maemo.org - Talk
Page 16 of 23 « First 61415 16 1718 Last »
Show 40 post(s) from this thread on one page

maemo.org - Talk (https://talk.maemo.org/index.php)
-   Maemo 5 / Fremantle (https://talk.maemo.org/forumdisplay.php?f=40)
-   -   [Development] kernel-power (https://talk.maemo.org/showthread.php?t=78371)

mr_pingu 2012-02-07 19:31
Re: [Development] kernel-power
 
I used maacruz kernel(pwck2) and experienced massive UI/graphical lags and even lock-ups, nothing is responding to input and needed to put my N900 away, till cpu/IO load was lessened . Then went back to omap1 (stock) and didn't experience anylags or lock-ups.

BTW
I used max freq = 900 mhz with SR on maacruz kernel. So maybe one of the patches was bugging. Trying pw49 now..

maacruz 2012-02-08 19:22
Re: [Development] kernel-power
 
Quote:
Originally Posted by mr_pingu (Post 1161826)
I used maacruz kernel(pwck2) and experienced massive UI/graphical lags and even lock-ups, nothing is responding to input and needed to put my N900 away, till cpu/IO load was lessened . Then went back to omap1 (stock) and didn't experience anylags or lock-ups.

BTW
I used max freq = 900 mhz with SR on maacruz kernel. So maybe one of the patches was bugging. Trying pw49 now..
Could you try this new build and check if the lags are gone?
https://garage.maemo.org/frs/downloa...wck3_armel.deb
https://garage.maemo.org/frs/downloa...wck3_armel.deb
https://garage.maemo.org/frs/downloa...wck3_armel.deb

freemangordon 2012-02-09 06:58
Re: [Development] kernel-power
 
Quote:
Originally Posted by maacruz (Post 1160607)
...About those patches, 0001 and 0002 change spinlocks to mutexes, so may be they cause trouble (look at 0002 comments).
Good luck!
With those 2 disabled there are no crashes so far. Will run that kernel for several days more to be 100% sure.

What did you change in the .debs attached to your last post?

mr_pingu 2012-02-09 07:38
Re: [Development] kernel-power
 
From now on I am testing these ;) Will report back...

freemangordon 2012-02-09 07:47
Re: [Development] kernel-power
 
Quote:
Originally Posted by mr_pingu (Post 1162525)
From now on I am testing these ;) Will report back...
@Pali - can you put some script in kernel-power-settings(kernel-config) to automatically replace 700,750 frequencies with 720 on profile load, using the voltage for 750? With some sort of a warning "you are using a deprecated frequencies, bla, bla". As I am afraid there will be an avalanche of reports "my battery, my CPU,..." once those go into extras.

Your scripting skills are better than mine, that's why it is better you to do it :) .

maacruz 2012-02-09 08:58
Re: [Development] kernel-power
 
Quote:
Originally Posted by freemangordon (Post 1162511)
With those 2 disabled there are no crashes so far. Will run that kernel for several days more to be 100% sure.

What did you change in the .debs attached to your last post?
Disabled the 0001-mtd and 0002-mtd patches as I think they were causing the slowdown during memory swapping.
For some reason they caused reboots in your device, while in mine only some slowdowns I failed to properly notice by putting the blame elsewere until mr-pingu opened my eyes.
Disabled thumb2 and mesh wireless too, as they have been reported here to cause instability.
It is also a 512hz kernel, as I have been running with that frequency since we first talked about it.

freemangordon 2012-02-09 09:36
Re: [Development] kernel-power
 
Quote:
Originally Posted by maacruz (Post 1162554)
Disabled the 0001-mtd and 0002-mtd patches as I think they were causing the slowdown during memory swapping.
For some reason they caused reboots in your device, while in mine only some slowdowns I failed to properly notice by putting the blame elsewere until mr-pingu opened my eyes.
Disabled thumb2 and mesh wireless too, as they have been reported here to cause instability.
It is also a 512hz kernel, as I have been running with that frequency since we first talked about it.
On the other hand those patches really seem like something useful, if there wasn't the problem they've introduced.

Is there a chance to build and test the kernel with those patches enabled and:

CONFIG_DEBUG_KERNEL=y
CONFIG_DETECT_SOFTLOCKUP=y
CONFIG_DEBUG_MUTEXES=y

and/or similar (can't remember the exact flags for spinlock/mutex debugging) to see what is going on. I have bridgedriver debugging session to do here :)

BTW I have compcache swap, that might be the reason for reboots on my device.

mr_pingu 2012-02-09 21:42
Re: [Development] kernel-power
 
Noticed not a single lag with this new kernel, maacruz. Thinking of adding 720p support, as I think it's smoother than kp49 or is that just me?

Need more info, logs of testing from me?

Am I the only one who always wondered why the omap1 kernel was running so smooth with only 600 mhz?

maacruz 2012-02-09 22:58
Re: [Development] kernel-power
 
Quote:
Originally Posted by freemangordon (Post 1162564)
On the other hand those patches really seem like something useful, if there wasn't the problem they've introduced.

Is there a chance to build and test the kernel with those patches enabled and:

CONFIG_DEBUG_KERNEL=y
CONFIG_DETECT_SOFTLOCKUP=y
CONFIG_DEBUG_MUTEXES=y

and/or similar (can't remember the exact flags for spinlock/mutex debugging) to see what is going on. I have bridgedriver debugging session to do here :)

BTW I have compcache swap, that might be the reason for reboots on my device.
I can try, but right now I'm looking at the patches and wondering what causes the problem:
0002-mtd should have no noticeable effect in the N900. If you look at the end of the patch...
Code:
/* This is used to handle contention on write/erase operations
    between partitions of the same physical chip. */
The N900 only uses one mtd partition, rootfs, so there can't be contention for accessing different partitions.

Now about 0001-mtd. It is a pretty straight forward patch, just replace all spinlocks with mutexes.
This patch shouldn't affect memory swapping because it takes place in mmc, not mtd, so the slowdown has to happen when a mmapped large file has been unmapped and then it is remapped again, a common scenario for executable files. Why the slowdown? The only reason I can think of is because the mutex has already been adquired.
If we look at the description text:
Quote:
I have checked the code of the drivers and there is no use of atomic
pathes like interrupt or timers. The mtdoops facility will also not be used
by this drivers. So it is dave to replace the spin_lock against mutex.

There is no performance regression since the mutex is normally not
acquired.
Looking at the Nokia patches (2.6.28-omap1.diff) I see a lot of changes to the mtd subsystem, including a new mtd driver, so it is quite possible that the assumptions taken for this patch are wrong in the N900, requiring substantial rework.
Since it is doubtful that this patch could provide any noticeable gain in the N900, and it is confirmed that it causes problems, and it would require a lot of work, I think it is better just to drop them.

maacruz 2012-02-09 23:11
Re: [Development] kernel-power
 
Quote:
Originally Posted by mr_pingu (Post 1162849)
Noticed not a single lag with this new kernel, maacruz. Thinking of adding 720p support, as I think it's smoother than kp49 or is that just me?
720p is already enabled in the kernel, although I haven't tested it yet, so you can go ahead and install the debs if you wish.
The ck patches are supposed to increase performance for desktop-like usage patterns, so it is possible that it were "smoother".
Quote:
Need more info, logs of testing from me?

Am I the only one who always wondered why the omap1 kernel was running so smooth with only 600 mhz?
No, but now that you already mentioned it, if you like testing it would be interesting to test if there is any difference at all in HD recording performance or jerkiness compared to kp49.


All times are GMT. The time now is 12:53.
Page 16 of 23 « First 61415 16 1718 Last »
Show 40 post(s) from this thread on one page

vBulletin® Version 3.8.8