source code changes required which I have no knowledge of.
I would be happy to provide any more information needed.
##############################################

I can deal with someone who has no experience, but I can't/won't deal with someone who made ignorance his/her life philosophy.

This post needs to be stickied everywhere! Lots of respect for you show-p. Loving your work since the sensation days.

Sent from my Nexus 10 using Tapatalk HD

 

[IImanuII]

hello in the future, you'll develop an app to control features?

 

[giantflash]

hello in the future, you'll develop an app to control features?

KControl will be on the market soon(tm) with some pretty neat configuration of mpdecision and thermald.
I decided that the kernel is more important than the app, so i put it on ice until the kernel has the most important features. (s2w, cmdline interface, fsync, cpuoc, cpu on/offline stats)

OP updated, 0.65 is the new stable.

Is it really that hard to use the search function? I think he has enough to do and you have the time to use it.

 

[batman368]

If you are a dev you have no time to pry the information out of somebody. You filter out messages that are incomplete or just plain stupid, ruthlessly. You are blunt, because you just have no frakking time to say please and maybe and could you every time you explain something, which happens 50 times a day though the frakking information is already in the first post or on google!

eg: Person X asks: "How do I get a kmsg?"
What you think about that person: "Ok, a newbie, np I will tell him."
What mods think about that person: "Why the frak do we have a search function?"
What hackers (aka devs) think about that person: *throws the post into the imaginary trash can*

eg: Person X says: "Crashed x times because y didn't function properly and it overall sucked! I reverted back to stock!"
What you think about that person: "Valid feedback."
What mods think about that person: "Somewhat lacking information but valid."
What hackers (aka devs) think about that person:
a) That person just assumes that feature y is broken while he has no clue what he is doing or about anything he just assumes would be broken. Strange dude #ignore or be blunt. (depending on time available)
b) That person has done some digging and found out that feature y is broken but didn't include the frakking log. #ignore or assume a) because if he had done some digging he surely would have included the log.
c) What the... #imaginary-trash-can

If I would be polite and pry out the necessary information out of everybody this wouldn't be a full time job, this would be impossible to accomplish on a 24 hour day.
I highly suggest that you read the link in my signature, which explains this situation and the misconception that someone is angry just because he is blunt quite good.

For example this would be a hole-in-one issue report which never, ever occurred (this issue is just a dummy):
##############################################
Let's start with the place: The ISSUE TRACKER, not the thread!
Device: mako
Android version: 4.2.1
Kernel version: Bricked 0.75
Settings: STOCK
Title: Kernel panic and hard reboot. Null pointer oops in do_sweep2wake()
Content:
Hi,
I recently experienced hard reboots due to kernel panics when using s2w on my mako with Bricked 0.75. It usually happens after resume from deep sleep and seems to be late_resume related. I have attached a log in which you can see a null pointer oops in do_sweep2wake().
[OPTIONAL, IF YOU HAVE PROGRAMMING KNOWLEDGE:]
My research showed that the power device pointer is not checked if it is valid and therefore is NULL sometimes.
I do not know why that pointer is not set on rare occasions, but I have added this patch:

#simplified

+ if (bad_device_ptr->bla)
  bad_device_ptr->bla();

This seems to fix the issue and keeps the function intact.

[IF YOU HAVE NO PROGRAMMING KNOWLEDGE:]
I googled this issues and it seems that there are source code changes required which I have no knowledge of.
I would be happy to provide any more information needed.
##############################################

I can deal with someone who has no experience, but I can't/won't deal with someone who made ignorance his/her life philosophy.

hang in there, showp life is tough, and to say it with wiser words:

Bullet Tooth Tony: You should never underestimate the predictability of stupidity. (Snatch)

keep up the great work, i know it's tough to stay motivated against all odds, but your work is very appreciated among a lot of people

 

[giantflash]

Don't want to be annoying and I know some others had mentioned this too. I lost 18% battery over about eight hours without doing anything just after a restart. I had fly mode on and GPS, bluetooth etc all deactivated.
BBS shows me that I had just two real wakelocks, "msm_hsic_host" and "suspend_backoff". I know that the first one is a known problem of our device, but I didn't actually thought that it will hold my device awake about 5/7 of screen off time. And I did some research on "suspend_backoff" but didn't find any helpfull solutions.
Because I didn't thought about it, I forgot to make a dump of this with BBS, just made a screenshot. After one and an half hour using it for media playback I made one but surely that isn't what I wanted to show.
I attached both of them so maybe you can do anything with it. I'm on AOKP and your kernel v.0.7 because with 0.75 I always got the SOD problem.
Thanks in advance for any help and if I forgot something than please ask me.
And because I don't think that it is really a bug of your kernel, show-p, I posted it here in the forum so maybe someone could help me. If I was wrong than please just say it.
I had installed this version of AOKP and your kernel after a full wipe about three days ago, so I didn't tried a full wipe like I do normal, because it existed directly after the fullwipe.
In small form:
Device: mako
Android version: 4.2.1 AOKP
Kernel version: Bricked 0.70
Settings: Undervolted about 150-200mv I think, ondemand, 192MHz-1,24GHz
Problem: Two wakelocks from the kernel, "msm_hsic_host" and "suspend_backoff", which hold my device awake 99% of screen off time

 

[sideeffects]

Don't want to be annoying and I know some others had mentioned this too. I lost 18% battery over about eight hours without doing anything just after a restart. I had fly mode on and GPS, bluetooth etc all deactivated.
BBS shows me that I had just two real wakelocks, "msm_hsic_host" and "suspend_backoff". I know that the first one is a known problem of our device, but I didn't actually thought that it will hold my device awake about 5/7 of screen off time. And I did some research on "suspend_backoff" but didn't find any helpfull solutions.
Because I didn't thought about it, I forgot to make a dump of this with BBS, just made a screenshot. After one and an half hour using it for media playback I made one but surely that isn't what I wanted to show.
I attached both of them so maybe you can do anything with it. I'm on AOKP and your kernel v.0.7 because with 0.75 I always got the SOD problem.
Thanks in advance for any help and if I forgot something than please ask me.
And because I don't think that it is really a bug of your kernel, show-p, I posted it here in the forum so maybe someone could help me. If I was wrong than please just say it.
I had installed this version of AOKP and your kernel after a full wipe about three days ago, so I didn't tried a full wipe like I do normal, because it existed directly after the fullwipe.
In small form:
Device: mako
Android version: 4.2.1 AOKP
Kernel version: Bricked 0.70
Settings: Undervolted about 150-200mv I think, ondemand, 192MHz-1,24GHz
Problem: Two wakelocks from the kernel, "msm_hsic_host" and "suspend_backoff", which hold my device awake 99% of screen off time

The msm_hsic wakelock is fixed in 0.75. Like you said it has a SOD when changing settings though. To avoid the SOD you can change voltages using Android tuner and set on boot then restart before the phone sleeps (changing anything other than voltages still gave me SOD after reboot).

I would recommend using another kernel though until a new bricked beta/stable is out to fix the SOD as many of the newer developments and fixes are not in 0.7.

 

[giantflash]

The msm_hsic wakelock is fixed in 0.75. Like you said it has a SOD when changing settings though. To avoid the SOD you can change voltages using Android tuner and set on boot then restart before the phone sleeps (changing anything other than voltages still gave me SOD after reboot).

I would recommend using another kernel though until a new bricked beta/stable is out to fix the SOD as many of the newer developments and fixes are not in 0.7.

oO read all the 76 pages but didn't remeber that it's fixed. Sorry for that.
I knew that with the voltage but that isn't really a solution for me because when I change for example the min frequence or the max frequence it gets the SOD anyway. Also I can't, like another user here, change the voltage for 192MHz lower then 800mV, not a big deal anyway but just to mention it.
I hope show-p gets enough time to work on the new version of this kernel, I just love his work, just incredible like on sensation
Thanks a lot for the help.

 

[barrybear]

-Deleted post after doing some hard searching-

 

[Mahdi786]

Re: [KERNEL][DEC.24]Bricked-Kernel|3.4|v0.7|thermalguard|mpdecision|stickyfre q |OC&U

-Deleted post after doing some hard searching-

If you come from Franco's kernel you have to flash a reset kernel you can find it in Motleys thread op.

 

[exadeci]

Is this app https://play.google.com/store/apps/details?id=com.grilledmonkey.lagfix useful on this kernel or is this kernel using -discard mount option ?

 

[show-p1984]

Is this app https://play.google.com/store/apps/details?id=com.grilledmonkey.lagfix useful on this kernel or is this kernel using -discard mount option ?

discard is a mount option = rom side.