Realtek Wireless firmware extention help?

[coreplayer2]

Thanks


this morning I created a new extension since both extensions made by myself and Juanito installed but failed to copy the file 8712u.ko.gz to it's desired location.

location for the driver is:
/usr/local/lib/modules/2.6.33.3-tinycore/kernel/drivers/net/wireless


So this morning I created another extension this time I cd to tmp (instead of /tmp/pkg  like yesterday)

 tmp/pkg/usr/local/lib/modules/2.6.33.3-tinycore/kernel/drivers/net/wireless/r8712u.ko.gz

cd tmp
mksquashfs pkg/ r8712u-2.6.33-tinycore.tcz

copied r8712u-2.6.33-tinycore.tcz  to optional directory and opened AppsAudit, On Boot Maintenance, then moved r8712u-2.6.33-tinycore.tcz to the onboot list, exited out of the OnBoot maintenance and rebooted.

After the reboot the extension is again loaded, it is listed in the installed list yet STILL the driver  r8712u.ko.gz can not be found anywhere on the system particularly is not listed in /usr/local/lib/modules/2.6.33.3-tinycore/kernel/drivers/net/wireless

back to square one, ok maybe two..

[coreplayer2]

Hi coreplayer2
Try this,  sudo find / -iname 8712u.ko.gz
If it's anywhere on your system, that should tell you where it is.

That's a cool command thanks


one more reboot, extension is installed (of this there is no question)  I can find only the development and backup copies of the driver now named r8712u.ko.gz   it is definitely not listed in any other location especially not in /usr/local/lib/modules/2.6.33.3-tinycore/kernel/drivers/net/wireless  where the extension should have copied it to.

but I can copy the driver manually to the proper location and load it up and works (or at least it did yesterday)

[coreplayer2]

Seems like every time I reboot there's is a different story

If I look back there is a familiar "file not found" when the file was present on the last boot.  eg, an earlier boot I was able to start the Wicd daemon now the command can't be found.  I uninstalled and reinstalled and loaded OnBoot wifi, now it can't be found either

When I run dmesg |tail -20  there is a WARNING maximal mount count reached

Checking all Md5 hashes turns up all OK.

I don't have any idea why the file system continuously gets corrupted, but this appears to be the root of all my issues.  and am not sure how to fix it either.

[Rich]

Hi coreplayer2

Considering the problems you are having I have a few questions/requests

1. How did you install TC? Did you use one of the install utilities?
2. Did you install to a hard drive or a USB memory stick?
3. When you reboot do you run a backup first or do just hit the reset or power button?
4. Could you please post the output of  showbootcodes.
5. Run the following commands:
           dmesg > dmesg.txt
           mount > mount.txt
           ps ax > ps.txt
           lsmod > lsmod.txt
6. When you go to reply, click on  Additional Options  and attach the files from step 5.

Maybe with some information about your setup someone can figure out what's going on.

[coreplayer2]

I can't believe it!!!  

I ran a file system check (which was clean) and set the max mount count to (0) I think??   and viola all my issues are resolved.   Booted with base norestore,  ran fsck &  tune2fs -i0 -c0 on /dev/hda1  & 2

tcloop is no longer empty  

all my apps function now  

no more "file not found" issues  

And the best thing is my new wireless extension loads at boot and copies the driver to the correct location and runs

double YAY!!

[Juanito]

I ran a file system check and set the max mount count to (-1) I think??   and viola all my issues are resolved.

No - setting the max count to -1 did not fix your issues.

Using fsck to repair the corruption in the file system fixed your issues.

max count triggers a warning to check the file system when the file system has been mounted {max count} times.

[coreplayer2]

Oh ok  but the fsck returned "clean"



@ Juanito
Note:  tcloop is no longer empty  

[coreplayer2]

It seems highly possible my issues were caused by file corruption, but how was this possible?  The physical drive has never had issues and is relatively new.  I'm thinking back to when these issue have occured previously (within the last few weeks) and I recollect introducing a USB thumb drive partially formatted with FAT32 as per the wicki to transfer files from a windows PC.   Could this be a cause?

[curaga]

Unclean shutdown, power cut, etc...

[coreplayer2]

Thanks,  but kinda obvious

Lately there have been several instances of similar issues, which I resolved  quickly with a wipe of the drive and a reinstall.    The more I think of this I am becoming increasingly more suspicious that each event was related to transferring files via a usb drive having a FAT32 partition  where corruption occurred shortly after the usb drives use.

Just a thought..

[gerald_clark]

It is extremely important to umount theUSB drives before unplugging.
This is true even for Windows ( Safely remove ).
FAT filesystem corruption may not be detected if the directory and FAT are
written, but not all of the data.

[coreplayer2]

It seems that all my compiling issues are resolved now after a fix of the file system, am still interested in other possible reasons (besides the obvious) for the file system corruption I experienced.   am still perusing other ideas on this.

Meanwhile the newly compiled wireless driver works great.   Only thing is the package included an early version of wpa_supplicant.  how this is used during compilation I don't know yet.    when compiling the driver what is the meaning of compiled with, or compiled against another  app source like wpa_supplicant??

The compiled driver whilst large does not apparently contain a usable wpa_supplicant as an additional wpa_supplicant extension is needed to function.

The issue I see is one wpa_supplicant version used with the driver source during compiling and another version used by a network manager?

any thoughts please?