![]() |
Re: [Announce] kernel-power v48
Could you please someone(Estel?) who experiences "Save As" reboot to try kernel from here and report if crashes continue.
Also I would like to ask if it is possible to install syslog (after installing kernel) apt-get install sysklogd and to post /var/log/syslog here after several "Save As" operations. There is additional logging enabled in the kernel, so I hope I will be able to have stack dump of buggy userland library/application Thanks EDIT: scratch the above, I was able to reproduce the crash, will investigate it by myself |
Re: [Announce] kernel-power v48
Aye, but if You need any help later, I'm more than willing to help :)
|
Re: [Announce] kernel-power v48
1 Attachment(s)
@colin.stephane - please copy the attached patch to debian/patches and add it to series file. it fixes leafpad "Save As" crash.
Anyway the root of the problem is still unknown. I suspect hildonfm or similar, someone is traversing all the filesystem, executing FAT_IOCTL_GET_ATTRIBUTES to all the files. It seems this has been reported several times (here for example) but noone has find the root cause. The attached patch will solve kernel crash but reboot will happen if you have watchdog disabled. In case watchdog is disabled you will have a reboot 30 seconds after "Save As" is pressed, because wd file is opened and then closed. When I have time I will find what is the root cause of the problem. |
Re: [Announce] kernel-power v48
but please colin.stephane consider to enable the injection drivers
just like the original kp 48 |
Re: [Announce] kernel-power v48
Quote:
A++ |
Re: [Announce] kernel-power v48
Quote:
http://bigbob.fun.free.fr/maemo-kernel-power/ Edit 01/09/2011 - 10:15 : pushed new one, this time I put 'fmtxpower' script in '/etc/event.d' properly, in the precedent build I made a mistake putting the file in '/etc', so for the ones that have used the precedent, please delete the file manually ... A++ |
Re: [Announce] kernel-power v48
how do i get back to kp47?
|
Re: [Announce] kernel-power v48
Quote:
@colin.stephane - it is your ramzswap module patch that is causing reboots, here is the backtrace Code:
[ 333.989776] CPU: 0 Not tainted (2.6.28.10-power48 #1) @colin.stephane - Will you please test the kernel at least once before uploading a link here. It is not only to put some patches and to leave others with reflashing and reboot loops. And BTW where is this patch? I mean .diff file. Here is the kernel with leafpad crash fixed an working ramzswap module. |
Re: [Announce] kernel-power v48
Quote:
Let me check where is located the problem ... A++ |
Re: [Announce] kernel-power v48
Which you would easily catch if you insmod ramzswap.ko at least once, thus my point.
|
Re: [Announce] kernel-power v48
Quote:
|
Re: [Announce] kernel-power v48
Quote:
Anyhow, I found my error quickly :D For something I don't really know, I have pushed bad 'ramzswap_drv.c' file inside my kernel source compilation tree ... Anyhow, right now it's FIXED !!! So let the current kernel compilation finishing with this time YOUR version of 'ramzswap' module properly backported, and I thing you regain smile (at least, I hope). Stay tunned ... A++ |
Re: [Announce] kernel-power v48
Buggy library leading to previous kernel crashes found. It is gnome-vfs-filechooser-backend. Fix implemented here. Will try to push it as a part of the next CSSU update.
|
Re: [Announce] kernel-power v48
Quote:
|
Re: [Announce] kernel-power v48
THIS TIME IT'S REALLY FIXED !!!
Fixed my brain, and tracked why reboots occurs when using 'ramzswap.ko' from my precedent integration of freemangordon working module (next time it's better if I don't mix source code :D). Download from here : This power-kernel contain the working freemangordon ramzswap.ko module directly, no more need to compile yourself under scratchbox or me to provide as external module ... Changelog : Latests commits by Pali before he have school obligations (see post #) : Commits '7ab052e2cf27b79aa8eea32b62f8cde752a87373' & '55fc617b021546d885dc413e15d99690488b3906' :
New changes by myself :
Quote:
A++ PS : freemangordon, sorry for all the bad things occurred ... |
Re: [Announce] kernel-power v48
Quote:
Do I have to leave in place 'dspbridge_ioctl_buffer_overrun.diff' patch inside my kernel-power source or does it no more required since you discovered Buggy library 'gnome-vfs-filechooser-backend' ??? A++ |
Re: [Announce] kernel-power v48
No problem, sh|t happens, but you lost your beer :p.
|
Re: [Announce] kernel-power v48
Quote:
|
Re: [Announce] kernel-power v48
Quote:
A++ |
Re: [Announce] kernel-power v48
@colin.stephane - ok, ramzswap seems to work, but it has statistics enabled, will you please consider downloading compcache.zip again and checking all the modifications. In latest Makefile from the above .zip statistics are disabled (for performance reasons). I have also add some #ifdefs in ramzswap_drv.c file. And please, upload .diif file here, so I can add it to my kernel tree and check what is going on, ok?
Thanks |
Re: [Announce] kernel-power v48
Quote:
I have disabled statistics, it was due to the backport of kernel module saying to activate statistics by default when you enable the 'ramzswap' kernel module compilation, nothing to do with your compcache.zip archive at all. Anyhow, here is the 'adding-ramzswap-driver.diff' file ... Obviously the compilation without statistics activated is running right now ... A++ |
Re: [Announce] kernel-power v48
Are you saying you are not using compcache from here?
EDIT: looking through the .diff it seems correct source is used. |
Re: [Announce] kernel-power v48
Quote:
So, to disable the statistics, it's on my side, like you can see in the diff file ;) ... A++ |
Re: [Announce] kernel-power v48
Quote:
|
Re: [Announce] kernel-power v48
Quote:
Done !!! I have updated the post # to reflect the disabling of statistics for the module 'ramzswap' ... A++ |
Re: [Announce] kernel-power v48
Hi.
Due several troubles experimented with the pk48 I had to go to pk47 version. These problems were, in my personal case: - Not able to video playback. Even the Nokia video on startup was gone. - Not able to video recording. - Some trouble with my "fonera" wifi router N900 not seeing the wifi signal from it. - Worst battery performance than over pk47. - Issue with battery recharging when sometimes the wall charger were identified by the machinne like an usb device asking me for pc suite or usb mode. So, going back to pk47 but really thanks for the effort. I will have an eye on this. I will like to know if I was the only one with all these issues. Maybe all these things came with some personal configurations? Even when I think I have a "standard" installation without any strange scripts doing strange things. Thanks again, guys, and keep the awesome work. |
Re: [Announce] kernel-power v48
Quote:
For sure your problems doesn't come from kp48 ... You need to investigate deeply to find what causing these troubles. A++ |
Re: [Announce] kernel-power v48
Quote:
But once back on kp47 all the mess gone away automagically. Anyway I will go on your advice to find out the problems. Thanks for your work. |
Re: [Announce] kernel-power v48
Quote:
A++ |
Re: [Announce] kernel-power v48
Quote:
I just played around the repos pk48 version. I do want to test it? :confused: |
Re: [Announce] kernel-power v48
Quote:
Code:
-bash-2.05b# kernel-config show Code:
clear && ID=$(id -un) ; if [ ${ID} != "root" ] ; then echo "You are not 'root', please use this command line as 'root' ..." ; else echo "Ok, you are 'root' ..." ; if grep -q "blacklist bq27x00_battery" /etc/modprobe.d/blacklist 2>/dev/null ; then echo "Patch already applyed, nothing to be done ..." ; echo "" ; else echo "Applying patch now ..." ; echo "blacklist bq27x00_battery" >>/etc/modprobe.d/blacklist ; echo "" ; echo "DONE, PLEASE REBOOT YOUR DEVICE ..." ; echo "" ; fi ; fi Finally, after that, try installing kp48 from the post #. Reboot, and step by step debug troubles ... A++ |
Re: [Announce] kernel-power v48
colin.stephane how much your battery lasts on continues usage time?
|
Re: [Announce] kernel-power v48
Quote:
BatteryGraph A++ |
Re: [Announce] kernel-power v48
@colin.stephane: Many thanks. I will go with your instructions. I will let you know with the results.
|
Re: [Announce] kernel-power v48
Guys I seem to be having a weird problem with microB after installing v48 from the repos. It doesn't delete my browsing history when I tell it to. Is this happening to anyone else? Also, does anyone know how to do it manually?
Thanks in advance. EDIT: Nevermind, guys. I just deleted places.sqlite and places.sqlite-journal from microB's folder and it seems fine again. Dunno what happened there. |
Re: [Announce] kernel-power v48
I'm running kp48 and deleting browser history works just fine (I've checked it 5 seconds ago). So, it seems unrelated.
|
Re: [Announce] kernel-power v48
A bit of a noobish question. I've had power kernel on for years for overclocking but not used many of its other features.
Now im trying to use cifs which I believe it supports, however upon trying to mount I get an error "No such device". Searching billy google it seems to suggest that cifs is not supported (the results were on about other devices/os's though) so I carried on searching on here but I've got a discrepancy. In FAM it reports my power kernel as v48, however running "uname -r" in xterm reports 47, so which is it? I wanted to figure this so i could search if there was any issues with cifs in that particular version, but I'm not sure which I've got. (Yes i've got devel repo enabled) Soooo, any ideas why cifs might be reporting that? Cheers |
Re: [Announce] kernel-power v48
Quote:
What about the results of these commands as 'root' inside a terminal console ? (copy/paste only the part in bold) : Code:
-bash-2.05b# modprobe cifs ; lsmod | grep cifs |
Re: [Announce] kernel-power v48
Quote:
Code:
/home/user # modprobe cifs ; lsmod | grep cifs Code:
mount -t cifs //LENOVO-DT/f$/ /home/user/MyDocs/KaosExt -o user=remote,pass=****,ip=192.168.2.2,direct |
Re: [Announce] kernel-power v48
Quote:
Well, to be able to mount samba shares (cifs) we need mount.cifs contained in package 'smbfs' that depend himself on package 'samba-common'. And there is error in postinstallation script of the package 'samba-common', leaving the packages unconfigured. I have tried to fix it but now the sed regex used in this script was mis-interpreted by busybox sed :-( I'm trying to do my best but I do not promise anything on this subject ... A++ |
All times are GMT. The time now is 16:39. |
vBulletin® Version 3.8.8