Hi,
dandiest:
> attached please find the package including config.gz, dmesg, traces.
Unfortunately I don't understand what you want to say by these files.
Particularly dmesg(last_kmsg) doesn't contain the aufs messages. It just
reports
- error or info from devices
- oom killer activated and some processes are killed which are
Killed process 638 (android.updater)
Killed process 897 (android.browser)
Killed process 785 (tc.RosieUtility)
Killed process 657 (.searchanywhere)
Killed process 770 (com.htc.wdm)
Killed process 628 (android.vending)
Killed process 607 (ndroid.settings)
Killed process 845 (goggles.smssync)
Killed process 715 (com.htc.sdm)
Killed process 763 (tc.googlereader)
Killed process 755 (MessageUploader)
Killed process 1527 (eatherWallpaper)
Killed process 876 (d.apps.uploader)
Killed process 617 (roid.worldclock)
Killed process 588 (d.process.media)
Killed process 578 (com.htc.bgp)
Killed process 852 (ogle.android.gm)
Killed process 951 (tc.android.mail)
Killed process 823 (LocationService)
Killed process 3192 (com.htc.sdm)
Killed process 3433 (htcime:provider)
Killed process 3481 (goggles.smssync)
Is this the main reason of your problem?
- these killed processes may affect all over your system. for example
Killed process 3192 (com.htc.sdm)
binder: 369:698 transaction failed 29189, size 116-0
:::
(loop2) bdi_queue_work: wake up flush-31:0 FAIL, retries 4
(loop2) bdi_queue_work: wake up flush-7:2 FAIL, retries 4
these messages looks as the system took a time to access the loopback
mounted filesystem.
Anyway I cannot do anything with these info you provided. There is no
messages about aufs.
For your information, I have experienced memory management trouble about
linux-2.6.32 series, also have recieved bug reports about it. Although I
have not digging into the .32 memory problem, I guess it depends upon
its sub-version and configurations. For example, .32.x with configA has
no problem, but .32.y with configA has. and vice verce.
> <http://www.linux4freak.com/gitweb/>our problem is, despite without yaffs2,
> we actually tested another scenario to copy the cache.sqsh and system.sqsh
> to sdcard, and generate a system.ext2 file for rw, the frequent system
> freezing still persist.
If the .32 memory problem is the root cause of your problem, you will
meet it regardless yaffs2, and I'd suggest you to throw .32 away.
J. R. Okajima
------------------------------------------------------------------------------
Create and publish websites with WebMatrix
Use the most popular FREE web apps or write code yourself;
WebMatrix provides all the features you need to develop and
publish your website. http://p.sf.net/sfu/ms-webmatrix-sf
