Reply
Thread Tools
Posts: 2,153 | Thanked: 8,462 times | Joined on May 2010
#201
@412b: standard rules for /opt is that every 3rd application will install data to /opt/<app_name>/. This comes from UNIX. But there is no special reason.
 
Posts: 462 | Thanked: 550 times | Joined on Sep 2008 @ Moscow
#202
Originally Posted by pali View Post
@412b: standard rules for /opt is that every 3rd application will install data to /opt/<app_name>/. This comes from UNIX.
Standard rules say:
/opt/<package> or /opt/<provider>
also
The directories /opt/bin, /opt/doc, /opt/include, /opt/info, /opt/lib, and /opt/man are reserved for local system administrator use. Packages may provide "front-end" files intended to be placed in (by linking or copying) these reserved directories by the local system administrator, but must function normally in the absence of these reserved directories.
Debian packaging
Originally Posted by pali View Post
But there is no special reason.
Ok. I just got curious.
 

The Following User Says Thank You to 412b For This Useful Post:
Posts: 31 | Thanked: 50 times | Joined on Jan 2012 @ Ireland
#203
Firstly, thank you for all your work on the power kernel, and on the functionalities which it enables, most recently KP51 and the new USB Mode Switch.

However, after installing and testing KP51, I have once again found myself reverting to the stock kernel, as I have done previously after installing KP49 and KP50. The reason is that the power kernels seem to degrade the quality of SIP and Skype calling, and this is a deal-breaker for me.

I have searched the forums and found that other users have reported similar experiences. Some of these users believe that the problem arises from the changes in the way the power kernel controls cpu scaling and power management.

Would it be possible for you to produce, or advise how to produce, an experimental kernel which follows the stock kernel's approach to cpu scaling and power management (i.e. no support for overclocking/undervolting), while retaining as many as possible of the other changes which make the power kernel so useful?

Thanks for any help you can give, and again, please accept my sincere appreciation of the work that you do.
 
Posts: 105 | Thanked: 87 times | Joined on Jun 2011 @ Unknown
#204
Originally Posted by zerox View Post
Firstly, thank you for all your work on the power kernel, and on the functionalities which it enables, most recently KP51 and the new USB Mode Switch.

However, after installing and testing KP51, I have once again found myself reverting to the stock kernel, as I have done previously after installing KP49 and KP50. The reason is that the power kernels seem to degrade the quality of SIP and Skype calling, and this is a deal-breaker for me.

I have searched the forums and found that other users have reported similar experiences. Some of these users believe that the problem arises from the changes in the way the power kernel controls cpu scaling and power management.

Would it be possible for you to produce, or advise how to produce, an experimental kernel which follows the stock kernel's approach to cpu scaling and power management (i.e. no support for overclocking/undervolting), while retaining as many as possible of the other changes which make the power kernel so useful?

Thanks for any help you can give, and again, please accept my sincere appreciation of the work that you do.
By "degrade the quality" if you refer to skype VIDEO CALL then I had similar problem (including viewing Tv channels via SPB TV) until I installed freemangordon's patches for 720p videos from here , after that I had no more problems making/receiving VIDEO CALLS via skype (or viewing the above tv channels usig SPB TV) also I am able to view mp4, flv, HD, etc videos on my device thanks to freemangordon.
I can confirm the above because I had to re-flash my device and forgot to instal the 720p patches, try it all the above with no success until I reinstalled the above 720p patches.
I am using CSSU-Stable and KP-51r1. (sorry for my bad english)

Edit: I am using the above without overclocking my device.

Last edited by g0r; 2012-07-30 at 19:31.
 

The Following User Says Thank You to g0r For This Useful Post:
Posts: 306 | Thanked: 106 times | Joined on Feb 2010
#205
Originally Posted by zerox View Post
Firstly, thank you for all your work on the power kernel, and on the functionalities which it enables, most recently KP51 and the new USB Mode Switch.

However, after installing and testing KP51, I have once again found myself reverting to the stock kernel, as I have done previously after installing KP49 and KP50. The reason is that the power kernels seem to degrade the quality of SIP and Skype calling, and this is a deal-breaker for me.

I have searched the forums and found that other users have reported similar experiences. Some of these users believe that the problem arises from the changes in the way the power kernel controls cpu scaling and power management.

Would it be possible for you to produce, or advise how to produce, an experimental kernel which follows the stock kernel's approach to cpu scaling and power management (i.e. no support for overclocking/undervolting), while retaining as many as possible of the other changes which make the power kernel so useful?

Thanks for any help you can give, and again, please accept my sincere appreciation of the work that you do.
You can set the frequency using dbus-scripts. Look here.

This is only for limiting frequency while making sip calls.
__________________
------------------------------------------------------------------
Voice choppy on sip calls
Please vote for bug number 10388
 

The Following User Says Thank You to rajil.s For This Useful Post:
Estel's Avatar
Posts: 5,028 | Thanked: 8,613 times | Joined on Mar 2011
#206
Originally Posted by zerox View Post
Firstly, thank you for all your work on the power kernel, and on the functionalities which it enables, most recently KP51 and the new USB Mode Switch.

However, after installing and testing KP51, I have once again found myself reverting to the stock kernel, as I have done previously after installing KP49 and KP50. The reason is that the power kernels seem to degrade the quality of SIP and Skype calling, and this is a deal-breaker for me.

I have searched the forums and found that other users have reported similar experiences. Some of these users believe that the problem arises from the changes in the way the power kernel controls cpu scaling and power management.

Would it be possible for you to produce, or advise how to produce, an experimental kernel which follows the stock kernel's approach to cpu scaling and power management (i.e. no support for overclocking/undervolting), while retaining as many as possible of the other changes which make the power kernel so useful?

Thanks for any help you can give, and again, please accept my sincere appreciation of the work that you do.
Unfortunately, it happens on stock kernel too, and many wise heads were shaked at it, without results. There is "decades" old bug 10388, which, surprisingly, wasn't marked won't fix by Nokians.... But, wasn't fixed either.

It's not simply about degrading quality. the problem is that randomly - sometimes never, sometimes after few seconds, and sometimes after a long as 30 minutes or more - audio for one side (usually, one on N900) starts to sound like experiencing packet loss. Sound become, gradually, more and more distorted, lacking parts of it, to the point of absolute silence on line (for party affected by this, 2nd one hear everything fine).

Surprisingly, suspending call and immediately activating it again, fixes it - until i re-appears again. In case, where it appears after 30 minutes, it's not much of a problem, but every 30 seconds definitely is irritating thing.

It was proved beyond any doubt, that it's not related to net settings, or any other "ambient" thing. It's for sure something on N900, and I would literally bless one, who would fix it. Hoever, it doesn't seem to be kernel-power related, and for sure isn't related to frequency management. Possible candidates are sophiasip & friends, but no one knows for sure. Reading comments on bug report I've provided earlier, you will see many "I got it!" ideas, that later were proved wrong, and bug is as healthy, as it was at the beginning.

As a notorious SIP-user, I'm equally pissed about it.

/Estel
__________________
N900's aluminum backcover / body replacement
-
N900's HDMI-Out
-
Camera cover MOD
-
Measure battery's real capacity on-device
-
TrueCrypt 7.1 | ereswap | bnf
-
Hardware's mods research is costly. To support my work, please consider donating. Thank You!
 

The Following 3 Users Say Thank You to Estel For This Useful Post:
Posts: 1,808 | Thanked: 4,272 times | Joined on Feb 2011 @ Germany
#207
Hi Pali or anyone who can answer this!

I'm now following your instructions to compile kernel-power
(cd kernel-power-2.6.28; dpkg-buildpackage -b -rfakeroot)

Since I'm not familiar with dpkg stuff, I thought I might just ask you:

If I want to change the .config (which only gets applied after/during dpkg-buildpackage), can I just run "make menuconfig" after having run dpkg-buildpackage and then run it again after having edited the .config?

(I'm home alone this weekend, so I'm going to try it anyway, but a quick confirmation would be nice).

My plan is to remove some stuff and add some other stuff. Will let you know if I get anywhere..

Edit: I managed to add a patch file to the right place so that the whole debian-circus worked OK.

Unfortunately I made a beginner mistake with my N900 (replaced getbootstate with a shell script just returning "USER". Forgot to chmod +x it. G*d*mnmot*erf*cki*gsh*t! My spare N900 is essentially unflashable. Linux doesn't consider it a valid device and the syslog spits a thousand errors (-110 and so) about the connected device. In-kernel charging works, sorta, but I cannot use the dedicated mode (with the wall charger, obviously), so it charges about 50mAh per hour, at most).

I guess it's time to take the second spare..

Last edited by reinob; 2012-08-03 at 23:50.
 

The Following 2 Users Say Thank You to reinob For This Useful Post:
Posts: 43 | Thanked: 45 times | Joined on May 2010
#208
Hello.
First post here, hope I don't step on any bear traps

I've been playing with using UDF instead of fat32/ntfs/ext234/whatever on my flash devices and external hard drives, since it's a filesystem that
a) has read-write support in linux and windows (7, I think XP doesn't like it) without extra software
b) is better than fat because it supports large files
c) is better than ntfs because it supports unix semantics (and for full ntfs on linux you need ntfs-3g and fuse)

Some more info:
http://serverfault.com/questions/550...d-drive-as-udf
http://superuser.com/questions/39942...sb-flash-drive

Anyway, UDF is listed as being supported by kernel-power, but at first it didn't work for me.
I then discovered that this fix is "needed":
https://git.kernel.org/?p=linux/kern...1d4d4b9781a555
Quoting from stackexchange:
Code:
Linux with an older kernel: needs the bs=512 option to mount, because it incorrectly used 2048 instead of the device sector size (fixed in commit 1197e4d).
Would it be possible for this patch to be included in KP?

Thanks for all your work!
 

The Following 7 Users Say Thank You to [Knuckles] For This Useful Post:
Posts: 2,153 | Thanked: 8,462 times | Joined on May 2010
#209
Originally Posted by reinob View Post
Hi Pali or anyone who can answer this!

I'm now following your instructions to compile kernel-power
(cd kernel-power-2.6.28; dpkg-buildpackage -b -rfakeroot)

Since I'm not familiar with dpkg stuff, I thought I might just ask you:

If I want to change the .config (which only gets applied after/during dpkg-buildpackage), can I just run "make menuconfig" after having run dpkg-buildpackage and then run it again after having edited the .config?

(I'm home alone this weekend, so I'm going to try it anyway, but a quick confirmation would be nice).

My plan is to remove some stuff and add some other stuff. Will let you know if I get anywhere..

Edit: I managed to add a patch file to the right place so that the whole debian-circus worked OK.

Unfortunately I made a beginner mistake with my N900 (replaced getbootstate with a shell script just returning "USER". Forgot to chmod +x it. G*d*mnmot*erf*cki*gsh*t! My spare N900 is essentially unflashable. Linux doesn't consider it a valid device and the syslog spits a thousand errors (-110 and so) about the connected device. In-kernel charging works, sorta, but I cannot use the dedicated mode (with the wall charger, obviously), so it charges about 50mAh per hour, at most).

I guess it's time to take the second spare..
Hi, .config file is generated by make. You want to edit file arch/arm/configs/rx51_defconfig (make take this file and place it to /.config). But kernel-power using quilt for patches and rx51_defconfig is also patched by quilt.

So you can edit directly defconfig diff file: debian/patches/rx51_defconfig.diff

Or use quilt to apply all diff files before debian/patches/rx51_defconfig.diff, then backup arch/arm/configs/rx51_defconfig, then apply rx51_defconfig.diff, make changes and generate diff agains original rx51_defconfig.
 

The Following 2 Users Say Thank You to pali For This Useful Post:
Posts: 2,153 | Thanked: 8,462 times | Joined on May 2010
#210
Originally Posted by [Knuckles] View Post
Hello.
First post here, hope I don't step on any bear traps

I've been playing with using UDF instead of fat32/ntfs/ext234/whatever on my flash devices and external hard drives, since it's a filesystem that
a) has read-write support in linux and windows (7, I think XP doesn't like it) without extra software
b) is better than fat because it supports large files
c) is better than ntfs because it supports unix semantics (and for full ntfs on linux you need ntfs-3g and fuse)

Some more info:
http://serverfault.com/questions/550...d-drive-as-udf
http://superuser.com/questions/39942...sb-flash-drive

Anyway, UDF is listed as being supported by kernel-power, but at first it didn't work for me.
I then discovered that this fix is "needed":
https://git.kernel.org/?p=linux/kern...1d4d4b9781a555
Quoting from stackexchange:
Code:
Linux with an older kernel: needs the bs=512 option to mount, because it incorrectly used 2048 instead of the device sector size (fixed in commit 1197e4d).
Would it be possible for this patch to be included in KP?

Thanks for all your work!
UDF support in Maemo should be usefull also for USB host mode. Patch is in upstream kernel so I can try it to backport to kernel-power...
 

The Following 7 Users Say Thank You to pali For This Useful Post:
Reply

Tags
development, kernel-power

Thread Tools

 
Forum Jump


All times are GMT. The time now is 12:24.